diff --git a/Documentation/dontdiff b/Documentation/dontdiff index 0c083c5..bf13011 100644 --- a/Documentation/dontdiff +++ b/Documentation/dontdiff @@ -2,9 +2,11 @@ *.aux *.bin *.bz2 +*.c.[012]*.* *.cis *.cpio *.csp +*.dbg *.dsp *.dvi *.elf @@ -14,6 +16,7 @@ *.gcov *.gen.S *.gif +*.gmo *.grep *.grp *.gz @@ -48,9 +51,11 @@ *.tab.h *.tex *.ver +*.vim *.xml *.xz *_MODULES +*_reg_safe.h *_vga16.c *~ \#*# @@ -69,6 +74,7 @@ Image Module.markers Module.symvers PENDING +PERF* SCCS System.map* TAGS @@ -92,19 +98,24 @@ bounds.h bsetup btfixupprep build +builtin-policy.h bvmlinux bzImage* capability_names.h capflags.c classlist.h* +clut_vga16.c +common-cmds.h comp*.log compile.h* conf config config-* config_data.h* +config.c config.mak config.mak.autogen +config.tmp conmakehash consolemap_deftbl.c* cpustr.h @@ -115,9 +126,11 @@ devlist.h* dnotify_test docproc dslm +dtc-lexer.lex.c elf2ecoff elfconfig.h* evergreen_reg_safe.h +exception_policy.conf fixdep flask.h fore200e_mkfirm @@ -125,12 +138,15 @@ fore200e_pca_fw.c* gconf gconf.glade.h gen-devlist +gen-kdb_cmds.c gen_crc32table gen_init_cpio generated genheaders genksyms *_gray256.c +hash +hid-example hpet_example hugepage-mmap hugepage-shm @@ -145,7 +161,7 @@ int32.c int4.c int8.c kallsyms -kconfig +kern_constants.h keywords.c ksym.c* ksym.h* @@ -153,7 +169,7 @@ kxgettext lkc_defs.h lex.c lex.*.c -linux +lib1funcs.S logo_*.c logo_*_clut224.c logo_*_mono.c @@ -165,14 +181,15 @@ machtypes.h map map_hugetlb maui_boot.h -media mconf +mdp miboot* mk_elfconfig mkboot mkbugboot mkcpustr mkdep +mkpiggy mkprep mkregtable mktables @@ -208,6 +225,7 @@ r300_reg_safe.h r420_reg_safe.h r600_reg_safe.h recordmcount +regdb.c relocs rlim_names.h rn50_reg_safe.h @@ -218,6 +236,7 @@ setup setup.bin setup.elf sImage +slabinfo sm_tbl* split-include syscalltab.h @@ -228,6 +247,7 @@ tftpboot.img timeconst.h times.h* trix_boot.h +user_constants.h utsrelease.h* vdso-syms.lds vdso.lds @@ -245,7 +265,9 @@ vmlinux vmlinux-* vmlinux.aout vmlinux.bin.all +vmlinux.bin.bz2 vmlinux.lds +vmlinux.relocs vmlinuz voffset.h vsyscall.lds @@ -253,9 +275,11 @@ vsyscall_32.lds wanxlfw.inc uImage unifdef +utsrelease.h wakeup.bin wakeup.elf wakeup.lds zImage* zconf.hash.c +zconf.lex.c zoffset.h diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index d99fd9c..8689fef 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1977,6 +1977,13 @@ bytes respectively. Such letter suffixes can also be entirely omitted. the specified number of seconds. This is to be used if your oopses keep scrolling off the screen. + pax_nouderef [X86] disables UDEREF. Most likely needed under certain + virtualization environments that don't cope well with the + expand down segment used by UDEREF on X86-32 or the frequent + page table updates on X86-64. + + pax_softmode= 0/1 to disable/enable PaX softmode on boot already. + pcbit= [HW,ISDN] pcd. [PARIDE] diff --git a/Makefile b/Makefile index 9cd6941..92e68ff 100644 --- a/Makefile +++ b/Makefile @@ -245,8 +245,9 @@ CONFIG_SHELL := $(shell if [ -x "$$BASH" ]; then echo $$BASH; \ HOSTCC = gcc HOSTCXX = g++ -HOSTCFLAGS = -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -fomit-frame-pointer -HOSTCXXFLAGS = -O2 +HOSTCFLAGS = -Wall -W -Wmissing-prototypes -Wstrict-prototypes -Wno-unused-parameter -Wno-missing-field-initializers -O2 -fomit-frame-pointer -fno-delete-null-pointer-checks +HOSTCLFAGS += $(call cc-option, -Wno-empty-body) +HOSTCXXFLAGS = -O2 -Wall -W -fno-delete-null-pointer-checks # Decide whether to build built-in, modular, or both. # Normally, just do built-in. @@ -407,8 +408,8 @@ export RCS_TAR_IGNORE := --exclude SCCS --exclude BitKeeper --exclude .svn --exc # Rules shared between *config targets and build targets # Basic helpers built in scripts/ -PHONY += scripts_basic -scripts_basic: +PHONY += scripts_basic gcc-plugins +scripts_basic: gcc-plugins $(Q)$(MAKE) $(build)=scripts/basic $(Q)rm -f .tmp_quiet_recordmcount @@ -564,6 +565,55 @@ else KBUILD_CFLAGS += -O2 endif +ifndef DISABLE_PAX_PLUGINS +ifeq ($(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-plugin.sh "$(HOSTCC)" "$(CC)"), y) +ifndef DISABLE_PAX_CONSTIFY_PLUGIN +ifndef CONFIG_UML +CONSTIFY_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/constify_plugin.so -DCONSTIFY_PLUGIN +endif +endif +ifdef CONFIG_PAX_MEMORY_STACKLEAK +STACKLEAK_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/stackleak_plugin.so -DSTACKLEAK_PLUGIN +STACKLEAK_PLUGIN_CFLAGS += -fplugin-arg-stackleak_plugin-track-lowest-sp=100 +endif +ifdef CONFIG_KALLOCSTAT_PLUGIN +KALLOCSTAT_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/kallocstat_plugin.so +endif +ifdef CONFIG_PAX_KERNEXEC_PLUGIN +KERNEXEC_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/kernexec_plugin.so +KERNEXEC_PLUGIN_CFLAGS += -fplugin-arg-kernexec_plugin-method=$(CONFIG_PAX_KERNEXEC_PLUGIN_METHOD) -DKERNEXEC_PLUGIN +KERNEXEC_PLUGIN_AFLAGS := -DKERNEXEC_PLUGIN +endif +ifdef CONFIG_CHECKER_PLUGIN +ifeq ($(call cc-ifversion, -ge, 0406, y), y) +CHECKER_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/checker_plugin.so -DCHECKER_PLUGIN +endif +endif +COLORIZE_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/colorize_plugin.so +ifdef CONFIG_PAX_SIZE_OVERFLOW +SIZE_OVERFLOW_PLUGIN := -fplugin=$(objtree)/tools/gcc/size_overflow_plugin.so -DSIZE_OVERFLOW_PLUGIN +endif +GCC_PLUGINS_CFLAGS := $(CONSTIFY_PLUGIN_CFLAGS) $(STACKLEAK_PLUGIN_CFLAGS) $(KALLOCSTAT_PLUGIN_CFLAGS) +GCC_PLUGINS_CFLAGS += $(KERNEXEC_PLUGIN_CFLAGS) $(CHECKER_PLUGIN_CFLAGS) $(COLORIZE_PLUGIN_CFLAGS) $(SIZE_OVERFLOW_PLUGIN) +GCC_PLUGINS_AFLAGS := $(KERNEXEC_PLUGIN_AFLAGS) +export CONSTIFY_PLUGIN STACKLEAK_PLUGIN KERNEXEC_PLUGIN CHECKER_PLUGIN SIZE_OVERFLOW_PLUGIN +ifeq ($(KBUILD_EXTMOD),) +gcc-plugins: + $(Q)$(MAKE) $(build)=tools/gcc +else +gcc-plugins: ; +endif +else +gcc-plugins: +ifeq ($(call cc-ifversion, -ge, 0405, y), y) + $(error Your gcc installation does not support plugins. If the necessary headers for plugin support are missing, they should be installed. On Debian, apt-get install gcc--plugin-dev. If you choose to ignore this error and lessen the improvements provided by this patch, re-run make with the DISABLE_PAX_PLUGINS=y argument.)) +else + $(Q)echo "warning, your gcc version does not support plugins, you should upgrade it to gcc 4.5 at least" +endif + $(Q)echo "PAX_MEMORY_STACKLEAK and other features will be less secure" +endif +endif + include $(srctree)/arch/$(SRCARCH)/Makefile ifneq ($(CONFIG_FRAME_WARN),0) @@ -708,7 +758,7 @@ export mod_strip_cmd ifeq ($(KBUILD_EXTMOD),) -core-y += kernel/ mm/ fs/ ipc/ security/ crypto/ block/ +core-y += kernel/ mm/ fs/ ipc/ security/ crypto/ block/ grsecurity/ vmlinux-dirs := $(patsubst %/,%,$(filter %/, $(init-y) $(init-m) \ $(core-y) $(core-m) $(drivers-y) $(drivers-m) \ @@ -932,6 +982,8 @@ vmlinux.o: $(modpost-init) $(vmlinux-main) FORCE # The actual objects are generated when descending, # make sure no implicit rule kicks in +$(sort $(vmlinux-init) $(vmlinux-main)) $(vmlinux-lds): KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +$(sort $(vmlinux-init) $(vmlinux-main)) $(vmlinux-lds): KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) $(sort $(vmlinux-init) $(vmlinux-main)) $(vmlinux-lds): $(vmlinux-dirs) ; # Handle descending into subdirectories listed in $(vmlinux-dirs) @@ -941,7 +993,7 @@ $(sort $(vmlinux-init) $(vmlinux-main)) $(vmlinux-lds): $(vmlinux-dirs) ; # Error messages still appears in the original language PHONY += $(vmlinux-dirs) -$(vmlinux-dirs): prepare scripts +$(vmlinux-dirs): gcc-plugins prepare scripts $(Q)$(MAKE) $(build)=$@ # Store (new) KERNELRELASE string in include/config/kernel.release @@ -985,6 +1037,7 @@ prepare0: archprepare FORCE $(Q)$(MAKE) $(build)=. # All the preparing.. +prepare: KBUILD_CFLAGS := $(filter-out $(GCC_PLUGINS_CFLAGS),$(KBUILD_CFLAGS)) prepare: prepare0 # Generate some files @@ -1089,6 +1142,8 @@ all: modules # using awk while concatenating to the final file. PHONY += modules +modules: KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +modules: KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) modules: $(vmlinux-dirs) $(if $(KBUILD_BUILTIN),vmlinux) modules.builtin $(Q)$(AWK) '!x[$$0]++' $(vmlinux-dirs:%=$(objtree)/%/modules.order) > $(objtree)/modules.order @$(kecho) ' Building modules, stage 2.'; @@ -1104,7 +1159,7 @@ modules.builtin: $(vmlinux-dirs:%=%/modules.builtin) # Target to prepare building external modules PHONY += modules_prepare -modules_prepare: prepare scripts +modules_prepare: gcc-plugins prepare scripts # Target to install modules PHONY += modules_install @@ -1201,6 +1256,7 @@ distclean: mrproper \( -name '*.orig' -o -name '*.rej' -o -name '*~' \ -o -name '*.bak' -o -name '#*#' -o -name '.*.orig' \ -o -name '.*.rej' \ + -o -name '.*.rej' -o -name '*.so' \ -o -name '*%' -o -name '.*.cmd' -o -name 'core' \) \ -type f -print | xargs rm -f @@ -1361,6 +1417,8 @@ PHONY += $(module-dirs) modules $(module-dirs): crmodverdir $(objtree)/Module.symvers $(Q)$(MAKE) $(build)=$(patsubst _module_%,%,$@) +modules: KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +modules: KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) modules: $(module-dirs) @$(kecho) ' Building modules, stage 2.'; $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modpost @@ -1487,17 +1545,21 @@ else target-dir = $(if $(KBUILD_EXTMOD),$(dir $<),$(dir $@)) endif -%.s: %.c prepare scripts FORCE +%.s: KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +%.s: KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) +%.s: %.c gcc-plugins prepare scripts FORCE $(Q)$(MAKE) $(build)=$(build-dir) $(target-dir)$(notdir $@) %.i: %.c prepare scripts FORCE $(Q)$(MAKE) $(build)=$(build-dir) $(target-dir)$(notdir $@) -%.o: %.c prepare scripts FORCE +%.o: KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +%.o: KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) +%.o: %.c gcc-plugins prepare scripts FORCE $(Q)$(MAKE) $(build)=$(build-dir) $(target-dir)$(notdir $@) %.lst: %.c prepare scripts FORCE $(Q)$(MAKE) $(build)=$(build-dir) $(target-dir)$(notdir $@) -%.s: %.S prepare scripts FORCE +%.s: %.S gcc-plugins prepare scripts FORCE $(Q)$(MAKE) $(build)=$(build-dir) $(target-dir)$(notdir $@) -%.o: %.S prepare scripts FORCE +%.o: %.S gcc-plugins prepare scripts FORCE $(Q)$(MAKE) $(build)=$(build-dir) $(target-dir)$(notdir $@) %.symtypes: %.c prepare scripts FORCE $(Q)$(MAKE) $(build)=$(build-dir) $(target-dir)$(notdir $@) @@ -1507,11 +1569,15 @@ endif $(cmd_crmodverdir) $(Q)$(MAKE) KBUILD_MODULES=$(if $(CONFIG_MODULES),1) \ $(build)=$(build-dir) -%/: prepare scripts FORCE +%/: KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +%/: KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) +%/: gcc-plugins prepare scripts FORCE $(cmd_crmodverdir) $(Q)$(MAKE) KBUILD_MODULES=$(if $(CONFIG_MODULES),1) \ $(build)=$(build-dir) -%.ko: prepare scripts FORCE +%.ko: KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +%.ko: KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) +%.ko: gcc-plugins prepare scripts FORCE $(cmd_crmodverdir) $(Q)$(MAKE) KBUILD_MODULES=$(if $(CONFIG_MODULES),1) \ $(build)=$(build-dir) $(@:.ko=.o) diff --git a/arch/alpha/include/asm/atomic.h b/arch/alpha/include/asm/atomic.h index 640f909..48b6597 100644 --- a/arch/alpha/include/asm/atomic.h +++ b/arch/alpha/include/asm/atomic.h @@ -250,6 +250,16 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) #define atomic_dec(v) atomic_sub(1,(v)) #define atomic64_dec(v) atomic64_sub(1,(v)) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + #define smp_mb__before_atomic_dec() smp_mb() #define smp_mb__after_atomic_dec() smp_mb() #define smp_mb__before_atomic_inc() smp_mb() diff --git a/arch/alpha/include/asm/cache.h b/arch/alpha/include/asm/cache.h index ad368a9..fbe0f25 100644 --- a/arch/alpha/include/asm/cache.h +++ b/arch/alpha/include/asm/cache.h @@ -4,19 +4,19 @@ #ifndef __ARCH_ALPHA_CACHE_H #define __ARCH_ALPHA_CACHE_H +#include /* Bytes per L1 (data) cache line. */ #if defined(CONFIG_ALPHA_GENERIC) || defined(CONFIG_ALPHA_EV6) -# define L1_CACHE_BYTES 64 # define L1_CACHE_SHIFT 6 #else /* Both EV4 and EV5 are write-through, read-allocate, direct-mapped, physical. */ -# define L1_CACHE_BYTES 32 # define L1_CACHE_SHIFT 5 #endif +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define SMP_CACHE_BYTES L1_CACHE_BYTES #endif diff --git a/arch/alpha/include/asm/elf.h b/arch/alpha/include/asm/elf.h index da5449e..7418343 100644 --- a/arch/alpha/include/asm/elf.h +++ b/arch/alpha/include/asm/elf.h @@ -90,6 +90,13 @@ typedef elf_fpreg_t elf_fpregset_t[ELF_NFPREG]; #define ELF_ET_DYN_BASE (TASK_UNMAPPED_BASE + 0x1000000) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (current->personality & ADDR_LIMIT_32BIT ? 0x10000 : 0x120000000UL) + +#define PAX_DELTA_MMAP_LEN (current->personality & ADDR_LIMIT_32BIT ? 14 : 28) +#define PAX_DELTA_STACK_LEN (current->personality & ADDR_LIMIT_32BIT ? 14 : 19) +#endif + /* $0 is set by ld.so to a pointer to a function which might be registered using atexit. This provides a mean for the dynamic linker to call DT_FINI functions for shared libraries that have diff --git a/arch/alpha/include/asm/pgalloc.h b/arch/alpha/include/asm/pgalloc.h index bc2a0da..8ad11ee 100644 --- a/arch/alpha/include/asm/pgalloc.h +++ b/arch/alpha/include/asm/pgalloc.h @@ -29,6 +29,12 @@ pgd_populate(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmd) pgd_set(pgd, pmd); } +static inline void +pgd_populate_kernel(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmd) +{ + pgd_populate(mm, pgd, pmd); +} + extern pgd_t *pgd_alloc(struct mm_struct *mm); static inline void diff --git a/arch/alpha/include/asm/pgtable.h b/arch/alpha/include/asm/pgtable.h index de98a73..bd4f1f8 100644 --- a/arch/alpha/include/asm/pgtable.h +++ b/arch/alpha/include/asm/pgtable.h @@ -101,6 +101,17 @@ struct vm_area_struct; #define PAGE_SHARED __pgprot(_PAGE_VALID | __ACCESS_BITS) #define PAGE_COPY __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW) #define PAGE_READONLY __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW) + +#ifdef CONFIG_PAX_PAGEEXEC +# define PAGE_SHARED_NOEXEC __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOE) +# define PAGE_COPY_NOEXEC __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW | _PAGE_FOE) +# define PAGE_READONLY_NOEXEC __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW | _PAGE_FOE) +#else +# define PAGE_SHARED_NOEXEC PAGE_SHARED +# define PAGE_COPY_NOEXEC PAGE_COPY +# define PAGE_READONLY_NOEXEC PAGE_READONLY +#endif + #define PAGE_KERNEL __pgprot(_PAGE_VALID | _PAGE_ASM | _PAGE_KRE | _PAGE_KWE) #define _PAGE_NORMAL(x) __pgprot(_PAGE_VALID | __ACCESS_BITS | (x)) diff --git a/arch/alpha/kernel/module.c b/arch/alpha/kernel/module.c index 2fd00b7..cfd5069 100644 --- a/arch/alpha/kernel/module.c +++ b/arch/alpha/kernel/module.c @@ -160,7 +160,7 @@ apply_relocate_add(Elf64_Shdr *sechdrs, const char *strtab, /* The small sections were sorted to the end of the segment. The following should definitely cover them. */ - gp = (u64)me->module_core + me->core_size - 0x8000; + gp = (u64)me->module_core_rw + me->core_size_rw - 0x8000; got = sechdrs[me->arch.gotsecindex].sh_addr; for (i = 0; i < n; i++) { diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c index 01e8715..be0e80f 100644 --- a/arch/alpha/kernel/osf_sys.c +++ b/arch/alpha/kernel/osf_sys.c @@ -1147,7 +1147,7 @@ arch_get_unmapped_area_1(unsigned long addr, unsigned long len, /* At this point: (!vma || addr < vma->vm_end). */ if (limit - len < addr) return -ENOMEM; - if (!vma || addr + len <= vma->vm_start) + if (check_heap_stack_gap(vma, addr, len)) return addr; addr = vma->vm_end; vma = vma->vm_next; @@ -1183,6 +1183,10 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, merely specific addresses, but regions of memory -- perhaps this feature should be incorporated into all ports? */ +#ifdef CONFIG_PAX_RANDMMAP + if (!(current->mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { addr = arch_get_unmapped_area_1 (PAGE_ALIGN(addr), len, limit); if (addr != (unsigned long) -ENOMEM) @@ -1190,8 +1194,8 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, } /* Next, try allocating at TASK_UNMAPPED_BASE. */ - addr = arch_get_unmapped_area_1 (PAGE_ALIGN(TASK_UNMAPPED_BASE), - len, limit); + addr = arch_get_unmapped_area_1 (PAGE_ALIGN(current->mm->mmap_base), len, limit); + if (addr != (unsigned long) -ENOMEM) return addr; diff --git a/arch/alpha/mm/fault.c b/arch/alpha/mm/fault.c index fadd5f8..904e73a 100644 --- a/arch/alpha/mm/fault.c +++ b/arch/alpha/mm/fault.c @@ -54,6 +54,124 @@ __load_new_mm_context(struct mm_struct *next_mm) __reload_thread(pcb); } +#ifdef CONFIG_PAX_PAGEEXEC +/* + * PaX: decide what to do with offenders (regs->pc = fault address) + * + * returns 1 when task should be killed + * 2 when patched PLT trampoline was detected + * 3 when unpatched PLT trampoline was detected + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + +#ifdef CONFIG_PAX_EMUPLT + int err; + + do { /* PaX: patched PLT emulation #1 */ + unsigned int ldah, ldq, jmp; + + err = get_user(ldah, (unsigned int *)regs->pc); + err |= get_user(ldq, (unsigned int *)(regs->pc+4)); + err |= get_user(jmp, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((ldah & 0xFFFF0000U) == 0x277B0000U && + (ldq & 0xFFFF0000U) == 0xA77B0000U && + jmp == 0x6BFB0000U) + { + unsigned long r27, addr; + unsigned long addrh = (ldah | 0xFFFFFFFFFFFF0000UL) << 16; + unsigned long addrl = ldq | 0xFFFFFFFFFFFF0000UL; + + addr = regs->r27 + ((addrh ^ 0x80000000UL) + 0x80000000UL) + ((addrl ^ 0x8000UL) + 0x8000UL); + err = get_user(r27, (unsigned long *)addr); + if (err) + break; + + regs->r27 = r27; + regs->pc = r27; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #2 */ + unsigned int ldah, lda, br; + + err = get_user(ldah, (unsigned int *)regs->pc); + err |= get_user(lda, (unsigned int *)(regs->pc+4)); + err |= get_user(br, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((ldah & 0xFFFF0000U) == 0x277B0000U && + (lda & 0xFFFF0000U) == 0xA77B0000U && + (br & 0xFFE00000U) == 0xC3E00000U) + { + unsigned long addr = br | 0xFFFFFFFFFFE00000UL; + unsigned long addrh = (ldah | 0xFFFFFFFFFFFF0000UL) << 16; + unsigned long addrl = lda | 0xFFFFFFFFFFFF0000UL; + + regs->r27 += ((addrh ^ 0x80000000UL) + 0x80000000UL) + ((addrl ^ 0x8000UL) + 0x8000UL); + regs->pc += 12 + (((addr ^ 0x00100000UL) + 0x00100000UL) << 2); + return 2; + } + } while (0); + + do { /* PaX: unpatched PLT emulation */ + unsigned int br; + + err = get_user(br, (unsigned int *)regs->pc); + + if (!err && (br & 0xFFE00000U) == 0xC3800000U) { + unsigned int br2, ldq, nop, jmp; + unsigned long addr = br | 0xFFFFFFFFFFE00000UL, resolver; + + addr = regs->pc + 4 + (((addr ^ 0x00100000UL) + 0x00100000UL) << 2); + err = get_user(br2, (unsigned int *)addr); + err |= get_user(ldq, (unsigned int *)(addr+4)); + err |= get_user(nop, (unsigned int *)(addr+8)); + err |= get_user(jmp, (unsigned int *)(addr+12)); + err |= get_user(resolver, (unsigned long *)(addr+16)); + + if (err) + break; + + if (br2 == 0xC3600000U && + ldq == 0xA77B000CU && + nop == 0x47FF041FU && + jmp == 0x6B7B0000U) + { + regs->r28 = regs->pc+4; + regs->r27 = addr+16; + regs->pc = resolver; + return 3; + } + } + } while (0); +#endif + + return 1; +} + +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 5; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif /* * This routine handles page faults. It determines the address, @@ -131,8 +249,29 @@ do_page_fault(unsigned long address, unsigned long mmcsr, good_area: si_code = SEGV_ACCERR; if (cause < 0) { - if (!(vma->vm_flags & VM_EXEC)) + if (!(vma->vm_flags & VM_EXEC)) { + +#ifdef CONFIG_PAX_PAGEEXEC + if (!(mm->pax_flags & MF_PAX_PAGEEXEC) || address != regs->pc) + goto bad_area; + + up_read(&mm->mmap_sem); + switch (pax_handle_fetch_fault(regs)) { + +#ifdef CONFIG_PAX_EMUPLT + case 2: + case 3: + return; +#endif + + } + pax_report_fault(regs, (void *)regs->pc, (void *)rdusp()); + do_group_exit(SIGKILL); +#else goto bad_area; +#endif + + } } else if (!cause) { /* Allow reads even for write-only mappings */ if (!(vma->vm_flags & (VM_READ | VM_WRITE))) diff --git a/arch/arm/include/asm/atomic.h b/arch/arm/include/asm/atomic.h index 86976d0..8e07f84 100644 --- a/arch/arm/include/asm/atomic.h +++ b/arch/arm/include/asm/atomic.h @@ -15,6 +15,10 @@ #include #include +#ifdef CONFIG_GENERIC_ATOMIC64 +#include +#endif + #define ATOMIC_INIT(i) { (i) } #ifdef __KERNEL__ @@ -25,7 +29,15 @@ * atomic_set() is the clrex or dummy strex done on every exception return. */ #define atomic_read(v) (*(volatile int *)&(v)->counter) +static inline int atomic_read_unchecked(const atomic_unchecked_t *v) +{ + return v->counter; +} #define atomic_set(v,i) (((v)->counter) = (i)) +static inline void atomic_set_unchecked(atomic_unchecked_t *v, int i) +{ + v->counter = i; +} #if __LINUX_ARM_ARCH__ >= 6 @@ -40,6 +52,35 @@ static inline void atomic_add(int i, atomic_t *v) int result; __asm__ __volatile__("@ atomic_add\n" +"1: ldrex %1, [%3]\n" +" adds %0, %1, %4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + +" strex %1, %0, [%3]\n" +" teq %1, #0\n" +" bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) + : "r" (&v->counter), "Ir" (i) + : "cc"); +} + +static inline void atomic_add_unchecked(int i, atomic_unchecked_t *v) +{ + unsigned long tmp; + int result; + + __asm__ __volatile__("@ atomic_add_unchecked\n" "1: ldrex %0, [%3]\n" " add %0, %0, %4\n" " strex %1, %0, [%3]\n" @@ -58,6 +99,42 @@ static inline int atomic_add_return(int i, atomic_t *v) smp_mb(); __asm__ __volatile__("@ atomic_add_return\n" +"1: ldrex %1, [%3]\n" +" adds %0, %1, %4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +" mov %0, %1\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + +" strex %1, %0, [%3]\n" +" teq %1, #0\n" +" bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) + : "r" (&v->counter), "Ir" (i) + : "cc"); + + smp_mb(); + + return result; +} + +static inline int atomic_add_return_unchecked(int i, atomic_unchecked_t *v) +{ + unsigned long tmp; + int result; + + smp_mb(); + + __asm__ __volatile__("@ atomic_add_return_unchecked\n" "1: ldrex %0, [%3]\n" " add %0, %0, %4\n" " strex %1, %0, [%3]\n" @@ -78,6 +155,35 @@ static inline void atomic_sub(int i, atomic_t *v) int result; __asm__ __volatile__("@ atomic_sub\n" +"1: ldrex %1, [%3]\n" +" subs %0, %1, %4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + +" strex %1, %0, [%3]\n" +" teq %1, #0\n" +" bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) + : "r" (&v->counter), "Ir" (i) + : "cc"); +} + +static inline void atomic_sub_unchecked(int i, atomic_unchecked_t *v) +{ + unsigned long tmp; + int result; + + __asm__ __volatile__("@ atomic_sub_unchecked\n" "1: ldrex %0, [%3]\n" " sub %0, %0, %4\n" " strex %1, %0, [%3]\n" @@ -96,11 +202,25 @@ static inline int atomic_sub_return(int i, atomic_t *v) smp_mb(); __asm__ __volatile__("@ atomic_sub_return\n" -"1: ldrex %0, [%3]\n" -" sub %0, %0, %4\n" +"1: ldrex %1, [%3]\n" +" sub %0, %1, %4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +" mov %0, %1\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + " strex %1, %0, [%3]\n" " teq %1, #0\n" " bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) : "r" (&v->counter), "Ir" (i) : "cc"); @@ -132,6 +252,28 @@ static inline int atomic_cmpxchg(atomic_t *ptr, int old, int new) return oldval; } +static inline int atomic_cmpxchg_unchecked(atomic_unchecked_t *ptr, int old, int new) +{ + unsigned long oldval, res; + + smp_mb(); + + do { + __asm__ __volatile__("@ atomic_cmpxchg_unchecked\n" + "ldrex %1, [%3]\n" + "mov %0, #0\n" + "teq %1, %4\n" + "strexeq %0, %5, [%3]\n" + : "=&r" (res), "=&r" (oldval), "+Qo" (ptr->counter) + : "r" (&ptr->counter), "Ir" (old), "r" (new) + : "cc"); + } while (res); + + smp_mb(); + + return oldval; +} + static inline void atomic_clear_mask(unsigned long mask, unsigned long *addr) { unsigned long tmp, tmp2; @@ -165,7 +307,9 @@ static inline int atomic_add_return(int i, atomic_t *v) return val; } +#define atomic_add_return_unchecked(i, v) atomic_add_return(i, v) #define atomic_add(i, v) (void) atomic_add_return(i, v) +#define atomic_add_unchecked(i, v) (void) atomic_add_return_unchecked(i, v) static inline int atomic_sub_return(int i, atomic_t *v) { @@ -179,7 +323,9 @@ static inline int atomic_sub_return(int i, atomic_t *v) return val; } +#define atomic_sub_return_unchecked(i, v) atomic_sub_return(i, v) #define atomic_sub(i, v) (void) atomic_sub_return(i, v) +#define atomic_sub_unchecked(i, v) (void) atomic_sub_return_unchecked(i, v) static inline int atomic_cmpxchg(atomic_t *v, int old, int new) { @@ -194,6 +340,7 @@ static inline int atomic_cmpxchg(atomic_t *v, int old, int new) return ret; } +#define atomic_cmpxchg_unchecked(v, o, n) atomic_cmpxchg(v, o, n) static inline void atomic_clear_mask(unsigned long mask, unsigned long *addr) { @@ -207,6 +354,10 @@ static inline void atomic_clear_mask(unsigned long mask, unsigned long *addr) #endif /* __LINUX_ARM_ARCH__ */ #define atomic_xchg(v, new) (xchg(&((v)->counter), new)) +static inline int atomic_xchg_unchecked(atomic_unchecked_t *v, int new) +{ + return xchg(&v->counter, new); +} static inline int __atomic_add_unless(atomic_t *v, int a, int u) { @@ -219,11 +370,27 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) } #define atomic_inc(v) atomic_add(1, v) +static inline void atomic_inc_unchecked(atomic_unchecked_t *v) +{ + atomic_add_unchecked(1, v); +} #define atomic_dec(v) atomic_sub(1, v) +static inline void atomic_dec_unchecked(atomic_unchecked_t *v) +{ + atomic_sub_unchecked(1, v); +} #define atomic_inc_and_test(v) (atomic_add_return(1, v) == 0) +static inline int atomic_inc_and_test_unchecked(atomic_unchecked_t *v) +{ + return atomic_add_return_unchecked(1, v) == 0; +} #define atomic_dec_and_test(v) (atomic_sub_return(1, v) == 0) #define atomic_inc_return(v) (atomic_add_return(1, v)) +static inline int atomic_inc_return_unchecked(atomic_unchecked_t *v) +{ + return atomic_add_return_unchecked(1, v); +} #define atomic_dec_return(v) (atomic_sub_return(1, v)) #define atomic_sub_and_test(i, v) (atomic_sub_return(i, v) == 0) @@ -239,6 +406,14 @@ typedef struct { u64 __aligned(8) counter; } atomic64_t; +#ifdef CONFIG_PAX_REFCOUNT +typedef struct { + u64 __aligned(8) counter; +} atomic64_unchecked_t; +#else +typedef atomic64_t atomic64_unchecked_t; +#endif + #define ATOMIC64_INIT(i) { (i) } static inline u64 atomic64_read(atomic64_t *v) @@ -254,6 +429,19 @@ static inline u64 atomic64_read(atomic64_t *v) return result; } +static inline u64 atomic64_read_unchecked(atomic64_unchecked_t *v) +{ + u64 result; + + __asm__ __volatile__("@ atomic64_read_unchecked\n" +" ldrexd %0, %H0, [%1]" + : "=&r" (result) + : "r" (&v->counter), "Qo" (v->counter) + ); + + return result; +} + static inline void atomic64_set(atomic64_t *v, u64 i) { u64 tmp; @@ -268,6 +456,20 @@ static inline void atomic64_set(atomic64_t *v, u64 i) : "cc"); } +static inline void atomic64_set_unchecked(atomic64_unchecked_t *v, u64 i) +{ + u64 tmp; + + __asm__ __volatile__("@ atomic64_set_unchecked\n" +"1: ldrexd %0, %H0, [%2]\n" +" strexd %0, %3, %H3, [%2]\n" +" teq %0, #0\n" +" bne 1b" + : "=&r" (tmp), "=Qo" (v->counter) + : "r" (&v->counter), "r" (i) + : "cc"); +} + static inline void atomic64_add(u64 i, atomic64_t *v) { u64 result; @@ -276,6 +478,36 @@ static inline void atomic64_add(u64 i, atomic64_t *v) __asm__ __volatile__("@ atomic64_add\n" "1: ldrexd %0, %H0, [%3]\n" " adds %0, %0, %4\n" +" adcs %H0, %H0, %H4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + +" strexd %1, %0, %H0, [%3]\n" +" teq %1, #0\n" +" bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) + : "r" (&v->counter), "r" (i) + : "cc"); +} + +static inline void atomic64_add_unchecked(u64 i, atomic64_unchecked_t *v) +{ + u64 result; + unsigned long tmp; + + __asm__ __volatile__("@ atomic64_add_unchecked\n" +"1: ldrexd %0, %H0, [%3]\n" +" adds %0, %0, %4\n" " adc %H0, %H0, %H4\n" " strexd %1, %0, %H0, [%3]\n" " teq %1, #0\n" @@ -287,12 +519,49 @@ static inline void atomic64_add(u64 i, atomic64_t *v) static inline u64 atomic64_add_return(u64 i, atomic64_t *v) { - u64 result; - unsigned long tmp; + u64 result, tmp; smp_mb(); __asm__ __volatile__("@ atomic64_add_return\n" +"1: ldrexd %1, %H1, [%3]\n" +" adds %0, %1, %4\n" +" adcs %H0, %H1, %H4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +" mov %0, %1\n" +" mov %H0, %H1\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + +" strexd %1, %0, %H0, [%3]\n" +" teq %1, #0\n" +" bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) + : "r" (&v->counter), "r" (i) + : "cc"); + + smp_mb(); + + return result; +} + +static inline u64 atomic64_add_return_unchecked(u64 i, atomic64_unchecked_t *v) +{ + u64 result; + unsigned long tmp; + + smp_mb(); + + __asm__ __volatile__("@ atomic64_add_return_unchecked\n" "1: ldrexd %0, %H0, [%3]\n" " adds %0, %0, %4\n" " adc %H0, %H0, %H4\n" @@ -316,6 +585,36 @@ static inline void atomic64_sub(u64 i, atomic64_t *v) __asm__ __volatile__("@ atomic64_sub\n" "1: ldrexd %0, %H0, [%3]\n" " subs %0, %0, %4\n" +" sbcs %H0, %H0, %H4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + +" strexd %1, %0, %H0, [%3]\n" +" teq %1, #0\n" +" bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) + : "r" (&v->counter), "r" (i) + : "cc"); +} + +static inline void atomic64_sub_unchecked(u64 i, atomic64_unchecked_t *v) +{ + u64 result; + unsigned long tmp; + + __asm__ __volatile__("@ atomic64_sub_unchecked\n" +"1: ldrexd %0, %H0, [%3]\n" +" subs %0, %0, %4\n" " sbc %H0, %H0, %H4\n" " strexd %1, %0, %H0, [%3]\n" " teq %1, #0\n" @@ -327,18 +626,32 @@ static inline void atomic64_sub(u64 i, atomic64_t *v) static inline u64 atomic64_sub_return(u64 i, atomic64_t *v) { - u64 result; - unsigned long tmp; + u64 result, tmp; smp_mb(); __asm__ __volatile__("@ atomic64_sub_return\n" -"1: ldrexd %0, %H0, [%3]\n" -" subs %0, %0, %4\n" -" sbc %H0, %H0, %H4\n" +"1: ldrexd %1, %H1, [%3]\n" +" subs %0, %1, %4\n" +" sbc %H0, %H1, %H4\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +" mov %0, %1\n" +" mov %H0, %H1\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + " strexd %1, %0, %H0, [%3]\n" " teq %1, #0\n" " bne 1b" + +#ifdef CONFIG_PAX_REFCOUNT +"\n4:\n" + _ASM_EXTABLE(2b, 4b) +#endif + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) : "r" (&v->counter), "r" (i) : "cc"); @@ -372,6 +685,30 @@ static inline u64 atomic64_cmpxchg(atomic64_t *ptr, u64 old, u64 new) return oldval; } +static inline u64 atomic64_cmpxchg_unchecked(atomic64_unchecked_t *ptr, u64 old, u64 new) +{ + u64 oldval; + unsigned long res; + + smp_mb(); + + do { + __asm__ __volatile__("@ atomic64_cmpxchg_unchecked\n" + "ldrexd %1, %H1, [%3]\n" + "mov %0, #0\n" + "teq %1, %4\n" + "teqeq %H1, %H4\n" + "strexdeq %0, %5, %H5, [%3]" + : "=&r" (res), "=&r" (oldval), "+Qo" (ptr->counter) + : "r" (&ptr->counter), "r" (old), "r" (new) + : "cc"); + } while (res); + + smp_mb(); + + return oldval; +} + static inline u64 atomic64_xchg(atomic64_t *ptr, u64 new) { u64 result; @@ -395,21 +732,34 @@ static inline u64 atomic64_xchg(atomic64_t *ptr, u64 new) static inline u64 atomic64_dec_if_positive(atomic64_t *v) { - u64 result; - unsigned long tmp; + u64 result, tmp; smp_mb(); __asm__ __volatile__("@ atomic64_dec_if_positive\n" -"1: ldrexd %0, %H0, [%3]\n" -" subs %0, %0, #1\n" -" sbc %H0, %H0, #0\n" +"1: ldrexd %1, %H1, [%3]\n" +" subs %0, %1, #1\n" +" sbc %H0, %H1, #0\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +" mov %0, %1\n" +" mov %H0, %H1\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + " teq %H0, #0\n" -" bmi 2f\n" +" bmi 4f\n" " strexd %1, %0, %H0, [%3]\n" " teq %1, #0\n" " bne 1b\n" -"2:" +"4:\n" + +#ifdef CONFIG_PAX_REFCOUNT + _ASM_EXTABLE(2b, 4b) +#endif + : "=&r" (result), "=&r" (tmp), "+Qo" (v->counter) : "r" (&v->counter) : "cc"); @@ -432,13 +782,25 @@ static inline int atomic64_add_unless(atomic64_t *v, u64 a, u64 u) " teq %0, %5\n" " teqeq %H0, %H5\n" " moveq %1, #0\n" -" beq 2f\n" +" beq 4f\n" " adds %0, %0, %6\n" " adc %H0, %H0, %H6\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + " strexd %2, %0, %H0, [%4]\n" " teq %2, #0\n" " bne 1b\n" -"2:" +"4:\n" + +#ifdef CONFIG_PAX_REFCOUNT + _ASM_EXTABLE(2b, 4b) +#endif + : "=&r" (val), "+r" (ret), "=&r" (tmp), "+Qo" (v->counter) : "r" (&v->counter), "r" (u), "r" (a) : "cc"); @@ -451,10 +813,13 @@ static inline int atomic64_add_unless(atomic64_t *v, u64 a, u64 u) #define atomic64_add_negative(a, v) (atomic64_add_return((a), (v)) < 0) #define atomic64_inc(v) atomic64_add(1LL, (v)) +#define atomic64_inc_unchecked(v) atomic64_add_unchecked(1LL, (v)) #define atomic64_inc_return(v) atomic64_add_return(1LL, (v)) +#define atomic64_inc_return_unchecked(v) atomic64_add_return_unchecked(1LL, (v)) #define atomic64_inc_and_test(v) (atomic64_inc_return(v) == 0) #define atomic64_sub_and_test(a, v) (atomic64_sub_return((a), (v)) == 0) #define atomic64_dec(v) atomic64_sub(1LL, (v)) +#define atomic64_dec_unchecked(v) atomic64_sub_unchecked(1LL, (v)) #define atomic64_dec_return(v) atomic64_sub_return(1LL, (v)) #define atomic64_dec_and_test(v) (atomic64_dec_return((v)) == 0) #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1LL, 0LL) diff --git a/arch/arm/include/asm/cache.h b/arch/arm/include/asm/cache.h index 75fe66b..2255c86 100644 --- a/arch/arm/include/asm/cache.h +++ b/arch/arm/include/asm/cache.h @@ -4,8 +4,10 @@ #ifndef __ASMARM_CACHE_H #define __ASMARM_CACHE_H +#include + #define L1_CACHE_SHIFT CONFIG_ARM_L1_CACHE_SHIFT -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) /* * Memory returned by kmalloc() may be used for DMA, so we must make diff --git a/arch/arm/include/asm/cacheflush.h b/arch/arm/include/asm/cacheflush.h index d5d8d5c..ad92c96 100644 --- a/arch/arm/include/asm/cacheflush.h +++ b/arch/arm/include/asm/cacheflush.h @@ -108,7 +108,7 @@ struct cpu_cache_fns { void (*dma_unmap_area)(const void *, size_t, int); void (*dma_flush_range)(const void *, const void *); -}; +} __no_const; /* * Select the calling method diff --git a/arch/arm/include/asm/elf.h b/arch/arm/include/asm/elf.h index 0e9ce8d..6ef1e03 100644 --- a/arch/arm/include/asm/elf.h +++ b/arch/arm/include/asm/elf.h @@ -116,7 +116,14 @@ int dump_task_regs(struct task_struct *t, elf_gregset_t *elfregs); the loader. We need to make sure that it is out of the way of the program that it will "exec", and that there is sufficient room for the brk. */ -#define ELF_ET_DYN_BASE (2 * TASK_SIZE / 3) +#define ELF_ET_DYN_BASE (TASK_SIZE / 3 * 2) + +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE 0x00008000UL + +#define PAX_DELTA_MMAP_LEN ((current->personality == PER_LINUX_32BIT) ? 16 : 10) +#define PAX_DELTA_STACK_LEN ((current->personality == PER_LINUX_32BIT) ? 16 : 10) +#endif /* When the program starts, a1 contains a pointer to a function to be registered with atexit, as per the SVR4 ABI. A value of 0 means we @@ -126,10 +133,6 @@ int dump_task_regs(struct task_struct *t, elf_gregset_t *elfregs); extern void elf_set_personality(const struct elf32_hdr *); #define SET_PERSONALITY(ex) elf_set_personality(&(ex)) -struct mm_struct; -extern unsigned long arch_randomize_brk(struct mm_struct *mm); -#define arch_randomize_brk arch_randomize_brk - extern int vectors_user_mapping(void); #define arch_setup_additional_pages(bprm, uses_interp) vectors_user_mapping() #define ARCH_HAS_SETUP_ADDITIONAL_PAGES diff --git a/arch/arm/include/asm/kmap_types.h b/arch/arm/include/asm/kmap_types.h index e51b1e8..32a3113 100644 --- a/arch/arm/include/asm/kmap_types.h +++ b/arch/arm/include/asm/kmap_types.h @@ -21,6 +21,7 @@ enum km_type { KM_L1_CACHE, KM_L2_CACHE, KM_KDB, + KM_CLEARPAGE, KM_TYPE_NR }; diff --git a/arch/arm/include/asm/outercache.h b/arch/arm/include/asm/outercache.h index 53426c6..c7baff3 100644 --- a/arch/arm/include/asm/outercache.h +++ b/arch/arm/include/asm/outercache.h @@ -35,7 +35,7 @@ struct outer_cache_fns { #endif void (*set_debug)(unsigned long); void (*resume)(void); -}; +} __no_const; #ifdef CONFIG_OUTER_CACHE diff --git a/arch/arm/include/asm/page.h b/arch/arm/include/asm/page.h index 97b440c..b7ff179 100644 --- a/arch/arm/include/asm/page.h +++ b/arch/arm/include/asm/page.h @@ -123,7 +123,7 @@ struct cpu_user_fns { void (*cpu_clear_user_highpage)(struct page *page, unsigned long vaddr); void (*cpu_copy_user_highpage)(struct page *to, struct page *from, unsigned long vaddr, struct vm_area_struct *vma); -}; +} __no_const; #ifdef MULTI_USER extern struct cpu_user_fns cpu_user; diff --git a/arch/arm/include/asm/pgalloc.h b/arch/arm/include/asm/pgalloc.h index 943504f..bf8d667 100644 --- a/arch/arm/include/asm/pgalloc.h +++ b/arch/arm/include/asm/pgalloc.h @@ -43,6 +43,11 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) set_pud(pud, __pud(__pa(pmd) | PMD_TYPE_TABLE)); } +static inline void pud_populate_kernel(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) +{ + pud_populate(mm, pud, pmd); +} + #else /* !CONFIG_ARM_LPAE */ /* @@ -51,6 +56,7 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) #define pmd_alloc_one(mm,addr) ({ BUG(); ((pmd_t *)2); }) #define pmd_free(mm, pmd) do { } while (0) #define pud_populate(mm,pmd,pte) BUG() +#define pud_populate_kernel(mm,pmd,pte) BUG() #endif /* CONFIG_ARM_LPAE */ diff --git a/arch/arm/include/asm/system.h b/arch/arm/include/asm/system.h index e4c96cc..1145653 100644 --- a/arch/arm/include/asm/system.h +++ b/arch/arm/include/asm/system.h @@ -98,6 +98,8 @@ void hook_ifault_code(int nr, int (*fn)(unsigned long, unsigned int, #define xchg(ptr,x) \ ((__typeof__(*(ptr)))__xchg((unsigned long)(x),(ptr),sizeof(*(ptr)))) +#define xchg_unchecked(ptr,x) \ + ((__typeof__(*(ptr)))__xchg((unsigned long)(x),(ptr),sizeof(*(ptr)))) extern asmlinkage void c_backtrace(unsigned long fp, int pmode); @@ -534,6 +536,13 @@ static inline unsigned long long __cmpxchg64_mb(volatile void *ptr, #endif /* __LINUX_ARM_ARCH__ >= 6 */ +#define _ASM_EXTABLE(from, to) \ +" .pushsection __ex_table,\"a\"\n"\ +" .align 3\n" \ +" .long " #from ", " #to"\n" \ +" .popsection" + + #endif /* __ASSEMBLY__ */ #define arch_align_stack(x) (x) diff --git a/arch/arm/include/asm/uaccess.h b/arch/arm/include/asm/uaccess.h index 2958976..12ccac4 100644 --- a/arch/arm/include/asm/uaccess.h +++ b/arch/arm/include/asm/uaccess.h @@ -22,6 +22,8 @@ #define VERIFY_READ 0 #define VERIFY_WRITE 1 +extern void check_object_size(const void *ptr, unsigned long n, bool to); + /* * The exception table consists of pairs of addresses: the first is the * address of an instruction that is allowed to fault, and the second is @@ -387,8 +389,23 @@ do { \ #ifdef CONFIG_MMU -extern unsigned long __must_check __copy_from_user(void *to, const void __user *from, unsigned long n); -extern unsigned long __must_check __copy_to_user(void __user *to, const void *from, unsigned long n); +extern unsigned long __must_check ___copy_from_user(void *to, const void __user *from, unsigned long n); +extern unsigned long __must_check ___copy_to_user(void __user *to, const void *from, unsigned long n); + +static inline unsigned long __must_check __copy_from_user(void *to, const void __user *from, unsigned long n) +{ + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); + return ___copy_from_user(to, from, n); +} + +static inline unsigned long __must_check __copy_to_user(void __user *to, const void *from, unsigned long n) +{ + if (!__builtin_constant_p(n)) + check_object_size(from, n, true); + return ___copy_to_user(to, from, n); +} + extern unsigned long __must_check __copy_to_user_std(void __user *to, const void *from, unsigned long n); extern unsigned long __must_check __clear_user(void __user *addr, unsigned long n); extern unsigned long __must_check __clear_user_std(void __user *addr, unsigned long n); @@ -403,6 +420,9 @@ extern unsigned long __must_check __strnlen_user(const char __user *s, long n); static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, unsigned long n) { + if ((long)n < 0) + return n; + if (access_ok(VERIFY_READ, from, n)) n = __copy_from_user(to, from, n); else /* security hole - plug it */ @@ -412,6 +432,9 @@ static inline unsigned long __must_check copy_from_user(void *to, const void __u static inline unsigned long __must_check copy_to_user(void __user *to, const void *from, unsigned long n) { + if ((long)n < 0) + return n; + if (access_ok(VERIFY_WRITE, to, n)) n = __copy_to_user(to, from, n); return n; diff --git a/arch/arm/kernel/armksyms.c b/arch/arm/kernel/armksyms.c index 5b0bce6..becd81c 100644 --- a/arch/arm/kernel/armksyms.c +++ b/arch/arm/kernel/armksyms.c @@ -95,8 +95,8 @@ EXPORT_SYMBOL(__strncpy_from_user); #ifdef CONFIG_MMU EXPORT_SYMBOL(copy_page); -EXPORT_SYMBOL(__copy_from_user); -EXPORT_SYMBOL(__copy_to_user); +EXPORT_SYMBOL(___copy_from_user); +EXPORT_SYMBOL(___copy_to_user); EXPORT_SYMBOL(__clear_user); EXPORT_SYMBOL(__get_user_1); diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c index 971d65c..cc936fb 100644 --- a/arch/arm/kernel/process.c +++ b/arch/arm/kernel/process.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include @@ -273,9 +272,10 @@ void machine_power_off(void) machine_shutdown(); if (pm_power_off) pm_power_off(); + BUG(); } -void machine_restart(char *cmd) +__noreturn void machine_restart(char *cmd) { machine_shutdown(); @@ -517,12 +517,6 @@ unsigned long get_wchan(struct task_struct *p) return 0; } -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long range_end = mm->brk + 0x02000000; - return randomize_range(mm->brk, range_end, 0) ? : mm->brk; -} - #ifdef CONFIG_MMU /* * The vectors page is always readable from user space for the diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index a255c39..4a19b25 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -109,13 +109,13 @@ struct processor processor __read_mostly; struct cpu_tlb_fns cpu_tlb __read_mostly; #endif #ifdef MULTI_USER -struct cpu_user_fns cpu_user __read_mostly; +struct cpu_user_fns cpu_user __read_only; #endif #ifdef MULTI_CACHE -struct cpu_cache_fns cpu_cache __read_mostly; +struct cpu_cache_fns cpu_cache __read_only; #endif #ifdef CONFIG_OUTER_CACHE -struct outer_cache_fns outer_cache __read_mostly; +struct outer_cache_fns outer_cache __read_only; EXPORT_SYMBOL(outer_cache); #endif diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index f84dfe6..13e94f7 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -259,6 +259,8 @@ static int __die(const char *str, int err, struct thread_info *thread, struct pt static DEFINE_RAW_SPINLOCK(die_lock); +extern void gr_handle_kernel_exploit(void); + /* * This function is protected against re-entrancy. */ @@ -291,6 +293,9 @@ void die(const char *str, struct pt_regs *regs, int err) panic("Fatal exception in interrupt"); if (panic_on_oops) panic("Fatal exception"); + + gr_handle_kernel_exploit(); + if (ret != NOTIFY_STOP) do_exit(SIGSEGV); } diff --git a/arch/arm/lib/copy_from_user.S b/arch/arm/lib/copy_from_user.S index 66a477a..bee61d3 100644 --- a/arch/arm/lib/copy_from_user.S +++ b/arch/arm/lib/copy_from_user.S @@ -16,7 +16,7 @@ /* * Prototype: * - * size_t __copy_from_user(void *to, const void *from, size_t n) + * size_t ___copy_from_user(void *to, const void *from, size_t n) * * Purpose: * @@ -84,11 +84,11 @@ .text -ENTRY(__copy_from_user) +ENTRY(___copy_from_user) #include "copy_template.S" -ENDPROC(__copy_from_user) +ENDPROC(___copy_from_user) .pushsection .fixup,"ax" .align 0 diff --git a/arch/arm/lib/copy_page.S b/arch/arm/lib/copy_page.S index 6ee2f67..d1cce76 100644 --- a/arch/arm/lib/copy_page.S +++ b/arch/arm/lib/copy_page.S @@ -10,6 +10,7 @@ * ASM optimised string functions */ #include +#include #include #include #include diff --git a/arch/arm/lib/copy_to_user.S b/arch/arm/lib/copy_to_user.S index d066df6..df28194 100644 --- a/arch/arm/lib/copy_to_user.S +++ b/arch/arm/lib/copy_to_user.S @@ -16,7 +16,7 @@ /* * Prototype: * - * size_t __copy_to_user(void *to, const void *from, size_t n) + * size_t ___copy_to_user(void *to, const void *from, size_t n) * * Purpose: * @@ -88,11 +88,11 @@ .text ENTRY(__copy_to_user_std) -WEAK(__copy_to_user) +WEAK(___copy_to_user) #include "copy_template.S" -ENDPROC(__copy_to_user) +ENDPROC(___copy_to_user) ENDPROC(__copy_to_user_std) .pushsection .fixup,"ax" diff --git a/arch/arm/lib/uaccess.S b/arch/arm/lib/uaccess.S index 5c908b1..e712687 100644 --- a/arch/arm/lib/uaccess.S +++ b/arch/arm/lib/uaccess.S @@ -20,7 +20,7 @@ #define PAGE_SHIFT 12 -/* Prototype: int __copy_to_user(void *to, const char *from, size_t n) +/* Prototype: int ___copy_to_user(void *to, const char *from, size_t n) * Purpose : copy a block to user memory from kernel memory * Params : to - user memory * : from - kernel memory @@ -40,7 +40,7 @@ USER( TUSER( strgtb) r3, [r0], #1) @ May fault sub r2, r2, ip b .Lc2u_dest_aligned -ENTRY(__copy_to_user) +ENTRY(___copy_to_user) stmfd sp!, {r2, r4 - r7, lr} cmp r2, #4 blt .Lc2u_not_enough @@ -278,14 +278,14 @@ USER( TUSER( strgeb) r3, [r0], #1) @ May fault ldrgtb r3, [r1], #0 USER( TUSER( strgtb) r3, [r0], #1) @ May fault b .Lc2u_finished -ENDPROC(__copy_to_user) +ENDPROC(___copy_to_user) .pushsection .fixup,"ax" .align 0 9001: ldmfd sp!, {r0, r4 - r7, pc} .popsection -/* Prototype: unsigned long __copy_from_user(void *to,const void *from,unsigned long n); +/* Prototype: unsigned long ___copy_from_user(void *to,const void *from,unsigned long n); * Purpose : copy a block from user memory to kernel memory * Params : to - kernel memory * : from - user memory @@ -304,7 +304,7 @@ USER( TUSER( ldrgtb) r3, [r1], #1) @ May fault sub r2, r2, ip b .Lcfu_dest_aligned -ENTRY(__copy_from_user) +ENTRY(___copy_from_user) stmfd sp!, {r0, r2, r4 - r7, lr} cmp r2, #4 blt .Lcfu_not_enough @@ -544,7 +544,7 @@ USER( TUSER( ldrgeb) r3, [r1], #1) @ May fault USER( TUSER( ldrgtb) r3, [r1], #1) @ May fault strgtb r3, [r0], #1 b .Lcfu_finished -ENDPROC(__copy_from_user) +ENDPROC(___copy_from_user) .pushsection .fixup,"ax" .align 0 diff --git a/arch/arm/lib/uaccess_with_memcpy.c b/arch/arm/lib/uaccess_with_memcpy.c index 025f742..8432b08 100644 --- a/arch/arm/lib/uaccess_with_memcpy.c +++ b/arch/arm/lib/uaccess_with_memcpy.c @@ -104,7 +104,7 @@ out: } unsigned long -__copy_to_user(void __user *to, const void *from, unsigned long n) +___copy_to_user(void __user *to, const void *from, unsigned long n) { /* * This test is stubbed out of the main function above to keep diff --git a/arch/arm/mach-omap2/board-n8x0.c b/arch/arm/mach-omap2/board-n8x0.c index 6722627..8f97548c 100644 --- a/arch/arm/mach-omap2/board-n8x0.c +++ b/arch/arm/mach-omap2/board-n8x0.c @@ -597,7 +597,7 @@ static int n8x0_menelaus_late_init(struct device *dev) } #endif -static struct menelaus_platform_data n8x0_menelaus_platform_data __initdata = { +static struct menelaus_platform_data n8x0_menelaus_platform_data __initconst = { .late_init = n8x0_menelaus_late_init, }; diff --git a/arch/arm/mach-ux500/mbox-db5500.c b/arch/arm/mach-ux500/mbox-db5500.c index 2b2d51c..0127490 100644 --- a/arch/arm/mach-ux500/mbox-db5500.c +++ b/arch/arm/mach-ux500/mbox-db5500.c @@ -168,7 +168,7 @@ static ssize_t mbox_read_fifo(struct device *dev, return sprintf(buf, "0x%X\n", mbox_value); } -static DEVICE_ATTR(fifo, S_IWUGO | S_IRUGO, mbox_read_fifo, mbox_write_fifo); +static DEVICE_ATTR(fifo, S_IWUSR | S_IRUGO, mbox_read_fifo, mbox_write_fifo); static int mbox_show(struct seq_file *s, void *data) { diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c index bb7eac3..3bade16 100644 --- a/arch/arm/mm/fault.c +++ b/arch/arm/mm/fault.c @@ -172,6 +172,13 @@ __do_user_fault(struct task_struct *tsk, unsigned long addr, } #endif +#ifdef CONFIG_PAX_PAGEEXEC + if (fsr & FSR_LNX_PF) { + pax_report_fault(regs, (void *)regs->ARM_pc, (void *)regs->ARM_sp); + do_group_exit(SIGKILL); + } +#endif + tsk->thread.address = addr; tsk->thread.error_code = fsr; tsk->thread.trap_no = 14; @@ -393,6 +400,33 @@ do_page_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) } #endif /* CONFIG_MMU */ +#ifdef CONFIG_PAX_PAGEEXEC +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 20; i++) { + unsigned char c; + if (get_user(c, (__force unsigned char __user *)pc+i)) + printk(KERN_CONT "?? "); + else + printk(KERN_CONT "%02x ", c); + } + printk("\n"); + + printk(KERN_ERR "PAX: bytes at SP-4: "); + for (i = -1; i < 20; i++) { + unsigned long c; + if (get_user(c, (__force unsigned long __user *)sp+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08lx ", c); + } + printk("\n"); +} +#endif + /* * First Level Translation Fault Handler * @@ -573,6 +607,20 @@ do_PrefetchAbort(unsigned long addr, unsigned int ifsr, struct pt_regs *regs) const struct fsr_info *inf = ifsr_info + fsr_fs(ifsr); struct siginfo info; +#ifdef CONFIG_PAX_REFCOUNT + if (fsr_fs(ifsr) == 2) { + unsigned int bkpt; + + if (!probe_kernel_address((unsigned int *)addr, bkpt) && bkpt == 0xe12f1073) { + current->thread.error_code = ifsr; + current->thread.trap_no = 0; + pax_report_refcount_overflow(regs); + fixup_exception(regs); + return; + } + } +#endif + if (!inf->fn(addr, ifsr | FSR_LNX_PF, regs)) return; diff --git a/arch/arm/mm/mmap.c b/arch/arm/mm/mmap.c index ce8cb19..3ec539d 100644 --- a/arch/arm/mm/mmap.c +++ b/arch/arm/mm/mmap.c @@ -93,6 +93,10 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, if (len > TASK_SIZE) return -ENOMEM; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_align) addr = COLOUR_ALIGN(addr, pgoff); @@ -100,15 +104,14 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, addr = PAGE_ALIGN(addr); vma = find_vma(mm, addr); - if (TASK_SIZE - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (TASK_SIZE - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } if (len > mm->cached_hole_size) { - start_addr = addr = mm->free_area_cache; + start_addr = addr = mm->free_area_cache; } else { - start_addr = addr = mm->mmap_base; - mm->cached_hole_size = 0; + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; } full_search: @@ -124,14 +127,14 @@ full_search: * Start a new search - just in case we missed * some holes. */ - if (start_addr != TASK_UNMAPPED_BASE) { - start_addr = addr = TASK_UNMAPPED_BASE; + if (start_addr != mm->mmap_base) { + start_addr = addr = mm->mmap_base; mm->cached_hole_size = 0; goto full_search; } return -ENOMEM; } - if (!vma || addr + len <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr, len)) { /* * Remember the place where we stopped the search: */ @@ -266,10 +269,22 @@ void arch_pick_mmap_layout(struct mm_struct *mm) if (mmap_is_legacy()) { mm->mmap_base = TASK_UNMAPPED_BASE + random_factor; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { mm->mmap_base = mmap_base(random_factor); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff --git a/arch/arm/plat-samsung/include/plat/dma-ops.h b/arch/arm/plat-samsung/include/plat/dma-ops.h index 71a6827..e7fbc23 100644 --- a/arch/arm/plat-samsung/include/plat/dma-ops.h +++ b/arch/arm/plat-samsung/include/plat/dma-ops.h @@ -43,7 +43,7 @@ struct samsung_dma_ops { int (*started)(unsigned ch); int (*flush)(unsigned ch); int (*stop)(unsigned ch); -}; +} __no_const; extern void *samsung_dmadev_get_ops(void); extern void *s3c_dma_get_ops(void); diff --git a/arch/arm/plat-samsung/include/plat/ehci.h b/arch/arm/plat-samsung/include/plat/ehci.h index 5f28cae..3d23723 100644 --- a/arch/arm/plat-samsung/include/plat/ehci.h +++ b/arch/arm/plat-samsung/include/plat/ehci.h @@ -14,7 +14,7 @@ struct s5p_ehci_platdata { int (*phy_init)(struct platform_device *pdev, int type); int (*phy_exit)(struct platform_device *pdev, int type); -}; +} __no_const; extern void s5p_ehci_set_platdata(struct s5p_ehci_platdata *pd); diff --git a/arch/avr32/include/asm/cache.h b/arch/avr32/include/asm/cache.h index c3a58a1..78fbf54 100644 --- a/arch/avr32/include/asm/cache.h +++ b/arch/avr32/include/asm/cache.h @@ -1,8 +1,10 @@ #ifndef __ASM_AVR32_CACHE_H #define __ASM_AVR32_CACHE_H +#include + #define L1_CACHE_SHIFT 5 -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) /* * Memory returned by kmalloc() may be used for DMA, so we must make diff --git a/arch/avr32/include/asm/elf.h b/arch/avr32/include/asm/elf.h index 3b3159b..425ea94 100644 --- a/arch/avr32/include/asm/elf.h +++ b/arch/avr32/include/asm/elf.h @@ -84,8 +84,14 @@ typedef struct user_fpu_struct elf_fpregset_t; the loader. We need to make sure that it is out of the way of the program that it will "exec", and that there is sufficient room for the brk. */ -#define ELF_ET_DYN_BASE (2 * TASK_SIZE / 3) +#define ELF_ET_DYN_BASE (TASK_SIZE / 3 * 2) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE 0x00001000UL + +#define PAX_DELTA_MMAP_LEN 15 +#define PAX_DELTA_STACK_LEN 15 +#endif /* This yields a mask that user programs can use to figure out what instruction set this CPU supports. This could be done in user space, diff --git a/arch/avr32/include/asm/kmap_types.h b/arch/avr32/include/asm/kmap_types.h index b7f5c68..556135c 100644 --- a/arch/avr32/include/asm/kmap_types.h +++ b/arch/avr32/include/asm/kmap_types.h @@ -22,7 +22,8 @@ D(10) KM_IRQ0, D(11) KM_IRQ1, D(12) KM_SOFTIRQ0, D(13) KM_SOFTIRQ1, -D(14) KM_TYPE_NR +D(14) KM_CLEARPAGE, +D(15) KM_TYPE_NR }; #undef D diff --git a/arch/avr32/mm/fault.c b/arch/avr32/mm/fault.c index f7040a1..db9f300 100644 --- a/arch/avr32/mm/fault.c +++ b/arch/avr32/mm/fault.c @@ -41,6 +41,23 @@ static inline int notify_page_fault(struct pt_regs *regs, int trap) int exception_trace = 1; +#ifdef CONFIG_PAX_PAGEEXEC +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 20; i++) { + unsigned char c; + if (get_user(c, (unsigned char *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%02x ", c); + } + printk("\n"); +} +#endif + /* * This routine handles page faults. It determines the address and the * problem, and then passes it off to one of the appropriate routines. @@ -156,6 +173,16 @@ bad_area: up_read(&mm->mmap_sem); if (user_mode(regs)) { + +#ifdef CONFIG_PAX_PAGEEXEC + if (mm->pax_flags & MF_PAX_PAGEEXEC) { + if (ecr == ECR_PROTECTION_X || ecr == ECR_TLB_MISS_X) { + pax_report_fault(regs, (void *)regs->pc, (void *)regs->sp); + do_group_exit(SIGKILL); + } + } +#endif + if (exception_trace && printk_ratelimit()) printk("%s%s[%d]: segfault at %08lx pc %08lx " "sp %08lx ecr %lu\n", diff --git a/arch/blackfin/include/asm/cache.h b/arch/blackfin/include/asm/cache.h index 568885a..f8008df 100644 --- a/arch/blackfin/include/asm/cache.h +++ b/arch/blackfin/include/asm/cache.h @@ -7,6 +7,7 @@ #ifndef __ARCH_BLACKFIN_CACHE_H #define __ARCH_BLACKFIN_CACHE_H +#include #include /* for asmlinkage */ /* @@ -14,7 +15,7 @@ * Blackfin loads 32 bytes for cache */ #define L1_CACHE_SHIFT 5 -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define SMP_CACHE_BYTES L1_CACHE_BYTES #define ARCH_DMA_MINALIGN L1_CACHE_BYTES diff --git a/arch/cris/include/arch-v10/arch/cache.h b/arch/cris/include/arch-v10/arch/cache.h index aea2718..3639a60 100644 --- a/arch/cris/include/arch-v10/arch/cache.h +++ b/arch/cris/include/arch-v10/arch/cache.h @@ -1,8 +1,9 @@ #ifndef _ASM_ARCH_CACHE_H #define _ASM_ARCH_CACHE_H +#include /* Etrax 100LX have 32-byte cache-lines. */ -#define L1_CACHE_BYTES 32 #define L1_CACHE_SHIFT 5 +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #endif /* _ASM_ARCH_CACHE_H */ diff --git a/arch/cris/include/arch-v32/arch/cache.h b/arch/cris/include/arch-v32/arch/cache.h index 1de779f..336fad3 100644 --- a/arch/cris/include/arch-v32/arch/cache.h +++ b/arch/cris/include/arch-v32/arch/cache.h @@ -1,11 +1,12 @@ #ifndef _ASM_CRIS_ARCH_CACHE_H #define _ASM_CRIS_ARCH_CACHE_H +#include #include /* A cache-line is 32 bytes. */ -#define L1_CACHE_BYTES 32 #define L1_CACHE_SHIFT 5 +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define __read_mostly __attribute__((__section__(".data.read_mostly"))) diff --git a/arch/frv/include/asm/atomic.h b/arch/frv/include/asm/atomic.h index 0d8a7d6..d0c9ff5 100644 --- a/arch/frv/include/asm/atomic.h +++ b/arch/frv/include/asm/atomic.h @@ -241,6 +241,16 @@ extern uint32_t __xchg_32(uint32_t i, volatile void *v); #define atomic64_cmpxchg(v, old, new) (__cmpxchg_64(old, new, &(v)->counter)) #define atomic64_xchg(v, new) (__xchg_64(new, &(v)->counter)) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u) { int c, old; diff --git a/arch/frv/include/asm/cache.h b/arch/frv/include/asm/cache.h index 2797163..c2a401d 100644 --- a/arch/frv/include/asm/cache.h +++ b/arch/frv/include/asm/cache.h @@ -12,10 +12,11 @@ #ifndef __ASM_CACHE_H #define __ASM_CACHE_H +#include /* bytes per L1 cache line */ #define L1_CACHE_SHIFT (CONFIG_FRV_L1_CACHE_SHIFT) -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define __cacheline_aligned __attribute__((aligned(L1_CACHE_BYTES))) #define ____cacheline_aligned __attribute__((aligned(L1_CACHE_BYTES))) diff --git a/arch/frv/include/asm/kmap_types.h b/arch/frv/include/asm/kmap_types.h index f8e16b2..c73ff79 100644 --- a/arch/frv/include/asm/kmap_types.h +++ b/arch/frv/include/asm/kmap_types.h @@ -23,6 +23,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff --git a/arch/frv/mm/elf-fdpic.c b/arch/frv/mm/elf-fdpic.c index 385fd30..6c3d97e 100644 --- a/arch/frv/mm/elf-fdpic.c +++ b/arch/frv/mm/elf-fdpic.c @@ -73,8 +73,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi if (addr) { addr = PAGE_ALIGN(addr); vma = find_vma(current->mm, addr); - if (TASK_SIZE - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (TASK_SIZE - len >= addr && check_heap_stack_gap(vma, addr, len)) goto success; } @@ -89,7 +88,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi for (; vma; vma = vma->vm_next) { if (addr > limit) break; - if (addr + len <= vma->vm_start) + if (check_heap_stack_gap(vma, addr, len)) goto success; addr = vma->vm_end; } @@ -104,7 +103,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi for (; vma; vma = vma->vm_next) { if (addr > limit) break; - if (addr + len <= vma->vm_start) + if (check_heap_stack_gap(vma, addr, len)) goto success; addr = vma->vm_end; } diff --git a/arch/h8300/include/asm/cache.h b/arch/h8300/include/asm/cache.h index c635028..6d9445a 100644 --- a/arch/h8300/include/asm/cache.h +++ b/arch/h8300/include/asm/cache.h @@ -1,8 +1,10 @@ #ifndef __ARCH_H8300_CACHE_H #define __ARCH_H8300_CACHE_H +#include + /* bytes per L1 cache line */ -#define L1_CACHE_BYTES 4 +#define L1_CACHE_BYTES _AC(4,UL) /* m68k-elf-gcc 2.95.2 doesn't like these */ diff --git a/arch/hexagon/include/asm/cache.h b/arch/hexagon/include/asm/cache.h index 0f01de2..d37d309 100644 --- a/arch/hexagon/include/asm/cache.h +++ b/arch/hexagon/include/asm/cache.h @@ -21,9 +21,11 @@ #ifndef __ASM_CACHE_H #define __ASM_CACHE_H +#include + /* Bytes per L1 cache line */ -#define L1_CACHE_SHIFT (5) -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_SHIFT 5 +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define __cacheline_aligned __aligned(L1_CACHE_BYTES) #define ____cacheline_aligned __aligned(L1_CACHE_BYTES) diff --git a/arch/ia64/include/asm/atomic.h b/arch/ia64/include/asm/atomic.h index 3fad89e..3047da5 100644 --- a/arch/ia64/include/asm/atomic.h +++ b/arch/ia64/include/asm/atomic.h @@ -209,6 +209,16 @@ atomic64_add_negative (__s64 i, atomic64_t *v) #define atomic64_inc(v) atomic64_add(1, (v)) #define atomic64_dec(v) atomic64_sub(1, (v)) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + /* Atomic operations are already serializing */ #define smp_mb__before_atomic_dec() barrier() #define smp_mb__after_atomic_dec() barrier() diff --git a/arch/ia64/include/asm/cache.h b/arch/ia64/include/asm/cache.h index 988254a..e1ee885 100644 --- a/arch/ia64/include/asm/cache.h +++ b/arch/ia64/include/asm/cache.h @@ -1,6 +1,7 @@ #ifndef _ASM_IA64_CACHE_H #define _ASM_IA64_CACHE_H +#include /* * Copyright (C) 1998-2000 Hewlett-Packard Co @@ -9,7 +10,7 @@ /* Bytes per L1 (data) cache line. */ #define L1_CACHE_SHIFT CONFIG_IA64_L1_CACHE_SHIFT -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #ifdef CONFIG_SMP # define SMP_CACHE_SHIFT L1_CACHE_SHIFT diff --git a/arch/ia64/include/asm/elf.h b/arch/ia64/include/asm/elf.h index b5298eb..67c6e62 100644 --- a/arch/ia64/include/asm/elf.h +++ b/arch/ia64/include/asm/elf.h @@ -42,6 +42,13 @@ */ #define ELF_ET_DYN_BASE (TASK_UNMAPPED_BASE + 0x800000000UL) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (current->personality == PER_LINUX32 ? 0x08048000UL : 0x4000000000000000UL) + +#define PAX_DELTA_MMAP_LEN (current->personality == PER_LINUX32 ? 16 : 3*PAGE_SHIFT - 13) +#define PAX_DELTA_STACK_LEN (current->personality == PER_LINUX32 ? 16 : 3*PAGE_SHIFT - 13) +#endif + #define PT_IA_64_UNWIND 0x70000001 /* IA-64 relocations: */ diff --git a/arch/ia64/include/asm/pgalloc.h b/arch/ia64/include/asm/pgalloc.h index 96a8d92..617a1cf 100644 --- a/arch/ia64/include/asm/pgalloc.h +++ b/arch/ia64/include/asm/pgalloc.h @@ -39,6 +39,12 @@ pgd_populate(struct mm_struct *mm, pgd_t * pgd_entry, pud_t * pud) pgd_val(*pgd_entry) = __pa(pud); } +static inline void +pgd_populate_kernel(struct mm_struct *mm, pgd_t * pgd_entry, pud_t * pud) +{ + pgd_populate(mm, pgd_entry, pud); +} + static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr) { return quicklist_alloc(0, GFP_KERNEL, NULL); @@ -57,6 +63,12 @@ pud_populate(struct mm_struct *mm, pud_t * pud_entry, pmd_t * pmd) pud_val(*pud_entry) = __pa(pmd); } +static inline void +pud_populate_kernel(struct mm_struct *mm, pud_t * pud_entry, pmd_t * pmd) +{ + pud_populate(mm, pud_entry, pmd); +} + static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr) { return quicklist_alloc(0, GFP_KERNEL, NULL); diff --git a/arch/ia64/include/asm/pgtable.h b/arch/ia64/include/asm/pgtable.h index 1a97af3..7529d31 100644 --- a/arch/ia64/include/asm/pgtable.h +++ b/arch/ia64/include/asm/pgtable.h @@ -12,7 +12,7 @@ * David Mosberger-Tang */ - +#include #include #include #include @@ -143,6 +143,17 @@ #define PAGE_READONLY __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) #define PAGE_COPY __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) #define PAGE_COPY_EXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RX) + +#ifdef CONFIG_PAX_PAGEEXEC +# define PAGE_SHARED_NOEXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RW) +# define PAGE_READONLY_NOEXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) +# define PAGE_COPY_NOEXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) +#else +# define PAGE_SHARED_NOEXEC PAGE_SHARED +# define PAGE_READONLY_NOEXEC PAGE_READONLY +# define PAGE_COPY_NOEXEC PAGE_COPY +#endif + #define PAGE_GATE __pgprot(__ACCESS_BITS | _PAGE_PL_0 | _PAGE_AR_X_RX) #define PAGE_KERNEL __pgprot(__DIRTY_BITS | _PAGE_PL_0 | _PAGE_AR_RWX) #define PAGE_KERNELRX __pgprot(__ACCESS_BITS | _PAGE_PL_0 | _PAGE_AR_RX) diff --git a/arch/ia64/include/asm/spinlock.h b/arch/ia64/include/asm/spinlock.h index b77768d..e0795eb 100644 --- a/arch/ia64/include/asm/spinlock.h +++ b/arch/ia64/include/asm/spinlock.h @@ -72,7 +72,7 @@ static __always_inline void __ticket_spin_unlock(arch_spinlock_t *lock) unsigned short *p = (unsigned short *)&lock->lock + 1, tmp; asm volatile ("ld2.bias %0=[%1]" : "=r"(tmp) : "r"(p)); - ACCESS_ONCE(*p) = (tmp + 2) & ~1; + ACCESS_ONCE_RW(*p) = (tmp + 2) & ~1; } static __always_inline void __ticket_spin_unlock_wait(arch_spinlock_t *lock) diff --git a/arch/ia64/include/asm/uaccess.h b/arch/ia64/include/asm/uaccess.h index 449c8c0..432a3d2 100644 --- a/arch/ia64/include/asm/uaccess.h +++ b/arch/ia64/include/asm/uaccess.h @@ -257,7 +257,7 @@ __copy_from_user (void *to, const void __user *from, unsigned long count) const void *__cu_from = (from); \ long __cu_len = (n); \ \ - if (__access_ok(__cu_to, __cu_len, get_fs())) \ + if (__cu_len > 0 && __cu_len <= INT_MAX && __access_ok(__cu_to, __cu_len, get_fs())) \ __cu_len = __copy_user(__cu_to, (__force void __user *) __cu_from, __cu_len); \ __cu_len; \ }) @@ -269,7 +269,7 @@ __copy_from_user (void *to, const void __user *from, unsigned long count) long __cu_len = (n); \ \ __chk_user_ptr(__cu_from); \ - if (__access_ok(__cu_from, __cu_len, get_fs())) \ + if (__cu_len > 0 && __cu_len <= INT_MAX && __access_ok(__cu_from, __cu_len, get_fs())) \ __cu_len = __copy_user((__force void __user *) __cu_to, __cu_from, __cu_len); \ __cu_len; \ }) diff --git a/arch/ia64/kernel/module.c b/arch/ia64/kernel/module.c index 24603be..948052d 100644 --- a/arch/ia64/kernel/module.c +++ b/arch/ia64/kernel/module.c @@ -307,8 +307,7 @@ plt_target (struct plt_entry *plt) void module_free (struct module *mod, void *module_region) { - if (mod && mod->arch.init_unw_table && - module_region == mod->module_init) { + if (mod && mod->arch.init_unw_table && module_region == mod->module_init_rx) { unw_remove_unwind_table(mod->arch.init_unw_table); mod->arch.init_unw_table = NULL; } @@ -494,15 +493,39 @@ module_frob_arch_sections (Elf_Ehdr *ehdr, Elf_Shdr *sechdrs, char *secstrings, } static inline int +in_init_rx (const struct module *mod, uint64_t addr) +{ + return addr - (uint64_t) mod->module_init_rx < mod->init_size_rx; +} + +static inline int +in_init_rw (const struct module *mod, uint64_t addr) +{ + return addr - (uint64_t) mod->module_init_rw < mod->init_size_rw; +} + +static inline int in_init (const struct module *mod, uint64_t addr) { - return addr - (uint64_t) mod->module_init < mod->init_size; + return in_init_rx(mod, addr) || in_init_rw(mod, addr); +} + +static inline int +in_core_rx (const struct module *mod, uint64_t addr) +{ + return addr - (uint64_t) mod->module_core_rx < mod->core_size_rx; +} + +static inline int +in_core_rw (const struct module *mod, uint64_t addr) +{ + return addr - (uint64_t) mod->module_core_rw < mod->core_size_rw; } static inline int in_core (const struct module *mod, uint64_t addr) { - return addr - (uint64_t) mod->module_core < mod->core_size; + return in_core_rx(mod, addr) || in_core_rw(mod, addr); } static inline int @@ -685,7 +708,14 @@ do_reloc (struct module *mod, uint8_t r_type, Elf64_Sym *sym, uint64_t addend, break; case RV_BDREL: - val -= (uint64_t) (in_init(mod, val) ? mod->module_init : mod->module_core); + if (in_init_rx(mod, val)) + val -= (uint64_t) mod->module_init_rx; + else if (in_init_rw(mod, val)) + val -= (uint64_t) mod->module_init_rw; + else if (in_core_rx(mod, val)) + val -= (uint64_t) mod->module_core_rx; + else if (in_core_rw(mod, val)) + val -= (uint64_t) mod->module_core_rw; break; case RV_LTV: @@ -820,15 +850,15 @@ apply_relocate_add (Elf64_Shdr *sechdrs, const char *strtab, unsigned int symind * addresses have been selected... */ uint64_t gp; - if (mod->core_size > MAX_LTOFF) + if (mod->core_size_rx + mod->core_size_rw > MAX_LTOFF) /* * This takes advantage of fact that SHF_ARCH_SMALL gets allocated * at the end of the module. */ - gp = mod->core_size - MAX_LTOFF / 2; + gp = mod->core_size_rx + mod->core_size_rw - MAX_LTOFF / 2; else - gp = mod->core_size / 2; - gp = (uint64_t) mod->module_core + ((gp + 7) & -8); + gp = (mod->core_size_rx + mod->core_size_rw) / 2; + gp = (uint64_t) mod->module_core_rx + ((gp + 7) & -8); mod->arch.gp = gp; DEBUGP("%s: placing gp at 0x%lx\n", __func__, gp); } diff --git a/arch/ia64/kernel/sys_ia64.c b/arch/ia64/kernel/sys_ia64.c index 609d500..7dde2a8 100644 --- a/arch/ia64/kernel/sys_ia64.c +++ b/arch/ia64/kernel/sys_ia64.c @@ -43,6 +43,13 @@ arch_get_unmapped_area (struct file *filp, unsigned long addr, unsigned long len if (REGION_NUMBER(addr) == RGN_HPAGE) addr = 0; #endif + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + addr = mm->free_area_cache; + else +#endif + if (!addr) addr = mm->free_area_cache; @@ -61,14 +68,14 @@ arch_get_unmapped_area (struct file *filp, unsigned long addr, unsigned long len for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { /* At this point: (!vma || addr < vma->vm_end). */ if (TASK_SIZE - len < addr || RGN_MAP_LIMIT - len < REGION_OFFSET(addr)) { - if (start_addr != TASK_UNMAPPED_BASE) { + if (start_addr != mm->mmap_base) { /* Start a new search --- just in case we missed some holes. */ - addr = TASK_UNMAPPED_BASE; + addr = mm->mmap_base; goto full_search; } return -ENOMEM; } - if (!vma || addr + len <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr, len)) { /* Remember the address where we stopped this search: */ mm->free_area_cache = addr + len; return addr; diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S index 53c0ba0..2accdde 100644 --- a/arch/ia64/kernel/vmlinux.lds.S +++ b/arch/ia64/kernel/vmlinux.lds.S @@ -199,7 +199,7 @@ SECTIONS { /* Per-cpu data: */ . = ALIGN(PERCPU_PAGE_SIZE); PERCPU_VADDR(SMP_CACHE_BYTES, PERCPU_ADDR, :percpu) - __phys_per_cpu_start = __per_cpu_load; + __phys_per_cpu_start = per_cpu_load; /* * ensure percpu data fits * into percpu page size diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c index 20b3593..1ce77f0 100644 --- a/arch/ia64/mm/fault.c +++ b/arch/ia64/mm/fault.c @@ -73,6 +73,23 @@ mapped_kernel_page_is_present (unsigned long address) return pte_present(pte); } +#ifdef CONFIG_PAX_PAGEEXEC +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 8; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + void __kprobes ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *regs) { @@ -146,9 +163,23 @@ ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *re mask = ( (((isr >> IA64_ISR_X_BIT) & 1UL) << VM_EXEC_BIT) | (((isr >> IA64_ISR_W_BIT) & 1UL) << VM_WRITE_BIT)); - if ((vma->vm_flags & mask) != mask) + if ((vma->vm_flags & mask) != mask) { + +#ifdef CONFIG_PAX_PAGEEXEC + if (!(vma->vm_flags & VM_EXEC) && (mask & VM_EXEC)) { + if (!(mm->pax_flags & MF_PAX_PAGEEXEC) || address != regs->cr_iip) + goto bad_area; + + up_read(&mm->mmap_sem); + pax_report_fault(regs, (void *)regs->cr_iip, (void *)regs->r12); + do_group_exit(SIGKILL); + } +#endif + goto bad_area; + } + /* * If for any reason at all we couldn't handle the fault, make * sure we exit gracefully rather than endlessly redo the diff --git a/arch/ia64/mm/hugetlbpage.c b/arch/ia64/mm/hugetlbpage.c index 5ca674b..e0e1b70 100644 --- a/arch/ia64/mm/hugetlbpage.c +++ b/arch/ia64/mm/hugetlbpage.c @@ -171,7 +171,7 @@ unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr, u /* At this point: (!vmm || addr < vmm->vm_end). */ if (REGION_OFFSET(addr) + len > RGN_MAP_LIMIT) return -ENOMEM; - if (!vmm || (addr + len) <= vmm->vm_start) + if (check_heap_stack_gap(vmm, addr, len)) return addr; addr = ALIGN(vmm->vm_end, HPAGE_SIZE); } diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 13df239d..cb52116 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -121,6 +121,19 @@ ia64_init_addr_space (void) vma->vm_start = current->thread.rbs_bot & PAGE_MASK; vma->vm_end = vma->vm_start + PAGE_SIZE; vma->vm_flags = VM_DATA_DEFAULT_FLAGS|VM_GROWSUP|VM_ACCOUNT; + +#ifdef CONFIG_PAX_PAGEEXEC + if (current->mm->pax_flags & MF_PAX_PAGEEXEC) { + vma->vm_flags &= ~VM_EXEC; + +#ifdef CONFIG_PAX_MPROTECT + if (current->mm->pax_flags & MF_PAX_MPROTECT) + vma->vm_flags &= ~VM_MAYEXEC; +#endif + + } +#endif + vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); down_write(¤t->mm->mmap_sem); if (insert_vm_struct(current->mm, vma)) { diff --git a/arch/m32r/include/asm/cache.h b/arch/m32r/include/asm/cache.h index 40b3ee9..8c2c112 100644 --- a/arch/m32r/include/asm/cache.h +++ b/arch/m32r/include/asm/cache.h @@ -1,8 +1,10 @@ #ifndef _ASM_M32R_CACHE_H #define _ASM_M32R_CACHE_H +#include + /* L1 cache line size */ #define L1_CACHE_SHIFT 4 -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #endif /* _ASM_M32R_CACHE_H */ diff --git a/arch/m32r/lib/usercopy.c b/arch/m32r/lib/usercopy.c index 82abd15..d95ae5d 100644 --- a/arch/m32r/lib/usercopy.c +++ b/arch/m32r/lib/usercopy.c @@ -14,6 +14,9 @@ unsigned long __generic_copy_to_user(void __user *to, const void *from, unsigned long n) { + if ((long)n < 0) + return n; + prefetch(from); if (access_ok(VERIFY_WRITE, to, n)) __copy_user(to,from,n); @@ -23,6 +26,9 @@ __generic_copy_to_user(void __user *to, const void *from, unsigned long n) unsigned long __generic_copy_from_user(void *to, const void __user *from, unsigned long n) { + if ((long)n < 0) + return n; + prefetchw(to); if (access_ok(VERIFY_READ, from, n)) __copy_user_zeroing(to,from,n); diff --git a/arch/m68k/include/asm/cache.h b/arch/m68k/include/asm/cache.h index 0395c51..5f26031 100644 --- a/arch/m68k/include/asm/cache.h +++ b/arch/m68k/include/asm/cache.h @@ -4,9 +4,11 @@ #ifndef __ARCH_M68K_CACHE_H #define __ARCH_M68K_CACHE_H +#include + /* bytes per L1 cache line */ #define L1_CACHE_SHIFT 4 -#define L1_CACHE_BYTES (1<< L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define ARCH_DMA_MINALIGN L1_CACHE_BYTES diff --git a/arch/microblaze/include/asm/cache.h b/arch/microblaze/include/asm/cache.h index 4efe96a..60e8699 100644 --- a/arch/microblaze/include/asm/cache.h +++ b/arch/microblaze/include/asm/cache.h @@ -13,11 +13,12 @@ #ifndef _ASM_MICROBLAZE_CACHE_H #define _ASM_MICROBLAZE_CACHE_H +#include #include #define L1_CACHE_SHIFT 5 /* word-granular cache in microblaze */ -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define SMP_CACHE_BYTES L1_CACHE_BYTES diff --git a/arch/mips/include/asm/atomic.h b/arch/mips/include/asm/atomic.h index 1d93f81..67794d0 100644 --- a/arch/mips/include/asm/atomic.h +++ b/arch/mips/include/asm/atomic.h @@ -21,6 +21,10 @@ #include #include +#ifdef CONFIG_GENERIC_ATOMIC64 +#include +#endif + #define ATOMIC_INIT(i) { (i) } /* @@ -765,6 +769,16 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) */ #define atomic64_add_negative(i, v) (atomic64_add_return(i, (v)) < 0) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + #endif /* CONFIG_64BIT */ /* diff --git a/arch/mips/include/asm/cache.h b/arch/mips/include/asm/cache.h index b4db69f..8f3b093 100644 --- a/arch/mips/include/asm/cache.h +++ b/arch/mips/include/asm/cache.h @@ -9,10 +9,11 @@ #ifndef _ASM_CACHE_H #define _ASM_CACHE_H +#include #include #define L1_CACHE_SHIFT CONFIG_MIPS_L1_CACHE_SHIFT -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define SMP_CACHE_SHIFT L1_CACHE_SHIFT #define SMP_CACHE_BYTES L1_CACHE_BYTES diff --git a/arch/mips/include/asm/elf.h b/arch/mips/include/asm/elf.h index 455c0ac..ad65fbe 100644 --- a/arch/mips/include/asm/elf.h +++ b/arch/mips/include/asm/elf.h @@ -372,13 +372,16 @@ extern const char *__elf_platform; #define ELF_ET_DYN_BASE (TASK_SIZE / 3 * 2) #endif +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (TASK_IS_32BIT_ADDR ? 0x00400000UL : 0x00400000UL) + +#define PAX_DELTA_MMAP_LEN (TASK_IS_32BIT_ADDR ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#define PAX_DELTA_STACK_LEN (TASK_IS_32BIT_ADDR ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#endif + #define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1 struct linux_binprm; extern int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp); -struct mm_struct; -extern unsigned long arch_randomize_brk(struct mm_struct *mm); -#define arch_randomize_brk arch_randomize_brk - #endif /* _ASM_ELF_H */ diff --git a/arch/mips/include/asm/page.h b/arch/mips/include/asm/page.h index da9bd7d..91aa7ab 100644 --- a/arch/mips/include/asm/page.h +++ b/arch/mips/include/asm/page.h @@ -98,7 +98,7 @@ extern void copy_user_highpage(struct page *to, struct page *from, #ifdef CONFIG_CPU_MIPS32 typedef struct { unsigned long pte_low, pte_high; } pte_t; #define pte_val(x) ((x).pte_low | ((unsigned long long)(x).pte_high << 32)) - #define __pte(x) ({ pte_t __pte = {(x), ((unsigned long long)(x)) >> 32}; __pte; }) + #define __pte(x) ({ pte_t __pte = {(x), (x) >> 32}; __pte; }) #else typedef struct { unsigned long long pte; } pte_t; #define pte_val(x) ((x).pte) diff --git a/arch/mips/include/asm/pgalloc.h b/arch/mips/include/asm/pgalloc.h index 881d18b..cea38bc 100644 --- a/arch/mips/include/asm/pgalloc.h +++ b/arch/mips/include/asm/pgalloc.h @@ -37,6 +37,11 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) { set_pud(pud, __pud((unsigned long)pmd)); } + +static inline void pud_populate_kernel(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) +{ + pud_populate(mm, pud, pmd); +} #endif /* diff --git a/arch/mips/include/asm/system.h b/arch/mips/include/asm/system.h index 6018c80..7c37203 100644 --- a/arch/mips/include/asm/system.h +++ b/arch/mips/include/asm/system.h @@ -230,6 +230,6 @@ extern void per_cpu_trap_init(void); */ #define __ARCH_WANT_UNLOCKED_CTXSW -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) #endif /* _ASM_SYSTEM_H */ diff --git a/arch/mips/kernel/binfmt_elfn32.c b/arch/mips/kernel/binfmt_elfn32.c index 9fdd8bc..4bd7f1a 100644 --- a/arch/mips/kernel/binfmt_elfn32.c +++ b/arch/mips/kernel/binfmt_elfn32.c @@ -50,6 +50,13 @@ typedef elf_fpreg_t elf_fpregset_t[ELF_NFPREG]; #undef ELF_ET_DYN_BASE #define ELF_ET_DYN_BASE (TASK32_SIZE / 3 * 2) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (TASK_IS_32BIT_ADDR ? 0x00400000UL : 0x00400000UL) + +#define PAX_DELTA_MMAP_LEN (TASK_IS_32BIT_ADDR ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#define PAX_DELTA_STACK_LEN (TASK_IS_32BIT_ADDR ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#endif + #include #include #include diff --git a/arch/mips/kernel/binfmt_elfo32.c b/arch/mips/kernel/binfmt_elfo32.c index ff44823..97f8906 100644 --- a/arch/mips/kernel/binfmt_elfo32.c +++ b/arch/mips/kernel/binfmt_elfo32.c @@ -52,6 +52,13 @@ typedef elf_fpreg_t elf_fpregset_t[ELF_NFPREG]; #undef ELF_ET_DYN_BASE #define ELF_ET_DYN_BASE (TASK32_SIZE / 3 * 2) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (TASK_IS_32BIT_ADDR ? 0x00400000UL : 0x00400000UL) + +#define PAX_DELTA_MMAP_LEN (TASK_IS_32BIT_ADDR ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#define PAX_DELTA_STACK_LEN (TASK_IS_32BIT_ADDR ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#endif + #include /* diff --git a/arch/mips/kernel/process.c b/arch/mips/kernel/process.c index 7955409..ceaea7c 100644 --- a/arch/mips/kernel/process.c +++ b/arch/mips/kernel/process.c @@ -483,15 +483,3 @@ unsigned long get_wchan(struct task_struct *task) out: return pc; } - -/* - * Don't forget that the stack pointer must be aligned on a 8 bytes - * boundary for 32-bits ABI and 16 bytes for 64-bits ABI. - */ -unsigned long arch_align_stack(unsigned long sp) -{ - if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) - sp -= get_random_int() & ~PAGE_MASK; - - return sp & ALMASK; -} diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c index 69ebd58..e4bff83 100644 --- a/arch/mips/mm/fault.c +++ b/arch/mips/mm/fault.c @@ -28,6 +28,23 @@ #include /* For VMALLOC_END */ #include +#ifdef CONFIG_PAX_PAGEEXEC +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 5; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + /* * This routine handles page faults. It determines the address, * and the problem, and then passes it off to one of the appropriate diff --git a/arch/mips/mm/mmap.c b/arch/mips/mm/mmap.c index 302d779..7d35bf8 100644 --- a/arch/mips/mm/mmap.c +++ b/arch/mips/mm/mmap.c @@ -95,6 +95,11 @@ static unsigned long arch_get_unmapped_area_common(struct file *filp, do_color_align = 1; /* requesting a specific address */ + +#ifdef CONFIG_PAX_RANDMMAP + if (!(current->mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_color_align) addr = COLOUR_ALIGN(addr, pgoff); @@ -102,8 +107,7 @@ static unsigned long arch_get_unmapped_area_common(struct file *filp, addr = PAGE_ALIGN(addr); vma = find_vma(mm, addr); - if (TASK_SIZE - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (TASK_SIZE - len >= addr && check_heap_stack_gap(vmm, addr, len)) return addr; } @@ -118,7 +122,7 @@ static unsigned long arch_get_unmapped_area_common(struct file *filp, /* At this point: (!vma || addr < vma->vm_end). */ if (TASK_SIZE - len < addr) return -ENOMEM; - if (!vma || addr + len <= vma->vm_start) + if (check_heap_stack_gap(vmm, addr, len)) return addr; addr = vma->vm_end; if (do_color_align) @@ -145,7 +149,7 @@ static unsigned long arch_get_unmapped_area_common(struct file *filp, /* make sure it can fit in the remaining address space */ if (likely(addr > len)) { vma = find_vma(mm, addr - len); - if (!vma || addr <= vma->vm_start) { + if (check_heap_stack_gap(vmm, addr - len, len)) /* cache the address as a hint for next time */ return mm->free_area_cache = addr - len; } @@ -165,7 +169,7 @@ static unsigned long arch_get_unmapped_area_common(struct file *filp, * return with success: */ vma = find_vma(mm, addr); - if (likely(!vma || addr + len <= vma->vm_start)) { + if (check_heap_stack_gap(vmm, addr, len)) { /* cache the address as a hint for next time */ return mm->free_area_cache = addr; } @@ -242,30 +246,3 @@ void arch_pick_mmap_layout(struct mm_struct *mm) mm->unmap_area = arch_unmap_area_topdown; } } - -static inline unsigned long brk_rnd(void) -{ - unsigned long rnd = get_random_int(); - - rnd = rnd << PAGE_SHIFT; - /* 8MB for 32bit, 256MB for 64bit */ - if (TASK_IS_32BIT_ADDR) - rnd = rnd & 0x7ffffful; - else - rnd = rnd & 0xffffffful; - - return rnd; -} - -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long base = mm->brk; - unsigned long ret; - - ret = PAGE_ALIGN(base + brk_rnd()); - - if (ret < mm->brk) - return mm->brk; - - return ret; -} diff --git a/arch/mn10300/proc-mn103e010/include/proc/cache.h b/arch/mn10300/proc-mn103e010/include/proc/cache.h index 967d144..db12197 100644 --- a/arch/mn10300/proc-mn103e010/include/proc/cache.h +++ b/arch/mn10300/proc-mn103e010/include/proc/cache.h @@ -11,12 +11,14 @@ #ifndef _ASM_PROC_CACHE_H #define _ASM_PROC_CACHE_H +#include + /* L1 cache */ #define L1_CACHE_NWAYS 4 /* number of ways in caches */ #define L1_CACHE_NENTRIES 256 /* number of entries in each way */ -#define L1_CACHE_BYTES 16 /* bytes per entry */ #define L1_CACHE_SHIFT 4 /* shift for bytes per entry */ +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) /* bytes per entry */ #define L1_CACHE_WAYDISP 0x1000 /* displacement of one way from the next */ #define L1_CACHE_TAG_VALID 0x00000001 /* cache tag valid bit */ diff --git a/arch/mn10300/proc-mn2ws0050/include/proc/cache.h b/arch/mn10300/proc-mn2ws0050/include/proc/cache.h index bcb5df2..84fabd2 100644 --- a/arch/mn10300/proc-mn2ws0050/include/proc/cache.h +++ b/arch/mn10300/proc-mn2ws0050/include/proc/cache.h @@ -16,13 +16,15 @@ #ifndef _ASM_PROC_CACHE_H #define _ASM_PROC_CACHE_H +#include + /* * L1 cache */ #define L1_CACHE_NWAYS 4 /* number of ways in caches */ #define L1_CACHE_NENTRIES 128 /* number of entries in each way */ -#define L1_CACHE_BYTES 32 /* bytes per entry */ #define L1_CACHE_SHIFT 5 /* shift for bytes per entry */ +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) /* bytes per entry */ #define L1_CACHE_WAYDISP 0x1000 /* distance from one way to the next */ #define L1_CACHE_TAG_VALID 0x00000001 /* cache tag valid bit */ diff --git a/arch/openrisc/include/asm/cache.h b/arch/openrisc/include/asm/cache.h index 4ce7a01..449202a 100644 --- a/arch/openrisc/include/asm/cache.h +++ b/arch/openrisc/include/asm/cache.h @@ -19,11 +19,13 @@ #ifndef __ASM_OPENRISC_CACHE_H #define __ASM_OPENRISC_CACHE_H +#include + /* FIXME: How can we replace these with values from the CPU... * they shouldn't be hard-coded! */ -#define L1_CACHE_BYTES 16 #define L1_CACHE_SHIFT 4 +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #endif /* __ASM_OPENRISC_CACHE_H */ diff --git a/arch/parisc/include/asm/atomic.h b/arch/parisc/include/asm/atomic.h index 4054b31..a10c105 100644 --- a/arch/parisc/include/asm/atomic.h +++ b/arch/parisc/include/asm/atomic.h @@ -335,6 +335,16 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + #endif /* !CONFIG_64BIT */ diff --git a/arch/parisc/include/asm/cache.h b/arch/parisc/include/asm/cache.h index 47f11c7..3420df2 100644 --- a/arch/parisc/include/asm/cache.h +++ b/arch/parisc/include/asm/cache.h @@ -5,6 +5,7 @@ #ifndef __ARCH_PARISC_CACHE_H #define __ARCH_PARISC_CACHE_H +#include /* * PA 2.0 processors have 64-byte cachelines; PA 1.1 processors have @@ -15,13 +16,13 @@ * just ruin performance. */ #ifdef CONFIG_PA20 -#define L1_CACHE_BYTES 64 #define L1_CACHE_SHIFT 6 #else -#define L1_CACHE_BYTES 32 #define L1_CACHE_SHIFT 5 #endif +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) + #ifndef __ASSEMBLY__ #define SMP_CACHE_BYTES L1_CACHE_BYTES diff --git a/arch/parisc/include/asm/elf.h b/arch/parisc/include/asm/elf.h index 19f6cb1..6c78cf2 100644 --- a/arch/parisc/include/asm/elf.h +++ b/arch/parisc/include/asm/elf.h @@ -342,6 +342,13 @@ struct pt_regs; /* forward declaration... */ #define ELF_ET_DYN_BASE (TASK_UNMAPPED_BASE + 0x01000000) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE 0x10000UL + +#define PAX_DELTA_MMAP_LEN 16 +#define PAX_DELTA_STACK_LEN 16 +#endif + /* This yields a mask that user programs can use to figure out what instruction set this CPU supports. This could be done in user space, but it's not easy, and we've already done it here. */ diff --git a/arch/parisc/include/asm/pgalloc.h b/arch/parisc/include/asm/pgalloc.h index fc987a1..6e068ef 100644 --- a/arch/parisc/include/asm/pgalloc.h +++ b/arch/parisc/include/asm/pgalloc.h @@ -61,6 +61,11 @@ static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmd) (__u32)(__pa((unsigned long)pmd) >> PxD_VALUE_SHIFT)); } +static inline void pgd_populate_kernel(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmd) +{ + pgd_populate(mm, pgd, pmd); +} + static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long address) { pmd_t *pmd = (pmd_t *)__get_free_pages(GFP_KERNEL|__GFP_REPEAT, @@ -93,6 +98,7 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) #define pmd_alloc_one(mm, addr) ({ BUG(); ((pmd_t *)2); }) #define pmd_free(mm, x) do { } while (0) #define pgd_populate(mm, pmd, pte) BUG() +#define pgd_populate_kernel(mm, pmd, pte) BUG() #endif diff --git a/arch/parisc/include/asm/pgtable.h b/arch/parisc/include/asm/pgtable.h index 22dadeb..f6c2be4 100644 --- a/arch/parisc/include/asm/pgtable.h +++ b/arch/parisc/include/asm/pgtable.h @@ -210,6 +210,17 @@ struct vm_area_struct; #define PAGE_EXECREAD __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | _PAGE_EXEC |_PAGE_ACCESSED) #define PAGE_COPY PAGE_EXECREAD #define PAGE_RWX __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | _PAGE_WRITE | _PAGE_EXEC |_PAGE_ACCESSED) + +#ifdef CONFIG_PAX_PAGEEXEC +# define PAGE_SHARED_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | _PAGE_WRITE | _PAGE_ACCESSED) +# define PAGE_COPY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | _PAGE_ACCESSED) +# define PAGE_READONLY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | _PAGE_ACCESSED) +#else +# define PAGE_SHARED_NOEXEC PAGE_SHARED +# define PAGE_COPY_NOEXEC PAGE_COPY +# define PAGE_READONLY_NOEXEC PAGE_READONLY +#endif + #define PAGE_KERNEL __pgprot(_PAGE_KERNEL) #define PAGE_KERNEL_EXEC __pgprot(_PAGE_KERNEL_EXEC) #define PAGE_KERNEL_RWX __pgprot(_PAGE_KERNEL_RWX) diff --git a/arch/parisc/kernel/module.c b/arch/parisc/kernel/module.c index 5e34ccf..672bc9c 100644 --- a/arch/parisc/kernel/module.c +++ b/arch/parisc/kernel/module.c @@ -98,16 +98,38 @@ /* three functions to determine where in the module core * or init pieces the location is */ +static inline int in_init_rx(struct module *me, void *loc) +{ + return (loc >= me->module_init_rx && + loc < (me->module_init_rx + me->init_size_rx)); +} + +static inline int in_init_rw(struct module *me, void *loc) +{ + return (loc >= me->module_init_rw && + loc < (me->module_init_rw + me->init_size_rw)); +} + static inline int in_init(struct module *me, void *loc) { - return (loc >= me->module_init && - loc <= (me->module_init + me->init_size)); + return in_init_rx(me, loc) || in_init_rw(me, loc); +} + +static inline int in_core_rx(struct module *me, void *loc) +{ + return (loc >= me->module_core_rx && + loc < (me->module_core_rx + me->core_size_rx)); +} + +static inline int in_core_rw(struct module *me, void *loc) +{ + return (loc >= me->module_core_rw && + loc < (me->module_core_rw + me->core_size_rw)); } static inline int in_core(struct module *me, void *loc) { - return (loc >= me->module_core && - loc <= (me->module_core + me->core_size)); + return in_core_rx(me, loc) || in_core_rw(me, loc); } static inline int in_local(struct module *me, void *loc) @@ -373,13 +395,13 @@ int module_frob_arch_sections(CONST Elf_Ehdr *hdr, } /* align things a bit */ - me->core_size = ALIGN(me->core_size, 16); - me->arch.got_offset = me->core_size; - me->core_size += gots * sizeof(struct got_entry); + me->core_size_rw = ALIGN(me->core_size_rw, 16); + me->arch.got_offset = me->core_size_rw; + me->core_size_rw += gots * sizeof(struct got_entry); - me->core_size = ALIGN(me->core_size, 16); - me->arch.fdesc_offset = me->core_size; - me->core_size += fdescs * sizeof(Elf_Fdesc); + me->core_size_rw = ALIGN(me->core_size_rw, 16); + me->arch.fdesc_offset = me->core_size_rw; + me->core_size_rw += fdescs * sizeof(Elf_Fdesc); me->arch.got_max = gots; me->arch.fdesc_max = fdescs; @@ -397,7 +419,7 @@ static Elf64_Word get_got(struct module *me, unsigned long value, long addend) BUG_ON(value == 0); - got = me->module_core + me->arch.got_offset; + got = me->module_core_rw + me->arch.got_offset; for (i = 0; got[i].addr; i++) if (got[i].addr == value) goto out; @@ -415,7 +437,7 @@ static Elf64_Word get_got(struct module *me, unsigned long value, long addend) #ifdef CONFIG_64BIT static Elf_Addr get_fdesc(struct module *me, unsigned long value) { - Elf_Fdesc *fdesc = me->module_core + me->arch.fdesc_offset; + Elf_Fdesc *fdesc = me->module_core_rw + me->arch.fdesc_offset; if (!value) { printk(KERN_ERR "%s: zero OPD requested!\n", me->name); @@ -433,7 +455,7 @@ static Elf_Addr get_fdesc(struct module *me, unsigned long value) /* Create new one */ fdesc->addr = value; - fdesc->gp = (Elf_Addr)me->module_core + me->arch.got_offset; + fdesc->gp = (Elf_Addr)me->module_core_rw + me->arch.got_offset; return (Elf_Addr)fdesc; } #endif /* CONFIG_64BIT */ @@ -845,7 +867,7 @@ register_unwind_table(struct module *me, table = (unsigned char *)sechdrs[me->arch.unwind_section].sh_addr; end = table + sechdrs[me->arch.unwind_section].sh_size; - gp = (Elf_Addr)me->module_core + me->arch.got_offset; + gp = (Elf_Addr)me->module_core_rw + me->arch.got_offset; DEBUGP("register_unwind_table(), sect = %d at 0x%p - 0x%p (gp=0x%lx)\n", me->arch.unwind_section, table, end, gp); diff --git a/arch/parisc/kernel/sys_parisc.c b/arch/parisc/kernel/sys_parisc.c index c9b9322..02d8940 100644 --- a/arch/parisc/kernel/sys_parisc.c +++ b/arch/parisc/kernel/sys_parisc.c @@ -43,7 +43,7 @@ static unsigned long get_unshared_area(unsigned long addr, unsigned long len) /* At this point: (!vma || addr < vma->vm_end). */ if (TASK_SIZE - len < addr) return -ENOMEM; - if (!vma || addr + len <= vma->vm_start) + if (check_heap_stack_gap(vma, addr, len)) return addr; addr = vma->vm_end; } @@ -79,7 +79,7 @@ static unsigned long get_shared_area(struct address_space *mapping, /* At this point: (!vma || addr < vma->vm_end). */ if (TASK_SIZE - len < addr) return -ENOMEM; - if (!vma || addr + len <= vma->vm_start) + if (check_heap_stack_gap(vma, addr, len)) return addr; addr = DCACHE_ALIGN(vma->vm_end - offset) + offset; if (addr < vma->vm_end) /* handle wraparound */ @@ -98,7 +98,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, if (flags & MAP_FIXED) return addr; if (!addr) - addr = TASK_UNMAPPED_BASE; + addr = current->mm->mmap_base; if (filp) { addr = get_shared_area(filp->f_mapping, addr, len, pgoff); diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c index f19e660..414fe24 100644 --- a/arch/parisc/kernel/traps.c +++ b/arch/parisc/kernel/traps.c @@ -733,9 +733,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) down_read(¤t->mm->mmap_sem); vma = find_vma(current->mm,regs->iaoq[0]); - if (vma && (regs->iaoq[0] >= vma->vm_start) - && (vma->vm_flags & VM_EXEC)) { - + if (vma && (regs->iaoq[0] >= vma->vm_start)) { fault_address = regs->iaoq[0]; fault_space = regs->iasq[0]; diff --git a/arch/parisc/mm/fault.c b/arch/parisc/mm/fault.c index 18162ce..94de376 100644 --- a/arch/parisc/mm/fault.c +++ b/arch/parisc/mm/fault.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -52,7 +53,7 @@ DEFINE_PER_CPU(struct exception_data, exception_data); static unsigned long parisc_acctyp(unsigned long code, unsigned int inst) { - if (code == 6 || code == 16) + if (code == 6 || code == 7 || code == 16) return VM_EXEC; switch (inst & 0xf0000000) { @@ -138,6 +139,116 @@ parisc_acctyp(unsigned long code, unsigned int inst) } #endif +#ifdef CONFIG_PAX_PAGEEXEC +/* + * PaX: decide what to do with offenders (instruction_pointer(regs) = fault address) + * + * returns 1 when task should be killed + * 2 when rt_sigreturn trampoline was detected + * 3 when unpatched PLT trampoline was detected + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + +#ifdef CONFIG_PAX_EMUPLT + int err; + + do { /* PaX: unpatched PLT emulation */ + unsigned int bl, depwi; + + err = get_user(bl, (unsigned int *)instruction_pointer(regs)); + err |= get_user(depwi, (unsigned int *)(instruction_pointer(regs)+4)); + + if (err) + break; + + if (bl == 0xEA9F1FDDU && depwi == 0xD6801C1EU) { + unsigned int ldw, bv, ldw2, addr = instruction_pointer(regs)-12; + + err = get_user(ldw, (unsigned int *)addr); + err |= get_user(bv, (unsigned int *)(addr+4)); + err |= get_user(ldw2, (unsigned int *)(addr+8)); + + if (err) + break; + + if (ldw == 0x0E801096U && + bv == 0xEAC0C000U && + ldw2 == 0x0E881095U) + { + unsigned int resolver, map; + + err = get_user(resolver, (unsigned int *)(instruction_pointer(regs)+8)); + err |= get_user(map, (unsigned int *)(instruction_pointer(regs)+12)); + if (err) + break; + + regs->gr[20] = instruction_pointer(regs)+8; + regs->gr[21] = map; + regs->gr[22] = resolver; + regs->iaoq[0] = resolver | 3UL; + regs->iaoq[1] = regs->iaoq[0] + 4; + return 3; + } + } + } while (0); +#endif + +#ifdef CONFIG_PAX_EMUTRAMP + +#ifndef CONFIG_PAX_EMUSIGRT + if (!(current->mm->pax_flags & MF_PAX_EMUTRAMP)) + return 1; +#endif + + do { /* PaX: rt_sigreturn emulation */ + unsigned int ldi1, ldi2, bel, nop; + + err = get_user(ldi1, (unsigned int *)instruction_pointer(regs)); + err |= get_user(ldi2, (unsigned int *)(instruction_pointer(regs)+4)); + err |= get_user(bel, (unsigned int *)(instruction_pointer(regs)+8)); + err |= get_user(nop, (unsigned int *)(instruction_pointer(regs)+12)); + + if (err) + break; + + if ((ldi1 == 0x34190000U || ldi1 == 0x34190002U) && + ldi2 == 0x3414015AU && + bel == 0xE4008200U && + nop == 0x08000240U) + { + regs->gr[25] = (ldi1 & 2) >> 1; + regs->gr[20] = __NR_rt_sigreturn; + regs->gr[31] = regs->iaoq[1] + 16; + regs->sr[0] = regs->iasq[1]; + regs->iaoq[0] = 0x100UL; + regs->iaoq[1] = regs->iaoq[0] + 4; + regs->iasq[0] = regs->sr[2]; + regs->iasq[1] = regs->sr[2]; + return 2; + } + } while (0); +#endif + + return 1; +} + +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 5; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + int fixup_exception(struct pt_regs *regs) { const struct exception_table_entry *fix; @@ -192,8 +303,33 @@ good_area: acc_type = parisc_acctyp(code,regs->iir); - if ((vma->vm_flags & acc_type) != acc_type) + if ((vma->vm_flags & acc_type) != acc_type) { + +#ifdef CONFIG_PAX_PAGEEXEC + if ((mm->pax_flags & MF_PAX_PAGEEXEC) && (acc_type & VM_EXEC) && + (address & ~3UL) == instruction_pointer(regs)) + { + up_read(&mm->mmap_sem); + switch (pax_handle_fetch_fault(regs)) { + +#ifdef CONFIG_PAX_EMUPLT + case 3: + return; +#endif + +#ifdef CONFIG_PAX_EMUTRAMP + case 2: + return; +#endif + + } + pax_report_fault(regs, (void *)instruction_pointer(regs), (void *)regs->gr[30]); + do_group_exit(SIGKILL); + } +#endif + goto bad_area; + } /* * If for any reason at all we couldn't handle the fault, make diff --git a/arch/powerpc/include/asm/atomic.h b/arch/powerpc/include/asm/atomic.h index 02e41b5..ec6e26c 100644 --- a/arch/powerpc/include/asm/atomic.h +++ b/arch/powerpc/include/asm/atomic.h @@ -469,6 +469,16 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + #endif /* __powerpc64__ */ #endif /* __KERNEL__ */ diff --git a/arch/powerpc/include/asm/cache.h b/arch/powerpc/include/asm/cache.h index 4b50941..5605819 100644 --- a/arch/powerpc/include/asm/cache.h +++ b/arch/powerpc/include/asm/cache.h @@ -3,6 +3,7 @@ #ifdef __KERNEL__ +#include /* bytes per L1 cache line */ #if defined(CONFIG_8xx) || defined(CONFIG_403GCX) @@ -22,7 +23,7 @@ #define L1_CACHE_SHIFT 7 #endif -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define SMP_CACHE_BYTES L1_CACHE_BYTES diff --git a/arch/powerpc/include/asm/elf.h b/arch/powerpc/include/asm/elf.h index 3bf9cca..e7457d0 100644 --- a/arch/powerpc/include/asm/elf.h +++ b/arch/powerpc/include/asm/elf.h @@ -178,8 +178,19 @@ typedef elf_fpreg_t elf_vsrreghalf_t32[ELF_NVSRHALFREG]; the loader. We need to make sure that it is out of the way of the program that it will "exec", and that there is sufficient room for the brk. */ -extern unsigned long randomize_et_dyn(unsigned long base); -#define ELF_ET_DYN_BASE (randomize_et_dyn(0x20000000)) +#define ELF_ET_DYN_BASE (0x20000000) + +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (0x10000000UL) + +#ifdef __powerpc64__ +#define PAX_DELTA_MMAP_LEN (is_32bit_task() ? 16 : 28) +#define PAX_DELTA_STACK_LEN (is_32bit_task() ? 16 : 28) +#else +#define PAX_DELTA_MMAP_LEN 15 +#define PAX_DELTA_STACK_LEN 15 +#endif +#endif /* * Our registers are always unsigned longs, whether we're a 32 bit @@ -274,9 +285,6 @@ extern int arch_setup_additional_pages(struct linux_binprm *bprm, (0x7ff >> (PAGE_SHIFT - 12)) : \ (0x3ffff >> (PAGE_SHIFT - 12))) -extern unsigned long arch_randomize_brk(struct mm_struct *mm); -#define arch_randomize_brk arch_randomize_brk - #endif /* __KERNEL__ */ /* diff --git a/arch/powerpc/include/asm/kmap_types.h b/arch/powerpc/include/asm/kmap_types.h index bca8fdc..61e9580 100644 --- a/arch/powerpc/include/asm/kmap_types.h +++ b/arch/powerpc/include/asm/kmap_types.h @@ -27,6 +27,7 @@ enum km_type { KM_PPC_SYNC_PAGE, KM_PPC_SYNC_ICACHE, KM_KDB, + KM_CLEARPAGE, KM_TYPE_NR }; diff --git a/arch/powerpc/include/asm/mman.h b/arch/powerpc/include/asm/mman.h index d4a7f64..451de1c 100644 --- a/arch/powerpc/include/asm/mman.h +++ b/arch/powerpc/include/asm/mman.h @@ -44,7 +44,7 @@ static inline unsigned long arch_calc_vm_prot_bits(unsigned long prot) } #define arch_calc_vm_prot_bits(prot) arch_calc_vm_prot_bits(prot) -static inline pgprot_t arch_vm_get_page_prot(unsigned long vm_flags) +static inline pgprot_t arch_vm_get_page_prot(vm_flags_t vm_flags) { return (vm_flags & VM_SAO) ? __pgprot(_PAGE_SAO) : __pgprot(0); } diff --git a/arch/powerpc/include/asm/page.h b/arch/powerpc/include/asm/page.h index f072e97..b436dee 100644 --- a/arch/powerpc/include/asm/page.h +++ b/arch/powerpc/include/asm/page.h @@ -220,8 +220,9 @@ extern long long virt_phys_offset; * and needs to be executable. This means the whole heap ends * up being executable. */ -#define VM_DATA_DEFAULT_FLAGS32 (VM_READ | VM_WRITE | VM_EXEC | \ - VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC) +#define VM_DATA_DEFAULT_FLAGS32 \ + (((current->personality & READ_IMPLIES_EXEC) ? VM_EXEC : 0) | \ + VM_READ | VM_WRITE | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC) #define VM_DATA_DEFAULT_FLAGS64 (VM_READ | VM_WRITE | \ VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC) @@ -249,6 +250,9 @@ extern long long virt_phys_offset; #define is_kernel_addr(x) ((x) >= PAGE_OFFSET) #endif +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) + /* * Use the top bit of the higher-level page table entries to indicate whether * the entries we point to contain hugepages. This works because we know that diff --git a/arch/powerpc/include/asm/page_64.h b/arch/powerpc/include/asm/page_64.h index fed85e6..da5c71b 100644 --- a/arch/powerpc/include/asm/page_64.h +++ b/arch/powerpc/include/asm/page_64.h @@ -146,15 +146,18 @@ do { \ * stack by default, so in the absence of a PT_GNU_STACK program header * we turn execute permission off. */ -#define VM_STACK_DEFAULT_FLAGS32 (VM_READ | VM_WRITE | VM_EXEC | \ - VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC) +#define VM_STACK_DEFAULT_FLAGS32 \ + (((current->personality & READ_IMPLIES_EXEC) ? VM_EXEC : 0) | \ + VM_READ | VM_WRITE | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC) #define VM_STACK_DEFAULT_FLAGS64 (VM_READ | VM_WRITE | \ VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC) +#ifndef CONFIG_PAX_PAGEEXEC #define VM_STACK_DEFAULT_FLAGS \ (is_32bit_task() ? \ VM_STACK_DEFAULT_FLAGS32 : VM_STACK_DEFAULT_FLAGS64) +#endif #include diff --git a/arch/powerpc/include/asm/pgalloc-64.h b/arch/powerpc/include/asm/pgalloc-64.h index 292725c..f87ae14 100644 --- a/arch/powerpc/include/asm/pgalloc-64.h +++ b/arch/powerpc/include/asm/pgalloc-64.h @@ -50,6 +50,7 @@ static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd) #ifndef CONFIG_PPC_64K_PAGES #define pgd_populate(MM, PGD, PUD) pgd_set(PGD, PUD) +#define pgd_populate_kernel(MM, PGD, PUD) pgd_populate((MM), (PGD), (PUD)) static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr) { @@ -67,6 +68,11 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) pud_set(pud, (unsigned long)pmd); } +static inline void pud_populate_kernel(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) +{ + pud_populate(mm, pud, pmd); +} + #define pmd_populate(mm, pmd, pte_page) \ pmd_populate_kernel(mm, pmd, page_address(pte_page)) #define pmd_populate_kernel(mm, pmd, pte) pmd_set(pmd, (unsigned long)(pte)) @@ -76,6 +82,7 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) #else /* CONFIG_PPC_64K_PAGES */ #define pud_populate(mm, pud, pmd) pud_set(pud, (unsigned long)pmd) +#define pud_populate_kernel(mm, pud, pmd) pud_populate((mm), (pud), (pmd)) static inline void pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte) diff --git a/arch/powerpc/include/asm/pgtable.h b/arch/powerpc/include/asm/pgtable.h index 2e0e411..7899c68 100644 --- a/arch/powerpc/include/asm/pgtable.h +++ b/arch/powerpc/include/asm/pgtable.h @@ -2,6 +2,7 @@ #define _ASM_POWERPC_PGTABLE_H #ifdef __KERNEL__ +#include #ifndef __ASSEMBLY__ #include /* For TASK_SIZE */ #include diff --git a/arch/powerpc/include/asm/pte-hash32.h b/arch/powerpc/include/asm/pte-hash32.h index 4aad413..85d86bf 100644 --- a/arch/powerpc/include/asm/pte-hash32.h +++ b/arch/powerpc/include/asm/pte-hash32.h @@ -21,6 +21,7 @@ #define _PAGE_FILE 0x004 /* when !present: nonlinear file mapping */ #define _PAGE_USER 0x004 /* usermode access allowed */ #define _PAGE_GUARDED 0x008 /* G: prohibit speculative access */ +#define _PAGE_EXEC _PAGE_GUARDED #define _PAGE_COHERENT 0x010 /* M: enforce memory coherence (SMP systems) */ #define _PAGE_NO_CACHE 0x020 /* I: cache inhibit */ #define _PAGE_WRITETHRU 0x040 /* W: cache write-through */ diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h index 7fdc2c0..e47a9b02d3 100644 --- a/arch/powerpc/include/asm/reg.h +++ b/arch/powerpc/include/asm/reg.h @@ -212,6 +212,7 @@ #define SPRN_DBCR 0x136 /* e300 Data Breakpoint Control Reg */ #define SPRN_DSISR 0x012 /* Data Storage Interrupt Status Register */ #define DSISR_NOHPTE 0x40000000 /* no translation found */ +#define DSISR_GUARDED 0x10000000 /* fetch from guarded storage */ #define DSISR_PROTFAULT 0x08000000 /* protection fault */ #define DSISR_ISSTORE 0x02000000 /* access was a store */ #define DSISR_DABRMATCH 0x00400000 /* hit data breakpoint */ diff --git a/arch/powerpc/include/asm/system.h b/arch/powerpc/include/asm/system.h index c377457..3c69fbc 100644 --- a/arch/powerpc/include/asm/system.h +++ b/arch/powerpc/include/asm/system.h @@ -539,7 +539,7 @@ __cmpxchg_local(volatile void *ptr, unsigned long old, unsigned long new, #define cmpxchg64_local(ptr, o, n) __cmpxchg64_local_generic((ptr), (o), (n)) #endif -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) /* Used in very early kernel initialization. */ extern unsigned long reloc_offset(void); diff --git a/arch/powerpc/include/asm/uaccess.h b/arch/powerpc/include/asm/uaccess.h index bd0fb84..a42a14b 100644 --- a/arch/powerpc/include/asm/uaccess.h +++ b/arch/powerpc/include/asm/uaccess.h @@ -13,6 +13,8 @@ #define VERIFY_READ 0 #define VERIFY_WRITE 1 +extern void check_object_size(const void *ptr, unsigned long n, bool to); + /* * The fs value determines whether argument validity checking should be * performed or not. If get_fs() == USER_DS, checking is performed, with @@ -327,52 +329,6 @@ do { \ extern unsigned long __copy_tofrom_user(void __user *to, const void __user *from, unsigned long size); -#ifndef __powerpc64__ - -static inline unsigned long copy_from_user(void *to, - const void __user *from, unsigned long n) -{ - unsigned long over; - - if (access_ok(VERIFY_READ, from, n)) - return __copy_tofrom_user((__force void __user *)to, from, n); - if ((unsigned long)from < TASK_SIZE) { - over = (unsigned long)from + n - TASK_SIZE; - return __copy_tofrom_user((__force void __user *)to, from, - n - over) + over; - } - return n; -} - -static inline unsigned long copy_to_user(void __user *to, - const void *from, unsigned long n) -{ - unsigned long over; - - if (access_ok(VERIFY_WRITE, to, n)) - return __copy_tofrom_user(to, (__force void __user *)from, n); - if ((unsigned long)to < TASK_SIZE) { - over = (unsigned long)to + n - TASK_SIZE; - return __copy_tofrom_user(to, (__force void __user *)from, - n - over) + over; - } - return n; -} - -#else /* __powerpc64__ */ - -#define __copy_in_user(to, from, size) \ - __copy_tofrom_user((to), (from), (size)) - -extern unsigned long copy_from_user(void *to, const void __user *from, - unsigned long n); -extern unsigned long copy_to_user(void __user *to, const void *from, - unsigned long n); -extern unsigned long copy_in_user(void __user *to, const void __user *from, - unsigned long n); - -#endif /* __powerpc64__ */ - static inline unsigned long __copy_from_user_inatomic(void *to, const void __user *from, unsigned long n) { @@ -396,6 +352,10 @@ static inline unsigned long __copy_from_user_inatomic(void *to, if (ret == 0) return 0; } + + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); + return __copy_tofrom_user((__force void __user *)to, from, n); } @@ -422,6 +382,10 @@ static inline unsigned long __copy_to_user_inatomic(void __user *to, if (ret == 0) return 0; } + + if (!__builtin_constant_p(n)) + check_object_size(from, n, true); + return __copy_tofrom_user(to, (__force const void __user *)from, n); } @@ -439,6 +403,92 @@ static inline unsigned long __copy_to_user(void __user *to, return __copy_to_user_inatomic(to, from, size); } +#ifndef __powerpc64__ + +static inline unsigned long __must_check copy_from_user(void *to, + const void __user *from, unsigned long n) +{ + unsigned long over; + + if ((long)n < 0) + return n; + + if (access_ok(VERIFY_READ, from, n)) { + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); + return __copy_tofrom_user((__force void __user *)to, from, n); + } + if ((unsigned long)from < TASK_SIZE) { + over = (unsigned long)from + n - TASK_SIZE; + if (!__builtin_constant_p(n - over)) + check_object_size(to, n - over, false); + return __copy_tofrom_user((__force void __user *)to, from, + n - over) + over; + } + return n; +} + +static inline unsigned long __must_check copy_to_user(void __user *to, + const void *from, unsigned long n) +{ + unsigned long over; + + if ((long)n < 0) + return n; + + if (access_ok(VERIFY_WRITE, to, n)) { + if (!__builtin_constant_p(n)) + check_object_size(from, n, true); + return __copy_tofrom_user(to, (__force void __user *)from, n); + } + if ((unsigned long)to < TASK_SIZE) { + over = (unsigned long)to + n - TASK_SIZE; + if (!__builtin_constant_p(n)) + check_object_size(from, n - over, true); + return __copy_tofrom_user(to, (__force void __user *)from, + n - over) + over; + } + return n; +} + +#else /* __powerpc64__ */ + +#define __copy_in_user(to, from, size) \ + __copy_tofrom_user((to), (from), (size)) + +static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, unsigned long n) +{ + if ((long)n < 0 || n > INT_MAX) + return n; + + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); + + if (likely(access_ok(VERIFY_READ, from, n))) + n = __copy_from_user(to, from, n); + else + memset(to, 0, n); + return n; +} + +static inline unsigned long __must_check copy_to_user(void __user *to, const void *from, unsigned long n) +{ + if ((long)n < 0 || n > INT_MAX) + return n; + + if (likely(access_ok(VERIFY_WRITE, to, n))) { + if (!__builtin_constant_p(n)) + check_object_size(from, n, true); + n = __copy_to_user(to, from, n); + } + return n; +} + +extern unsigned long copy_in_user(void __user *to, const void __user *from, + unsigned long n); + +#endif /* __powerpc64__ */ + extern unsigned long __clear_user(void __user *addr, unsigned long size); static inline unsigned long clear_user(void __user *addr, unsigned long size) diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S index 429983c..7af363b 100644 --- a/arch/powerpc/kernel/exceptions-64e.S +++ b/arch/powerpc/kernel/exceptions-64e.S @@ -587,6 +587,7 @@ storage_fault_common: std r14,_DAR(r1) std r15,_DSISR(r1) addi r3,r1,STACK_FRAME_OVERHEAD + bl .save_nvgprs mr r4,r14 mr r5,r15 ld r14,PACA_EXGEN+EX_R14(r13) @@ -596,8 +597,7 @@ storage_fault_common: cmpdi r3,0 bne- 1f b .ret_from_except_lite -1: bl .save_nvgprs - mr r5,r3 +1: mr r5,r3 addi r3,r1,STACK_FRAME_OVERHEAD ld r4,_DAR(r1) bl .bad_page_fault diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S index 15c5a4f..22a4000 100644 --- a/arch/powerpc/kernel/exceptions-64s.S +++ b/arch/powerpc/kernel/exceptions-64s.S @@ -1004,10 +1004,10 @@ handle_page_fault: 11: ld r4,_DAR(r1) ld r5,_DSISR(r1) addi r3,r1,STACK_FRAME_OVERHEAD + bl .save_nvgprs bl .do_page_fault cmpdi r3,0 beq+ 13f - bl .save_nvgprs mr r5,r3 addi r3,r1,STACK_FRAME_OVERHEAD lwz r4,_DAR(r1) diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c index 01e2877..a1ba360 100644 --- a/arch/powerpc/kernel/irq.c +++ b/arch/powerpc/kernel/irq.c @@ -560,9 +560,6 @@ struct irq_host *irq_alloc_host(struct device_node *of_node, host->ops = ops; host->of_node = of_node_get(of_node); - if (host->ops->match == NULL) - host->ops->match = default_irq_host_match; - raw_spin_lock_irqsave(&irq_big_lock, flags); /* If it's a legacy controller, check for duplicates and @@ -635,7 +632,12 @@ struct irq_host *irq_find_host(struct device_node *node) */ raw_spin_lock_irqsave(&irq_big_lock, flags); list_for_each_entry(h, &irq_hosts, link) - if (h->ops->match(h, node)) { + if (h->ops->match) { + if (h->ops->match(h, node)) { + found = h; + break; + } + } else if (default_irq_host_match(h, node)) { found = h; break; } diff --git a/arch/powerpc/kernel/module_32.c b/arch/powerpc/kernel/module_32.c index 0b6d796..d760ddb 100644 --- a/arch/powerpc/kernel/module_32.c +++ b/arch/powerpc/kernel/module_32.c @@ -162,7 +162,7 @@ int module_frob_arch_sections(Elf32_Ehdr *hdr, me->arch.core_plt_section = i; } if (!me->arch.core_plt_section || !me->arch.init_plt_section) { - printk("Module doesn't contain .plt or .init.plt sections.\n"); + printk("Module %s doesn't contain .plt or .init.plt sections.\n", me->name); return -ENOEXEC; } @@ -192,11 +192,16 @@ static uint32_t do_plt_call(void *location, DEBUGP("Doing plt for call to 0x%x at 0x%x\n", val, (unsigned int)location); /* Init, or core PLT? */ - if (location >= mod->module_core - && location < mod->module_core + mod->core_size) + if ((location >= mod->module_core_rx && location < mod->module_core_rx + mod->core_size_rx) || + (location >= mod->module_core_rw && location < mod->module_core_rw + mod->core_size_rw)) entry = (void *)sechdrs[mod->arch.core_plt_section].sh_addr; - else + else if ((location >= mod->module_init_rx && location < mod->module_init_rx + mod->init_size_rx) || + (location >= mod->module_init_rw && location < mod->module_init_rw + mod->init_size_rw)) entry = (void *)sechdrs[mod->arch.init_plt_section].sh_addr; + else { + printk(KERN_ERR "%s: invalid R_PPC_REL24 entry found\n", mod->name); + return ~0UL; + } /* Find this entry, or if that fails, the next avail. entry */ while (entry->jump[0]) { diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index d817ab0..b23b18e 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -676,8 +676,8 @@ void show_regs(struct pt_regs * regs) * Lookup NIP late so we have the best change of getting the * above info out without failing */ - printk("NIP ["REG"] %pS\n", regs->nip, (void *)regs->nip); - printk("LR ["REG"] %pS\n", regs->link, (void *)regs->link); + printk("NIP ["REG"] %pA\n", regs->nip, (void *)regs->nip); + printk("LR ["REG"] %pA\n", regs->link, (void *)regs->link); #endif show_stack(current, (unsigned long *) regs->gpr[1]); if (!user_mode(regs)) @@ -1181,10 +1181,10 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) newsp = stack[0]; ip = stack[STACK_FRAME_LR_SAVE]; if (!firstframe || ip != lr) { - printk("["REG"] ["REG"] %pS", sp, ip, (void *)ip); + printk("["REG"] ["REG"] %pA", sp, ip, (void *)ip); #ifdef CONFIG_FUNCTION_GRAPH_TRACER if ((ip == rth || ip == mrth) && curr_frame >= 0) { - printk(" (%pS)", + printk(" (%pA)", (void *)current->ret_stack[curr_frame].ret); curr_frame--; } @@ -1204,7 +1204,7 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) struct pt_regs *regs = (struct pt_regs *) (sp + STACK_FRAME_OVERHEAD); lr = regs->link; - printk("--- Exception: %lx at %pS\n LR = %pS\n", + printk("--- Exception: %lx at %pA\n LR = %pA\n", regs->trap, (void *)regs->nip, (void *)lr); firstframe = 1; } @@ -1279,58 +1279,3 @@ void thread_info_cache_init(void) } #endif /* THREAD_SHIFT < PAGE_SHIFT */ - -unsigned long arch_align_stack(unsigned long sp) -{ - if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) - sp -= get_random_int() & ~PAGE_MASK; - return sp & ~0xf; -} - -static inline unsigned long brk_rnd(void) -{ - unsigned long rnd = 0; - - /* 8MB for 32bit, 1GB for 64bit */ - if (is_32bit_task()) - rnd = (long)(get_random_int() % (1<<(23-PAGE_SHIFT))); - else - rnd = (long)(get_random_int() % (1<<(30-PAGE_SHIFT))); - - return rnd << PAGE_SHIFT; -} - -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long base = mm->brk; - unsigned long ret; - -#ifdef CONFIG_PPC_STD_MMU_64 - /* - * If we are using 1TB segments and we are allowed to randomise - * the heap, we can put it above 1TB so it is backed by a 1TB - * segment. Otherwise the heap will be in the bottom 1TB - * which always uses 256MB segments and this may result in a - * performance penalty. - */ - if (!is_32bit_task() && (mmu_highuser_ssize == MMU_SEGSIZE_1T)) - base = max_t(unsigned long, mm->brk, 1UL << SID_SHIFT_1T); -#endif - - ret = PAGE_ALIGN(base + brk_rnd()); - - if (ret < mm->brk) - return mm->brk; - - return ret; -} - -unsigned long randomize_et_dyn(unsigned long base) -{ - unsigned long ret = PAGE_ALIGN(base + brk_rnd()); - - if (ret < base) - return base; - - return ret; -} diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c index 836a5a1..27289a3 100644 --- a/arch/powerpc/kernel/signal_32.c +++ b/arch/powerpc/kernel/signal_32.c @@ -859,7 +859,7 @@ int handle_rt_signal32(unsigned long sig, struct k_sigaction *ka, /* Save user registers on the stack */ frame = &rt_sf->uc.uc_mcontext; addr = frame; - if (vdso32_rt_sigtramp && current->mm->context.vdso_base) { + if (vdso32_rt_sigtramp && current->mm->context.vdso_base != ~0UL) { if (save_user_regs(regs, frame, 0, 1)) goto badframe; regs->link = current->mm->context.vdso_base + vdso32_rt_sigtramp; diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c index a50b5ec..547078a 100644 --- a/arch/powerpc/kernel/signal_64.c +++ b/arch/powerpc/kernel/signal_64.c @@ -429,7 +429,7 @@ int handle_rt_signal64(int signr, struct k_sigaction *ka, siginfo_t *info, current->thread.fpscr.val = 0; /* Set up to return from userspace. */ - if (vdso64_rt_sigtramp && current->mm->context.vdso_base) { + if (vdso64_rt_sigtramp && current->mm->context.vdso_base != ~0UL) { regs->link = current->mm->context.vdso_base + vdso64_rt_sigtramp; } else { err |= setup_trampoline(__NR_rt_sigreturn, &frame->tramp[0]); diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c index c091527..5592625 100644 --- a/arch/powerpc/kernel/traps.c +++ b/arch/powerpc/kernel/traps.c @@ -131,6 +131,8 @@ static unsigned __kprobes long oops_begin(struct pt_regs *regs) return flags; } +extern void gr_handle_kernel_exploit(void); + static void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr) { @@ -178,6 +180,9 @@ static void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, panic("Fatal exception in interrupt"); if (panic_on_oops) panic("Fatal exception"); + + gr_handle_kernel_exploit(); + do_exit(signr); } diff --git a/arch/powerpc/kernel/vdso.c b/arch/powerpc/kernel/vdso.c index 7d14bb6..1305601 100644 --- a/arch/powerpc/kernel/vdso.c +++ b/arch/powerpc/kernel/vdso.c @@ -35,6 +35,7 @@ #include #include #include +#include #include "setup.h" @@ -219,7 +220,7 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) vdso_base = VDSO32_MBASE; #endif - current->mm->context.vdso_base = 0; + current->mm->context.vdso_base = ~0UL; /* vDSO has a problem and was disabled, just don't "enable" it for the * process @@ -239,7 +240,7 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) vdso_base = get_unmapped_area(NULL, vdso_base, (vdso_pages << PAGE_SHIFT) + ((VDSO_ALIGNMENT - 1) & PAGE_MASK), - 0, 0); + 0, MAP_PRIVATE | MAP_EXECUTABLE); if (IS_ERR_VALUE(vdso_base)) { rc = vdso_base; goto fail_mmapsem; diff --git a/arch/powerpc/lib/usercopy_64.c b/arch/powerpc/lib/usercopy_64.c index 5eea6f3..5d10396 100644 --- a/arch/powerpc/lib/usercopy_64.c +++ b/arch/powerpc/lib/usercopy_64.c @@ -9,22 +9,6 @@ #include #include -unsigned long copy_from_user(void *to, const void __user *from, unsigned long n) -{ - if (likely(access_ok(VERIFY_READ, from, n))) - n = __copy_from_user(to, from, n); - else - memset(to, 0, n); - return n; -} - -unsigned long copy_to_user(void __user *to, const void *from, unsigned long n) -{ - if (likely(access_ok(VERIFY_WRITE, to, n))) - n = __copy_to_user(to, from, n); - return n; -} - unsigned long copy_in_user(void __user *to, const void __user *from, unsigned long n) { @@ -35,7 +19,5 @@ unsigned long copy_in_user(void __user *to, const void __user *from, return n; } -EXPORT_SYMBOL(copy_from_user); -EXPORT_SYMBOL(copy_to_user); EXPORT_SYMBOL(copy_in_user); diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c index 2f0d1b0..36fb5cc 100644 --- a/arch/powerpc/mm/fault.c +++ b/arch/powerpc/mm/fault.c @@ -32,6 +32,10 @@ #include #include #include +#include +#include +#include +#include #include #include @@ -43,6 +47,7 @@ #include #include #include +#include #include "icswx.h" @@ -68,6 +73,33 @@ static inline int notify_page_fault(struct pt_regs *regs) } #endif +#ifdef CONFIG_PAX_PAGEEXEC +/* + * PaX: decide what to do with offenders (regs->nip = fault address) + * + * returns 1 when task should be killed + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + return 1; +} + +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 5; i++) { + unsigned int c; + if (get_user(c, (unsigned int __user *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + /* * Check whether the instruction at regs->nip is a store using * an update addressing form which will update r1. @@ -138,7 +170,7 @@ int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address, * indicate errors in DSISR but can validly be set in SRR1. */ if (trap == 0x400) - error_code &= 0x48200000; + error_code &= 0x58200000; else is_write = error_code & DSISR_ISSTORE; #else @@ -276,7 +308,7 @@ good_area: * "undefined". Of those that can be set, this is the only * one which seems bad. */ - if (error_code & 0x10000000) + if (error_code & DSISR_GUARDED) /* Guarded storage error. */ goto bad_area; #endif /* CONFIG_8xx */ @@ -291,7 +323,7 @@ good_area: * processors use the same I/D cache coherency mechanism * as embedded. */ - if (error_code & DSISR_PROTFAULT) + if (error_code & (DSISR_PROTFAULT | DSISR_GUARDED)) goto bad_area; #endif /* CONFIG_PPC_STD_MMU */ @@ -360,6 +392,23 @@ bad_area: bad_area_nosemaphore: /* User mode accesses cause a SIGSEGV */ if (user_mode(regs)) { + +#ifdef CONFIG_PAX_PAGEEXEC + if (mm->pax_flags & MF_PAX_PAGEEXEC) { +#ifdef CONFIG_PPC_STD_MMU + if (is_exec && (error_code & (DSISR_PROTFAULT | DSISR_GUARDED))) { +#else + if (is_exec && regs->nip == address) { +#endif + switch (pax_handle_fetch_fault(regs)) { + } + + pax_report_fault(regs, (void *)regs->nip, (void *)regs->gpr[PT_R1]); + do_group_exit(SIGKILL); + } + } +#endif + _exception(SIGSEGV, regs, code, address); return 0; } diff --git a/arch/powerpc/mm/mmap_64.c b/arch/powerpc/mm/mmap_64.c index 67a42ed..1c7210c 100644 --- a/arch/powerpc/mm/mmap_64.c +++ b/arch/powerpc/mm/mmap_64.c @@ -91,10 +91,22 @@ void arch_pick_mmap_layout(struct mm_struct *mm) */ if (mmap_is_legacy()) { mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { mm->mmap_base = mmap_base(); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff --git a/arch/powerpc/mm/slice.c b/arch/powerpc/mm/slice.c index 73709f7..6b90313 100644 --- a/arch/powerpc/mm/slice.c +++ b/arch/powerpc/mm/slice.c @@ -98,7 +98,7 @@ static int slice_area_is_free(struct mm_struct *mm, unsigned long addr, if ((mm->task_size - len) < addr) return 0; vma = find_vma(mm, addr); - return (!vma || (addr + len) <= vma->vm_start); + return check_heap_stack_gap(vma, addr, len); } static int slice_low_has_vma(struct mm_struct *mm, unsigned long slice) @@ -256,7 +256,7 @@ full_search: addr = _ALIGN_UP(addr + 1, 1ul << SLICE_HIGH_SHIFT); continue; } - if (!vma || addr + len <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr, len)) { /* * Remember the place where we stopped the search: */ @@ -313,10 +313,14 @@ static unsigned long slice_find_area_topdown(struct mm_struct *mm, } } - addr = mm->mmap_base; - while (addr > len) { + if (mm->mmap_base < len) + addr = -ENOMEM; + else + addr = mm->mmap_base - len; + + while (!IS_ERR_VALUE(addr)) { /* Go down by chunk size */ - addr = _ALIGN_DOWN(addr - len, 1ul << pshift); + addr = _ALIGN_DOWN(addr, 1ul << pshift); /* Check for hit with different page size */ mask = slice_range_to_mask(addr, len); @@ -336,7 +340,7 @@ static unsigned long slice_find_area_topdown(struct mm_struct *mm, * return with success: */ vma = find_vma(mm, addr); - if (!vma || (addr + len) <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr, len)) { /* remember the address as a hint for next time */ if (use_cache) mm->free_area_cache = addr; @@ -348,7 +352,7 @@ static unsigned long slice_find_area_topdown(struct mm_struct *mm, mm->cached_hole_size = vma->vm_start - addr; /* try just below the current vma->vm_start */ - addr = vma->vm_start; + addr = skip_heap_stack_gap(vma, len); } /* @@ -426,6 +430,11 @@ unsigned long slice_get_unmapped_area(unsigned long addr, unsigned long len, if (fixed && addr > (mm->task_size - len)) return -EINVAL; +#ifdef CONFIG_PAX_RANDMMAP + if (!fixed && (mm->pax_flags & MF_PAX_RANDMMAP)) + addr = 0; +#endif + /* If hint, make sure it matches our alignment restrictions */ if (!fixed && addr) { addr = _ALIGN_UP(addr, 1ul << pshift); diff --git a/arch/s390/include/asm/atomic.h b/arch/s390/include/asm/atomic.h index 8517d2a..d2738d4 100644 --- a/arch/s390/include/asm/atomic.h +++ b/arch/s390/include/asm/atomic.h @@ -326,6 +326,16 @@ static inline long long atomic64_dec_if_positive(atomic64_t *v) #define atomic64_dec_and_test(_v) (atomic64_sub_return(1, _v) == 0) #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + #define smp_mb__before_atomic_dec() smp_mb() #define smp_mb__after_atomic_dec() smp_mb() #define smp_mb__before_atomic_inc() smp_mb() diff --git a/arch/s390/include/asm/cache.h b/arch/s390/include/asm/cache.h index 2a30d5a..5e5586f 100644 --- a/arch/s390/include/asm/cache.h +++ b/arch/s390/include/asm/cache.h @@ -11,8 +11,10 @@ #ifndef __ARCH_S390_CACHE_H #define __ARCH_S390_CACHE_H -#define L1_CACHE_BYTES 256 +#include + #define L1_CACHE_SHIFT 8 +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define NET_SKB_PAD 32 #define __read_mostly __attribute__((__section__(".data..read_mostly"))) diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h index 547f1a6..0b22b53 100644 --- a/arch/s390/include/asm/elf.h +++ b/arch/s390/include/asm/elf.h @@ -162,8 +162,14 @@ extern unsigned int vdso_enabled; the loader. We need to make sure that it is out of the way of the program that it will "exec", and that there is sufficient room for the brk. */ -extern unsigned long randomize_et_dyn(unsigned long base); -#define ELF_ET_DYN_BASE (randomize_et_dyn(STACK_TOP / 3 * 2)) +#define ELF_ET_DYN_BASE (STACK_TOP / 3 * 2) + +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (test_thread_flag(TIF_31BIT) ? 0x10000UL : 0x80000000UL) + +#define PAX_DELTA_MMAP_LEN (test_thread_flag(TIF_31BIT) ? 15 : 26) +#define PAX_DELTA_STACK_LEN (test_thread_flag(TIF_31BIT) ? 15 : 26) +#endif /* This yields a mask that user programs can use to figure out what instruction set this CPU supports. */ @@ -211,7 +217,4 @@ struct linux_binprm; #define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1 int arch_setup_additional_pages(struct linux_binprm *, int); -extern unsigned long arch_randomize_brk(struct mm_struct *mm); -#define arch_randomize_brk arch_randomize_brk - #endif diff --git a/arch/s390/include/asm/system.h b/arch/s390/include/asm/system.h index d73cc6b..1a296ad 100644 --- a/arch/s390/include/asm/system.h +++ b/arch/s390/include/asm/system.h @@ -260,7 +260,7 @@ extern void (*_machine_restart)(char *command); extern void (*_machine_halt)(void); extern void (*_machine_power_off)(void); -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) static inline int tprot(unsigned long addr) { diff --git a/arch/s390/include/asm/uaccess.h b/arch/s390/include/asm/uaccess.h index 2b23885..e136e31 100644 --- a/arch/s390/include/asm/uaccess.h +++ b/arch/s390/include/asm/uaccess.h @@ -235,6 +235,10 @@ static inline unsigned long __must_check copy_to_user(void __user *to, const void *from, unsigned long n) { might_fault(); + + if ((long)n < 0) + return n; + if (access_ok(VERIFY_WRITE, to, n)) n = __copy_to_user(to, from, n); return n; @@ -260,6 +264,9 @@ copy_to_user(void __user *to, const void *from, unsigned long n) static inline unsigned long __must_check __copy_from_user(void *to, const void __user *from, unsigned long n) { + if ((long)n < 0) + return n; + if (__builtin_constant_p(n) && (n <= 256)) return uaccess.copy_from_user_small(n, from, to); else @@ -294,6 +301,10 @@ copy_from_user(void *to, const void __user *from, unsigned long n) unsigned int sz = __compiletime_object_size(to); might_fault(); + + if ((long)n < 0) + return n; + if (unlikely(sz != -1 && sz < n)) { copy_from_user_overflow(); return n; diff --git a/arch/s390/kernel/module.c b/arch/s390/kernel/module.c index dfcb343..eda788a 100644 --- a/arch/s390/kernel/module.c +++ b/arch/s390/kernel/module.c @@ -161,11 +161,11 @@ module_frob_arch_sections(Elf_Ehdr *hdr, Elf_Shdr *sechdrs, /* Increase core size by size of got & plt and set start offsets for got and plt. */ - me->core_size = ALIGN(me->core_size, 4); - me->arch.got_offset = me->core_size; - me->core_size += me->arch.got_size; - me->arch.plt_offset = me->core_size; - me->core_size += me->arch.plt_size; + me->core_size_rw = ALIGN(me->core_size_rw, 4); + me->arch.got_offset = me->core_size_rw; + me->core_size_rw += me->arch.got_size; + me->arch.plt_offset = me->core_size_rx; + me->core_size_rx += me->arch.plt_size; return 0; } @@ -242,7 +242,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base, Elf_Sym *symtab, if (info->got_initialized == 0) { Elf_Addr *gotent; - gotent = me->module_core + me->arch.got_offset + + gotent = me->module_core_rw + me->arch.got_offset + info->got_offset; *gotent = val; info->got_initialized = 1; @@ -266,7 +266,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base, Elf_Sym *symtab, else if (r_type == R_390_GOTENT || r_type == R_390_GOTPLTENT) *(unsigned int *) loc = - (val + (Elf_Addr) me->module_core - loc) >> 1; + (val + (Elf_Addr) me->module_core_rw - loc) >> 1; else if (r_type == R_390_GOT64 || r_type == R_390_GOTPLT64) *(unsigned long *) loc = val; @@ -280,7 +280,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base, Elf_Sym *symtab, case R_390_PLTOFF64: /* 16 bit offset from GOT to PLT. */ if (info->plt_initialized == 0) { unsigned int *ip; - ip = me->module_core + me->arch.plt_offset + + ip = me->module_core_rx + me->arch.plt_offset + info->plt_offset; #ifndef CONFIG_64BIT ip[0] = 0x0d105810; /* basr 1,0; l 1,6(1); br 1 */ @@ -305,7 +305,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base, Elf_Sym *symtab, val - loc + 0xffffUL < 0x1ffffeUL) || (r_type == R_390_PLT32DBL && val - loc + 0xffffffffULL < 0x1fffffffeULL))) - val = (Elf_Addr) me->module_core + + val = (Elf_Addr) me->module_core_rx + me->arch.plt_offset + info->plt_offset; val += rela->r_addend - loc; @@ -327,7 +327,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base, Elf_Sym *symtab, case R_390_GOTOFF32: /* 32 bit offset to GOT. */ case R_390_GOTOFF64: /* 64 bit offset to GOT. */ val = val + rela->r_addend - - ((Elf_Addr) me->module_core + me->arch.got_offset); + ((Elf_Addr) me->module_core_rw + me->arch.got_offset); if (r_type == R_390_GOTOFF16) *(unsigned short *) loc = val; else if (r_type == R_390_GOTOFF32) @@ -337,7 +337,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base, Elf_Sym *symtab, break; case R_390_GOTPC: /* 32 bit PC relative offset to GOT. */ case R_390_GOTPCDBL: /* 32 bit PC rel. off. to GOT shifted by 1. */ - val = (Elf_Addr) me->module_core + me->arch.got_offset + + val = (Elf_Addr) me->module_core_rw + me->arch.got_offset + rela->r_addend - loc; if (r_type == R_390_GOTPC) *(unsigned int *) loc = val; diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index e795933..b32563c 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -323,39 +323,3 @@ unsigned long get_wchan(struct task_struct *p) } return 0; } - -unsigned long arch_align_stack(unsigned long sp) -{ - if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) - sp -= get_random_int() & ~PAGE_MASK; - return sp & ~0xf; -} - -static inline unsigned long brk_rnd(void) -{ - /* 8MB for 32bit, 1GB for 64bit */ - if (is_32bit_task()) - return (get_random_int() & 0x7ffUL) << PAGE_SHIFT; - else - return (get_random_int() & 0x3ffffUL) << PAGE_SHIFT; -} - -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long ret = PAGE_ALIGN(mm->brk + brk_rnd()); - - if (ret < mm->brk) - return mm->brk; - return ret; -} - -unsigned long randomize_et_dyn(unsigned long base) -{ - unsigned long ret = PAGE_ALIGN(base + brk_rnd()); - - if (!(current->flags & PF_RANDOMIZE)) - return base; - if (ret < base) - return base; - return ret; -} diff --git a/arch/s390/mm/mmap.c b/arch/s390/mm/mmap.c index a0155c0..34cc491 100644 --- a/arch/s390/mm/mmap.c +++ b/arch/s390/mm/mmap.c @@ -92,10 +92,22 @@ void arch_pick_mmap_layout(struct mm_struct *mm) */ if (mmap_is_legacy()) { mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { mm->mmap_base = mmap_base(); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } @@ -167,10 +179,22 @@ void arch_pick_mmap_layout(struct mm_struct *mm) */ if (mmap_is_legacy()) { mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = s390_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { mm->mmap_base = mmap_base(); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = s390_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff --git a/arch/score/include/asm/cache.h b/arch/score/include/asm/cache.h index ae3d59f..f65f075 100644 --- a/arch/score/include/asm/cache.h +++ b/arch/score/include/asm/cache.h @@ -1,7 +1,9 @@ #ifndef _ASM_SCORE_CACHE_H #define _ASM_SCORE_CACHE_H +#include + #define L1_CACHE_SHIFT 4 -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #endif /* _ASM_SCORE_CACHE_H */ diff --git a/arch/score/include/asm/system.h b/arch/score/include/asm/system.h index 589d5c7..669e274 100644 --- a/arch/score/include/asm/system.h +++ b/arch/score/include/asm/system.h @@ -17,7 +17,7 @@ do { \ #define finish_arch_switch(prev) do {} while (0) typedef void (*vi_handler_t)(void); -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) (x) #define mb() barrier() #define rmb() barrier() diff --git a/arch/score/kernel/process.c b/arch/score/kernel/process.c index 25d0803..d6c8e36 100644 --- a/arch/score/kernel/process.c +++ b/arch/score/kernel/process.c @@ -161,8 +161,3 @@ unsigned long get_wchan(struct task_struct *task) return task_pt_regs(task)->cp0_epc; } - -unsigned long arch_align_stack(unsigned long sp) -{ - return sp; -} diff --git a/arch/sh/include/asm/cache.h b/arch/sh/include/asm/cache.h index ef9e555..331bd29 100644 --- a/arch/sh/include/asm/cache.h +++ b/arch/sh/include/asm/cache.h @@ -9,10 +9,11 @@ #define __ASM_SH_CACHE_H #ifdef __KERNEL__ +#include #include #include -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define __read_mostly __attribute__((__section__(".data..read_mostly"))) diff --git a/arch/sh/mm/mmap.c b/arch/sh/mm/mmap.c index afeb710..d1d1289 100644 --- a/arch/sh/mm/mmap.c +++ b/arch/sh/mm/mmap.c @@ -74,8 +74,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, addr = PAGE_ALIGN(addr); vma = find_vma(mm, addr); - if (TASK_SIZE - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (TASK_SIZE - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } @@ -106,7 +105,7 @@ full_search: } return -ENOMEM; } - if (likely(!vma || addr + len <= vma->vm_start)) { + if (likely(check_heap_stack_gap(vma, addr, len))) { /* * Remember the place where we stopped the search: */ @@ -157,8 +156,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, addr = PAGE_ALIGN(addr); vma = find_vma(mm, addr); - if (TASK_SIZE - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (TASK_SIZE - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } @@ -179,7 +177,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, /* make sure it can fit in the remaining address space */ if (likely(addr > len)) { vma = find_vma(mm, addr-len); - if (!vma || addr <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr - len, len)) { /* remember the address as a hint for next time */ return (mm->free_area_cache = addr-len); } @@ -188,18 +186,18 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (unlikely(mm->mmap_base < len)) goto bottomup; - addr = mm->mmap_base-len; - if (do_colour_align) - addr = COLOUR_ALIGN_DOWN(addr, pgoff); + addr = mm->mmap_base - len; do { + if (do_colour_align) + addr = COLOUR_ALIGN_DOWN(addr, pgoff); /* * Lookup failure means no vma is above this address, * else if new region fits below vma->vm_start, * return with success: */ vma = find_vma(mm, addr); - if (likely(!vma || addr+len <= vma->vm_start)) { + if (likely(check_heap_stack_gap(vma, addr, len))) { /* remember the address as a hint for next time */ return (mm->free_area_cache = addr); } @@ -209,10 +207,8 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, mm->cached_hole_size = vma->vm_start - addr; /* try just below the current vma->vm_start */ - addr = vma->vm_start-len; - if (do_colour_align) - addr = COLOUR_ALIGN_DOWN(addr, pgoff); - } while (likely(len < vma->vm_start)); + addr = skip_heap_stack_gap(vma, len); + } while (!IS_ERR_VALUE(addr)); bottomup: /* diff --git a/arch/sparc/Makefile b/arch/sparc/Makefile index eddcfb3..b117d90 100644 --- a/arch/sparc/Makefile +++ b/arch/sparc/Makefile @@ -75,7 +75,7 @@ drivers-$(CONFIG_OPROFILE) += arch/sparc/oprofile/ # Export what is needed by arch/sparc/boot/Makefile export VMLINUX_INIT VMLINUX_MAIN VMLINUX_INIT := $(head-y) $(init-y) -VMLINUX_MAIN := $(core-y) kernel/ mm/ fs/ ipc/ security/ crypto/ block/ +VMLINUX_MAIN := $(core-y) kernel/ mm/ fs/ ipc/ security/ crypto/ block/ grsecurity/ VMLINUX_MAIN += $(patsubst %/, %/lib.a, $(libs-y)) $(libs-y) VMLINUX_MAIN += $(drivers-y) $(net-y) diff --git a/arch/sparc/include/asm/atomic_64.h b/arch/sparc/include/asm/atomic_64.h index 9f421df..b81fc12 100644 --- a/arch/sparc/include/asm/atomic_64.h +++ b/arch/sparc/include/asm/atomic_64.h @@ -14,18 +14,40 @@ #define ATOMIC64_INIT(i) { (i) } #define atomic_read(v) (*(volatile int *)&(v)->counter) +static inline int atomic_read_unchecked(const atomic_unchecked_t *v) +{ + return v->counter; +} #define atomic64_read(v) (*(volatile long *)&(v)->counter) +static inline long atomic64_read_unchecked(const atomic64_unchecked_t *v) +{ + return v->counter; +} #define atomic_set(v, i) (((v)->counter) = i) +static inline void atomic_set_unchecked(atomic_unchecked_t *v, int i) +{ + v->counter = i; +} #define atomic64_set(v, i) (((v)->counter) = i) +static inline void atomic64_set_unchecked(atomic64_unchecked_t *v, long i) +{ + v->counter = i; +} extern void atomic_add(int, atomic_t *); +extern void atomic_add_unchecked(int, atomic_unchecked_t *); extern void atomic64_add(long, atomic64_t *); +extern void atomic64_add_unchecked(long, atomic64_unchecked_t *); extern void atomic_sub(int, atomic_t *); +extern void atomic_sub_unchecked(int, atomic_unchecked_t *); extern void atomic64_sub(long, atomic64_t *); +extern void atomic64_sub_unchecked(long, atomic64_unchecked_t *); extern int atomic_add_ret(int, atomic_t *); +extern int atomic_add_ret_unchecked(int, atomic_unchecked_t *); extern long atomic64_add_ret(long, atomic64_t *); +extern long atomic64_add_ret_unchecked(long, atomic64_unchecked_t *); extern int atomic_sub_ret(int, atomic_t *); extern long atomic64_sub_ret(long, atomic64_t *); @@ -33,13 +55,29 @@ extern long atomic64_sub_ret(long, atomic64_t *); #define atomic64_dec_return(v) atomic64_sub_ret(1, v) #define atomic_inc_return(v) atomic_add_ret(1, v) +static inline int atomic_inc_return_unchecked(atomic_unchecked_t *v) +{ + return atomic_add_ret_unchecked(1, v); +} #define atomic64_inc_return(v) atomic64_add_ret(1, v) +static inline long atomic64_inc_return_unchecked(atomic64_unchecked_t *v) +{ + return atomic64_add_ret_unchecked(1, v); +} #define atomic_sub_return(i, v) atomic_sub_ret(i, v) #define atomic64_sub_return(i, v) atomic64_sub_ret(i, v) #define atomic_add_return(i, v) atomic_add_ret(i, v) +static inline int atomic_add_return_unchecked(int i, atomic_unchecked_t *v) +{ + return atomic_add_ret_unchecked(i, v); +} #define atomic64_add_return(i, v) atomic64_add_ret(i, v) +static inline long atomic64_add_return_unchecked(long i, atomic64_unchecked_t *v) +{ + return atomic64_add_ret_unchecked(i, v); +} /* * atomic_inc_and_test - increment and test @@ -50,6 +88,10 @@ extern long atomic64_sub_ret(long, atomic64_t *); * other cases. */ #define atomic_inc_and_test(v) (atomic_inc_return(v) == 0) +static inline int atomic_inc_and_test_unchecked(atomic_unchecked_t *v) +{ + return atomic_inc_return_unchecked(v) == 0; +} #define atomic64_inc_and_test(v) (atomic64_inc_return(v) == 0) #define atomic_sub_and_test(i, v) (atomic_sub_ret(i, v) == 0) @@ -59,25 +101,60 @@ extern long atomic64_sub_ret(long, atomic64_t *); #define atomic64_dec_and_test(v) (atomic64_sub_ret(1, v) == 0) #define atomic_inc(v) atomic_add(1, v) +static inline void atomic_inc_unchecked(atomic_unchecked_t *v) +{ + atomic_add_unchecked(1, v); +} #define atomic64_inc(v) atomic64_add(1, v) +static inline void atomic64_inc_unchecked(atomic64_unchecked_t *v) +{ + atomic64_add_unchecked(1, v); +} #define atomic_dec(v) atomic_sub(1, v) +static inline void atomic_dec_unchecked(atomic_unchecked_t *v) +{ + atomic_sub_unchecked(1, v); +} #define atomic64_dec(v) atomic64_sub(1, v) +static inline void atomic64_dec_unchecked(atomic64_unchecked_t *v) +{ + atomic64_sub_unchecked(1, v); +} #define atomic_add_negative(i, v) (atomic_add_ret(i, v) < 0) #define atomic64_add_negative(i, v) (atomic64_add_ret(i, v) < 0) #define atomic_cmpxchg(v, o, n) (cmpxchg(&((v)->counter), (o), (n))) +static inline int atomic_cmpxchg_unchecked(atomic_unchecked_t *v, int old, int new) +{ + return cmpxchg(&v->counter, old, new); +} #define atomic_xchg(v, new) (xchg(&((v)->counter), new)) +static inline int atomic_xchg_unchecked(atomic_unchecked_t *v, int new) +{ + return xchg(&v->counter, new); +} static inline int __atomic_add_unless(atomic_t *v, int a, int u) { - int c, old; + int c, old, new; c = atomic_read(v); for (;;) { - if (unlikely(c == (u))) + if (unlikely(c == u)) break; - old = atomic_cmpxchg((v), c, c + (a)); + + asm volatile("addcc %2, %0, %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "tvs %%icc, 6\n" +#endif + + : "=r" (new) + : "0" (c), "ir" (a) + : "cc"); + + old = atomic_cmpxchg(v, c, new); if (likely(old == c)) break; c = old; @@ -89,20 +166,35 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) #define atomic64_cmpxchg(v, o, n) \ ((__typeof__((v)->counter))cmpxchg(&((v)->counter), (o), (n))) #define atomic64_xchg(v, new) (xchg(&((v)->counter), new)) +static inline long atomic64_xchg_unchecked(atomic64_unchecked_t *v, long new) +{ + return xchg(&v->counter, new); +} static inline long atomic64_add_unless(atomic64_t *v, long a, long u) { - long c, old; + long c, old, new; c = atomic64_read(v); for (;;) { - if (unlikely(c == (u))) + if (unlikely(c == u)) break; - old = atomic64_cmpxchg((v), c, c + (a)); + + asm volatile("addcc %2, %0, %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "tvs %%xcc, 6\n" +#endif + + : "=r" (new) + : "0" (c), "ir" (a) + : "cc"); + + old = atomic64_cmpxchg(v, c, new); if (likely(old == c)) break; c = old; } - return c != (u); + return c != u; } #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) diff --git a/arch/sparc/include/asm/cache.h b/arch/sparc/include/asm/cache.h index 69358b5..9d0d492 100644 --- a/arch/sparc/include/asm/cache.h +++ b/arch/sparc/include/asm/cache.h @@ -7,10 +7,12 @@ #ifndef _SPARC_CACHE_H #define _SPARC_CACHE_H +#include + #define ARCH_SLAB_MINALIGN __alignof__(unsigned long long) #define L1_CACHE_SHIFT 5 -#define L1_CACHE_BYTES 32 +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #ifdef CONFIG_SPARC32 #define SMP_CACHE_BYTES_SHIFT 5 diff --git a/arch/sparc/include/asm/elf_32.h b/arch/sparc/include/asm/elf_32.h index 4269ca6..e3da77f 100644 --- a/arch/sparc/include/asm/elf_32.h +++ b/arch/sparc/include/asm/elf_32.h @@ -114,6 +114,13 @@ typedef struct { #define ELF_ET_DYN_BASE (TASK_UNMAPPED_BASE) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE 0x10000UL + +#define PAX_DELTA_MMAP_LEN 16 +#define PAX_DELTA_STACK_LEN 16 +#endif + /* This yields a mask that user programs can use to figure out what instruction set this cpu supports. This can NOT be done in userspace on Sparc. */ diff --git a/arch/sparc/include/asm/elf_64.h b/arch/sparc/include/asm/elf_64.h index 7df8b7f..4946269 100644 --- a/arch/sparc/include/asm/elf_64.h +++ b/arch/sparc/include/asm/elf_64.h @@ -180,6 +180,13 @@ typedef struct { #define ELF_ET_DYN_BASE 0x0000010000000000UL #define COMPAT_ELF_ET_DYN_BASE 0x0000000070000000UL +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (test_thread_flag(TIF_32BIT) ? 0x10000UL : 0x100000UL) + +#define PAX_DELTA_MMAP_LEN (test_thread_flag(TIF_32BIT) ? 14 : 28) +#define PAX_DELTA_STACK_LEN (test_thread_flag(TIF_32BIT) ? 15 : 29) +#endif + extern unsigned long sparc64_elf_hwcap; #define ELF_HWCAP sparc64_elf_hwcap diff --git a/arch/sparc/include/asm/pgalloc_32.h b/arch/sparc/include/asm/pgalloc_32.h index ca2b344..c6084f89 100644 --- a/arch/sparc/include/asm/pgalloc_32.h +++ b/arch/sparc/include/asm/pgalloc_32.h @@ -37,6 +37,7 @@ BTFIXUPDEF_CALL(void, free_pgd_fast, pgd_t *) BTFIXUPDEF_CALL(void, pgd_set, pgd_t *, pmd_t *) #define pgd_set(pgdp,pmdp) BTFIXUP_CALL(pgd_set)(pgdp,pmdp) #define pgd_populate(MM, PGD, PMD) pgd_set(PGD, PMD) +#define pgd_populate_kernel(MM, PGD, PMD) pgd_populate((MM), (PGD), (PMD)) BTFIXUPDEF_CALL(pmd_t *, pmd_alloc_one, struct mm_struct *, unsigned long) #define pmd_alloc_one(mm, address) BTFIXUP_CALL(pmd_alloc_one)(mm, address) diff --git a/arch/sparc/include/asm/pgalloc_64.h b/arch/sparc/include/asm/pgalloc_64.h index 40b2d7a..22a665b 100644 --- a/arch/sparc/include/asm/pgalloc_64.h +++ b/arch/sparc/include/asm/pgalloc_64.h @@ -26,6 +26,7 @@ static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd) } #define pud_populate(MM, PUD, PMD) pud_set(PUD, PMD) +#define pud_populate_kernel(MM, PUD, PMD) pud_populate((MM), (PUD), (PMD)) static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr) { diff --git a/arch/sparc/include/asm/pgtable_32.h b/arch/sparc/include/asm/pgtable_32.h index a790cc6..091ed94 100644 --- a/arch/sparc/include/asm/pgtable_32.h +++ b/arch/sparc/include/asm/pgtable_32.h @@ -45,6 +45,13 @@ BTFIXUPDEF_SIMM13(user_ptrs_per_pgd) BTFIXUPDEF_INT(page_none) BTFIXUPDEF_INT(page_copy) BTFIXUPDEF_INT(page_readonly) + +#ifdef CONFIG_PAX_PAGEEXEC +BTFIXUPDEF_INT(page_shared_noexec) +BTFIXUPDEF_INT(page_copy_noexec) +BTFIXUPDEF_INT(page_readonly_noexec) +#endif + BTFIXUPDEF_INT(page_kernel) #define PMD_SHIFT SUN4C_PMD_SHIFT @@ -66,6 +73,16 @@ extern pgprot_t PAGE_SHARED; #define PAGE_COPY __pgprot(BTFIXUP_INT(page_copy)) #define PAGE_READONLY __pgprot(BTFIXUP_INT(page_readonly)) +#ifdef CONFIG_PAX_PAGEEXEC +extern pgprot_t PAGE_SHARED_NOEXEC; +# define PAGE_COPY_NOEXEC __pgprot(BTFIXUP_INT(page_copy_noexec)) +# define PAGE_READONLY_NOEXEC __pgprot(BTFIXUP_INT(page_readonly_noexec)) +#else +# define PAGE_SHARED_NOEXEC PAGE_SHARED +# define PAGE_COPY_NOEXEC PAGE_COPY +# define PAGE_READONLY_NOEXEC PAGE_READONLY +#endif + extern unsigned long page_kernel; #ifdef MODULE diff --git a/arch/sparc/include/asm/pgtsrmmu.h b/arch/sparc/include/asm/pgtsrmmu.h index f6ae2b2..b03ffc7 100644 --- a/arch/sparc/include/asm/pgtsrmmu.h +++ b/arch/sparc/include/asm/pgtsrmmu.h @@ -115,6 +115,13 @@ SRMMU_EXEC | SRMMU_REF) #define SRMMU_PAGE_RDONLY __pgprot(SRMMU_VALID | SRMMU_CACHE | \ SRMMU_EXEC | SRMMU_REF) + +#ifdef CONFIG_PAX_PAGEEXEC +#define SRMMU_PAGE_SHARED_NOEXEC __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_WRITE | SRMMU_REF) +#define SRMMU_PAGE_COPY_NOEXEC __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_REF) +#define SRMMU_PAGE_RDONLY_NOEXEC __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_REF) +#endif + #define SRMMU_PAGE_KERNEL __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_PRIV | \ SRMMU_DIRTY | SRMMU_REF) diff --git a/arch/sparc/include/asm/spinlock_64.h b/arch/sparc/include/asm/spinlock_64.h index 9689176..63c18ea 100644 --- a/arch/sparc/include/asm/spinlock_64.h +++ b/arch/sparc/include/asm/spinlock_64.h @@ -92,14 +92,19 @@ static inline void arch_spin_lock_flags(arch_spinlock_t *lock, unsigned long fla /* Multi-reader locks, these are much saner than the 32-bit Sparc ones... */ -static void inline arch_read_lock(arch_rwlock_t *lock) +static inline void arch_read_lock(arch_rwlock_t *lock) { unsigned long tmp1, tmp2; __asm__ __volatile__ ( "1: ldsw [%2], %0\n" " brlz,pn %0, 2f\n" -"4: add %0, 1, %1\n" +"4: addcc %0, 1, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT +" tvs %%icc, 6\n" +#endif + " cas [%2], %0, %1\n" " cmp %0, %1\n" " bne,pn %%icc, 1b\n" @@ -112,10 +117,10 @@ static void inline arch_read_lock(arch_rwlock_t *lock) " .previous" : "=&r" (tmp1), "=&r" (tmp2) : "r" (lock) - : "memory"); + : "memory", "cc"); } -static int inline arch_read_trylock(arch_rwlock_t *lock) +static inline int arch_read_trylock(arch_rwlock_t *lock) { int tmp1, tmp2; @@ -123,7 +128,12 @@ static int inline arch_read_trylock(arch_rwlock_t *lock) "1: ldsw [%2], %0\n" " brlz,a,pn %0, 2f\n" " mov 0, %0\n" -" add %0, 1, %1\n" +" addcc %0, 1, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT +" tvs %%icc, 6\n" +#endif + " cas [%2], %0, %1\n" " cmp %0, %1\n" " bne,pn %%icc, 1b\n" @@ -136,13 +146,18 @@ static int inline arch_read_trylock(arch_rwlock_t *lock) return tmp1; } -static void inline arch_read_unlock(arch_rwlock_t *lock) +static inline void arch_read_unlock(arch_rwlock_t *lock) { unsigned long tmp1, tmp2; __asm__ __volatile__( "1: lduw [%2], %0\n" -" sub %0, 1, %1\n" +" subcc %0, 1, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT +" tvs %%icc, 6\n" +#endif + " cas [%2], %0, %1\n" " cmp %0, %1\n" " bne,pn %%xcc, 1b\n" @@ -152,7 +167,7 @@ static void inline arch_read_unlock(arch_rwlock_t *lock) : "memory"); } -static void inline arch_write_lock(arch_rwlock_t *lock) +static inline void arch_write_lock(arch_rwlock_t *lock) { unsigned long mask, tmp1, tmp2; @@ -177,7 +192,7 @@ static void inline arch_write_lock(arch_rwlock_t *lock) : "memory"); } -static void inline arch_write_unlock(arch_rwlock_t *lock) +static inline void arch_write_unlock(arch_rwlock_t *lock) { __asm__ __volatile__( " stw %%g0, [%0]" @@ -186,7 +201,7 @@ static void inline arch_write_unlock(arch_rwlock_t *lock) : "memory"); } -static int inline arch_write_trylock(arch_rwlock_t *lock) +static inline int arch_write_trylock(arch_rwlock_t *lock) { unsigned long mask, tmp1, tmp2, result; diff --git a/arch/sparc/include/asm/thread_info_32.h b/arch/sparc/include/asm/thread_info_32.h index c2a1080..21ed218 100644 --- a/arch/sparc/include/asm/thread_info_32.h +++ b/arch/sparc/include/asm/thread_info_32.h @@ -50,6 +50,8 @@ struct thread_info { unsigned long w_saved; struct restart_block restart_block; + + unsigned long lowest_stack; }; /* diff --git a/arch/sparc/include/asm/thread_info_64.h b/arch/sparc/include/asm/thread_info_64.h index 01d057f..0a02f7e 100644 --- a/arch/sparc/include/asm/thread_info_64.h +++ b/arch/sparc/include/asm/thread_info_64.h @@ -63,6 +63,8 @@ struct thread_info { struct pt_regs *kern_una_regs; unsigned int kern_una_insn; + unsigned long lowest_stack; + unsigned long fpregs[0] __attribute__ ((aligned(64))); }; diff --git a/arch/sparc/include/asm/uaccess.h b/arch/sparc/include/asm/uaccess.h index e88fbe5..96b0ce5 100644 --- a/arch/sparc/include/asm/uaccess.h +++ b/arch/sparc/include/asm/uaccess.h @@ -1,5 +1,13 @@ #ifndef ___ASM_SPARC_UACCESS_H #define ___ASM_SPARC_UACCESS_H + +#ifdef __KERNEL__ +#ifndef __ASSEMBLY__ +#include +extern void check_object_size(const void *ptr, unsigned long n, bool to); +#endif +#endif + #if defined(__sparc__) && defined(__arch64__) #include #else diff --git a/arch/sparc/include/asm/uaccess_32.h b/arch/sparc/include/asm/uaccess_32.h index 8303ac4..07f333d 100644 --- a/arch/sparc/include/asm/uaccess_32.h +++ b/arch/sparc/include/asm/uaccess_32.h @@ -249,27 +249,46 @@ extern unsigned long __copy_user(void __user *to, const void __user *from, unsig static inline unsigned long copy_to_user(void __user *to, const void *from, unsigned long n) { - if (n && __access_ok((unsigned long) to, n)) + if ((long)n < 0) + return n; + + if (n && __access_ok((unsigned long) to, n)) { + if (!__builtin_constant_p(n)) + check_object_size(from, n, true); return __copy_user(to, (__force void __user *) from, n); - else + } else return n; } static inline unsigned long __copy_to_user(void __user *to, const void *from, unsigned long n) { + if ((long)n < 0) + return n; + + if (!__builtin_constant_p(n)) + check_object_size(from, n, true); + return __copy_user(to, (__force void __user *) from, n); } static inline unsigned long copy_from_user(void *to, const void __user *from, unsigned long n) { - if (n && __access_ok((unsigned long) from, n)) + if ((long)n < 0) + return n; + + if (n && __access_ok((unsigned long) from, n)) { + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); return __copy_user((__force void __user *) to, from, n); - else + } else return n; } static inline unsigned long __copy_from_user(void *to, const void __user *from, unsigned long n) { + if ((long)n < 0) + return n; + return __copy_user((__force void __user *) to, from, n); } diff --git a/arch/sparc/include/asm/uaccess_64.h b/arch/sparc/include/asm/uaccess_64.h index 3e1449f..5293a0e 100644 --- a/arch/sparc/include/asm/uaccess_64.h +++ b/arch/sparc/include/asm/uaccess_64.h @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -213,8 +214,15 @@ extern unsigned long copy_from_user_fixup(void *to, const void __user *from, static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, unsigned long size) { - unsigned long ret = ___copy_from_user(to, from, size); + unsigned long ret; + if ((long)size < 0 || size > INT_MAX) + return size; + + if (!__builtin_constant_p(size)) + check_object_size(to, size, false); + + ret = ___copy_from_user(to, from, size); if (unlikely(ret)) ret = copy_from_user_fixup(to, from, size); @@ -230,8 +238,15 @@ extern unsigned long copy_to_user_fixup(void __user *to, const void *from, static inline unsigned long __must_check copy_to_user(void __user *to, const void *from, unsigned long size) { - unsigned long ret = ___copy_to_user(to, from, size); + unsigned long ret; + if ((long)size < 0 || size > INT_MAX) + return size; + + if (!__builtin_constant_p(size)) + check_object_size(from, size, true); + + ret = ___copy_to_user(to, from, size); if (unlikely(ret)) ret = copy_to_user_fixup(to, from, size); return ret; diff --git a/arch/sparc/kernel/Makefile b/arch/sparc/kernel/Makefile index cb85458..e063f17 100644 --- a/arch/sparc/kernel/Makefile +++ b/arch/sparc/kernel/Makefile @@ -3,7 +3,7 @@ # asflags-y := -ansi -ccflags-y := -Werror +#ccflags-y := -Werror extra-y := head_$(BITS).o extra-y += init_task.o diff --git a/arch/sparc/kernel/process_32.c b/arch/sparc/kernel/process_32.c index f793742..4d880af 100644 --- a/arch/sparc/kernel/process_32.c +++ b/arch/sparc/kernel/process_32.c @@ -204,7 +204,7 @@ void __show_backtrace(unsigned long fp) rw->ins[4], rw->ins[5], rw->ins[6], rw->ins[7]); - printk("%pS\n", (void *) rw->ins[7]); + printk("%pA\n", (void *) rw->ins[7]); rw = (struct reg_window32 *) rw->ins[6]; } spin_unlock_irqrestore(&sparc_backtrace_lock, flags); @@ -271,14 +271,14 @@ void show_regs(struct pt_regs *r) printk("PSR: %08lx PC: %08lx NPC: %08lx Y: %08lx %s\n", r->psr, r->pc, r->npc, r->y, print_tainted()); - printk("PC: <%pS>\n", (void *) r->pc); + printk("PC: <%pA>\n", (void *) r->pc); printk("%%G: %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx\n", r->u_regs[0], r->u_regs[1], r->u_regs[2], r->u_regs[3], r->u_regs[4], r->u_regs[5], r->u_regs[6], r->u_regs[7]); printk("%%O: %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx\n", r->u_regs[8], r->u_regs[9], r->u_regs[10], r->u_regs[11], r->u_regs[12], r->u_regs[13], r->u_regs[14], r->u_regs[15]); - printk("RPC: <%pS>\n", (void *) r->u_regs[15]); + printk("RPC: <%pA>\n", (void *) r->u_regs[15]); printk("%%L: %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx\n", rw->locals[0], rw->locals[1], rw->locals[2], rw->locals[3], @@ -313,7 +313,7 @@ void show_stack(struct task_struct *tsk, unsigned long *_ksp) rw = (struct reg_window32 *) fp; pc = rw->ins[7]; printk("[%08lx : ", pc); - printk("%pS ] ", (void *) pc); + printk("%pA ] ", (void *) pc); fp = rw->ins[6]; } while (++count < 16); printk("\n"); diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c index 39d8b05..d1a7d90 100644 --- a/arch/sparc/kernel/process_64.c +++ b/arch/sparc/kernel/process_64.c @@ -182,14 +182,14 @@ static void show_regwindow(struct pt_regs *regs) printk("i4: %016lx i5: %016lx i6: %016lx i7: %016lx\n", rwk->ins[4], rwk->ins[5], rwk->ins[6], rwk->ins[7]); if (regs->tstate & TSTATE_PRIV) - printk("I7: <%pS>\n", (void *) rwk->ins[7]); + printk("I7: <%pA>\n", (void *) rwk->ins[7]); } void show_regs(struct pt_regs *regs) { printk("TSTATE: %016lx TPC: %016lx TNPC: %016lx Y: %08x %s\n", regs->tstate, regs->tpc, regs->tnpc, regs->y, print_tainted()); - printk("TPC: <%pS>\n", (void *) regs->tpc); + printk("TPC: <%pA>\n", (void *) regs->tpc); printk("g0: %016lx g1: %016lx g2: %016lx g3: %016lx\n", regs->u_regs[0], regs->u_regs[1], regs->u_regs[2], regs->u_regs[3]); @@ -202,7 +202,7 @@ void show_regs(struct pt_regs *regs) printk("o4: %016lx o5: %016lx sp: %016lx ret_pc: %016lx\n", regs->u_regs[12], regs->u_regs[13], regs->u_regs[14], regs->u_regs[15]); - printk("RPC: <%pS>\n", (void *) regs->u_regs[15]); + printk("RPC: <%pA>\n", (void *) regs->u_regs[15]); show_regwindow(regs); show_stack(current, (unsigned long *) regs->u_regs[UREG_FP]); } @@ -287,7 +287,7 @@ void arch_trigger_all_cpu_backtrace(void) ((tp && tp->task) ? tp->task->pid : -1)); if (gp->tstate & TSTATE_PRIV) { - printk(" TPC[%pS] O7[%pS] I7[%pS] RPC[%pS]\n", + printk(" TPC[%pA] O7[%pA] I7[%pA] RPC[%pA]\n", (void *) gp->tpc, (void *) gp->o7, (void *) gp->i7, diff --git a/arch/sparc/kernel/sys_sparc_32.c b/arch/sparc/kernel/sys_sparc_32.c index 42b282f..28ce9f2 100644 --- a/arch/sparc/kernel/sys_sparc_32.c +++ b/arch/sparc/kernel/sys_sparc_32.c @@ -56,7 +56,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi if (ARCH_SUN4C && len > 0x20000000) return -ENOMEM; if (!addr) - addr = TASK_UNMAPPED_BASE; + addr = current->mm->mmap_base; if (flags & MAP_SHARED) addr = COLOUR_ALIGN(addr); @@ -71,7 +71,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi } if (TASK_SIZE - PAGE_SIZE - len < addr) return -ENOMEM; - if (!vmm || addr + len <= vmm->vm_start) + if (check_heap_stack_gap(vmm, addr, len)) return addr; addr = vmm->vm_end; if (flags & MAP_SHARED) diff --git a/arch/sparc/kernel/sys_sparc_64.c b/arch/sparc/kernel/sys_sparc_64.c index 232df99..cee1f9c 100644 --- a/arch/sparc/kernel/sys_sparc_64.c +++ b/arch/sparc/kernel/sys_sparc_64.c @@ -124,7 +124,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi /* We do not accept a shared mapping if it would violate * cache aliasing constraints. */ - if ((flags & MAP_SHARED) && + if ((filp || (flags & MAP_SHARED)) && ((addr - (pgoff << PAGE_SHIFT)) & (SHMLBA - 1))) return -EINVAL; return addr; @@ -139,6 +139,10 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi if (filp || (flags & MAP_SHARED)) do_color_align = 1; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_color_align) addr = COLOUR_ALIGN(addr, pgoff); @@ -146,15 +150,14 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi addr = PAGE_ALIGN(addr); vma = find_vma(mm, addr); - if (task_size - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (task_size - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } if (len > mm->cached_hole_size) { - start_addr = addr = mm->free_area_cache; + start_addr = addr = mm->free_area_cache; } else { - start_addr = addr = TASK_UNMAPPED_BASE; + start_addr = addr = mm->mmap_base; mm->cached_hole_size = 0; } @@ -174,14 +177,14 @@ full_search: vma = find_vma(mm, VA_EXCLUDE_END); } if (unlikely(task_size < addr)) { - if (start_addr != TASK_UNMAPPED_BASE) { - start_addr = addr = TASK_UNMAPPED_BASE; + if (start_addr != mm->mmap_base) { + start_addr = addr = mm->mmap_base; mm->cached_hole_size = 0; goto full_search; } return -ENOMEM; } - if (likely(!vma || addr + len <= vma->vm_start)) { + if (likely(check_heap_stack_gap(vma, addr, len))) { /* * Remember the place where we stopped the search: */ @@ -215,7 +218,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, /* We do not accept a shared mapping if it would violate * cache aliasing constraints. */ - if ((flags & MAP_SHARED) && + if ((filp || (flags & MAP_SHARED)) && ((addr - (pgoff << PAGE_SHIFT)) & (SHMLBA - 1))) return -EINVAL; return addr; @@ -236,8 +239,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, addr = PAGE_ALIGN(addr); vma = find_vma(mm, addr); - if (task_size - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (task_size - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } @@ -258,7 +260,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, /* make sure it can fit in the remaining address space */ if (likely(addr > len)) { vma = find_vma(mm, addr-len); - if (!vma || addr <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr - len, len)) { /* remember the address as a hint for next time */ return (mm->free_area_cache = addr-len); } @@ -267,18 +269,18 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (unlikely(mm->mmap_base < len)) goto bottomup; - addr = mm->mmap_base-len; - if (do_color_align) - addr = COLOUR_ALIGN_DOWN(addr, pgoff); + addr = mm->mmap_base - len; do { + if (do_color_align) + addr = COLOUR_ALIGN_DOWN(addr, pgoff); /* * Lookup failure means no vma is above this address, * else if new region fits below vma->vm_start, * return with success: */ vma = find_vma(mm, addr); - if (likely(!vma || addr+len <= vma->vm_start)) { + if (likely(check_heap_stack_gap(vma, addr, len))) { /* remember the address as a hint for next time */ return (mm->free_area_cache = addr); } @@ -288,10 +290,8 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, mm->cached_hole_size = vma->vm_start - addr; /* try just below the current vma->vm_start */ - addr = vma->vm_start-len; - if (do_color_align) - addr = COLOUR_ALIGN_DOWN(addr, pgoff); - } while (likely(len < vma->vm_start)); + addr = skip_heap_stack_gap(vma, len); + } while (!IS_ERR_VALUE(addr)); bottomup: /* @@ -390,6 +390,12 @@ void arch_pick_mmap_layout(struct mm_struct *mm) gap == RLIM_INFINITY || sysctl_legacy_va_layout) { mm->mmap_base = TASK_UNMAPPED_BASE + random_factor; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { @@ -402,6 +408,12 @@ void arch_pick_mmap_layout(struct mm_struct *mm) gap = (task_size / 6 * 5); mm->mmap_base = PAGE_ALIGN(task_size - gap - random_factor); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff --git a/arch/sparc/kernel/traps_32.c b/arch/sparc/kernel/traps_32.c index 591f20c..0f1b925 100644 --- a/arch/sparc/kernel/traps_32.c +++ b/arch/sparc/kernel/traps_32.c @@ -45,6 +45,8 @@ static void instruction_dump(unsigned long *pc) #define __SAVE __asm__ __volatile__("save %sp, -0x40, %sp\n\t") #define __RESTORE __asm__ __volatile__("restore %g0, %g0, %g0\n\t") +extern void gr_handle_kernel_exploit(void); + void die_if_kernel(char *str, struct pt_regs *regs) { static int die_counter; @@ -77,15 +79,17 @@ void die_if_kernel(char *str, struct pt_regs *regs) count++ < 30 && (((unsigned long) rw) >= PAGE_OFFSET) && !(((unsigned long) rw) & 0x7)) { - printk("Caller[%08lx]: %pS\n", rw->ins[7], + printk("Caller[%08lx]: %pA\n", rw->ins[7], (void *) rw->ins[7]); rw = (struct reg_window32 *)rw->ins[6]; } } printk("Instruction DUMP:"); instruction_dump ((unsigned long *) regs->pc); - if(regs->psr & PSR_PS) + if(regs->psr & PSR_PS) { + gr_handle_kernel_exploit(); do_exit(SIGKILL); + } do_exit(SIGSEGV); } diff --git a/arch/sparc/kernel/traps_64.c b/arch/sparc/kernel/traps_64.c index 0cbdaa4..438e4c9 100644 --- a/arch/sparc/kernel/traps_64.c +++ b/arch/sparc/kernel/traps_64.c @@ -75,7 +75,7 @@ static void dump_tl1_traplog(struct tl1_traplog *p) i + 1, p->trapstack[i].tstate, p->trapstack[i].tpc, p->trapstack[i].tnpc, p->trapstack[i].tt); - printk("TRAPLOG: TPC<%pS>\n", (void *) p->trapstack[i].tpc); + printk("TRAPLOG: TPC<%pA>\n", (void *) p->trapstack[i].tpc); } } @@ -95,6 +95,12 @@ void bad_trap(struct pt_regs *regs, long lvl) lvl -= 0x100; if (regs->tstate & TSTATE_PRIV) { + +#ifdef CONFIG_PAX_REFCOUNT + if (lvl == 6) + pax_report_refcount_overflow(regs); +#endif + sprintf(buffer, "Kernel bad sw trap %lx", lvl); die_if_kernel(buffer, regs); } @@ -113,11 +119,16 @@ void bad_trap(struct pt_regs *regs, long lvl) void bad_trap_tl1(struct pt_regs *regs, long lvl) { char buffer[32]; - + if (notify_die(DIE_TRAP_TL1, "bad trap tl1", regs, 0, lvl, SIGTRAP) == NOTIFY_STOP) return; +#ifdef CONFIG_PAX_REFCOUNT + if (lvl == 6) + pax_report_refcount_overflow(regs); +#endif + dump_tl1_traplog((struct tl1_traplog *)(regs + 1)); sprintf (buffer, "Bad trap %lx at tl>0", lvl); @@ -1141,7 +1152,7 @@ static void cheetah_log_errors(struct pt_regs *regs, struct cheetah_err_info *in regs->tpc, regs->tnpc, regs->u_regs[UREG_I7], regs->tstate); printk("%s" "ERROR(%d): ", (recoverable ? KERN_WARNING : KERN_CRIT), smp_processor_id()); - printk("TPC<%pS>\n", (void *) regs->tpc); + printk("TPC<%pA>\n", (void *) regs->tpc); printk("%s" "ERROR(%d): M_SYND(%lx), E_SYND(%lx)%s%s\n", (recoverable ? KERN_WARNING : KERN_CRIT), smp_processor_id(), (afsr & CHAFSR_M_SYNDROME) >> CHAFSR_M_SYNDROME_SHIFT, @@ -1748,7 +1759,7 @@ void cheetah_plus_parity_error(int type, struct pt_regs *regs) smp_processor_id(), (type & 0x1) ? 'I' : 'D', regs->tpc); - printk(KERN_EMERG "TPC<%pS>\n", (void *) regs->tpc); + printk(KERN_EMERG "TPC<%pA>\n", (void *) regs->tpc); panic("Irrecoverable Cheetah+ parity error."); } @@ -1756,7 +1767,7 @@ void cheetah_plus_parity_error(int type, struct pt_regs *regs) smp_processor_id(), (type & 0x1) ? 'I' : 'D', regs->tpc); - printk(KERN_WARNING "TPC<%pS>\n", (void *) regs->tpc); + printk(KERN_WARNING "TPC<%pA>\n", (void *) regs->tpc); } struct sun4v_error_entry { @@ -1963,9 +1974,9 @@ void sun4v_itlb_error_report(struct pt_regs *regs, int tl) printk(KERN_EMERG "SUN4V-ITLB: Error at TPC[%lx], tl %d\n", regs->tpc, tl); - printk(KERN_EMERG "SUN4V-ITLB: TPC<%pS>\n", (void *) regs->tpc); + printk(KERN_EMERG "SUN4V-ITLB: TPC<%pA>\n", (void *) regs->tpc); printk(KERN_EMERG "SUN4V-ITLB: O7[%lx]\n", regs->u_regs[UREG_I7]); - printk(KERN_EMERG "SUN4V-ITLB: O7<%pS>\n", + printk(KERN_EMERG "SUN4V-ITLB: O7<%pA>\n", (void *) regs->u_regs[UREG_I7]); printk(KERN_EMERG "SUN4V-ITLB: vaddr[%lx] ctx[%lx] " "pte[%lx] error[%lx]\n", @@ -1987,9 +1998,9 @@ void sun4v_dtlb_error_report(struct pt_regs *regs, int tl) printk(KERN_EMERG "SUN4V-DTLB: Error at TPC[%lx], tl %d\n", regs->tpc, tl); - printk(KERN_EMERG "SUN4V-DTLB: TPC<%pS>\n", (void *) regs->tpc); + printk(KERN_EMERG "SUN4V-DTLB: TPC<%pA>\n", (void *) regs->tpc); printk(KERN_EMERG "SUN4V-DTLB: O7[%lx]\n", regs->u_regs[UREG_I7]); - printk(KERN_EMERG "SUN4V-DTLB: O7<%pS>\n", + printk(KERN_EMERG "SUN4V-DTLB: O7<%pA>\n", (void *) regs->u_regs[UREG_I7]); printk(KERN_EMERG "SUN4V-DTLB: vaddr[%lx] ctx[%lx] " "pte[%lx] error[%lx]\n", @@ -2195,13 +2206,13 @@ void show_stack(struct task_struct *tsk, unsigned long *_ksp) fp = (unsigned long)sf->fp + STACK_BIAS; } - printk(" [%016lx] %pS\n", pc, (void *) pc); + printk(" [%016lx] %pA\n", pc, (void *) pc); #ifdef CONFIG_FUNCTION_GRAPH_TRACER if ((pc + 8UL) == (unsigned long) &return_to_handler) { int index = tsk->curr_ret_stack; if (tsk->ret_stack && index >= graph) { pc = tsk->ret_stack[index - graph].ret; - printk(" [%016lx] %pS\n", pc, (void *) pc); + printk(" [%016lx] %pA\n", pc, (void *) pc); graph++; } } @@ -2226,6 +2237,8 @@ static inline struct reg_window *kernel_stack_up(struct reg_window *rw) return (struct reg_window *) (fp + STACK_BIAS); } +extern void gr_handle_kernel_exploit(void); + void die_if_kernel(char *str, struct pt_regs *regs) { static int die_counter; @@ -2254,7 +2267,7 @@ void die_if_kernel(char *str, struct pt_regs *regs) while (rw && count++ < 30 && kstack_valid(tp, (unsigned long) rw)) { - printk("Caller[%016lx]: %pS\n", rw->ins[7], + printk("Caller[%016lx]: %pA\n", rw->ins[7], (void *) rw->ins[7]); rw = kernel_stack_up(rw); @@ -2267,8 +2280,10 @@ void die_if_kernel(char *str, struct pt_regs *regs) } user_instruction_dump ((unsigned int __user *) regs->tpc); } - if (regs->tstate & TSTATE_PRIV) + if (regs->tstate & TSTATE_PRIV) { + gr_handle_kernel_exploit(); do_exit(SIGKILL); + } do_exit(SIGSEGV); } EXPORT_SYMBOL(die_if_kernel); diff --git a/arch/sparc/kernel/unaligned_64.c b/arch/sparc/kernel/unaligned_64.c index 76e4ac1..78f8bb1 100644 --- a/arch/sparc/kernel/unaligned_64.c +++ b/arch/sparc/kernel/unaligned_64.c @@ -279,7 +279,7 @@ static void log_unaligned(struct pt_regs *regs) static DEFINE_RATELIMIT_STATE(ratelimit, 5 * HZ, 5); if (__ratelimit(&ratelimit)) { - printk("Kernel unaligned access at TPC[%lx] %pS\n", + printk("Kernel unaligned access at TPC[%lx] %pA\n", regs->tpc, (void *) regs->tpc); } } diff --git a/arch/sparc/lib/Makefile b/arch/sparc/lib/Makefile index a3fc437..fea9957 100644 --- a/arch/sparc/lib/Makefile +++ b/arch/sparc/lib/Makefile @@ -2,7 +2,7 @@ # asflags-y := -ansi -DST_DIV0=0x02 -ccflags-y := -Werror +#ccflags-y := -Werror lib-$(CONFIG_SPARC32) += mul.o rem.o sdiv.o udiv.o umul.o urem.o ashrdi3.o lib-$(CONFIG_SPARC32) += memcpy.o memset.o diff --git a/arch/sparc/lib/atomic_64.S b/arch/sparc/lib/atomic_64.S index 59186e0..f747d7a 100644 --- a/arch/sparc/lib/atomic_64.S +++ b/arch/sparc/lib/atomic_64.S @@ -18,7 +18,12 @@ atomic_add: /* %o0 = increment, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: lduw [%o1], %g1 - add %g1, %o0, %g7 + addcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %icc, 6 +#endif + cas [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %icc, BACKOFF_LABEL(2f, 1b) @@ -28,12 +33,32 @@ atomic_add: /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) .size atomic_add, .-atomic_add + .globl atomic_add_unchecked + .type atomic_add_unchecked,#function +atomic_add_unchecked: /* %o0 = increment, %o1 = atomic_ptr */ + BACKOFF_SETUP(%o2) +1: lduw [%o1], %g1 + add %g1, %o0, %g7 + cas [%o1], %g1, %g7 + cmp %g1, %g7 + bne,pn %icc, 2f + nop + retl + nop +2: BACKOFF_SPIN(%o2, %o3, 1b) + .size atomic_add_unchecked, .-atomic_add_unchecked + .globl atomic_sub .type atomic_sub,#function atomic_sub: /* %o0 = decrement, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: lduw [%o1], %g1 - sub %g1, %o0, %g7 + subcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %icc, 6 +#endif + cas [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %icc, BACKOFF_LABEL(2f, 1b) @@ -43,12 +68,32 @@ atomic_sub: /* %o0 = decrement, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) .size atomic_sub, .-atomic_sub + .globl atomic_sub_unchecked + .type atomic_sub_unchecked,#function +atomic_sub_unchecked: /* %o0 = decrement, %o1 = atomic_ptr */ + BACKOFF_SETUP(%o2) +1: lduw [%o1], %g1 + sub %g1, %o0, %g7 + cas [%o1], %g1, %g7 + cmp %g1, %g7 + bne,pn %icc, 2f + nop + retl + nop +2: BACKOFF_SPIN(%o2, %o3, 1b) + .size atomic_sub_unchecked, .-atomic_sub_unchecked + .globl atomic_add_ret .type atomic_add_ret,#function atomic_add_ret: /* %o0 = increment, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: lduw [%o1], %g1 - add %g1, %o0, %g7 + addcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %icc, 6 +#endif + cas [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %icc, BACKOFF_LABEL(2f, 1b) @@ -58,12 +103,33 @@ atomic_add_ret: /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) .size atomic_add_ret, .-atomic_add_ret + .globl atomic_add_ret_unchecked + .type atomic_add_ret_unchecked,#function +atomic_add_ret_unchecked: /* %o0 = increment, %o1 = atomic_ptr */ + BACKOFF_SETUP(%o2) +1: lduw [%o1], %g1 + addcc %g1, %o0, %g7 + cas [%o1], %g1, %g7 + cmp %g1, %g7 + bne,pn %icc, 2f + add %g7, %o0, %g7 + sra %g7, 0, %o0 + retl + nop +2: BACKOFF_SPIN(%o2, %o3, 1b) + .size atomic_add_ret_unchecked, .-atomic_add_ret_unchecked + .globl atomic_sub_ret .type atomic_sub_ret,#function atomic_sub_ret: /* %o0 = decrement, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: lduw [%o1], %g1 - sub %g1, %o0, %g7 + subcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %icc, 6 +#endif + cas [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %icc, BACKOFF_LABEL(2f, 1b) @@ -78,7 +144,12 @@ atomic_sub_ret: /* %o0 = decrement, %o1 = atomic_ptr */ atomic64_add: /* %o0 = increment, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: ldx [%o1], %g1 - add %g1, %o0, %g7 + addcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %xcc, 6 +#endif + casx [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %xcc, BACKOFF_LABEL(2f, 1b) @@ -88,12 +159,32 @@ atomic64_add: /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) .size atomic64_add, .-atomic64_add + .globl atomic64_add_unchecked + .type atomic64_add_unchecked,#function +atomic64_add_unchecked: /* %o0 = increment, %o1 = atomic_ptr */ + BACKOFF_SETUP(%o2) +1: ldx [%o1], %g1 + addcc %g1, %o0, %g7 + casx [%o1], %g1, %g7 + cmp %g1, %g7 + bne,pn %xcc, 2f + nop + retl + nop +2: BACKOFF_SPIN(%o2, %o3, 1b) + .size atomic64_add_unchecked, .-atomic64_add_unchecked + .globl atomic64_sub .type atomic64_sub,#function atomic64_sub: /* %o0 = decrement, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: ldx [%o1], %g1 - sub %g1, %o0, %g7 + subcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %xcc, 6 +#endif + casx [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %xcc, BACKOFF_LABEL(2f, 1b) @@ -103,12 +194,32 @@ atomic64_sub: /* %o0 = decrement, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) .size atomic64_sub, .-atomic64_sub + .globl atomic64_sub_unchecked + .type atomic64_sub_unchecked,#function +atomic64_sub_unchecked: /* %o0 = decrement, %o1 = atomic_ptr */ + BACKOFF_SETUP(%o2) +1: ldx [%o1], %g1 + subcc %g1, %o0, %g7 + casx [%o1], %g1, %g7 + cmp %g1, %g7 + bne,pn %xcc, 2f + nop + retl + nop +2: BACKOFF_SPIN(%o2, %o3, 1b) + .size atomic64_sub_unchecked, .-atomic64_sub_unchecked + .globl atomic64_add_ret .type atomic64_add_ret,#function atomic64_add_ret: /* %o0 = increment, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: ldx [%o1], %g1 - add %g1, %o0, %g7 + addcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %xcc, 6 +#endif + casx [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %xcc, BACKOFF_LABEL(2f, 1b) @@ -118,12 +229,33 @@ atomic64_add_ret: /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) .size atomic64_add_ret, .-atomic64_add_ret + .globl atomic64_add_ret_unchecked + .type atomic64_add_ret_unchecked,#function +atomic64_add_ret_unchecked: /* %o0 = increment, %o1 = atomic_ptr */ + BACKOFF_SETUP(%o2) +1: ldx [%o1], %g1 + addcc %g1, %o0, %g7 + casx [%o1], %g1, %g7 + cmp %g1, %g7 + bne,pn %xcc, 2f + add %g7, %o0, %g7 + mov %g7, %o0 + retl + nop +2: BACKOFF_SPIN(%o2, %o3, 1b) + .size atomic64_add_ret_unchecked, .-atomic64_add_ret_unchecked + .globl atomic64_sub_ret .type atomic64_sub_ret,#function atomic64_sub_ret: /* %o0 = decrement, %o1 = atomic_ptr */ BACKOFF_SETUP(%o2) 1: ldx [%o1], %g1 - sub %g1, %o0, %g7 + subcc %g1, %o0, %g7 + +#ifdef CONFIG_PAX_REFCOUNT + tvs %xcc, 6 +#endif + casx [%o1], %g1, %g7 cmp %g1, %g7 bne,pn %xcc, BACKOFF_LABEL(2f, 1b) diff --git a/arch/sparc/lib/ksyms.c b/arch/sparc/lib/ksyms.c index f73c224..662af10 100644 --- a/arch/sparc/lib/ksyms.c +++ b/arch/sparc/lib/ksyms.c @@ -136,12 +136,18 @@ EXPORT_SYMBOL(__downgrade_write); /* Atomic counter implementation. */ EXPORT_SYMBOL(atomic_add); +EXPORT_SYMBOL(atomic_add_unchecked); EXPORT_SYMBOL(atomic_add_ret); +EXPORT_SYMBOL(atomic_add_ret_unchecked); EXPORT_SYMBOL(atomic_sub); +EXPORT_SYMBOL(atomic_sub_unchecked); EXPORT_SYMBOL(atomic_sub_ret); EXPORT_SYMBOL(atomic64_add); +EXPORT_SYMBOL(atomic64_add_unchecked); EXPORT_SYMBOL(atomic64_add_ret); +EXPORT_SYMBOL(atomic64_add_ret_unchecked); EXPORT_SYMBOL(atomic64_sub); +EXPORT_SYMBOL(atomic64_sub_unchecked); EXPORT_SYMBOL(atomic64_sub_ret); /* Atomic bit operations. */ diff --git a/arch/sparc/mm/Makefile b/arch/sparc/mm/Makefile index 301421c..e2535d1 100644 --- a/arch/sparc/mm/Makefile +++ b/arch/sparc/mm/Makefile @@ -2,7 +2,7 @@ # asflags-y := -ansi -ccflags-y := -Werror +#ccflags-y := -Werror obj-$(CONFIG_SPARC64) += ultra.o tlb.o tsb.o gup.o obj-y += fault_$(BITS).o diff --git a/arch/sparc/mm/fault_32.c b/arch/sparc/mm/fault_32.c index 8023fd7..c8e89e9 100644 --- a/arch/sparc/mm/fault_32.c +++ b/arch/sparc/mm/fault_32.c @@ -21,6 +21,9 @@ #include #include #include +#include +#include +#include #include #include @@ -208,6 +211,268 @@ static unsigned long compute_si_addr(struct pt_regs *regs, int text_fault) return safe_compute_effective_address(regs, insn); } +#ifdef CONFIG_PAX_PAGEEXEC +#ifdef CONFIG_PAX_DLRESOLVE +static void pax_emuplt_close(struct vm_area_struct *vma) +{ + vma->vm_mm->call_dl_resolve = 0UL; +} + +static int pax_emuplt_fault(struct vm_area_struct *vma, struct vm_fault *vmf) +{ + unsigned int *kaddr; + + vmf->page = alloc_page(GFP_HIGHUSER); + if (!vmf->page) + return VM_FAULT_OOM; + + kaddr = kmap(vmf->page); + memset(kaddr, 0, PAGE_SIZE); + kaddr[0] = 0x9DE3BFA8U; /* save */ + flush_dcache_page(vmf->page); + kunmap(vmf->page); + return VM_FAULT_MAJOR; +} + +static const struct vm_operations_struct pax_vm_ops = { + .close = pax_emuplt_close, + .fault = pax_emuplt_fault +}; + +static int pax_insert_vma(struct vm_area_struct *vma, unsigned long addr) +{ + int ret; + + INIT_LIST_HEAD(&vma->anon_vma_chain); + vma->vm_mm = current->mm; + vma->vm_start = addr; + vma->vm_end = addr + PAGE_SIZE; + vma->vm_flags = VM_READ | VM_EXEC | VM_MAYREAD | VM_MAYEXEC; + vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); + vma->vm_ops = &pax_vm_ops; + + ret = insert_vm_struct(current->mm, vma); + if (ret) + return ret; + + ++current->mm->total_vm; + return 0; +} +#endif + +/* + * PaX: decide what to do with offenders (regs->pc = fault address) + * + * returns 1 when task should be killed + * 2 when patched PLT trampoline was detected + * 3 when unpatched PLT trampoline was detected + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + +#ifdef CONFIG_PAX_EMUPLT + int err; + + do { /* PaX: patched PLT emulation #1 */ + unsigned int sethi1, sethi2, jmpl; + + err = get_user(sethi1, (unsigned int *)regs->pc); + err |= get_user(sethi2, (unsigned int *)(regs->pc+4)); + err |= get_user(jmpl, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((sethi1 & 0xFFC00000U) == 0x03000000U && + (sethi2 & 0xFFC00000U) == 0x03000000U && + (jmpl & 0xFFFFE000U) == 0x81C06000U) + { + unsigned int addr; + + regs->u_regs[UREG_G1] = (sethi2 & 0x003FFFFFU) << 10; + addr = regs->u_regs[UREG_G1]; + addr += (((jmpl | 0xFFFFE000U) ^ 0x00001000U) + 0x00001000U); + regs->pc = addr; + regs->npc = addr+4; + return 2; + } + } while (0); + + { /* PaX: patched PLT emulation #2 */ + unsigned int ba; + + err = get_user(ba, (unsigned int *)regs->pc); + + if (!err && (ba & 0xFFC00000U) == 0x30800000U) { + unsigned int addr; + + addr = regs->pc + ((((ba | 0xFFC00000U) ^ 0x00200000U) + 0x00200000U) << 2); + regs->pc = addr; + regs->npc = addr+4; + return 2; + } + } + + do { /* PaX: patched PLT emulation #3 */ + unsigned int sethi, jmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->pc); + err |= get_user(jmpl, (unsigned int *)(regs->pc+4)); + err |= get_user(nop, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + (jmpl & 0xFFFFE000U) == 0x81C06000U && + nop == 0x01000000U) + { + unsigned int addr; + + addr = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] = addr; + addr += (((jmpl | 0xFFFFE000U) ^ 0x00001000U) + 0x00001000U); + regs->pc = addr; + regs->npc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: unpatched PLT emulation step 1 */ + unsigned int sethi, ba, nop; + + err = get_user(sethi, (unsigned int *)regs->pc); + err |= get_user(ba, (unsigned int *)(regs->pc+4)); + err |= get_user(nop, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + ((ba & 0xFFC00000U) == 0x30800000U || (ba & 0xFFF80000U) == 0x30680000U) && + nop == 0x01000000U) + { + unsigned int addr, save, call; + + if ((ba & 0xFFC00000U) == 0x30800000U) + addr = regs->pc + 4 + ((((ba | 0xFFC00000U) ^ 0x00200000U) + 0x00200000U) << 2); + else + addr = regs->pc + 4 + ((((ba | 0xFFF80000U) ^ 0x00040000U) + 0x00040000U) << 2); + + err = get_user(save, (unsigned int *)addr); + err |= get_user(call, (unsigned int *)(addr+4)); + err |= get_user(nop, (unsigned int *)(addr+8)); + if (err) + break; + +#ifdef CONFIG_PAX_DLRESOLVE + if (save == 0x9DE3BFA8U && + (call & 0xC0000000U) == 0x40000000U && + nop == 0x01000000U) + { + struct vm_area_struct *vma; + unsigned long call_dl_resolve; + + down_read(¤t->mm->mmap_sem); + call_dl_resolve = current->mm->call_dl_resolve; + up_read(¤t->mm->mmap_sem); + if (likely(call_dl_resolve)) + goto emulate; + + vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); + + down_write(¤t->mm->mmap_sem); + if (current->mm->call_dl_resolve) { + call_dl_resolve = current->mm->call_dl_resolve; + up_write(¤t->mm->mmap_sem); + if (vma) + kmem_cache_free(vm_area_cachep, vma); + goto emulate; + } + + call_dl_resolve = get_unmapped_area(NULL, 0UL, PAGE_SIZE, 0UL, MAP_PRIVATE); + if (!vma || (call_dl_resolve & ~PAGE_MASK)) { + up_write(¤t->mm->mmap_sem); + if (vma) + kmem_cache_free(vm_area_cachep, vma); + return 1; + } + + if (pax_insert_vma(vma, call_dl_resolve)) { + up_write(¤t->mm->mmap_sem); + kmem_cache_free(vm_area_cachep, vma); + return 1; + } + + current->mm->call_dl_resolve = call_dl_resolve; + up_write(¤t->mm->mmap_sem); + +emulate: + regs->u_regs[UREG_G1] = (sethi & 0x003FFFFFU) << 10; + regs->pc = call_dl_resolve; + regs->npc = addr+4; + return 3; + } +#endif + + /* PaX: glibc 2.4+ generates sethi/jmpl instead of save/call */ + if ((save & 0xFFC00000U) == 0x05000000U && + (call & 0xFFFFE000U) == 0x85C0A000U && + nop == 0x01000000U) + { + regs->u_regs[UREG_G1] = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G2] = addr + 4; + addr = (save & 0x003FFFFFU) << 10; + addr += (((call | 0xFFFFE000U) ^ 0x00001000U) + 0x00001000U); + regs->pc = addr; + regs->npc = addr+4; + return 3; + } + } + } while (0); + + do { /* PaX: unpatched PLT emulation step 2 */ + unsigned int save, call, nop; + + err = get_user(save, (unsigned int *)(regs->pc-4)); + err |= get_user(call, (unsigned int *)regs->pc); + err |= get_user(nop, (unsigned int *)(regs->pc+4)); + if (err) + break; + + if (save == 0x9DE3BFA8U && + (call & 0xC0000000U) == 0x40000000U && + nop == 0x01000000U) + { + unsigned int dl_resolve = regs->pc + ((((call | 0xC0000000U) ^ 0x20000000U) + 0x20000000U) << 2); + + regs->u_regs[UREG_RETPC] = regs->pc; + regs->pc = dl_resolve; + regs->npc = dl_resolve+4; + return 3; + } + } while (0); +#endif + + return 1; +} + +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 8; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + static noinline void do_fault_siginfo(int code, int sig, struct pt_regs *regs, int text_fault) { @@ -280,6 +545,24 @@ good_area: if(!(vma->vm_flags & VM_WRITE)) goto bad_area; } else { + +#ifdef CONFIG_PAX_PAGEEXEC + if ((mm->pax_flags & MF_PAX_PAGEEXEC) && text_fault && !(vma->vm_flags & VM_EXEC)) { + up_read(&mm->mmap_sem); + switch (pax_handle_fetch_fault(regs)) { + +#ifdef CONFIG_PAX_EMUPLT + case 2: + case 3: + return; +#endif + + } + pax_report_fault(regs, (void *)regs->pc, (void *)regs->u_regs[UREG_FP]); + do_group_exit(SIGKILL); + } +#endif + /* Allow reads even for write-only mappings */ if(!(vma->vm_flags & (VM_READ | VM_EXEC))) goto bad_area; diff --git a/arch/sparc/mm/fault_64.c b/arch/sparc/mm/fault_64.c index 504c062..6fcb9c6 100644 --- a/arch/sparc/mm/fault_64.c +++ b/arch/sparc/mm/fault_64.c @@ -21,6 +21,9 @@ #include #include #include +#include +#include +#include #include #include @@ -74,7 +77,7 @@ static void __kprobes bad_kernel_pc(struct pt_regs *regs, unsigned long vaddr) printk(KERN_CRIT "OOPS: Bogus kernel PC [%016lx] in fault handler\n", regs->tpc); printk(KERN_CRIT "OOPS: RPC [%016lx]\n", regs->u_regs[15]); - printk("OOPS: RPC <%pS>\n", (void *) regs->u_regs[15]); + printk("OOPS: RPC <%pA>\n", (void *) regs->u_regs[15]); printk(KERN_CRIT "OOPS: Fault was to vaddr[%lx]\n", vaddr); dump_stack(); unhandled_fault(regs->tpc, current, regs); @@ -272,6 +275,457 @@ static void noinline __kprobes bogus_32bit_fault_address(struct pt_regs *regs, show_regs(regs); } +#ifdef CONFIG_PAX_PAGEEXEC +#ifdef CONFIG_PAX_DLRESOLVE +static void pax_emuplt_close(struct vm_area_struct *vma) +{ + vma->vm_mm->call_dl_resolve = 0UL; +} + +static int pax_emuplt_fault(struct vm_area_struct *vma, struct vm_fault *vmf) +{ + unsigned int *kaddr; + + vmf->page = alloc_page(GFP_HIGHUSER); + if (!vmf->page) + return VM_FAULT_OOM; + + kaddr = kmap(vmf->page); + memset(kaddr, 0, PAGE_SIZE); + kaddr[0] = 0x9DE3BFA8U; /* save */ + flush_dcache_page(vmf->page); + kunmap(vmf->page); + return VM_FAULT_MAJOR; +} + +static const struct vm_operations_struct pax_vm_ops = { + .close = pax_emuplt_close, + .fault = pax_emuplt_fault +}; + +static int pax_insert_vma(struct vm_area_struct *vma, unsigned long addr) +{ + int ret; + + INIT_LIST_HEAD(&vma->anon_vma_chain); + vma->vm_mm = current->mm; + vma->vm_start = addr; + vma->vm_end = addr + PAGE_SIZE; + vma->vm_flags = VM_READ | VM_EXEC | VM_MAYREAD | VM_MAYEXEC; + vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); + vma->vm_ops = &pax_vm_ops; + + ret = insert_vm_struct(current->mm, vma); + if (ret) + return ret; + + ++current->mm->total_vm; + return 0; +} +#endif + +/* + * PaX: decide what to do with offenders (regs->tpc = fault address) + * + * returns 1 when task should be killed + * 2 when patched PLT trampoline was detected + * 3 when unpatched PLT trampoline was detected + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + +#ifdef CONFIG_PAX_EMUPLT + int err; + + do { /* PaX: patched PLT emulation #1 */ + unsigned int sethi1, sethi2, jmpl; + + err = get_user(sethi1, (unsigned int *)regs->tpc); + err |= get_user(sethi2, (unsigned int *)(regs->tpc+4)); + err |= get_user(jmpl, (unsigned int *)(regs->tpc+8)); + + if (err) + break; + + if ((sethi1 & 0xFFC00000U) == 0x03000000U && + (sethi2 & 0xFFC00000U) == 0x03000000U && + (jmpl & 0xFFFFE000U) == 0x81C06000U) + { + unsigned long addr; + + regs->u_regs[UREG_G1] = (sethi2 & 0x003FFFFFU) << 10; + addr = regs->u_regs[UREG_G1]; + addr += (((jmpl | 0xFFFFFFFFFFFFE000UL) ^ 0x00001000UL) + 0x00001000UL); + + if (test_thread_flag(TIF_32BIT)) + addr &= 0xFFFFFFFFUL; + + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + { /* PaX: patched PLT emulation #2 */ + unsigned int ba; + + err = get_user(ba, (unsigned int *)regs->tpc); + + if (!err && (ba & 0xFFC00000U) == 0x30800000U) { + unsigned long addr; + + addr = regs->tpc + ((((ba | 0xFFFFFFFFFFC00000UL) ^ 0x00200000UL) + 0x00200000UL) << 2); + + if (test_thread_flag(TIF_32BIT)) + addr &= 0xFFFFFFFFUL; + + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } + + do { /* PaX: patched PLT emulation #3 */ + unsigned int sethi, jmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(jmpl, (unsigned int *)(regs->tpc+4)); + err |= get_user(nop, (unsigned int *)(regs->tpc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + (jmpl & 0xFFFFE000U) == 0x81C06000U && + nop == 0x01000000U) + { + unsigned long addr; + + addr = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] = addr; + addr += (((jmpl | 0xFFFFFFFFFFFFE000UL) ^ 0x00001000UL) + 0x00001000UL); + + if (test_thread_flag(TIF_32BIT)) + addr &= 0xFFFFFFFFUL; + + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #4 */ + unsigned int sethi, mov1, call, mov2; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(mov1, (unsigned int *)(regs->tpc+4)); + err |= get_user(call, (unsigned int *)(regs->tpc+8)); + err |= get_user(mov2, (unsigned int *)(regs->tpc+12)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + mov1 == 0x8210000FU && + (call & 0xC0000000U) == 0x40000000U && + mov2 == 0x9E100001U) + { + unsigned long addr; + + regs->u_regs[UREG_G1] = regs->u_regs[UREG_RETPC]; + addr = regs->tpc + 4 + ((((call | 0xFFFFFFFFC0000000UL) ^ 0x20000000UL) + 0x20000000UL) << 2); + + if (test_thread_flag(TIF_32BIT)) + addr &= 0xFFFFFFFFUL; + + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #5 */ + unsigned int sethi, sethi1, sethi2, or1, or2, sllx, jmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(sethi1, (unsigned int *)(regs->tpc+4)); + err |= get_user(sethi2, (unsigned int *)(regs->tpc+8)); + err |= get_user(or1, (unsigned int *)(regs->tpc+12)); + err |= get_user(or2, (unsigned int *)(regs->tpc+16)); + err |= get_user(sllx, (unsigned int *)(regs->tpc+20)); + err |= get_user(jmpl, (unsigned int *)(regs->tpc+24)); + err |= get_user(nop, (unsigned int *)(regs->tpc+28)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + (sethi1 & 0xFFC00000U) == 0x03000000U && + (sethi2 & 0xFFC00000U) == 0x0B000000U && + (or1 & 0xFFFFE000U) == 0x82106000U && + (or2 & 0xFFFFE000U) == 0x8A116000U && + sllx == 0x83287020U && + jmpl == 0x81C04005U && + nop == 0x01000000U) + { + unsigned long addr; + + regs->u_regs[UREG_G1] = ((sethi1 & 0x003FFFFFU) << 10) | (or1 & 0x000003FFU); + regs->u_regs[UREG_G1] <<= 32; + regs->u_regs[UREG_G5] = ((sethi2 & 0x003FFFFFU) << 10) | (or2 & 0x000003FFU); + addr = regs->u_regs[UREG_G1] + regs->u_regs[UREG_G5]; + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #6 */ + unsigned int sethi, sethi1, sethi2, sllx, or, jmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(sethi1, (unsigned int *)(regs->tpc+4)); + err |= get_user(sethi2, (unsigned int *)(regs->tpc+8)); + err |= get_user(sllx, (unsigned int *)(regs->tpc+12)); + err |= get_user(or, (unsigned int *)(regs->tpc+16)); + err |= get_user(jmpl, (unsigned int *)(regs->tpc+20)); + err |= get_user(nop, (unsigned int *)(regs->tpc+24)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + (sethi1 & 0xFFC00000U) == 0x03000000U && + (sethi2 & 0xFFC00000U) == 0x0B000000U && + sllx == 0x83287020U && + (or & 0xFFFFE000U) == 0x8A116000U && + jmpl == 0x81C04005U && + nop == 0x01000000U) + { + unsigned long addr; + + regs->u_regs[UREG_G1] = (sethi1 & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] <<= 32; + regs->u_regs[UREG_G5] = ((sethi2 & 0x003FFFFFU) << 10) | (or & 0x3FFU); + addr = regs->u_regs[UREG_G1] + regs->u_regs[UREG_G5]; + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: unpatched PLT emulation step 1 */ + unsigned int sethi, ba, nop; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(ba, (unsigned int *)(regs->tpc+4)); + err |= get_user(nop, (unsigned int *)(regs->tpc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + ((ba & 0xFFC00000U) == 0x30800000U || (ba & 0xFFF80000U) == 0x30680000U) && + nop == 0x01000000U) + { + unsigned long addr; + unsigned int save, call; + unsigned int sethi1, sethi2, or1, or2, sllx, add, jmpl; + + if ((ba & 0xFFC00000U) == 0x30800000U) + addr = regs->tpc + 4 + ((((ba | 0xFFFFFFFFFFC00000UL) ^ 0x00200000UL) + 0x00200000UL) << 2); + else + addr = regs->tpc + 4 + ((((ba | 0xFFFFFFFFFFF80000UL) ^ 0x00040000UL) + 0x00040000UL) << 2); + + if (test_thread_flag(TIF_32BIT)) + addr &= 0xFFFFFFFFUL; + + err = get_user(save, (unsigned int *)addr); + err |= get_user(call, (unsigned int *)(addr+4)); + err |= get_user(nop, (unsigned int *)(addr+8)); + if (err) + break; + +#ifdef CONFIG_PAX_DLRESOLVE + if (save == 0x9DE3BFA8U && + (call & 0xC0000000U) == 0x40000000U && + nop == 0x01000000U) + { + struct vm_area_struct *vma; + unsigned long call_dl_resolve; + + down_read(¤t->mm->mmap_sem); + call_dl_resolve = current->mm->call_dl_resolve; + up_read(¤t->mm->mmap_sem); + if (likely(call_dl_resolve)) + goto emulate; + + vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); + + down_write(¤t->mm->mmap_sem); + if (current->mm->call_dl_resolve) { + call_dl_resolve = current->mm->call_dl_resolve; + up_write(¤t->mm->mmap_sem); + if (vma) + kmem_cache_free(vm_area_cachep, vma); + goto emulate; + } + + call_dl_resolve = get_unmapped_area(NULL, 0UL, PAGE_SIZE, 0UL, MAP_PRIVATE); + if (!vma || (call_dl_resolve & ~PAGE_MASK)) { + up_write(¤t->mm->mmap_sem); + if (vma) + kmem_cache_free(vm_area_cachep, vma); + return 1; + } + + if (pax_insert_vma(vma, call_dl_resolve)) { + up_write(¤t->mm->mmap_sem); + kmem_cache_free(vm_area_cachep, vma); + return 1; + } + + current->mm->call_dl_resolve = call_dl_resolve; + up_write(¤t->mm->mmap_sem); + +emulate: + regs->u_regs[UREG_G1] = (sethi & 0x003FFFFFU) << 10; + regs->tpc = call_dl_resolve; + regs->tnpc = addr+4; + return 3; + } +#endif + + /* PaX: glibc 2.4+ generates sethi/jmpl instead of save/call */ + if ((save & 0xFFC00000U) == 0x05000000U && + (call & 0xFFFFE000U) == 0x85C0A000U && + nop == 0x01000000U) + { + regs->u_regs[UREG_G1] = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G2] = addr + 4; + addr = (save & 0x003FFFFFU) << 10; + addr += (((call | 0xFFFFFFFFFFFFE000UL) ^ 0x00001000UL) + 0x00001000UL); + + if (test_thread_flag(TIF_32BIT)) + addr &= 0xFFFFFFFFUL; + + regs->tpc = addr; + regs->tnpc = addr+4; + return 3; + } + + /* PaX: 64-bit PLT stub */ + err = get_user(sethi1, (unsigned int *)addr); + err |= get_user(sethi2, (unsigned int *)(addr+4)); + err |= get_user(or1, (unsigned int *)(addr+8)); + err |= get_user(or2, (unsigned int *)(addr+12)); + err |= get_user(sllx, (unsigned int *)(addr+16)); + err |= get_user(add, (unsigned int *)(addr+20)); + err |= get_user(jmpl, (unsigned int *)(addr+24)); + err |= get_user(nop, (unsigned int *)(addr+28)); + if (err) + break; + + if ((sethi1 & 0xFFC00000U) == 0x09000000U && + (sethi2 & 0xFFC00000U) == 0x0B000000U && + (or1 & 0xFFFFE000U) == 0x88112000U && + (or2 & 0xFFFFE000U) == 0x8A116000U && + sllx == 0x89293020U && + add == 0x8A010005U && + jmpl == 0x89C14000U && + nop == 0x01000000U) + { + regs->u_regs[UREG_G1] = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G4] = ((sethi1 & 0x003FFFFFU) << 10) | (or1 & 0x000003FFU); + regs->u_regs[UREG_G4] <<= 32; + regs->u_regs[UREG_G5] = ((sethi2 & 0x003FFFFFU) << 10) | (or2 & 0x000003FFU); + regs->u_regs[UREG_G5] += regs->u_regs[UREG_G4]; + regs->u_regs[UREG_G4] = addr + 24; + addr = regs->u_regs[UREG_G5]; + regs->tpc = addr; + regs->tnpc = addr+4; + return 3; + } + } + } while (0); + +#ifdef CONFIG_PAX_DLRESOLVE + do { /* PaX: unpatched PLT emulation step 2 */ + unsigned int save, call, nop; + + err = get_user(save, (unsigned int *)(regs->tpc-4)); + err |= get_user(call, (unsigned int *)regs->tpc); + err |= get_user(nop, (unsigned int *)(regs->tpc+4)); + if (err) + break; + + if (save == 0x9DE3BFA8U && + (call & 0xC0000000U) == 0x40000000U && + nop == 0x01000000U) + { + unsigned long dl_resolve = regs->tpc + ((((call | 0xFFFFFFFFC0000000UL) ^ 0x20000000UL) + 0x20000000UL) << 2); + + if (test_thread_flag(TIF_32BIT)) + dl_resolve &= 0xFFFFFFFFUL; + + regs->u_regs[UREG_RETPC] = regs->tpc; + regs->tpc = dl_resolve; + regs->tnpc = dl_resolve+4; + return 3; + } + } while (0); +#endif + + do { /* PaX: patched PLT emulation #7, must be AFTER the unpatched PLT emulation */ + unsigned int sethi, ba, nop; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(ba, (unsigned int *)(regs->tpc+4)); + err |= get_user(nop, (unsigned int *)(regs->tpc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + (ba & 0xFFF00000U) == 0x30600000U && + nop == 0x01000000U) + { + unsigned long addr; + + addr = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] = addr; + addr = regs->tpc + ((((ba | 0xFFFFFFFFFFF80000UL) ^ 0x00040000UL) + 0x00040000UL) << 2); + + if (test_thread_flag(TIF_32BIT)) + addr &= 0xFFFFFFFFUL; + + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + +#endif + + return 1; +} + +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 8; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs) { struct mm_struct *mm = current->mm; @@ -340,6 +794,29 @@ asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs) if (!vma) goto bad_area; +#ifdef CONFIG_PAX_PAGEEXEC + /* PaX: detect ITLB misses on non-exec pages */ + if ((mm->pax_flags & MF_PAX_PAGEEXEC) && vma->vm_start <= address && + !(vma->vm_flags & VM_EXEC) && (fault_code & FAULT_CODE_ITLB)) + { + if (address != regs->tpc) + goto good_area; + + up_read(&mm->mmap_sem); + switch (pax_handle_fetch_fault(regs)) { + +#ifdef CONFIG_PAX_EMUPLT + case 2: + case 3: + return; +#endif + + } + pax_report_fault(regs, (void *)regs->tpc, (void *)(regs->u_regs[UREG_FP] + STACK_BIAS)); + do_group_exit(SIGKILL); + } +#endif + /* Pure DTLB misses do not tell us whether the fault causing * load/store/atomic was a write or not, it only says that there * was no match. So in such a case we (carefully) read the diff --git a/arch/sparc/mm/hugetlbpage.c b/arch/sparc/mm/hugetlbpage.c index 07e1453..0a7d9e9 100644 --- a/arch/sparc/mm/hugetlbpage.c +++ b/arch/sparc/mm/hugetlbpage.c @@ -67,7 +67,7 @@ full_search: } return -ENOMEM; } - if (likely(!vma || addr + len <= vma->vm_start)) { + if (likely(check_heap_stack_gap(vma, addr, len))) { /* * Remember the place where we stopped the search: */ @@ -106,7 +106,7 @@ hugetlb_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, /* make sure it can fit in the remaining address space */ if (likely(addr > len)) { vma = find_vma(mm, addr-len); - if (!vma || addr <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr - len, len)) { /* remember the address as a hint for next time */ return (mm->free_area_cache = addr-len); } @@ -115,16 +115,17 @@ hugetlb_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (unlikely(mm->mmap_base < len)) goto bottomup; - addr = (mm->mmap_base-len) & HPAGE_MASK; + addr = mm->mmap_base - len; do { + addr &= HPAGE_MASK; /* * Lookup failure means no vma is above this address, * else if new region fits below vma->vm_start, * return with success: */ vma = find_vma(mm, addr); - if (likely(!vma || addr+len <= vma->vm_start)) { + if (likely(check_heap_stack_gap(vma, addr, len))) { /* remember the address as a hint for next time */ return (mm->free_area_cache = addr); } @@ -134,8 +135,8 @@ hugetlb_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, mm->cached_hole_size = vma->vm_start - addr; /* try just below the current vma->vm_start */ - addr = (vma->vm_start-len) & HPAGE_MASK; - } while (likely(len < vma->vm_start)); + addr = skip_heap_stack_gap(vma, len); + } while (!IS_ERR_VALUE(addr)); bottomup: /* @@ -181,8 +182,7 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, if (addr) { addr = ALIGN(addr, HPAGE_SIZE); vma = find_vma(mm, addr); - if (task_size - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (task_size - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } if (mm->get_unmapped_area == arch_get_unmapped_area) diff --git a/arch/sparc/mm/init_32.c b/arch/sparc/mm/init_32.c index 7b00de6..78239f4 100644 --- a/arch/sparc/mm/init_32.c +++ b/arch/sparc/mm/init_32.c @@ -316,6 +316,9 @@ extern void device_scan(void); pgprot_t PAGE_SHARED __read_mostly; EXPORT_SYMBOL(PAGE_SHARED); +pgprot_t PAGE_SHARED_NOEXEC __read_mostly; +EXPORT_SYMBOL(PAGE_SHARED_NOEXEC); + void __init paging_init(void) { switch(sparc_cpu_model) { @@ -344,17 +347,17 @@ void __init paging_init(void) /* Initialize the protection map with non-constant, MMU dependent values. */ protection_map[0] = PAGE_NONE; - protection_map[1] = PAGE_READONLY; - protection_map[2] = PAGE_COPY; - protection_map[3] = PAGE_COPY; + protection_map[1] = PAGE_READONLY_NOEXEC; + protection_map[2] = PAGE_COPY_NOEXEC; + protection_map[3] = PAGE_COPY_NOEXEC; protection_map[4] = PAGE_READONLY; protection_map[5] = PAGE_READONLY; protection_map[6] = PAGE_COPY; protection_map[7] = PAGE_COPY; protection_map[8] = PAGE_NONE; - protection_map[9] = PAGE_READONLY; - protection_map[10] = PAGE_SHARED; - protection_map[11] = PAGE_SHARED; + protection_map[9] = PAGE_READONLY_NOEXEC; + protection_map[10] = PAGE_SHARED_NOEXEC; + protection_map[11] = PAGE_SHARED_NOEXEC; protection_map[12] = PAGE_READONLY; protection_map[13] = PAGE_READONLY; protection_map[14] = PAGE_SHARED; diff --git a/arch/sparc/mm/srmmu.c b/arch/sparc/mm/srmmu.c index cbef74e..c38fead 100644 --- a/arch/sparc/mm/srmmu.c +++ b/arch/sparc/mm/srmmu.c @@ -2200,6 +2200,13 @@ void __init ld_mmu_srmmu(void) PAGE_SHARED = pgprot_val(SRMMU_PAGE_SHARED); BTFIXUPSET_INT(page_copy, pgprot_val(SRMMU_PAGE_COPY)); BTFIXUPSET_INT(page_readonly, pgprot_val(SRMMU_PAGE_RDONLY)); + +#ifdef CONFIG_PAX_PAGEEXEC + PAGE_SHARED_NOEXEC = pgprot_val(SRMMU_PAGE_SHARED_NOEXEC); + BTFIXUPSET_INT(page_copy_noexec, pgprot_val(SRMMU_PAGE_COPY_NOEXEC)); + BTFIXUPSET_INT(page_readonly_noexec, pgprot_val(SRMMU_PAGE_RDONLY_NOEXEC)); +#endif + BTFIXUPSET_INT(page_kernel, pgprot_val(SRMMU_PAGE_KERNEL)); page_kernel = pgprot_val(SRMMU_PAGE_KERNEL); diff --git a/arch/tile/include/asm/atomic_64.h b/arch/tile/include/asm/atomic_64.h index 27fe667..36d474c 100644 --- a/arch/tile/include/asm/atomic_64.h +++ b/arch/tile/include/asm/atomic_64.h @@ -142,6 +142,16 @@ static inline long atomic64_add_unless(atomic64_t *v, long a, long u) #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) +#define atomic64_read_unchecked(v) atomic64_read(v) +#define atomic64_set_unchecked(v, i) atomic64_set((v), (i)) +#define atomic64_add_unchecked(a, v) atomic64_add((a), (v)) +#define atomic64_add_return_unchecked(a, v) atomic64_add_return((a), (v)) +#define atomic64_sub_unchecked(a, v) atomic64_sub((a), (v)) +#define atomic64_inc_unchecked(v) atomic64_inc(v) +#define atomic64_inc_return_unchecked(v) atomic64_inc_return(v) +#define atomic64_dec_unchecked(v) atomic64_dec(v) +#define atomic64_cmpxchg_unchecked(v, o, n) atomic64_cmpxchg((v), (o), (n)) + /* Atomic dec and inc don't implement barrier, so provide them if needed. */ #define smp_mb__before_atomic_dec() smp_mb() #define smp_mb__after_atomic_dec() smp_mb() diff --git a/arch/tile/include/asm/cache.h b/arch/tile/include/asm/cache.h index 392e533..536b092 100644 --- a/arch/tile/include/asm/cache.h +++ b/arch/tile/include/asm/cache.h @@ -15,11 +15,12 @@ #ifndef _ASM_TILE_CACHE_H #define _ASM_TILE_CACHE_H +#include #include /* bytes per L1 data cache line */ #define L1_CACHE_SHIFT CHIP_L1D_LOG_LINE_SIZE() -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) /* bytes per L2 cache line */ #define L2_CACHE_SHIFT CHIP_L2_LOG_LINE_SIZE() diff --git a/arch/um/Makefile b/arch/um/Makefile index 28688e6..4c0aa1c 100644 --- a/arch/um/Makefile +++ b/arch/um/Makefile @@ -61,6 +61,10 @@ USER_CFLAGS = $(patsubst $(KERNEL_DEFINES),,$(patsubst -D__KERNEL__,,\ $(patsubst -I%,,$(KBUILD_CFLAGS)))) $(ARCH_INCLUDE) $(MODE_INCLUDE) \ $(filter -I%,$(CFLAGS)) -D_FILE_OFFSET_BITS=64 -idirafter include +ifdef CONSTIFY_PLUGIN +USER_CFLAGS += $(CONSTIFY_PLUGIN) -fplugin-arg-constify_plugin-no-constify +endif + #This will adjust *FLAGS accordingly to the platform. include $(srctree)/$(ARCH_DIR)/Makefile-os-$(OS) diff --git a/arch/um/include/asm/cache.h b/arch/um/include/asm/cache.h index 19e1bdd..3665b77 100644 --- a/arch/um/include/asm/cache.h +++ b/arch/um/include/asm/cache.h @@ -1,6 +1,7 @@ #ifndef __UM_CACHE_H #define __UM_CACHE_H +#include #if defined(CONFIG_UML_X86) && !defined(CONFIG_64BIT) # define L1_CACHE_SHIFT (CONFIG_X86_L1_CACHE_SHIFT) @@ -12,6 +13,6 @@ # define L1_CACHE_SHIFT 5 #endif -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #endif diff --git a/arch/um/include/asm/kmap_types.h b/arch/um/include/asm/kmap_types.h index 6c03acd..a5e0215 100644 --- a/arch/um/include/asm/kmap_types.h +++ b/arch/um/include/asm/kmap_types.h @@ -23,6 +23,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff --git a/arch/um/include/asm/page.h b/arch/um/include/asm/page.h index 7cfc3ce..cbd1a58 100644 --- a/arch/um/include/asm/page.h +++ b/arch/um/include/asm/page.h @@ -14,6 +14,9 @@ #define PAGE_SIZE (_AC(1, UL) << PAGE_SHIFT) #define PAGE_MASK (~(PAGE_SIZE-1)) +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) + #ifndef __ASSEMBLY__ struct page; diff --git a/arch/um/include/asm/pgtable-3level.h b/arch/um/include/asm/pgtable-3level.h index 0032f92..cd151e0 100644 --- a/arch/um/include/asm/pgtable-3level.h +++ b/arch/um/include/asm/pgtable-3level.h @@ -58,6 +58,7 @@ #define pud_present(x) (pud_val(x) & _PAGE_PRESENT) #define pud_populate(mm, pud, pmd) \ set_pud(pud, __pud(_PAGE_TABLE + __pa(pmd))) +#define pud_populate_kernel(mm, pud, pmd) pud_populate((mm), (pud), (pmd)) #ifdef CONFIG_64BIT #define set_pud(pudptr, pudval) set_64bit((u64 *) (pudptr), pud_val(pudval)) diff --git a/arch/um/kernel/process.c b/arch/um/kernel/process.c index 69f2490..2634831 100644 --- a/arch/um/kernel/process.c +++ b/arch/um/kernel/process.c @@ -408,22 +408,6 @@ int singlestepping(void * t) return 2; } -/* - * Only x86 and x86_64 have an arch_align_stack(). - * All other arches have "#define arch_align_stack(x) (x)" - * in their asm/system.h - * As this is included in UML from asm-um/system-generic.h, - * we can use it to behave as the subarch does. - */ -#ifndef arch_align_stack -unsigned long arch_align_stack(unsigned long sp) -{ - if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) - sp -= get_random_int() % 8192; - return sp & ~0xf; -} -#endif - unsigned long get_wchan(struct task_struct *p) { unsigned long stack_page, sp, ip; diff --git a/arch/unicore32/include/asm/cache.h b/arch/unicore32/include/asm/cache.h index ad8f795..2c7eec6 100644 --- a/arch/unicore32/include/asm/cache.h +++ b/arch/unicore32/include/asm/cache.h @@ -12,8 +12,10 @@ #ifndef __UNICORE_CACHE_H__ #define __UNICORE_CACHE_H__ -#define L1_CACHE_SHIFT (5) -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#include + +#define L1_CACHE_SHIFT 5 +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) /* * Memory returned by kmalloc() may be used for DMA, so we must make diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 5bed94e..fbcf200 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -226,7 +226,7 @@ config X86_HT config X86_32_LAZY_GS def_bool y - depends on X86_32 && !CC_STACKPROTECTOR + depends on X86_32 && !CC_STACKPROTECTOR && !PAX_MEMORY_UDEREF config ARCH_HWEIGHT_CFLAGS string @@ -1058,7 +1058,7 @@ choice config NOHIGHMEM bool "off" - depends on !X86_NUMAQ + depends on !X86_NUMAQ && !(PAX_PAGEEXEC && PAX_ENABLE_PAE) ---help--- Linux can use up to 64 Gigabytes of physical memory on x86 systems. However, the address space of 32-bit x86 processors is only 4 @@ -1095,7 +1095,7 @@ config NOHIGHMEM config HIGHMEM4G bool "4GB" - depends on !X86_NUMAQ + depends on !X86_NUMAQ && !(PAX_PAGEEXEC && PAX_ENABLE_PAE) ---help--- Select this if you have a 32-bit processor and between 1 and 4 gigabytes of physical RAM. @@ -1149,7 +1149,7 @@ config PAGE_OFFSET hex default 0xB0000000 if VMSPLIT_3G_OPT default 0x80000000 if VMSPLIT_2G - default 0x78000000 if VMSPLIT_2G_OPT + default 0x70000000 if VMSPLIT_2G_OPT default 0x40000000 if VMSPLIT_1G default 0xC0000000 depends on X86_32 @@ -1539,6 +1539,7 @@ config SECCOMP config CC_STACKPROTECTOR bool "Enable -fstack-protector buffer overflow detection (EXPERIMENTAL)" + depends on X86_64 || !PAX_MEMORY_UDEREF ---help--- This option turns on the -fstack-protector GCC feature. This feature puts, at the beginning of functions, a canary value on @@ -1596,6 +1597,7 @@ config KEXEC_JUMP config PHYSICAL_START hex "Physical address where the kernel is loaded" if (EXPERT || CRASH_DUMP) default "0x1000000" + range 0x400000 0x40000000 ---help--- This gives the physical address where the kernel is loaded. @@ -1659,6 +1661,7 @@ config X86_NEED_RELOCS config PHYSICAL_ALIGN hex "Alignment value to which kernel should be aligned" if X86_32 default "0x1000000" + range 0x400000 0x1000000 if PAX_KERNEXEC range 0x2000 0x1000000 ---help--- This value puts the alignment restrictions on physical address @@ -1690,9 +1693,10 @@ config HOTPLUG_CPU Say N if you want to disable CPU hotplug. config COMPAT_VDSO - def_bool y + def_bool n prompt "Compat VDSO support" depends on X86_32 || IA32_EMULATION + depends on !PAX_NOEXEC && !PAX_MEMORY_UDEREF ---help--- Map the 32-bit VDSO to the predictable old-style address too. diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index 3c57033..22d44aa 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -335,7 +335,7 @@ config X86_PPRO_FENCE config X86_F00F_BUG def_bool y - depends on M586MMX || M586TSC || M586 || M486 || M386 + depends on (M586MMX || M586TSC || M586 || M486 || M386) && !PAX_KERNEXEC config X86_INVD_BUG def_bool y @@ -359,7 +359,7 @@ config X86_POPAD_OK config X86_ALIGNMENT_16 def_bool y - depends on MWINCHIP3D || MWINCHIPC6 || MCYRIXIII || MELAN || MK6 || M586MMX || M586TSC || M586 || M486 || MVIAC3_2 || MGEODEGX1 + depends on MWINCHIP3D || MWINCHIPC6 || MCYRIXIII || X86_ELAN || MK8 || MK7 || MK6 || MCORE2 || MPENTIUM4 || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || M586 || M486 || MVIAC3_2 || MGEODEGX1 config X86_INTEL_USERCOPY def_bool y @@ -405,7 +405,7 @@ config X86_CMPXCHG64 # generates cmov. config X86_CMOV def_bool y - depends on (MK8 || MK7 || MCORE2 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || MVIAC3_2 || MVIAC7 || MCRUSOE || MEFFICEON || X86_64 || MATOM || MGEODE_LX) + depends on (MK8 || MK7 || MCORE2 || MPSC || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || MVIAC3_2 || MVIAC7 || MCRUSOE || MEFFICEON || X86_64 || MATOM || MGEODE_LX) config X86_MINIMUM_CPU_FAMILY int diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug index e46c214..7c72b55 100644 --- a/arch/x86/Kconfig.debug +++ b/arch/x86/Kconfig.debug @@ -84,7 +84,7 @@ config X86_PTDUMP config DEBUG_RODATA bool "Write protect kernel read-only data structures" default y - depends on DEBUG_KERNEL + depends on DEBUG_KERNEL && BROKEN ---help--- Mark the kernel read-only data as write-protected in the pagetables, in order to catch accidental (and incorrect) writes to such const @@ -102,7 +102,7 @@ config DEBUG_RODATA_TEST config DEBUG_SET_MODULE_RONX bool "Set loadable kernel module data as NX and text as RO" - depends on MODULES + depends on MODULES && BROKEN ---help--- This option helps catch unintended modifications to loadable kernel module's text and read-only data. It also prevents execution diff --git a/arch/x86/Makefile b/arch/x86/Makefile index 209ba12..15140db 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -46,6 +46,7 @@ else UTS_MACHINE := x86_64 CHECKFLAGS += -D__x86_64__ -m64 + biarch := $(call cc-option,-m64) KBUILD_AFLAGS += -m64 KBUILD_CFLAGS += -m64 @@ -201,3 +202,12 @@ define archhelp echo ' FDARGS="..." arguments for the booted kernel' echo ' FDINITRD=file initrd for the booted kernel' endef + +define OLD_LD + +*** ${VERSION}.${PATCHLEVEL} PaX kernels no longer build correctly with old versions of binutils. +*** Please upgrade your binutils to 2.18 or newer +endef + +archprepare: + $(if $(LDFLAGS_BUILD_ID),,$(error $(OLD_LD))) diff --git a/arch/x86/boot/Makefile b/arch/x86/boot/Makefile index 5a747dd..ff7b12c 100644 --- a/arch/x86/boot/Makefile +++ b/arch/x86/boot/Makefile @@ -64,6 +64,9 @@ KBUILD_CFLAGS := $(LINUXINCLUDE) -g -Os -D_SETUP -D__KERNEL__ \ $(call cc-option, -fno-stack-protector) \ $(call cc-option, -mpreferred-stack-boundary=2) KBUILD_CFLAGS += $(call cc-option, -m32) +ifdef CONSTIFY_PLUGIN +KBUILD_CFLAGS += $(CONSTIFY_PLUGIN) -fplugin-arg-constify_plugin-no-constify +endif KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ GCOV_PROFILE := n diff --git a/arch/x86/boot/bitops.h b/arch/x86/boot/bitops.h index 878e4b9..20537ab 100644 --- a/arch/x86/boot/bitops.h +++ b/arch/x86/boot/bitops.h @@ -26,7 +26,7 @@ static inline int variable_test_bit(int nr, const void *addr) u8 v; const u32 *p = (const u32 *)addr; - asm("btl %2,%1; setc %0" : "=qm" (v) : "m" (*p), "Ir" (nr)); + asm volatile("btl %2,%1; setc %0" : "=qm" (v) : "m" (*p), "Ir" (nr)); return v; } @@ -37,7 +37,7 @@ static inline int variable_test_bit(int nr, const void *addr) static inline void set_bit(int nr, void *addr) { - asm("btsl %1,%0" : "+m" (*(u32 *)addr) : "Ir" (nr)); + asm volatile("btsl %1,%0" : "+m" (*(u32 *)addr) : "Ir" (nr)); } #endif /* BOOT_BITOPS_H */ diff --git a/arch/x86/boot/boot.h b/arch/x86/boot/boot.h index c7093bd..d4247ffe0 100644 --- a/arch/x86/boot/boot.h +++ b/arch/x86/boot/boot.h @@ -85,7 +85,7 @@ static inline void io_delay(void) static inline u16 ds(void) { u16 seg; - asm("movw %%ds,%0" : "=rm" (seg)); + asm volatile("movw %%ds,%0" : "=rm" (seg)); return seg; } @@ -181,7 +181,7 @@ static inline void wrgs32(u32 v, addr_t addr) static inline int memcmp(const void *s1, const void *s2, size_t len) { u8 diff; - asm("repe; cmpsb; setnz %0" + asm volatile("repe; cmpsb; setnz %0" : "=qm" (diff), "+D" (s1), "+S" (s2), "+c" (len)); return diff; } diff --git a/arch/x86/boot/compressed/Makefile b/arch/x86/boot/compressed/Makefile index fd55a2f..217b501 100644 --- a/arch/x86/boot/compressed/Makefile +++ b/arch/x86/boot/compressed/Makefile @@ -14,6 +14,9 @@ cflags-$(CONFIG_X86_64) := -mcmodel=small KBUILD_CFLAGS += $(cflags-y) KBUILD_CFLAGS += $(call cc-option,-ffreestanding) KBUILD_CFLAGS += $(call cc-option,-fno-stack-protector) +ifdef CONSTIFY_PLUGIN +KBUILD_CFLAGS += $(CONSTIFY_PLUGIN) -fplugin-arg-constify_plugin-no-constify +endif KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ GCOV_PROFILE := n diff --git a/arch/x86/boot/compressed/head_32.S b/arch/x86/boot/compressed/head_32.S index c85e3ac..6f5aa80 100644 --- a/arch/x86/boot/compressed/head_32.S +++ b/arch/x86/boot/compressed/head_32.S @@ -106,7 +106,7 @@ preferred_addr: notl %eax andl %eax, %ebx #else - movl $LOAD_PHYSICAL_ADDR, %ebx + movl $____LOAD_PHYSICAL_ADDR, %ebx #endif /* Target address to relocate to for decompression */ @@ -192,7 +192,7 @@ relocated: * and where it was actually loaded. */ movl %ebp, %ebx - subl $LOAD_PHYSICAL_ADDR, %ebx + subl $____LOAD_PHYSICAL_ADDR, %ebx jz 2f /* Nothing to be done if loaded at compiled addr. */ /* * Process relocations. @@ -200,8 +200,7 @@ relocated: 1: subl $4, %edi movl (%edi), %ecx - testl %ecx, %ecx - jz 2f + jecxz 2f addl %ebx, -__PAGE_OFFSET(%ebx, %ecx) jmp 1b 2: diff --git a/arch/x86/boot/compressed/head_64.S b/arch/x86/boot/compressed/head_64.S index 87e03a1..0d94c76 100644 --- a/arch/x86/boot/compressed/head_64.S +++ b/arch/x86/boot/compressed/head_64.S @@ -91,7 +91,7 @@ ENTRY(startup_32) notl %eax andl %eax, %ebx #else - movl $LOAD_PHYSICAL_ADDR, %ebx + movl $____LOAD_PHYSICAL_ADDR, %ebx #endif /* Target address to relocate to for decompression */ @@ -263,7 +263,7 @@ preferred_addr: notq %rax andq %rax, %rbp #else - movq $LOAD_PHYSICAL_ADDR, %rbp + movq $____LOAD_PHYSICAL_ADDR, %rbp #endif /* Target address to relocate to for decompression */ diff --git a/arch/x86/boot/compressed/misc.c b/arch/x86/boot/compressed/misc.c index 7116dcb..d9ae1d7 100644 --- a/arch/x86/boot/compressed/misc.c +++ b/arch/x86/boot/compressed/misc.c @@ -310,7 +310,7 @@ static void parse_elf(void *output) case PT_LOAD: #ifdef CONFIG_RELOCATABLE dest = output; - dest += (phdr->p_paddr - LOAD_PHYSICAL_ADDR); + dest += (phdr->p_paddr - ____LOAD_PHYSICAL_ADDR); #else dest = (void *)(phdr->p_paddr); #endif @@ -365,7 +365,7 @@ asmlinkage void decompress_kernel(void *rmode, memptr heap, error("Destination address too large"); #endif #ifndef CONFIG_RELOCATABLE - if ((unsigned long)output != LOAD_PHYSICAL_ADDR) + if ((unsigned long)output != ____LOAD_PHYSICAL_ADDR) error("Wrong destination address"); #endif diff --git a/arch/x86/boot/compressed/relocs.c b/arch/x86/boot/compressed/relocs.c index e77f4e4..17e511f 100644 --- a/arch/x86/boot/compressed/relocs.c +++ b/arch/x86/boot/compressed/relocs.c @@ -13,8 +13,11 @@ static void die(char *fmt, ...); +#include "../../../../include/generated/autoconf.h" + #define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0])) static Elf32_Ehdr ehdr; +static Elf32_Phdr *phdr; static unsigned long reloc_count, reloc_idx; static unsigned long *relocs; @@ -270,9 +273,39 @@ static void read_ehdr(FILE *fp) } } +static void read_phdrs(FILE *fp) +{ + unsigned int i; + + phdr = calloc(ehdr.e_phnum, sizeof(Elf32_Phdr)); + if (!phdr) { + die("Unable to allocate %d program headers\n", + ehdr.e_phnum); + } + if (fseek(fp, ehdr.e_phoff, SEEK_SET) < 0) { + die("Seek to %d failed: %s\n", + ehdr.e_phoff, strerror(errno)); + } + if (fread(phdr, sizeof(*phdr), ehdr.e_phnum, fp) != ehdr.e_phnum) { + die("Cannot read ELF program headers: %s\n", + strerror(errno)); + } + for(i = 0; i < ehdr.e_phnum; i++) { + phdr[i].p_type = elf32_to_cpu(phdr[i].p_type); + phdr[i].p_offset = elf32_to_cpu(phdr[i].p_offset); + phdr[i].p_vaddr = elf32_to_cpu(phdr[i].p_vaddr); + phdr[i].p_paddr = elf32_to_cpu(phdr[i].p_paddr); + phdr[i].p_filesz = elf32_to_cpu(phdr[i].p_filesz); + phdr[i].p_memsz = elf32_to_cpu(phdr[i].p_memsz); + phdr[i].p_flags = elf32_to_cpu(phdr[i].p_flags); + phdr[i].p_align = elf32_to_cpu(phdr[i].p_align); + } + +} + static void read_shdrs(FILE *fp) { - int i; + unsigned int i; Elf32_Shdr shdr; secs = calloc(ehdr.e_shnum, sizeof(struct section)); @@ -307,7 +340,7 @@ static void read_shdrs(FILE *fp) static void read_strtabs(FILE *fp) { - int i; + unsigned int i; for (i = 0; i < ehdr.e_shnum; i++) { struct section *sec = &secs[i]; if (sec->shdr.sh_type != SHT_STRTAB) { @@ -332,7 +365,7 @@ static void read_strtabs(FILE *fp) static void read_symtabs(FILE *fp) { - int i,j; + unsigned int i,j; for (i = 0; i < ehdr.e_shnum; i++) { struct section *sec = &secs[i]; if (sec->shdr.sh_type != SHT_SYMTAB) { @@ -365,7 +398,9 @@ static void read_symtabs(FILE *fp) static void read_relocs(FILE *fp) { - int i,j; + unsigned int i,j; + uint32_t base; + for (i = 0; i < ehdr.e_shnum; i++) { struct section *sec = &secs[i]; if (sec->shdr.sh_type != SHT_REL) { @@ -385,9 +420,18 @@ static void read_relocs(FILE *fp) die("Cannot read symbol table: %s\n", strerror(errno)); } + base = 0; + for (j = 0; j < ehdr.e_phnum; j++) { + if (phdr[j].p_type != PT_LOAD ) + continue; + if (secs[sec->shdr.sh_info].shdr.sh_offset < phdr[j].p_offset || secs[sec->shdr.sh_info].shdr.sh_offset >= phdr[j].p_offset + phdr[j].p_filesz) + continue; + base = CONFIG_PAGE_OFFSET + phdr[j].p_paddr - phdr[j].p_vaddr; + break; + } for (j = 0; j < sec->shdr.sh_size/sizeof(Elf32_Rel); j++) { Elf32_Rel *rel = &sec->reltab[j]; - rel->r_offset = elf32_to_cpu(rel->r_offset); + rel->r_offset = elf32_to_cpu(rel->r_offset) + base; rel->r_info = elf32_to_cpu(rel->r_info); } } @@ -396,13 +440,13 @@ static void read_relocs(FILE *fp) static void print_absolute_symbols(void) { - int i; + unsigned int i; printf("Absolute symbols\n"); printf(" Num: Value Size Type Bind Visibility Name\n"); for (i = 0; i < ehdr.e_shnum; i++) { struct section *sec = &secs[i]; char *sym_strtab; - int j; + unsigned int j; if (sec->shdr.sh_type != SHT_SYMTAB) { continue; @@ -429,14 +473,14 @@ static void print_absolute_symbols(void) static void print_absolute_relocs(void) { - int i, printed = 0; + unsigned int i, printed = 0; for (i = 0; i < ehdr.e_shnum; i++) { struct section *sec = &secs[i]; struct section *sec_applies, *sec_symtab; char *sym_strtab; Elf32_Sym *sh_symtab; - int j; + unsigned int j; if (sec->shdr.sh_type != SHT_REL) { continue; } @@ -497,13 +541,13 @@ static void print_absolute_relocs(void) static void walk_relocs(void (*visit)(Elf32_Rel *rel, Elf32_Sym *sym)) { - int i; + unsigned int i; /* Walk through the relocations */ for (i = 0; i < ehdr.e_shnum; i++) { char *sym_strtab; Elf32_Sym *sh_symtab; struct section *sec_applies, *sec_symtab; - int j; + unsigned int j; struct section *sec = &secs[i]; if (sec->shdr.sh_type != SHT_REL) { @@ -528,6 +572,22 @@ static void walk_relocs(void (*visit)(Elf32_Rel *rel, Elf32_Sym *sym)) !is_rel_reloc(sym_name(sym_strtab, sym))) { continue; } + /* Don't relocate actual per-cpu variables, they are absolute indices, not addresses */ + if (!strcmp(sec_name(sym->st_shndx), ".data..percpu") && strcmp(sym_name(sym_strtab, sym), "__per_cpu_load")) + continue; + +#if defined(CONFIG_PAX_KERNEXEC) && defined(CONFIG_X86_32) + /* Don't relocate actual code, they are relocated implicitly by the base address of KERNEL_CS */ + if (!strcmp(sec_name(sym->st_shndx), ".module.text") && !strcmp(sym_name(sym_strtab, sym), "_etext")) + continue; + if (!strcmp(sec_name(sym->st_shndx), ".init.text")) + continue; + if (!strcmp(sec_name(sym->st_shndx), ".exit.text")) + continue; + if (!strcmp(sec_name(sym->st_shndx), ".text") && strcmp(sym_name(sym_strtab, sym), "__LOAD_PHYSICAL_ADDR")) + continue; +#endif + switch (r_type) { case R_386_NONE: case R_386_PC32: @@ -569,7 +629,7 @@ static int cmp_relocs(const void *va, const void *vb) static void emit_relocs(int as_text) { - int i; + unsigned int i; /* Count how many relocations I have and allocate space for them. */ reloc_count = 0; walk_relocs(count_reloc); @@ -663,6 +723,7 @@ int main(int argc, char **argv) fname, strerror(errno)); } read_ehdr(fp); + read_phdrs(fp); read_shdrs(fp); read_strtabs(fp); read_symtabs(fp); diff --git a/arch/x86/boot/cpucheck.c b/arch/x86/boot/cpucheck.c index 4d3ff03..e4972ff 100644 --- a/arch/x86/boot/cpucheck.c +++ b/arch/x86/boot/cpucheck.c @@ -74,7 +74,7 @@ static int has_fpu(void) u16 fcw = -1, fsw = -1; u32 cr0; - asm("movl %%cr0,%0" : "=r" (cr0)); + asm volatile("movl %%cr0,%0" : "=r" (cr0)); if (cr0 & (X86_CR0_EM|X86_CR0_TS)) { cr0 &= ~(X86_CR0_EM|X86_CR0_TS); asm volatile("movl %0,%%cr0" : : "r" (cr0)); @@ -90,7 +90,7 @@ static int has_eflag(u32 mask) { u32 f0, f1; - asm("pushfl ; " + asm volatile("pushfl ; " "pushfl ; " "popl %0 ; " "movl %0,%1 ; " @@ -115,7 +115,7 @@ static void get_flags(void) set_bit(X86_FEATURE_FPU, cpu.flags); if (has_eflag(X86_EFLAGS_ID)) { - asm("cpuid" + asm volatile("cpuid" : "=a" (max_intel_level), "=b" (cpu_vendor[0]), "=d" (cpu_vendor[1]), @@ -124,7 +124,7 @@ static void get_flags(void) if (max_intel_level >= 0x00000001 && max_intel_level <= 0x0000ffff) { - asm("cpuid" + asm volatile("cpuid" : "=a" (tfms), "=c" (cpu.flags[4]), "=d" (cpu.flags[0]) @@ -136,7 +136,7 @@ static void get_flags(void) cpu.model += ((tfms >> 16) & 0xf) << 4; } - asm("cpuid" + asm volatile("cpuid" : "=a" (max_amd_level) : "a" (0x80000000) : "ebx", "ecx", "edx"); @@ -144,7 +144,7 @@ static void get_flags(void) if (max_amd_level >= 0x80000001 && max_amd_level <= 0x8000ffff) { u32 eax = 0x80000001; - asm("cpuid" + asm volatile("cpuid" : "+a" (eax), "=c" (cpu.flags[6]), "=d" (cpu.flags[1]) @@ -203,9 +203,9 @@ int check_cpu(int *cpu_level_ptr, int *req_level_ptr, u32 **err_flags_ptr) u32 ecx = MSR_K7_HWCR; u32 eax, edx; - asm("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); + asm volatile("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); eax &= ~(1 << 15); - asm("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); + asm volatile("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); get_flags(); /* Make sure it really did something */ err = check_flags(); @@ -218,9 +218,9 @@ int check_cpu(int *cpu_level_ptr, int *req_level_ptr, u32 **err_flags_ptr) u32 ecx = MSR_VIA_FCR; u32 eax, edx; - asm("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); + asm volatile("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); eax |= (1<<1)|(1<<7); - asm("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); + asm volatile("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); set_bit(X86_FEATURE_CX8, cpu.flags); err = check_flags(); @@ -231,12 +231,12 @@ int check_cpu(int *cpu_level_ptr, int *req_level_ptr, u32 **err_flags_ptr) u32 eax, edx; u32 level = 1; - asm("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); - asm("wrmsr" : : "a" (~0), "d" (edx), "c" (ecx)); - asm("cpuid" + asm volatile("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); + asm volatile("wrmsr" : : "a" (~0), "d" (edx), "c" (ecx)); + asm volatile("cpuid" : "+a" (level), "=d" (cpu.flags[0]) : : "ecx", "ebx"); - asm("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); + asm volatile("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); err = check_flags(); } diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S index f1bbeeb..aff09cb 100644 --- a/arch/x86/boot/header.S +++ b/arch/x86/boot/header.S @@ -372,7 +372,7 @@ setup_data: .quad 0 # 64-bit physical pointer to # single linked list of # struct setup_data -pref_address: .quad LOAD_PHYSICAL_ADDR # preferred load addr +pref_address: .quad ____LOAD_PHYSICAL_ADDR # preferred load addr #define ZO_INIT_SIZE (ZO__end - ZO_startup_32 + ZO_z_extract_offset) #define VO_INIT_SIZE (VO__end - VO__text) diff --git a/arch/x86/boot/memory.c b/arch/x86/boot/memory.c index db75d07..8e6d0af 100644 --- a/arch/x86/boot/memory.c +++ b/arch/x86/boot/memory.c @@ -19,7 +19,7 @@ static int detect_memory_e820(void) { - int count = 0; + unsigned int count = 0; struct biosregs ireg, oreg; struct e820entry *desc = boot_params.e820_map; static struct e820entry buf; /* static so it is zeroed */ diff --git a/arch/x86/boot/video-vesa.c b/arch/x86/boot/video-vesa.c index 11e8c6e..fdbb1ed 100644 --- a/arch/x86/boot/video-vesa.c +++ b/arch/x86/boot/video-vesa.c @@ -200,6 +200,7 @@ static void vesa_store_pm_info(void) boot_params.screen_info.vesapm_seg = oreg.es; boot_params.screen_info.vesapm_off = oreg.di; + boot_params.screen_info.vesapm_size = oreg.cx; } /* diff --git a/arch/x86/boot/video.c b/arch/x86/boot/video.c index 43eda28..5ab5fdb 100644 --- a/arch/x86/boot/video.c +++ b/arch/x86/boot/video.c @@ -96,7 +96,7 @@ static void store_mode_params(void) static unsigned int get_entry(void) { char entry_buf[4]; - int i, len = 0; + unsigned int i, len = 0; int key; unsigned int v; diff --git a/arch/x86/crypto/aes-x86_64-asm_64.S b/arch/x86/crypto/aes-x86_64-asm_64.S index 5b577d5..3c1fed4 100644 --- a/arch/x86/crypto/aes-x86_64-asm_64.S +++ b/arch/x86/crypto/aes-x86_64-asm_64.S @@ -8,6 +8,8 @@ * including this sentence is retained in full. */ +#include + .extern crypto_ft_tab .extern crypto_it_tab .extern crypto_fl_tab @@ -71,6 +73,8 @@ FUNC: movq r1,r2; \ je B192; \ leaq 32(r9),r9; +#define ret pax_force_retaddr 0, 1; ret + #define epilogue(r1,r2,r3,r4,r5,r6,r7,r8,r9) \ movq r1,r2; \ movq r3,r4; \ diff --git a/arch/x86/crypto/aesni-intel_asm.S b/arch/x86/crypto/aesni-intel_asm.S index be6d9e3..21fbbca 100644 --- a/arch/x86/crypto/aesni-intel_asm.S +++ b/arch/x86/crypto/aesni-intel_asm.S @@ -31,6 +31,7 @@ #include #include +#include #ifdef __x86_64__ .data @@ -1436,7 +1437,9 @@ _return_T_done_decrypt: pop %r14 pop %r13 pop %r12 + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_gcm_dec) /***************************************************************************** @@ -1699,7 +1702,9 @@ _return_T_done_encrypt: pop %r14 pop %r13 pop %r12 + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_gcm_enc) #endif @@ -1714,6 +1719,7 @@ _key_expansion_256a: pxor %xmm1, %xmm0 movaps %xmm0, (TKEYP) add $0x10, TKEYP + pax_force_retaddr_bts ret .align 4 @@ -1738,6 +1744,7 @@ _key_expansion_192a: shufps $0b01001110, %xmm2, %xmm1 movaps %xmm1, 0x10(TKEYP) add $0x20, TKEYP + pax_force_retaddr_bts ret .align 4 @@ -1757,6 +1764,7 @@ _key_expansion_192b: movaps %xmm0, (TKEYP) add $0x10, TKEYP + pax_force_retaddr_bts ret .align 4 @@ -1769,6 +1777,7 @@ _key_expansion_256b: pxor %xmm1, %xmm2 movaps %xmm2, (TKEYP) add $0x10, TKEYP + pax_force_retaddr_bts ret /* @@ -1881,7 +1890,9 @@ ENTRY(aesni_set_key) #ifndef __x86_64__ popl KEYP #endif + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_set_key) /* * void aesni_enc(struct crypto_aes_ctx *ctx, u8 *dst, const u8 *src) @@ -1902,7 +1913,9 @@ ENTRY(aesni_enc) popl KLEN popl KEYP #endif + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_enc) /* * _aesni_enc1: internal ABI @@ -1959,6 +1972,7 @@ _aesni_enc1: AESENC KEY STATE movaps 0x70(TKEYP), KEY AESENCLAST KEY STATE + pax_force_retaddr_bts ret /* @@ -2067,6 +2081,7 @@ _aesni_enc4: AESENCLAST KEY STATE2 AESENCLAST KEY STATE3 AESENCLAST KEY STATE4 + pax_force_retaddr_bts ret /* @@ -2089,7 +2104,9 @@ ENTRY(aesni_dec) popl KLEN popl KEYP #endif + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_dec) /* * _aesni_dec1: internal ABI @@ -2146,6 +2163,7 @@ _aesni_dec1: AESDEC KEY STATE movaps 0x70(TKEYP), KEY AESDECLAST KEY STATE + pax_force_retaddr_bts ret /* @@ -2254,6 +2272,7 @@ _aesni_dec4: AESDECLAST KEY STATE2 AESDECLAST KEY STATE3 AESDECLAST KEY STATE4 + pax_force_retaddr_bts ret /* @@ -2311,7 +2330,9 @@ ENTRY(aesni_ecb_enc) popl KEYP popl LEN #endif + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_ecb_enc) /* * void aesni_ecb_dec(struct crypto_aes_ctx *ctx, const u8 *dst, u8 *src, @@ -2369,7 +2390,9 @@ ENTRY(aesni_ecb_dec) popl KEYP popl LEN #endif + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_ecb_dec) /* * void aesni_cbc_enc(struct crypto_aes_ctx *ctx, const u8 *dst, u8 *src, @@ -2410,7 +2433,9 @@ ENTRY(aesni_cbc_enc) popl LEN popl IVP #endif + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_cbc_enc) /* * void aesni_cbc_dec(struct crypto_aes_ctx *ctx, const u8 *dst, u8 *src, @@ -2498,7 +2523,9 @@ ENTRY(aesni_cbc_dec) popl LEN popl IVP #endif + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_cbc_dec) #ifdef __x86_64__ .align 16 @@ -2524,6 +2551,7 @@ _aesni_inc_init: mov $1, TCTR_LOW MOVQ_R64_XMM TCTR_LOW INC MOVQ_R64_XMM CTR TCTR_LOW + pax_force_retaddr_bts ret /* @@ -2552,6 +2580,7 @@ _aesni_inc: .Linc_low: movaps CTR, IV PSHUFB_XMM BSWAP_MASK IV + pax_force_retaddr_bts ret /* @@ -2612,5 +2641,7 @@ ENTRY(aesni_ctr_enc) .Lctr_enc_ret: movups IV, (IVP) .Lctr_enc_just_ret: + pax_force_retaddr 0, 1 ret +ENDPROC(aesni_ctr_enc) #endif diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c index 545d0ce..14841a6 100644 --- a/arch/x86/crypto/aesni-intel_glue.c +++ b/arch/x86/crypto/aesni-intel_glue.c @@ -929,6 +929,8 @@ out_free_ablkcipher: } static int rfc4106_set_key(struct crypto_aead *parent, const u8 *key, + unsigned int key_len) __size_overflow(3); +static int rfc4106_set_key(struct crypto_aead *parent, const u8 *key, unsigned int key_len) { int ret = 0; diff --git a/arch/x86/crypto/blowfish-x86_64-asm_64.S b/arch/x86/crypto/blowfish-x86_64-asm_64.S index 391d245..67f35c2 100644 --- a/arch/x86/crypto/blowfish-x86_64-asm_64.S +++ b/arch/x86/crypto/blowfish-x86_64-asm_64.S @@ -20,6 +20,8 @@ * */ +#include + .file "blowfish-x86_64-asm.S" .text @@ -151,9 +153,11 @@ __blowfish_enc_blk: jnz __enc_xor; write_block(); + pax_force_retaddr 0, 1 ret; __enc_xor: xor_block(); + pax_force_retaddr 0, 1 ret; .align 8 @@ -188,6 +192,7 @@ blowfish_dec_blk: movq %r11, %rbp; + pax_force_retaddr 0, 1 ret; /********************************************************************** @@ -342,6 +347,7 @@ __blowfish_enc_blk_4way: popq %rbx; popq %rbp; + pax_force_retaddr 0, 1 ret; __enc_xor4: @@ -349,6 +355,7 @@ __enc_xor4: popq %rbx; popq %rbp; + pax_force_retaddr 0, 1 ret; .align 8 @@ -386,5 +393,6 @@ blowfish_dec_blk_4way: popq %rbx; popq %rbp; + pax_force_retaddr 0, 1 ret; diff --git a/arch/x86/crypto/salsa20-x86_64-asm_64.S b/arch/x86/crypto/salsa20-x86_64-asm_64.S index 6214a9b..1f4fc9a 100644 --- a/arch/x86/crypto/salsa20-x86_64-asm_64.S +++ b/arch/x86/crypto/salsa20-x86_64-asm_64.S @@ -1,3 +1,5 @@ +#include + # enter ECRYPT_encrypt_bytes .text .p2align 5 @@ -790,6 +792,7 @@ ECRYPT_encrypt_bytes: add %r11,%rsp mov %rdi,%rax mov %rsi,%rdx + pax_force_retaddr 0, 1 ret # bytesatleast65: ._bytesatleast65: @@ -891,6 +894,7 @@ ECRYPT_keysetup: add %r11,%rsp mov %rdi,%rax mov %rsi,%rdx + pax_force_retaddr ret # enter ECRYPT_ivsetup .text @@ -917,4 +921,5 @@ ECRYPT_ivsetup: add %r11,%rsp mov %rdi,%rax mov %rsi,%rdx + pax_force_retaddr ret diff --git a/arch/x86/crypto/serpent-sse2-x86_64-asm_64.S b/arch/x86/crypto/serpent-sse2-x86_64-asm_64.S index 7f24a15..9cd3ffe 100644 --- a/arch/x86/crypto/serpent-sse2-x86_64-asm_64.S +++ b/arch/x86/crypto/serpent-sse2-x86_64-asm_64.S @@ -24,6 +24,8 @@ * */ +#include + .file "serpent-sse2-x86_64-asm_64.S" .text @@ -695,12 +697,14 @@ __serpent_enc_blk_8way: write_blocks(%rsi, RA1, RB1, RC1, RD1, RK0, RK1, RK2); write_blocks(%rax, RA2, RB2, RC2, RD2, RK0, RK1, RK2); + pax_force_retaddr ret; __enc_xor8: xor_blocks(%rsi, RA1, RB1, RC1, RD1, RK0, RK1, RK2); xor_blocks(%rax, RA2, RB2, RC2, RD2, RK0, RK1, RK2); + pax_force_retaddr ret; .align 8 @@ -758,4 +762,5 @@ serpent_dec_blk_8way: write_blocks(%rsi, RC1, RD1, RB1, RE1, RK0, RK1, RK2); write_blocks(%rax, RC2, RD2, RB2, RE2, RK0, RK1, RK2); + pax_force_retaddr ret; diff --git a/arch/x86/crypto/sha1_ssse3_asm.S b/arch/x86/crypto/sha1_ssse3_asm.S index b2c2f57..8470cab 100644 --- a/arch/x86/crypto/sha1_ssse3_asm.S +++ b/arch/x86/crypto/sha1_ssse3_asm.S @@ -28,6 +28,8 @@ * (at your option) any later version. */ +#include + #define CTX %rdi // arg1 #define BUF %rsi // arg2 #define CNT %rdx // arg3 @@ -104,6 +106,7 @@ pop %r12 pop %rbp pop %rbx + pax_force_retaddr 0, 1 ret .size \name, .-\name diff --git a/arch/x86/crypto/twofish-x86_64-asm_64-3way.S b/arch/x86/crypto/twofish-x86_64-asm_64-3way.S index 5b012a2..36d5364 100644 --- a/arch/x86/crypto/twofish-x86_64-asm_64-3way.S +++ b/arch/x86/crypto/twofish-x86_64-asm_64-3way.S @@ -20,6 +20,8 @@ * */ +#include + .file "twofish-x86_64-asm-3way.S" .text @@ -260,6 +262,7 @@ __twofish_enc_blk_3way: popq %r13; popq %r14; popq %r15; + pax_force_retaddr 0, 1 ret; __enc_xor3: @@ -271,6 +274,7 @@ __enc_xor3: popq %r13; popq %r14; popq %r15; + pax_force_retaddr 0, 1 ret; .global twofish_dec_blk_3way @@ -312,5 +316,6 @@ twofish_dec_blk_3way: popq %r13; popq %r14; popq %r15; + pax_force_retaddr 0, 1 ret; diff --git a/arch/x86/crypto/twofish-x86_64-asm_64.S b/arch/x86/crypto/twofish-x86_64-asm_64.S index 7bcf3fc..f53832f 100644 --- a/arch/x86/crypto/twofish-x86_64-asm_64.S +++ b/arch/x86/crypto/twofish-x86_64-asm_64.S @@ -21,6 +21,7 @@ .text #include +#include #define a_offset 0 #define b_offset 4 @@ -268,6 +269,7 @@ twofish_enc_blk: popq R1 movq $1,%rax + pax_force_retaddr 0, 1 ret twofish_dec_blk: @@ -319,4 +321,5 @@ twofish_dec_blk: popq R1 movq $1,%rax + pax_force_retaddr 0, 1 ret diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c index 39e4909..887aa7e 100644 --- a/arch/x86/ia32/ia32_aout.c +++ b/arch/x86/ia32/ia32_aout.c @@ -162,6 +162,8 @@ static int aout_core_dump(long signr, struct pt_regs *regs, struct file *file, unsigned long dump_start, dump_size; struct user32 dump; + memset(&dump, 0, sizeof(dump)); + fs = get_fs(); set_fs(KERNEL_DS); has_dumped = 1; diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c index 6557769..ef6ae89 100644 --- a/arch/x86/ia32/ia32_signal.c +++ b/arch/x86/ia32/ia32_signal.c @@ -169,7 +169,7 @@ asmlinkage long sys32_sigaltstack(const stack_ia32_t __user *uss_ptr, } seg = get_fs(); set_fs(KERNEL_DS); - ret = do_sigaltstack(uss_ptr ? &uss : NULL, &uoss, regs->sp); + ret = do_sigaltstack(uss_ptr ? (const stack_t __force_user *)&uss : NULL, (stack_t __force_user *)&uoss, regs->sp); set_fs(seg); if (ret >= 0 && uoss_ptr) { if (!access_ok(VERIFY_WRITE, uoss_ptr, sizeof(stack_ia32_t))) @@ -370,7 +370,7 @@ static int ia32_setup_sigcontext(struct sigcontext_ia32 __user *sc, */ static void __user *get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size, - void **fpstate) + void __user **fpstate) { unsigned long sp; @@ -391,7 +391,7 @@ static void __user *get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, if (used_math()) { sp = sp - sig_xstate_ia32_size; - *fpstate = (struct _fpstate_ia32 *) sp; + *fpstate = (struct _fpstate_ia32 __user *) sp; if (save_i387_xstate_ia32(*fpstate) < 0) return (void __user *) -1L; } @@ -399,7 +399,7 @@ static void __user *get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, sp -= frame_size; /* Align the stack pointer according to the i386 ABI, * i.e. so that on function entry ((sp + 4) & 15) == 0. */ - sp = ((sp + 4) & -16ul) - 4; + sp = ((sp - 12) & -16ul) - 4; return (void __user *) sp; } @@ -457,7 +457,7 @@ int ia32_setup_frame(int sig, struct k_sigaction *ka, * These are actually not used anymore, but left because some * gdb versions depend on them as a marker. */ - put_user_ex(*((u64 *)&code), (u64 *)frame->retcode); + put_user_ex(*((const u64 *)&code), (u64 __user *)frame->retcode); } put_user_catch(err); if (err) @@ -499,7 +499,7 @@ int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, 0xb8, __NR_ia32_rt_sigreturn, 0x80cd, - 0, + 0 }; frame = get_sigframe(ka, regs, sizeof(*frame), &fpstate); @@ -529,16 +529,18 @@ int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, if (ka->sa.sa_flags & SA_RESTORER) restorer = ka->sa.sa_restorer; + else if (current->mm->context.vdso) + /* Return stub is in 32bit vsyscall page */ + restorer = VDSO32_SYMBOL(current->mm->context.vdso, rt_sigreturn); else - restorer = VDSO32_SYMBOL(current->mm->context.vdso, - rt_sigreturn); + restorer = &frame->retcode; put_user_ex(ptr_to_compat(restorer), &frame->pretcode); /* * Not actually used anymore, but left because some gdb * versions need it. */ - put_user_ex(*((u64 *)&code), (u64 *)frame->retcode); + put_user_ex(*((const u64 *)&code), (u64 __user *)frame->retcode); } put_user_catch(err); if (err) diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index e3e7340..05ed805 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -13,8 +13,10 @@ #include #include #include +#include #include #include +#include /* Avoid __ASSEMBLER__'ifying just for this. */ #include @@ -94,6 +96,32 @@ ENTRY(native_irq_enable_sysexit) ENDPROC(native_irq_enable_sysexit) #endif + .macro pax_enter_kernel_user + pax_set_fptr_mask +#ifdef CONFIG_PAX_MEMORY_UDEREF + call pax_enter_kernel_user +#endif + .endm + + .macro pax_exit_kernel_user +#ifdef CONFIG_PAX_MEMORY_UDEREF + call pax_exit_kernel_user +#endif +#ifdef CONFIG_PAX_RANDKSTACK + pushq %rax + pushq %r11 + call pax_randomize_kstack + popq %r11 + popq %rax +#endif + .endm + +.macro pax_erase_kstack +#ifdef CONFIG_PAX_MEMORY_STACKLEAK + call pax_erase_kstack +#endif +.endm + /* * 32bit SYSENTER instruction entry. * @@ -120,12 +148,6 @@ ENTRY(ia32_sysenter_target) CFI_REGISTER rsp,rbp SWAPGS_UNSAFE_STACK movq PER_CPU_VAR(kernel_stack), %rsp - addq $(KERNEL_STACK_OFFSET),%rsp - /* - * No need to follow this irqs on/off section: the syscall - * disabled irqs, here we enable it straight after entry: - */ - ENABLE_INTERRUPTS(CLBR_NONE) movl %ebp,%ebp /* zero extension */ pushq_cfi $__USER32_DS /*CFI_REL_OFFSET ss,0*/ @@ -133,24 +155,39 @@ ENTRY(ia32_sysenter_target) CFI_REL_OFFSET rsp,0 pushfq_cfi /*CFI_REL_OFFSET rflags,0*/ - movl TI_sysenter_return+THREAD_INFO(%rsp,3*8-KERNEL_STACK_OFFSET),%r10d - CFI_REGISTER rip,r10 + orl $X86_EFLAGS_IF,(%rsp) + GET_THREAD_INFO(%r11) + movl TI_sysenter_return(%r11), %r11d + CFI_REGISTER rip,r11 pushq_cfi $__USER32_CS /*CFI_REL_OFFSET cs,0*/ movl %eax, %eax - pushq_cfi %r10 + pushq_cfi %r11 CFI_REL_OFFSET rip,0 pushq_cfi %rax cld SAVE_ARGS 0,1,0 + pax_enter_kernel_user + /* + * No need to follow this irqs on/off section: the syscall + * disabled irqs, here we enable it straight after entry: + */ + ENABLE_INTERRUPTS(CLBR_NONE) /* no need to do an access_ok check here because rbp has been 32bit zero extended */ + +#ifdef CONFIG_PAX_MEMORY_UDEREF + mov $PAX_USER_SHADOW_BASE,%r11 + add %r11,%rbp +#endif + 1: movl (%rbp),%ebp .section __ex_table,"a" .quad 1b,ia32_badarg .previous - orl $TS_COMPAT,TI_status+THREAD_INFO(%rsp,RIP-ARGOFFSET) - testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + GET_THREAD_INFO(%r11) + orl $TS_COMPAT,TI_status(%r11) + testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r11) CFI_REMEMBER_STATE jnz sysenter_tracesys cmpq $(IA32_NR_syscalls-1),%rax @@ -160,12 +197,15 @@ sysenter_do_call: sysenter_dispatch: call *ia32_sys_call_table(,%rax,8) movq %rax,RAX-ARGOFFSET(%rsp) + GET_THREAD_INFO(%r11) DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF - testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + testl $_TIF_ALLWORK_MASK,TI_flags(%r11) jnz sysexit_audit sysexit_from_sys_call: - andl $~TS_COMPAT,TI_status+THREAD_INFO(%rsp,RIP-ARGOFFSET) + pax_exit_kernel_user + pax_erase_kstack + andl $~TS_COMPAT,TI_status(%r11) /* clear IF, that popfq doesn't enable interrupts early */ andl $~0x200,EFLAGS-R11(%rsp) movl RIP-R11(%rsp),%edx /* User %eip */ @@ -191,6 +231,9 @@ sysexit_from_sys_call: movl %eax,%esi /* 2nd arg: syscall number */ movl $AUDIT_ARCH_I386,%edi /* 1st arg: audit arch */ call __audit_syscall_entry + + pax_erase_kstack + movl RAX-ARGOFFSET(%rsp),%eax /* reload syscall number */ cmpq $(IA32_NR_syscalls-1),%rax ja ia32_badsys @@ -202,7 +245,7 @@ sysexit_from_sys_call: .endm .macro auditsys_exit exit - testl $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT),TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + testl $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT),TI_flags(%r11) jnz ia32_ret_from_sys_call TRACE_IRQS_ON sti @@ -213,11 +256,12 @@ sysexit_from_sys_call: 1: setbe %al /* 1 if error, 0 if not */ movzbl %al,%edi /* zero-extend that into %edi */ call __audit_syscall_exit + GET_THREAD_INFO(%r11) movq RAX-ARGOFFSET(%rsp),%rax /* reload syscall return value */ movl $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT),%edi cli TRACE_IRQS_OFF - testl %edi,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + testl %edi,TI_flags(%r11) jz \exit CLEAR_RREGS -ARGOFFSET jmp int_with_check @@ -235,7 +279,7 @@ sysexit_audit: sysenter_tracesys: #ifdef CONFIG_AUDITSYSCALL - testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags(%r11) jz sysenter_auditsys #endif SAVE_REST @@ -243,6 +287,9 @@ sysenter_tracesys: movq $-ENOSYS,RAX(%rsp)/* ptrace can change this for a bad syscall */ movq %rsp,%rdi /* &pt_regs -> arg1 */ call syscall_trace_enter + + pax_erase_kstack + LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */ RESTORE_REST cmpq $(IA32_NR_syscalls-1),%rax @@ -274,19 +321,20 @@ ENDPROC(ia32_sysenter_target) ENTRY(ia32_cstar_target) CFI_STARTPROC32 simple CFI_SIGNAL_FRAME - CFI_DEF_CFA rsp,KERNEL_STACK_OFFSET + CFI_DEF_CFA rsp,0 CFI_REGISTER rip,rcx /*CFI_REGISTER rflags,r11*/ SWAPGS_UNSAFE_STACK movl %esp,%r8d CFI_REGISTER rsp,r8 movq PER_CPU_VAR(kernel_stack),%rsp + SAVE_ARGS 8*6,0,0 + pax_enter_kernel_user /* * No need to follow this irqs on/off section: the syscall * disabled irqs and here we enable it straight after entry: */ ENABLE_INTERRUPTS(CLBR_NONE) - SAVE_ARGS 8,0,0 movl %eax,%eax /* zero extension */ movq %rax,ORIG_RAX-ARGOFFSET(%rsp) movq %rcx,RIP-ARGOFFSET(%rsp) @@ -302,12 +350,19 @@ ENTRY(ia32_cstar_target) /* no need to do an access_ok check here because r8 has been 32bit zero extended */ /* hardware stack frame is complete now */ + +#ifdef CONFIG_PAX_MEMORY_UDEREF + mov $PAX_USER_SHADOW_BASE,%r11 + add %r11,%r8 +#endif + 1: movl (%r8),%r9d .section __ex_table,"a" .quad 1b,ia32_badarg .previous - orl $TS_COMPAT,TI_status+THREAD_INFO(%rsp,RIP-ARGOFFSET) - testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + GET_THREAD_INFO(%r11) + orl $TS_COMPAT,TI_status(%r11) + testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r11) CFI_REMEMBER_STATE jnz cstar_tracesys cmpq $IA32_NR_syscalls-1,%rax @@ -317,12 +372,15 @@ cstar_do_call: cstar_dispatch: call *ia32_sys_call_table(,%rax,8) movq %rax,RAX-ARGOFFSET(%rsp) + GET_THREAD_INFO(%r11) DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF - testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + testl $_TIF_ALLWORK_MASK,TI_flags(%r11) jnz sysretl_audit sysretl_from_sys_call: - andl $~TS_COMPAT,TI_status+THREAD_INFO(%rsp,RIP-ARGOFFSET) + pax_exit_kernel_user + pax_erase_kstack + andl $~TS_COMPAT,TI_status(%r11) RESTORE_ARGS 0,-ARG_SKIP,0,0,0 movl RIP-ARGOFFSET(%rsp),%ecx CFI_REGISTER rip,rcx @@ -350,7 +408,7 @@ sysretl_audit: cstar_tracesys: #ifdef CONFIG_AUDITSYSCALL - testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags(%r11) jz cstar_auditsys #endif xchgl %r9d,%ebp @@ -359,6 +417,9 @@ cstar_tracesys: movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */ movq %rsp,%rdi /* &pt_regs -> arg1 */ call syscall_trace_enter + + pax_erase_kstack + LOAD_ARGS32 ARGOFFSET, 1 /* reload args from stack in case ptrace changed it */ RESTORE_REST xchgl %ebp,%r9d @@ -404,19 +465,21 @@ ENTRY(ia32_syscall) CFI_REL_OFFSET rip,RIP-RIP PARAVIRT_ADJUST_EXCEPTION_FRAME SWAPGS - /* - * No need to follow this irqs on/off section: the syscall - * disabled irqs and here we enable it straight after entry: - */ - ENABLE_INTERRUPTS(CLBR_NONE) movl %eax,%eax pushq_cfi %rax cld /* note the registers are not zero extended to the sf. this could be a problem. */ SAVE_ARGS 0,1,0 - orl $TS_COMPAT,TI_status+THREAD_INFO(%rsp,RIP-ARGOFFSET) - testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + pax_enter_kernel_user + /* + * No need to follow this irqs on/off section: the syscall + * disabled irqs and here we enable it straight after entry: + */ + ENABLE_INTERRUPTS(CLBR_NONE) + GET_THREAD_INFO(%r11) + orl $TS_COMPAT,TI_status(%r11) + testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r11) jnz ia32_tracesys cmpq $(IA32_NR_syscalls-1),%rax ja ia32_badsys @@ -435,6 +498,9 @@ ia32_tracesys: movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */ movq %rsp,%rdi /* &pt_regs -> arg1 */ call syscall_trace_enter + + pax_erase_kstack + LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */ RESTORE_REST cmpq $(IA32_NR_syscalls-1),%rax diff --git a/arch/x86/ia32/sys_ia32.c b/arch/x86/ia32/sys_ia32.c index f6f5c53..b358b28 100644 --- a/arch/x86/ia32/sys_ia32.c +++ b/arch/x86/ia32/sys_ia32.c @@ -69,8 +69,8 @@ asmlinkage long sys32_ftruncate64(unsigned int fd, unsigned long offset_low, */ static int cp_stat64(struct stat64 __user *ubuf, struct kstat *stat) { - typeof(ubuf->st_uid) uid = 0; - typeof(ubuf->st_gid) gid = 0; + typeof(((struct stat64 *)0)->st_uid) uid = 0; + typeof(((struct stat64 *)0)->st_gid) gid = 0; SET_UID(uid, stat->uid); SET_GID(gid, stat->gid); if (!access_ok(VERIFY_WRITE, ubuf, sizeof(struct stat64)) || @@ -308,8 +308,8 @@ asmlinkage long sys32_rt_sigprocmask(int how, compat_sigset_t __user *set, } set_fs(KERNEL_DS); ret = sys_rt_sigprocmask(how, - set ? (sigset_t __user *)&s : NULL, - oset ? (sigset_t __user *)&s : NULL, + set ? (sigset_t __force_user *)&s : NULL, + oset ? (sigset_t __force_user *)&s : NULL, sigsetsize); set_fs(old_fs); if (ret) @@ -332,7 +332,7 @@ asmlinkage long sys32_alarm(unsigned int seconds) return alarm_setitimer(seconds); } -asmlinkage long sys32_waitpid(compat_pid_t pid, unsigned int *stat_addr, +asmlinkage long sys32_waitpid(compat_pid_t pid, unsigned int __user *stat_addr, int options) { return compat_sys_wait4(pid, stat_addr, options, NULL); @@ -353,7 +353,7 @@ asmlinkage long sys32_sched_rr_get_interval(compat_pid_t pid, mm_segment_t old_fs = get_fs(); set_fs(KERNEL_DS); - ret = sys_sched_rr_get_interval(pid, (struct timespec __user *)&t); + ret = sys_sched_rr_get_interval(pid, (struct timespec __force_user *)&t); set_fs(old_fs); if (put_compat_timespec(&t, interval)) return -EFAULT; @@ -369,7 +369,7 @@ asmlinkage long sys32_rt_sigpending(compat_sigset_t __user *set, mm_segment_t old_fs = get_fs(); set_fs(KERNEL_DS); - ret = sys_rt_sigpending((sigset_t __user *)&s, sigsetsize); + ret = sys_rt_sigpending((sigset_t __force_user *)&s, sigsetsize); set_fs(old_fs); if (!ret) { switch (_NSIG_WORDS) { @@ -394,7 +394,7 @@ asmlinkage long sys32_rt_sigqueueinfo(int pid, int sig, if (copy_siginfo_from_user32(&info, uinfo)) return -EFAULT; set_fs(KERNEL_DS); - ret = sys_rt_sigqueueinfo(pid, sig, (siginfo_t __user *)&info); + ret = sys_rt_sigqueueinfo(pid, sig, (siginfo_t __force_user *)&info); set_fs(old_fs); return ret; } @@ -439,7 +439,7 @@ asmlinkage long sys32_sendfile(int out_fd, int in_fd, return -EFAULT; set_fs(KERNEL_DS); - ret = sys_sendfile(out_fd, in_fd, offset ? (off_t __user *)&of : NULL, + ret = sys_sendfile(out_fd, in_fd, offset ? (off_t __force_user *)&of : NULL, count); set_fs(old_fs); diff --git a/arch/x86/include/asm/alternative-asm.h b/arch/x86/include/asm/alternative-asm.h index 952bd01..7692c6f 100644 --- a/arch/x86/include/asm/alternative-asm.h +++ b/arch/x86/include/asm/alternative-asm.h @@ -15,6 +15,45 @@ .endm #endif +#ifdef KERNEXEC_PLUGIN + .macro pax_force_retaddr_bts rip=0 + btsq $63,\rip(%rsp) + .endm +#ifdef CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_BTS + .macro pax_force_retaddr rip=0, reload=0 + btsq $63,\rip(%rsp) + .endm + .macro pax_force_fptr ptr + btsq $63,\ptr + .endm + .macro pax_set_fptr_mask + .endm +#endif +#ifdef CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_OR + .macro pax_force_retaddr rip=0, reload=0 + .if \reload + pax_set_fptr_mask + .endif + orq %r10,\rip(%rsp) + .endm + .macro pax_force_fptr ptr + orq %r10,\ptr + .endm + .macro pax_set_fptr_mask + movabs $0x8000000000000000,%r10 + .endm +#endif +#else + .macro pax_force_retaddr rip=0, reload=0 + .endm + .macro pax_force_fptr ptr + .endm + .macro pax_force_retaddr_bts rip=0 + .endm + .macro pax_set_fptr_mask + .endm +#endif + .macro altinstruction_entry orig alt feature orig_len alt_len .long \orig - . .long \alt - . diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h index 37ad100..7d47faa 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -89,7 +89,7 @@ static inline int alternatives_text_reserved(void *start, void *end) ".section .discard,\"aw\",@progbits\n" \ " .byte 0xff + (664f-663f) - (662b-661b)\n" /* rlen <= slen */ \ ".previous\n" \ - ".section .altinstr_replacement, \"ax\"\n" \ + ".section .altinstr_replacement, \"a\"\n" \ "663:\n\t" newinstr "\n664:\n" /* replacement */ \ ".previous" diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index 3ab9bdd..238033e 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -45,7 +45,7 @@ static inline void generic_apic_probe(void) #ifdef CONFIG_X86_LOCAL_APIC -extern unsigned int apic_verbosity; +extern int apic_verbosity; extern int local_apic_timer_c2_ok; extern int disable_apic; diff --git a/arch/x86/include/asm/apm.h b/arch/x86/include/asm/apm.h index 20370c6..a2eb9b0 100644 --- a/arch/x86/include/asm/apm.h +++ b/arch/x86/include/asm/apm.h @@ -34,7 +34,7 @@ static inline void apm_bios_call_asm(u32 func, u32 ebx_in, u32 ecx_in, __asm__ __volatile__(APM_DO_ZERO_SEGS "pushl %%edi\n\t" "pushl %%ebp\n\t" - "lcall *%%cs:apm_bios_entry\n\t" + "lcall *%%ss:apm_bios_entry\n\t" "setc %%al\n\t" "popl %%ebp\n\t" "popl %%edi\n\t" @@ -58,7 +58,7 @@ static inline u8 apm_bios_call_simple_asm(u32 func, u32 ebx_in, __asm__ __volatile__(APM_DO_ZERO_SEGS "pushl %%edi\n\t" "pushl %%ebp\n\t" - "lcall *%%cs:apm_bios_entry\n\t" + "lcall *%%ss:apm_bios_entry\n\t" "setc %%bl\n\t" "popl %%ebp\n\t" "popl %%edi\n\t" diff --git a/arch/x86/include/asm/atomic.h b/arch/x86/include/asm/atomic.h index 58cb6d4..ca9010d 100644 --- a/arch/x86/include/asm/atomic.h +++ b/arch/x86/include/asm/atomic.h @@ -22,7 +22,18 @@ */ static inline int atomic_read(const atomic_t *v) { - return (*(volatile int *)&(v)->counter); + return (*(volatile const int *)&(v)->counter); +} + +/** + * atomic_read_unchecked - read atomic variable + * @v: pointer of type atomic_unchecked_t + * + * Atomically reads the value of @v. + */ +static inline int atomic_read_unchecked(const atomic_unchecked_t *v) +{ + return (*(volatile const int *)&(v)->counter); } /** @@ -38,6 +49,18 @@ static inline void atomic_set(atomic_t *v, int i) } /** + * atomic_set_unchecked - set atomic variable + * @v: pointer of type atomic_unchecked_t + * @i: required value + * + * Atomically sets the value of @v to @i. + */ +static inline void atomic_set_unchecked(atomic_unchecked_t *v, int i) +{ + v->counter = i; +} + +/** * atomic_add - add integer to atomic variable * @i: integer value to add * @v: pointer of type atomic_t @@ -46,7 +69,29 @@ static inline void atomic_set(atomic_t *v, int i) */ static inline void atomic_add(int i, atomic_t *v) { - asm volatile(LOCK_PREFIX "addl %1,%0" + asm volatile(LOCK_PREFIX "addl %1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "subl %1,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "+m" (v->counter) + : "ir" (i)); +} + +/** + * atomic_add_unchecked - add integer to atomic variable + * @i: integer value to add + * @v: pointer of type atomic_unchecked_t + * + * Atomically adds @i to @v. + */ +static inline void atomic_add_unchecked(int i, atomic_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "addl %1,%0\n" : "+m" (v->counter) : "ir" (i)); } @@ -60,7 +105,29 @@ static inline void atomic_add(int i, atomic_t *v) */ static inline void atomic_sub(int i, atomic_t *v) { - asm volatile(LOCK_PREFIX "subl %1,%0" + asm volatile(LOCK_PREFIX "subl %1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "addl %1,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "+m" (v->counter) + : "ir" (i)); +} + +/** + * atomic_sub_unchecked - subtract integer from atomic variable + * @i: integer value to subtract + * @v: pointer of type atomic_unchecked_t + * + * Atomically subtracts @i from @v. + */ +static inline void atomic_sub_unchecked(int i, atomic_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "subl %1,%0\n" : "+m" (v->counter) : "ir" (i)); } @@ -78,7 +145,16 @@ static inline int atomic_sub_and_test(int i, atomic_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "subl %2,%0; sete %1" + asm volatile(LOCK_PREFIX "subl %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "addl %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "+m" (v->counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -92,7 +168,27 @@ static inline int atomic_sub_and_test(int i, atomic_t *v) */ static inline void atomic_inc(atomic_t *v) { - asm volatile(LOCK_PREFIX "incl %0" + asm volatile(LOCK_PREFIX "incl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "decl %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "+m" (v->counter)); +} + +/** + * atomic_inc_unchecked - increment atomic variable + * @v: pointer of type atomic_unchecked_t + * + * Atomically increments @v by 1. + */ +static inline void atomic_inc_unchecked(atomic_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "incl %0\n" : "+m" (v->counter)); } @@ -104,7 +200,27 @@ static inline void atomic_inc(atomic_t *v) */ static inline void atomic_dec(atomic_t *v) { - asm volatile(LOCK_PREFIX "decl %0" + asm volatile(LOCK_PREFIX "decl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "incl %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "+m" (v->counter)); +} + +/** + * atomic_dec_unchecked - decrement atomic variable + * @v: pointer of type atomic_unchecked_t + * + * Atomically decrements @v by 1. + */ +static inline void atomic_dec_unchecked(atomic_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "decl %0\n" : "+m" (v->counter)); } @@ -120,7 +236,16 @@ static inline int atomic_dec_and_test(atomic_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "decl %0; sete %1" + asm volatile(LOCK_PREFIX "decl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "incl %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "+m" (v->counter), "=qm" (c) : : "memory"); return c != 0; @@ -138,7 +263,35 @@ static inline int atomic_inc_and_test(atomic_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "incl %0; sete %1" + asm volatile(LOCK_PREFIX "incl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "decl %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" + : "+m" (v->counter), "=qm" (c) + : : "memory"); + return c != 0; +} + +/** + * atomic_inc_and_test_unchecked - increment and test + * @v: pointer of type atomic_unchecked_t + * + * Atomically increments @v by 1 + * and returns true if the result is zero, or false for all + * other cases. + */ +static inline int atomic_inc_and_test_unchecked(atomic_unchecked_t *v) +{ + unsigned char c; + + asm volatile(LOCK_PREFIX "incl %0\n" + "sete %1\n" : "+m" (v->counter), "=qm" (c) : : "memory"); return c != 0; @@ -157,7 +310,16 @@ static inline int atomic_add_negative(int i, atomic_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "addl %2,%0; sets %1" + asm volatile(LOCK_PREFIX "addl %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "subl %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sets %1\n" : "+m" (v->counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -179,7 +341,7 @@ static inline int atomic_add_return(int i, atomic_t *v) goto no_xadd; #endif /* Modern 486+ processor */ - return i + xadd(&v->counter, i); + return i + xadd_check_overflow(&v->counter, i); #ifdef CONFIG_M386 no_xadd: /* Legacy 386 processor */ @@ -192,6 +354,34 @@ no_xadd: /* Legacy 386 processor */ } /** + * atomic_add_return_unchecked - add integer and return + * @i: integer value to add + * @v: pointer of type atomic_unchecked_t + * + * Atomically adds @i to @v and returns @i + @v + */ +static inline int atomic_add_return_unchecked(int i, atomic_unchecked_t *v) +{ +#ifdef CONFIG_M386 + int __i; + unsigned long flags; + if (unlikely(boot_cpu_data.x86 <= 3)) + goto no_xadd; +#endif + /* Modern 486+ processor */ + return i + xadd(&v->counter, i); + +#ifdef CONFIG_M386 +no_xadd: /* Legacy 386 processor */ + raw_local_irq_save(flags); + __i = atomic_read_unchecked(v); + atomic_set_unchecked(v, i + __i); + raw_local_irq_restore(flags); + return i + __i; +#endif +} + +/** * atomic_sub_return - subtract integer and return * @v: pointer of type atomic_t * @i: integer value to subtract @@ -204,6 +394,10 @@ static inline int atomic_sub_return(int i, atomic_t *v) } #define atomic_inc_return(v) (atomic_add_return(1, v)) +static inline int atomic_inc_return_unchecked(atomic_unchecked_t *v) +{ + return atomic_add_return_unchecked(1, v); +} #define atomic_dec_return(v) (atomic_sub_return(1, v)) static inline int atomic_cmpxchg(atomic_t *v, int old, int new) @@ -211,11 +405,21 @@ static inline int atomic_cmpxchg(atomic_t *v, int old, int new) return cmpxchg(&v->counter, old, new); } +static inline int atomic_cmpxchg_unchecked(atomic_unchecked_t *v, int old, int new) +{ + return cmpxchg(&v->counter, old, new); +} + static inline int atomic_xchg(atomic_t *v, int new) { return xchg(&v->counter, new); } +static inline int atomic_xchg_unchecked(atomic_unchecked_t *v, int new) +{ + return xchg(&v->counter, new); +} + /** * __atomic_add_unless - add unless the number is already a given value * @v: pointer of type atomic_t @@ -227,12 +431,25 @@ static inline int atomic_xchg(atomic_t *v, int new) */ static inline int __atomic_add_unless(atomic_t *v, int a, int u) { - int c, old; + int c, old, new; c = atomic_read(v); for (;;) { - if (unlikely(c == (u))) + if (unlikely(c == u)) break; - old = atomic_cmpxchg((v), c, c + (a)); + + asm volatile("addl %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "subl %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "=r" (new) + : "0" (c), "ir" (a)); + + old = atomic_cmpxchg(v, c, new); if (likely(old == c)) break; c = old; @@ -240,6 +457,48 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) return c; } +/** + * atomic_inc_not_zero_hint - increment if not null + * @v: pointer of type atomic_t + * @hint: probable value of the atomic before the increment + * + * This version of atomic_inc_not_zero() gives a hint of probable + * value of the atomic. This helps processor to not read the memory + * before doing the atomic read/modify/write cycle, lowering + * number of bus transactions on some arches. + * + * Returns: 0 if increment was not done, 1 otherwise. + */ +#define atomic_inc_not_zero_hint atomic_inc_not_zero_hint +static inline int atomic_inc_not_zero_hint(atomic_t *v, int hint) +{ + int val, c = hint, new; + + /* sanity test, should be removed by compiler if hint is a constant */ + if (!hint) + return __atomic_add_unless(v, 1, 0); + + do { + asm volatile("incl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "decl %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "=r" (new) + : "0" (c)); + + val = atomic_cmpxchg(v, c, new); + if (val == c) + return 1; + c = val; + } while (c); + + return 0; +} /* * atomic_dec_if_positive - decrement by 1 if old value positive diff --git a/arch/x86/include/asm/atomic64_32.h b/arch/x86/include/asm/atomic64_32.h index fa13f0e..27c2e08 100644 --- a/arch/x86/include/asm/atomic64_32.h +++ b/arch/x86/include/asm/atomic64_32.h @@ -12,6 +12,14 @@ typedef struct { u64 __aligned(8) counter; } atomic64_t; +#ifdef CONFIG_PAX_REFCOUNT +typedef struct { + u64 __aligned(8) counter; +} atomic64_unchecked_t; +#else +typedef atomic64_t atomic64_unchecked_t; +#endif + #define ATOMIC64_INIT(val) { (val) } #ifdef CONFIG_X86_CMPXCHG64 @@ -38,6 +46,21 @@ static inline long long atomic64_cmpxchg(atomic64_t *v, long long o, long long n } /** + * atomic64_cmpxchg_unchecked - cmpxchg atomic64 variable + * @p: pointer to type atomic64_unchecked_t + * @o: expected value + * @n: new value + * + * Atomically sets @v to @n if it was equal to @o and returns + * the old value. + */ + +static inline long long atomic64_cmpxchg_unchecked(atomic64_unchecked_t *v, long long o, long long n) +{ + return cmpxchg64(&v->counter, o, n); +} + +/** * atomic64_xchg - xchg atomic64 variable * @v: pointer to type atomic64_t * @n: value to assign @@ -77,6 +100,24 @@ static inline void atomic64_set(atomic64_t *v, long long i) } /** + * atomic64_set_unchecked - set atomic64 variable + * @v: pointer to type atomic64_unchecked_t + * @n: value to assign + * + * Atomically sets the value of @v to @n. + */ +static inline void atomic64_set_unchecked(atomic64_unchecked_t *v, long long i) +{ + unsigned high = (unsigned)(i >> 32); + unsigned low = (unsigned)i; + asm volatile(ATOMIC64_ALTERNATIVE(set) + : "+b" (low), "+c" (high) + : "S" (v) + : "eax", "edx", "memory" + ); +} + +/** * atomic64_read - read atomic64 variable * @v: pointer to type atomic64_t * @@ -93,6 +134,22 @@ static inline long long atomic64_read(const atomic64_t *v) } /** + * atomic64_read_unchecked - read atomic64 variable + * @v: pointer to type atomic64_unchecked_t + * + * Atomically reads the value of @v and returns it. + */ +static inline long long atomic64_read_unchecked(atomic64_unchecked_t *v) +{ + long long r; + asm volatile(ATOMIC64_ALTERNATIVE(read_unchecked) + : "=A" (r), "+c" (v) + : : "memory" + ); + return r; + } + +/** * atomic64_add_return - add and return * @i: integer value to add * @v: pointer to type atomic64_t @@ -108,6 +165,22 @@ static inline long long atomic64_add_return(long long i, atomic64_t *v) return i; } +/** + * atomic64_add_return_unchecked - add and return + * @i: integer value to add + * @v: pointer to type atomic64_unchecked_t + * + * Atomically adds @i to @v and returns @i + *@v + */ +static inline long long atomic64_add_return_unchecked(long long i, atomic64_unchecked_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE(add_return_unchecked) + : "+A" (i), "+c" (v) + : : "memory" + ); + return i; +} + /* * Other variants with different arithmetic operators: */ @@ -131,6 +204,17 @@ static inline long long atomic64_inc_return(atomic64_t *v) return a; } +static inline long long atomic64_inc_return_unchecked(atomic64_unchecked_t *v) +{ + long long a; + asm volatile(ATOMIC64_ALTERNATIVE(inc_return_unchecked) + : "=A" (a) + : "S" (v) + : "memory", "ecx" + ); + return a; +} + static inline long long atomic64_dec_return(atomic64_t *v) { long long a; @@ -159,6 +243,22 @@ static inline long long atomic64_add(long long i, atomic64_t *v) } /** + * atomic64_add_unchecked - add integer to atomic64 variable + * @i: integer value to add + * @v: pointer to type atomic64_unchecked_t + * + * Atomically adds @i to @v. + */ +static inline long long atomic64_add_unchecked(long long i, atomic64_unchecked_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE_(add_unchecked, add_return_unchecked) + : "+A" (i), "+c" (v) + : : "memory" + ); + return i; +} + +/** * atomic64_sub - subtract the atomic64 variable * @i: integer value to subtract * @v: pointer to type atomic64_t diff --git a/arch/x86/include/asm/atomic64_64.h b/arch/x86/include/asm/atomic64_64.h index 0e1cbfc..5623683 100644 --- a/arch/x86/include/asm/atomic64_64.h +++ b/arch/x86/include/asm/atomic64_64.h @@ -18,7 +18,19 @@ */ static inline long atomic64_read(const atomic64_t *v) { - return (*(volatile long *)&(v)->counter); + return (*(volatile const long *)&(v)->counter); +} + +/** + * atomic64_read_unchecked - read atomic64 variable + * @v: pointer of type atomic64_unchecked_t + * + * Atomically reads the value of @v. + * Doesn't imply a read memory barrier. + */ +static inline long atomic64_read_unchecked(const atomic64_unchecked_t *v) +{ + return (*(volatile const long *)&(v)->counter); } /** @@ -34,6 +46,18 @@ static inline void atomic64_set(atomic64_t *v, long i) } /** + * atomic64_set_unchecked - set atomic64 variable + * @v: pointer to type atomic64_unchecked_t + * @i: required value + * + * Atomically sets the value of @v to @i. + */ +static inline void atomic64_set_unchecked(atomic64_unchecked_t *v, long i) +{ + v->counter = i; +} + +/** * atomic64_add - add integer to atomic64 variable * @i: integer value to add * @v: pointer to type atomic64_t @@ -42,6 +66,28 @@ static inline void atomic64_set(atomic64_t *v, long i) */ static inline void atomic64_add(long i, atomic64_t *v) { + asm volatile(LOCK_PREFIX "addq %1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "subq %1,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "=m" (v->counter) + : "er" (i), "m" (v->counter)); +} + +/** + * atomic64_add_unchecked - add integer to atomic64 variable + * @i: integer value to add + * @v: pointer to type atomic64_unchecked_t + * + * Atomically adds @i to @v. + */ +static inline void atomic64_add_unchecked(long i, atomic64_unchecked_t *v) +{ asm volatile(LOCK_PREFIX "addq %1,%0" : "=m" (v->counter) : "er" (i), "m" (v->counter)); @@ -56,7 +102,29 @@ static inline void atomic64_add(long i, atomic64_t *v) */ static inline void atomic64_sub(long i, atomic64_t *v) { - asm volatile(LOCK_PREFIX "subq %1,%0" + asm volatile(LOCK_PREFIX "subq %1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "addq %1,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "=m" (v->counter) + : "er" (i), "m" (v->counter)); +} + +/** + * atomic64_sub_unchecked - subtract the atomic64 variable + * @i: integer value to subtract + * @v: pointer to type atomic64_unchecked_t + * + * Atomically subtracts @i from @v. + */ +static inline void atomic64_sub_unchecked(long i, atomic64_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "subq %1,%0\n" : "=m" (v->counter) : "er" (i), "m" (v->counter)); } @@ -74,7 +142,16 @@ static inline int atomic64_sub_and_test(long i, atomic64_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "subq %2,%0; sete %1" + asm volatile(LOCK_PREFIX "subq %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "addq %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "er" (i), "m" (v->counter) : "memory"); return c; @@ -88,6 +165,27 @@ static inline int atomic64_sub_and_test(long i, atomic64_t *v) */ static inline void atomic64_inc(atomic64_t *v) { + asm volatile(LOCK_PREFIX "incq %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "decq %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "=m" (v->counter) + : "m" (v->counter)); +} + +/** + * atomic64_inc_unchecked - increment atomic64 variable + * @v: pointer to type atomic64_unchecked_t + * + * Atomically increments @v by 1. + */ +static inline void atomic64_inc_unchecked(atomic64_unchecked_t *v) +{ asm volatile(LOCK_PREFIX "incq %0" : "=m" (v->counter) : "m" (v->counter)); @@ -101,7 +199,28 @@ static inline void atomic64_inc(atomic64_t *v) */ static inline void atomic64_dec(atomic64_t *v) { - asm volatile(LOCK_PREFIX "decq %0" + asm volatile(LOCK_PREFIX "decq %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "incq %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "=m" (v->counter) + : "m" (v->counter)); +} + +/** + * atomic64_dec_unchecked - decrement atomic64 variable + * @v: pointer to type atomic64_t + * + * Atomically decrements @v by 1. + */ +static inline void atomic64_dec_unchecked(atomic64_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "decq %0\n" : "=m" (v->counter) : "m" (v->counter)); } @@ -118,7 +237,16 @@ static inline int atomic64_dec_and_test(atomic64_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "decq %0; sete %1" + asm volatile(LOCK_PREFIX "decq %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "incq %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "m" (v->counter) : "memory"); return c != 0; @@ -136,7 +264,16 @@ static inline int atomic64_inc_and_test(atomic64_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "incq %0; sete %1" + asm volatile(LOCK_PREFIX "incq %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "decq %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "m" (v->counter) : "memory"); return c != 0; @@ -155,7 +292,16 @@ static inline int atomic64_add_negative(long i, atomic64_t *v) { unsigned char c; - asm volatile(LOCK_PREFIX "addq %2,%0; sets %1" + asm volatile(LOCK_PREFIX "addq %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "subq %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sets %1\n" : "=m" (v->counter), "=qm" (c) : "er" (i), "m" (v->counter) : "memory"); return c; @@ -170,6 +316,18 @@ static inline int atomic64_add_negative(long i, atomic64_t *v) */ static inline long atomic64_add_return(long i, atomic64_t *v) { + return i + xadd_check_overflow(&v->counter, i); +} + +/** + * atomic64_add_return_unchecked - add and return + * @i: integer value to add + * @v: pointer to type atomic64_unchecked_t + * + * Atomically adds @i to @v and returns @i + @v + */ +static inline long atomic64_add_return_unchecked(long i, atomic64_unchecked_t *v) +{ return i + xadd(&v->counter, i); } @@ -179,6 +337,10 @@ static inline long atomic64_sub_return(long i, atomic64_t *v) } #define atomic64_inc_return(v) (atomic64_add_return(1, (v))) +static inline long atomic64_inc_return_unchecked(atomic64_unchecked_t *v) +{ + return atomic64_add_return_unchecked(1, v); +} #define atomic64_dec_return(v) (atomic64_sub_return(1, (v))) static inline long atomic64_cmpxchg(atomic64_t *v, long old, long new) @@ -186,6 +348,11 @@ static inline long atomic64_cmpxchg(atomic64_t *v, long old, long new) return cmpxchg(&v->counter, old, new); } +static inline long atomic64_cmpxchg_unchecked(atomic64_unchecked_t *v, long old, long new) +{ + return cmpxchg(&v->counter, old, new); +} + static inline long atomic64_xchg(atomic64_t *v, long new) { return xchg(&v->counter, new); @@ -202,17 +369,30 @@ static inline long atomic64_xchg(atomic64_t *v, long new) */ static inline int atomic64_add_unless(atomic64_t *v, long a, long u) { - long c, old; + long c, old, new; c = atomic64_read(v); for (;;) { - if (unlikely(c == (u))) + if (unlikely(c == u)) break; - old = atomic64_cmpxchg((v), c, c + (a)); + + asm volatile("add %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "sub %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + : "=r" (new) + : "0" (c), "ir" (a)); + + old = atomic64_cmpxchg(v, c, new); if (likely(old == c)) break; c = old; } - return c != (u); + return c != u; } #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) diff --git a/arch/x86/include/asm/bitops.h b/arch/x86/include/asm/bitops.h index b97596e..9bd48b06 100644 --- a/arch/x86/include/asm/bitops.h +++ b/arch/x86/include/asm/bitops.h @@ -38,7 +38,7 @@ * a mask operation on a byte. */ #define IS_IMMEDIATE(nr) (__builtin_constant_p(nr)) -#define CONST_MASK_ADDR(nr, addr) BITOP_ADDR((void *)(addr) + ((nr)>>3)) +#define CONST_MASK_ADDR(nr, addr) BITOP_ADDR((volatile void *)(addr) + ((nr)>>3)) #define CONST_MASK(nr) (1 << ((nr) & 7)) /** diff --git a/arch/x86/include/asm/boot.h b/arch/x86/include/asm/boot.h index 5e1a2ee..c9f9533 100644 --- a/arch/x86/include/asm/boot.h +++ b/arch/x86/include/asm/boot.h @@ -11,10 +11,15 @@ #include /* Physical address where kernel should be loaded. */ -#define LOAD_PHYSICAL_ADDR ((CONFIG_PHYSICAL_START \ +#define ____LOAD_PHYSICAL_ADDR ((CONFIG_PHYSICAL_START \ + (CONFIG_PHYSICAL_ALIGN - 1)) \ & ~(CONFIG_PHYSICAL_ALIGN - 1)) +#ifndef __ASSEMBLY__ +extern unsigned char __LOAD_PHYSICAL_ADDR[]; +#define LOAD_PHYSICAL_ADDR ((unsigned long)__LOAD_PHYSICAL_ADDR) +#endif + /* Minimum kernel alignment, as a power of two */ #ifdef CONFIG_X86_64 #define MIN_KERNEL_ALIGN_LG2 PMD_SHIFT diff --git a/arch/x86/include/asm/cache.h b/arch/x86/include/asm/cache.h index 48f99f1..d78ebf9 100644 --- a/arch/x86/include/asm/cache.h +++ b/arch/x86/include/asm/cache.h @@ -5,12 +5,13 @@ /* L1 cache line size */ #define L1_CACHE_SHIFT (CONFIG_X86_L1_CACHE_SHIFT) -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) +#define L1_CACHE_BYTES (_AC(1,UL) << L1_CACHE_SHIFT) #define __read_mostly __attribute__((__section__(".data..read_mostly"))) +#define __read_only __attribute__((__section__(".data..read_only"))) #define INTERNODE_CACHE_SHIFT CONFIG_X86_INTERNODE_CACHE_SHIFT -#define INTERNODE_CACHE_BYTES (1 << INTERNODE_CACHE_SHIFT) +#define INTERNODE_CACHE_BYTES (_AC(1,UL) << INTERNODE_CACHE_SHIFT) #ifdef CONFIG_X86_VSMP #ifdef CONFIG_SMP diff --git a/arch/x86/include/asm/cacheflush.h b/arch/x86/include/asm/cacheflush.h index 4e12668..501d239 100644 --- a/arch/x86/include/asm/cacheflush.h +++ b/arch/x86/include/asm/cacheflush.h @@ -26,7 +26,7 @@ static inline unsigned long get_page_memtype(struct page *pg) unsigned long pg_flags = pg->flags & _PGMT_MASK; if (pg_flags == _PGMT_DEFAULT) - return -1; + return ~0UL; else if (pg_flags == _PGMT_WC) return _PAGE_CACHE_WC; else if (pg_flags == _PGMT_UC_MINUS) diff --git a/arch/x86/include/asm/checksum_32.h b/arch/x86/include/asm/checksum_32.h index 46fc474..b02b0f9 100644 --- a/arch/x86/include/asm/checksum_32.h +++ b/arch/x86/include/asm/checksum_32.h @@ -31,6 +31,14 @@ asmlinkage __wsum csum_partial_copy_generic(const void *src, void *dst, int len, __wsum sum, int *src_err_ptr, int *dst_err_ptr); +asmlinkage __wsum csum_partial_copy_generic_to_user(const void *src, void *dst, + int len, __wsum sum, + int *src_err_ptr, int *dst_err_ptr); + +asmlinkage __wsum csum_partial_copy_generic_from_user(const void *src, void *dst, + int len, __wsum sum, + int *src_err_ptr, int *dst_err_ptr); + /* * Note: when you get a NULL pointer exception here this means someone * passed in an incorrect kernel address to one of these functions. @@ -50,7 +58,7 @@ static inline __wsum csum_partial_copy_from_user(const void __user *src, int *err_ptr) { might_sleep(); - return csum_partial_copy_generic((__force void *)src, dst, + return csum_partial_copy_generic_from_user((__force void *)src, dst, len, sum, err_ptr, NULL); } @@ -178,7 +186,7 @@ static inline __wsum csum_and_copy_to_user(const void *src, { might_sleep(); if (access_ok(VERIFY_WRITE, dst, len)) - return csum_partial_copy_generic(src, (__force void *)dst, + return csum_partial_copy_generic_to_user(src, (__force void *)dst, len, sum, NULL, err_ptr); if (len) diff --git a/arch/x86/include/asm/cmpxchg.h b/arch/x86/include/asm/cmpxchg.h index 99480e5..d81165b 100644 --- a/arch/x86/include/asm/cmpxchg.h +++ b/arch/x86/include/asm/cmpxchg.h @@ -14,8 +14,12 @@ extern void __cmpxchg_wrong_size(void) __compiletime_error("Bad argument size for cmpxchg"); extern void __xadd_wrong_size(void) __compiletime_error("Bad argument size for xadd"); +extern void __xadd_check_overflow_wrong_size(void) + __compiletime_error("Bad argument size for xadd_check_overflow"); extern void __add_wrong_size(void) __compiletime_error("Bad argument size for add"); +extern void __add_check_overflow_wrong_size(void) + __compiletime_error("Bad argument size for add_check_overflow"); /* * Constants for operation sizes. On 32-bit, the 64-bit size it set to @@ -67,6 +71,34 @@ extern void __add_wrong_size(void) __ret; \ }) +#define __xchg_op_check_overflow(ptr, arg, op, lock) \ + ({ \ + __typeof__ (*(ptr)) __ret = (arg); \ + switch (sizeof(*(ptr))) { \ + case __X86_CASE_L: \ + asm volatile (lock #op "l %0, %1\n" \ + "jno 0f\n" \ + "mov %0,%1\n" \ + "int $4\n0:\n" \ + _ASM_EXTABLE(0b, 0b) \ + : "+r" (__ret), "+m" (*(ptr)) \ + : : "memory", "cc"); \ + break; \ + case __X86_CASE_Q: \ + asm volatile (lock #op "q %q0, %1\n" \ + "jno 0f\n" \ + "mov %0,%1\n" \ + "int $4\n0:\n" \ + _ASM_EXTABLE(0b, 0b) \ + : "+r" (__ret), "+m" (*(ptr)) \ + : : "memory", "cc"); \ + break; \ + default: \ + __ ## op ## _check_overflow_wrong_size(); \ + } \ + __ret; \ + }) + /* * Note: no "lock" prefix even on SMP: xchg always implies lock anyway. * Since this is generally used to protect other memory information, we @@ -167,6 +199,9 @@ extern void __add_wrong_size(void) #define xadd_sync(ptr, inc) __xadd((ptr), (inc), "lock; ") #define xadd_local(ptr, inc) __xadd((ptr), (inc), "") +#define __xadd_check_overflow(ptr, inc, lock) __xchg_op_check_overflow((ptr), (inc), xadd, lock) +#define xadd_check_overflow(ptr, inc) __xadd_check_overflow((ptr), (inc), LOCK_PREFIX) + #define __add(ptr, inc, lock) \ ({ \ __typeof__ (*(ptr)) __ret = (inc); \ diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index 8d67d42..183d0eb 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -367,7 +367,7 @@ static __always_inline __pure bool __static_cpu_has(u16 bit) ".section .discard,\"aw\",@progbits\n" " .byte 0xff + (4f-3f) - (2b-1b)\n" /* size check */ ".previous\n" - ".section .altinstr_replacement,\"ax\"\n" + ".section .altinstr_replacement,\"a\"\n" "3: movb $1,%0\n" "4:\n" ".previous\n" diff --git a/arch/x86/include/asm/desc.h b/arch/x86/include/asm/desc.h index e95822d..a90010e 100644 --- a/arch/x86/include/asm/desc.h +++ b/arch/x86/include/asm/desc.h @@ -4,6 +4,7 @@ #include #include #include +#include #include @@ -16,6 +17,7 @@ static inline void fill_ldt(struct desc_struct *desc, const struct user_desc *in desc->type = (info->read_exec_only ^ 1) << 1; desc->type |= info->contents << 2; + desc->type |= info->seg_not_present ^ 1; desc->s = 1; desc->dpl = 0x3; @@ -34,19 +36,14 @@ static inline void fill_ldt(struct desc_struct *desc, const struct user_desc *in } extern struct desc_ptr idt_descr; -extern gate_desc idt_table[]; extern struct desc_ptr nmi_idt_descr; -extern gate_desc nmi_idt_table[]; - -struct gdt_page { - struct desc_struct gdt[GDT_ENTRIES]; -} __attribute__((aligned(PAGE_SIZE))); - -DECLARE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page); +extern gate_desc idt_table[256]; +extern gate_desc nmi_idt_table[256]; +extern struct desc_struct cpu_gdt_table[NR_CPUS][PAGE_SIZE / sizeof(struct desc_struct)]; static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu) { - return per_cpu(gdt_page, cpu).gdt; + return cpu_gdt_table[cpu]; } #ifdef CONFIG_X86_64 @@ -71,8 +68,14 @@ static inline void pack_gate(gate_desc *gate, unsigned char type, unsigned long base, unsigned dpl, unsigned flags, unsigned short seg) { - gate->a = (seg << 16) | (base & 0xffff); - gate->b = (base & 0xffff0000) | (((0x80 | type | (dpl << 5)) & 0xff) << 8); + gate->gate.offset_low = base; + gate->gate.seg = seg; + gate->gate.reserved = 0; + gate->gate.type = type; + gate->gate.s = 0; + gate->gate.dpl = dpl; + gate->gate.p = 1; + gate->gate.offset_high = base >> 16; } #endif @@ -117,12 +120,16 @@ static inline void paravirt_free_ldt(struct desc_struct *ldt, unsigned entries) static inline void native_write_idt_entry(gate_desc *idt, int entry, const gate_desc *gate) { + pax_open_kernel(); memcpy(&idt[entry], gate, sizeof(*gate)); + pax_close_kernel(); } static inline void native_write_ldt_entry(struct desc_struct *ldt, int entry, const void *desc) { + pax_open_kernel(); memcpy(&ldt[entry], desc, 8); + pax_close_kernel(); } static inline void @@ -136,7 +143,9 @@ native_write_gdt_entry(struct desc_struct *gdt, int entry, const void *desc, int default: size = sizeof(*gdt); break; } + pax_open_kernel(); memcpy(&gdt[entry], desc, size); + pax_close_kernel(); } static inline void pack_descriptor(struct desc_struct *desc, unsigned long base, @@ -209,7 +218,9 @@ static inline void native_set_ldt(const void *addr, unsigned int entries) static inline void native_load_tr_desc(void) { + pax_open_kernel(); asm volatile("ltr %w0"::"q" (GDT_ENTRY_TSS*8)); + pax_close_kernel(); } static inline void native_load_gdt(const struct desc_ptr *dtr) @@ -246,8 +257,10 @@ static inline void native_load_tls(struct thread_struct *t, unsigned int cpu) struct desc_struct *gdt = get_cpu_gdt_table(cpu); unsigned int i; + pax_open_kernel(); for (i = 0; i < GDT_ENTRY_TLS_ENTRIES; i++) gdt[GDT_ENTRY_TLS_MIN + i] = t->tls_array[i]; + pax_close_kernel(); } #define _LDT_empty(info) \ @@ -310,7 +323,7 @@ static inline void set_desc_limit(struct desc_struct *desc, unsigned long limit) } #ifdef CONFIG_X86_64 -static inline void set_nmi_gate(int gate, void *addr) +static inline void set_nmi_gate(int gate, const void *addr) { gate_desc s; @@ -319,7 +332,7 @@ static inline void set_nmi_gate(int gate, void *addr) } #endif -static inline void _set_gate(int gate, unsigned type, void *addr, +static inline void _set_gate(int gate, unsigned type, const void *addr, unsigned dpl, unsigned ist, unsigned seg) { gate_desc s; @@ -338,7 +351,7 @@ static inline void _set_gate(int gate, unsigned type, void *addr, * Pentium F0 0F bugfix can have resulted in the mapped * IDT being write-protected. */ -static inline void set_intr_gate(unsigned int n, void *addr) +static inline void set_intr_gate(unsigned int n, const void *addr) { BUG_ON((unsigned)n > 0xFF); _set_gate(n, GATE_INTERRUPT, addr, 0, 0, __KERNEL_CS); @@ -368,19 +381,19 @@ static inline void alloc_intr_gate(unsigned int n, void *addr) /* * This routine sets up an interrupt gate at directory privilege level 3. */ -static inline void set_system_intr_gate(unsigned int n, void *addr) +static inline void set_system_intr_gate(unsigned int n, const void *addr) { BUG_ON((unsigned)n > 0xFF); _set_gate(n, GATE_INTERRUPT, addr, 0x3, 0, __KERNEL_CS); } -static inline void set_system_trap_gate(unsigned int n, void *addr) +static inline void set_system_trap_gate(unsigned int n, const void *addr) { BUG_ON((unsigned)n > 0xFF); _set_gate(n, GATE_TRAP, addr, 0x3, 0, __KERNEL_CS); } -static inline void set_trap_gate(unsigned int n, void *addr) +static inline void set_trap_gate(unsigned int n, const void *addr) { BUG_ON((unsigned)n > 0xFF); _set_gate(n, GATE_TRAP, addr, 0, 0, __KERNEL_CS); @@ -389,19 +402,31 @@ static inline void set_trap_gate(unsigned int n, void *addr) static inline void set_task_gate(unsigned int n, unsigned int gdt_entry) { BUG_ON((unsigned)n > 0xFF); - _set_gate(n, GATE_TASK, (void *)0, 0, 0, (gdt_entry<<3)); + _set_gate(n, GATE_TASK, (const void *)0, 0, 0, (gdt_entry<<3)); } -static inline void set_intr_gate_ist(int n, void *addr, unsigned ist) +static inline void set_intr_gate_ist(int n, const void *addr, unsigned ist) { BUG_ON((unsigned)n > 0xFF); _set_gate(n, GATE_INTERRUPT, addr, 0, ist, __KERNEL_CS); } -static inline void set_system_intr_gate_ist(int n, void *addr, unsigned ist) +static inline void set_system_intr_gate_ist(int n, const void *addr, unsigned ist) { BUG_ON((unsigned)n > 0xFF); _set_gate(n, GATE_INTERRUPT, addr, 0x3, ist, __KERNEL_CS); } +#ifdef CONFIG_X86_32 +static inline void set_user_cs(unsigned long base, unsigned long limit, int cpu) +{ + struct desc_struct d; + + if (likely(limit)) + limit = (limit - 1UL) >> PAGE_SHIFT; + pack_descriptor(&d, base, limit, 0xFB, 0xC); + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_DEFAULT_USER_CS, &d, DESCTYPE_S); +} +#endif + #endif /* _ASM_X86_DESC_H */ diff --git a/arch/x86/include/asm/desc_defs.h b/arch/x86/include/asm/desc_defs.h index 278441f..b95a174 100644 --- a/arch/x86/include/asm/desc_defs.h +++ b/arch/x86/include/asm/desc_defs.h @@ -31,6 +31,12 @@ struct desc_struct { unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1; unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8; }; + struct { + u16 offset_low; + u16 seg; + unsigned reserved: 8, type: 4, s: 1, dpl: 2, p: 1; + unsigned offset_high: 16; + } gate; }; } __attribute__((packed)); diff --git a/arch/x86/include/asm/e820.h b/arch/x86/include/asm/e820.h index 3778256..c5d4fce 100644 --- a/arch/x86/include/asm/e820.h +++ b/arch/x86/include/asm/e820.h @@ -69,7 +69,7 @@ struct e820map { #define ISA_START_ADDRESS 0xa0000 #define ISA_END_ADDRESS 0x100000 -#define BIOS_BEGIN 0x000a0000 +#define BIOS_BEGIN 0x000c0000 #define BIOS_END 0x00100000 #define BIOS_ROM_BASE 0xffe00000 diff --git a/arch/x86/include/asm/elf.h b/arch/x86/include/asm/elf.h index 5f962df..7289f09 100644 --- a/arch/x86/include/asm/elf.h +++ b/arch/x86/include/asm/elf.h @@ -238,7 +238,25 @@ extern int force_personality32; the loader. We need to make sure that it is out of the way of the program that it will "exec", and that there is sufficient room for the brk. */ +#ifdef CONFIG_PAX_SEGMEXEC +#define ELF_ET_DYN_BASE ((current->mm->pax_flags & MF_PAX_SEGMEXEC) ? SEGMEXEC_TASK_SIZE/3*2 : TASK_SIZE/3*2) +#else #define ELF_ET_DYN_BASE (TASK_SIZE / 3 * 2) +#endif + +#ifdef CONFIG_PAX_ASLR +#ifdef CONFIG_X86_32 +#define PAX_ELF_ET_DYN_BASE 0x10000000UL + +#define PAX_DELTA_MMAP_LEN (current->mm->pax_flags & MF_PAX_SEGMEXEC ? 15 : 16) +#define PAX_DELTA_STACK_LEN (current->mm->pax_flags & MF_PAX_SEGMEXEC ? 15 : 16) +#else +#define PAX_ELF_ET_DYN_BASE 0x400000UL + +#define PAX_DELTA_MMAP_LEN ((test_thread_flag(TIF_IA32)) ? 16 : TASK_SIZE_MAX_SHIFT - PAGE_SHIFT - 3) +#define PAX_DELTA_STACK_LEN ((test_thread_flag(TIF_IA32)) ? 16 : TASK_SIZE_MAX_SHIFT - PAGE_SHIFT - 3) +#endif +#endif /* This yields a mask that user programs can use to figure out what instruction set this CPU supports. This could be done in user space, @@ -291,9 +309,7 @@ do { \ #define ARCH_DLINFO \ do { \ - if (vdso_enabled) \ - NEW_AUX_ENT(AT_SYSINFO_EHDR, \ - (unsigned long)current->mm->context.vdso); \ + NEW_AUX_ENT(AT_SYSINFO_EHDR, current->mm->context.vdso); \ } while (0) #define AT_SYSINFO 32 @@ -304,7 +320,7 @@ do { \ #endif /* !CONFIG_X86_32 */ -#define VDSO_CURRENT_BASE ((unsigned long)current->mm->context.vdso) +#define VDSO_CURRENT_BASE (current->mm->context.vdso) #define VDSO_ENTRY \ ((unsigned long)VDSO32_SYMBOL(VDSO_CURRENT_BASE, vsyscall)) @@ -318,9 +334,6 @@ extern int arch_setup_additional_pages(struct linux_binprm *bprm, extern int syscall32_setup_pages(struct linux_binprm *, int exstack); #define compat_arch_setup_additional_pages syscall32_setup_pages -extern unsigned long arch_randomize_brk(struct mm_struct *mm); -#define arch_randomize_brk arch_randomize_brk - /* * True on X86_32 or when emulating IA32 on X86_64 */ diff --git a/arch/x86/include/asm/emergency-restart.h b/arch/x86/include/asm/emergency-restart.h index cc70c1c..d96d011 100644 --- a/arch/x86/include/asm/emergency-restart.h +++ b/arch/x86/include/asm/emergency-restart.h @@ -15,6 +15,6 @@ enum reboot_type { extern enum reboot_type reboot_type; -extern void machine_emergency_restart(void); +extern void machine_emergency_restart(void) __noreturn; #endif /* _ASM_X86_EMERGENCY_RESTART_H */ diff --git a/arch/x86/include/asm/floppy.h b/arch/x86/include/asm/floppy.h index dbe82a5..c6d8a00 100644 --- a/arch/x86/include/asm/floppy.h +++ b/arch/x86/include/asm/floppy.h @@ -157,6 +157,7 @@ static unsigned long dma_mem_alloc(unsigned long size) } +static unsigned long vdma_mem_alloc(unsigned long size) __size_overflow(1); static unsigned long vdma_mem_alloc(unsigned long size) { return (unsigned long)vmalloc(size); diff --git a/arch/x86/include/asm/futex.h b/arch/x86/include/asm/futex.h index d09bb03..4ea4194 100644 --- a/arch/x86/include/asm/futex.h +++ b/arch/x86/include/asm/futex.h @@ -12,16 +12,18 @@ #include #define __futex_atomic_op1(insn, ret, oldval, uaddr, oparg) \ + typecheck(u32 __user *, uaddr); \ asm volatile("1:\t" insn "\n" \ "2:\t.section .fixup,\"ax\"\n" \ "3:\tmov\t%3, %1\n" \ "\tjmp\t2b\n" \ "\t.previous\n" \ _ASM_EXTABLE(1b, 3b) \ - : "=r" (oldval), "=r" (ret), "+m" (*uaddr) \ + : "=r" (oldval), "=r" (ret), "+m" (*(u32 __user *)____m(uaddr))\ : "i" (-EFAULT), "0" (oparg), "1" (0)) #define __futex_atomic_op2(insn, ret, oldval, uaddr, oparg) \ + typecheck(u32 __user *, uaddr); \ asm volatile("1:\tmovl %2, %0\n" \ "\tmovl\t%0, %3\n" \ "\t" insn "\n" \ @@ -34,7 +36,7 @@ _ASM_EXTABLE(1b, 4b) \ _ASM_EXTABLE(2b, 4b) \ : "=&a" (oldval), "=&r" (ret), \ - "+m" (*uaddr), "=&r" (tem) \ + "+m" (*(u32 __user *)____m(uaddr)), "=&r" (tem) \ : "r" (oparg), "i" (-EFAULT), "1" (0)) static inline int futex_atomic_op_inuser(int encoded_op, u32 __user *uaddr) @@ -61,10 +63,10 @@ static inline int futex_atomic_op_inuser(int encoded_op, u32 __user *uaddr) switch (op) { case FUTEX_OP_SET: - __futex_atomic_op1("xchgl %0, %2", ret, oldval, uaddr, oparg); + __futex_atomic_op1(__copyuser_seg"xchgl %0, %2", ret, oldval, uaddr, oparg); break; case FUTEX_OP_ADD: - __futex_atomic_op1(LOCK_PREFIX "xaddl %0, %2", ret, oldval, + __futex_atomic_op1(LOCK_PREFIX __copyuser_seg"xaddl %0, %2", ret, oldval, uaddr, oparg); break; case FUTEX_OP_OR: @@ -123,13 +125,13 @@ static inline int futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr, if (!access_ok(VERIFY_WRITE, uaddr, sizeof(u32))) return -EFAULT; - asm volatile("1:\t" LOCK_PREFIX "cmpxchgl %4, %2\n" + asm volatile("1:\t" LOCK_PREFIX __copyuser_seg"cmpxchgl %4, %2\n" "2:\t.section .fixup, \"ax\"\n" "3:\tmov %3, %0\n" "\tjmp 2b\n" "\t.previous\n" _ASM_EXTABLE(1b, 3b) - : "+r" (ret), "=a" (oldval), "+m" (*uaddr) + : "+r" (ret), "=a" (oldval), "+m" (*(u32 __user *)____m(uaddr)) : "i" (-EFAULT), "r" (newval), "1" (oldval) : "memory" ); diff --git a/arch/x86/include/asm/hw_irq.h b/arch/x86/include/asm/hw_irq.h index eb92a6e..b98b2f4 100644 --- a/arch/x86/include/asm/hw_irq.h +++ b/arch/x86/include/asm/hw_irq.h @@ -136,8 +136,8 @@ extern void setup_ioapic_dest(void); extern void enable_IO_APIC(void); /* Statistics */ -extern atomic_t irq_err_count; -extern atomic_t irq_mis_count; +extern atomic_unchecked_t irq_err_count; +extern atomic_unchecked_t irq_mis_count; /* EISA */ extern void eisa_set_level_irq(unsigned int irq); diff --git a/arch/x86/include/asm/i387.h b/arch/x86/include/asm/i387.h index 2479049..3fb9795 100644 --- a/arch/x86/include/asm/i387.h +++ b/arch/x86/include/asm/i387.h @@ -93,6 +93,11 @@ static inline int fxrstor_checking(struct i387_fxsave_struct *fx) { int err; +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if ((unsigned long)fx < PAX_USER_SHADOW_BASE) + fx = (struct i387_fxsave_struct __user *)((void *)fx + PAX_USER_SHADOW_BASE); +#endif + /* See comment in fxsave() below. */ #ifdef CONFIG_AS_FXSAVEQ asm volatile("1: fxrstorq %[fx]\n\t" @@ -122,6 +127,11 @@ static inline int fxsave_user(struct i387_fxsave_struct __user *fx) { int err; +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if ((unsigned long)fx < PAX_USER_SHADOW_BASE) + fx = (struct i387_fxsave_struct __user *)((void __user *)fx + PAX_USER_SHADOW_BASE); +#endif + /* * Clear the bytes not touched by the fxsave and reserved * for the SW usage. @@ -278,7 +288,7 @@ static inline int restore_fpu_checking(struct task_struct *tsk) "emms\n\t" /* clear stack tags */ "fildl %P[addr]", /* set F?P to defined value */ X86_FEATURE_FXSAVE_LEAK, - [addr] "m" (tsk->thread.fpu.has_fpu)); + [addr] "m" (init_tss[smp_processor_id()].x86_tss.sp0)); return fpu_restore_checking(&tsk->thread.fpu); } @@ -445,7 +455,7 @@ static inline bool interrupted_kernel_fpu_idle(void) static inline bool interrupted_user_mode(void) { struct pt_regs *regs = get_irq_regs(); - return regs && user_mode_vm(regs); + return regs && user_mode(regs); } /* diff --git a/arch/x86/include/asm/io.h b/arch/x86/include/asm/io.h index d8e8eef..99f81ae 100644 --- a/arch/x86/include/asm/io.h +++ b/arch/x86/include/asm/io.h @@ -194,6 +194,17 @@ extern void set_iounmap_nonlazy(void); #include +#define ARCH_HAS_VALID_PHYS_ADDR_RANGE +static inline int valid_phys_addr_range(unsigned long addr, size_t count) +{ + return ((addr + count + PAGE_SIZE - 1) >> PAGE_SHIFT) < (1ULL << (boot_cpu_data.x86_phys_bits - PAGE_SHIFT)) ? 1 : 0; +} + +static inline int valid_mmap_phys_addr_range(unsigned long pfn, size_t count) +{ + return (pfn + (count >> PAGE_SHIFT)) < (1ULL << (boot_cpu_data.x86_phys_bits - PAGE_SHIFT)) ? 1 : 0; +} + /* * Convert a virtual cached pointer to an uncached pointer */ diff --git a/arch/x86/include/asm/irqflags.h b/arch/x86/include/asm/irqflags.h index bba3cf8..06bc8da 100644 --- a/arch/x86/include/asm/irqflags.h +++ b/arch/x86/include/asm/irqflags.h @@ -141,6 +141,11 @@ static inline notrace unsigned long arch_local_irq_save(void) sti; \ sysexit +#define GET_CR0_INTO_RDI mov %cr0, %rdi +#define SET_RDI_INTO_CR0 mov %rdi, %cr0 +#define GET_CR3_INTO_RDI mov %cr3, %rdi +#define SET_RDI_INTO_CR3 mov %rdi, %cr3 + #else #define INTERRUPT_RETURN iret #define ENABLE_INTERRUPTS_SYSEXIT sti; sysexit diff --git a/arch/x86/include/asm/kprobes.h b/arch/x86/include/asm/kprobes.h index 5478825..839e88c 100644 --- a/arch/x86/include/asm/kprobes.h +++ b/arch/x86/include/asm/kprobes.h @@ -37,13 +37,8 @@ typedef u8 kprobe_opcode_t; #define RELATIVEJUMP_SIZE 5 #define RELATIVECALL_OPCODE 0xe8 #define RELATIVE_ADDR_SIZE 4 -#define MAX_STACK_SIZE 64 -#define MIN_STACK_SIZE(ADDR) \ - (((MAX_STACK_SIZE) < (((unsigned long)current_thread_info()) + \ - THREAD_SIZE - (unsigned long)(ADDR))) \ - ? (MAX_STACK_SIZE) \ - : (((unsigned long)current_thread_info()) + \ - THREAD_SIZE - (unsigned long)(ADDR))) +#define MAX_STACK_SIZE 64UL +#define MIN_STACK_SIZE(ADDR) min(MAX_STACK_SIZE, current->thread.sp0 - (unsigned long)(ADDR)) #define flush_insn_slot(p) do { } while (0) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 52d6640..136b3bd 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -663,7 +663,7 @@ struct kvm_x86_ops { int (*check_intercept)(struct kvm_vcpu *vcpu, struct x86_instruction_info *info, enum x86_intercept_stage stage); -}; +} __do_const; struct kvm_arch_async_pf { u32 token; @@ -694,7 +694,7 @@ void kvm_mmu_change_mmu_pages(struct kvm *kvm, unsigned int kvm_nr_mmu_pages); int load_pdptrs(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, unsigned long cr3); int emulator_write_phys(struct kvm_vcpu *vcpu, gpa_t gpa, - const void *val, int bytes); + const void *val, int bytes) __size_overflow(2); u8 kvm_get_guest_memory_type(struct kvm_vcpu *vcpu, gfn_t gfn); extern bool tdp_enabled; @@ -781,7 +781,7 @@ int fx_init(struct kvm_vcpu *vcpu); void kvm_mmu_flush_tlb(struct kvm_vcpu *vcpu); void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa, - const u8 *new, int bytes); + const u8 *new, int bytes) __size_overflow(2); int kvm_mmu_unprotect_page(struct kvm *kvm, gfn_t gfn); int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva); void __kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu); diff --git a/arch/x86/include/asm/local.h b/arch/x86/include/asm/local.h index 9cdae5d..300d20f 100644 --- a/arch/x86/include/asm/local.h +++ b/arch/x86/include/asm/local.h @@ -18,26 +18,58 @@ typedef struct { static inline void local_inc(local_t *l) { - asm volatile(_ASM_INC "%0" + asm volatile(_ASM_INC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_DEC "%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+m" (l->a.counter)); } static inline void local_dec(local_t *l) { - asm volatile(_ASM_DEC "%0" + asm volatile(_ASM_DEC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_INC "%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+m" (l->a.counter)); } static inline void local_add(long i, local_t *l) { - asm volatile(_ASM_ADD "%1,%0" + asm volatile(_ASM_ADD "%1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_SUB "%1,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+m" (l->a.counter) : "ir" (i)); } static inline void local_sub(long i, local_t *l) { - asm volatile(_ASM_SUB "%1,%0" + asm volatile(_ASM_SUB "%1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_ADD "%1,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+m" (l->a.counter) : "ir" (i)); } @@ -55,7 +87,16 @@ static inline int local_sub_and_test(long i, local_t *l) { unsigned char c; - asm volatile(_ASM_SUB "%2,%0; sete %1" + asm volatile(_ASM_SUB "%2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_ADD "%2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "+m" (l->a.counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -73,7 +114,16 @@ static inline int local_dec_and_test(local_t *l) { unsigned char c; - asm volatile(_ASM_DEC "%0; sete %1" + asm volatile(_ASM_DEC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_INC "%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "+m" (l->a.counter), "=qm" (c) : : "memory"); return c != 0; @@ -91,7 +141,16 @@ static inline int local_inc_and_test(local_t *l) { unsigned char c; - asm volatile(_ASM_INC "%0; sete %1" + asm volatile(_ASM_INC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_DEC "%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "+m" (l->a.counter), "=qm" (c) : : "memory"); return c != 0; @@ -110,7 +169,16 @@ static inline int local_add_negative(long i, local_t *l) { unsigned char c; - asm volatile(_ASM_ADD "%2,%0; sets %1" + asm volatile(_ASM_ADD "%2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_SUB "%2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sets %1\n" : "+m" (l->a.counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -133,7 +201,15 @@ static inline long local_add_return(long i, local_t *l) #endif /* Modern 486+ processor */ __i = i; - asm volatile(_ASM_XADD "%0, %1;" + asm volatile(_ASM_XADD "%0, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + _ASM_MOV "%0,%1\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+r" (i), "+m" (l->a.counter) : : "memory"); return i + __i; diff --git a/arch/x86/include/asm/mman.h b/arch/x86/include/asm/mman.h index 593e51d..fa69c9a 100644 --- a/arch/x86/include/asm/mman.h +++ b/arch/x86/include/asm/mman.h @@ -5,4 +5,14 @@ #include +#ifdef __KERNEL__ +#ifndef __ASSEMBLY__ +#ifdef CONFIG_X86_32 +#define arch_mmap_check i386_mmap_check +int i386_mmap_check(unsigned long addr, unsigned long len, + unsigned long flags); +#endif +#endif +#endif + #endif /* _ASM_X86_MMAN_H */ diff --git a/arch/x86/include/asm/mmu.h b/arch/x86/include/asm/mmu.h index 5f55e69..e20bfb1 100644 --- a/arch/x86/include/asm/mmu.h +++ b/arch/x86/include/asm/mmu.h @@ -9,7 +9,7 @@ * we put the segment information here. */ typedef struct { - void *ldt; + struct desc_struct *ldt; int size; #ifdef CONFIG_X86_64 @@ -18,7 +18,19 @@ typedef struct { #endif struct mutex lock; - void *vdso; + unsigned long vdso; + +#ifdef CONFIG_X86_32 +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) + unsigned long user_cs_base; + unsigned long user_cs_limit; + +#if defined(CONFIG_PAX_PAGEEXEC) && defined(CONFIG_SMP) + cpumask_t cpu_user_cs_mask; +#endif + +#endif +#endif } mm_context_t; #ifdef CONFIG_SMP diff --git a/arch/x86/include/asm/mmu_context.h b/arch/x86/include/asm/mmu_context.h index 6902152..399f3a2 100644 --- a/arch/x86/include/asm/mmu_context.h +++ b/arch/x86/include/asm/mmu_context.h @@ -24,6 +24,18 @@ void destroy_context(struct mm_struct *mm); static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk) { + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + unsigned int i; + pgd_t *pgd; + + pax_open_kernel(); + pgd = get_cpu_pgd(smp_processor_id()); + for (i = USER_PGD_PTRS; i < 2 * USER_PGD_PTRS; ++i) + set_pgd_batched(pgd+i, native_make_pgd(0)); + pax_close_kernel(); +#endif + #ifdef CONFIG_SMP if (percpu_read(cpu_tlbstate.state) == TLBSTATE_OK) percpu_write(cpu_tlbstate.state, TLBSTATE_LAZY); @@ -34,16 +46,30 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, struct task_struct *tsk) { unsigned cpu = smp_processor_id(); +#if defined(CONFIG_X86_32) && defined(CONFIG_SMP) && (defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC)) + int tlbstate = TLBSTATE_OK; +#endif if (likely(prev != next)) { #ifdef CONFIG_SMP +#if defined(CONFIG_X86_32) && (defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC)) + tlbstate = percpu_read(cpu_tlbstate.state); +#endif percpu_write(cpu_tlbstate.state, TLBSTATE_OK); percpu_write(cpu_tlbstate.active_mm, next); #endif cpumask_set_cpu(cpu, mm_cpumask(next)); /* Re-load page tables */ +#ifdef CONFIG_PAX_PER_CPU_PGD + pax_open_kernel(); + __clone_user_pgds(get_cpu_pgd(cpu), next->pgd, USER_PGD_PTRS); + __shadow_user_pgds(get_cpu_pgd(cpu) + USER_PGD_PTRS, next->pgd, USER_PGD_PTRS); + pax_close_kernel(); + load_cr3(get_cpu_pgd(cpu)); +#else load_cr3(next->pgd); +#endif /* stop flush ipis for the previous mm */ cpumask_clear_cpu(cpu, mm_cpumask(prev)); @@ -53,9 +79,38 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, */ if (unlikely(prev->context.ldt != next->context.ldt)) load_LDT_nolock(&next->context); - } + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) && defined(CONFIG_SMP) + if (!(__supported_pte_mask & _PAGE_NX)) { + smp_mb__before_clear_bit(); + cpu_clear(cpu, prev->context.cpu_user_cs_mask); + smp_mb__after_clear_bit(); + cpu_set(cpu, next->context.cpu_user_cs_mask); + } +#endif + +#if defined(CONFIG_X86_32) && (defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC)) + if (unlikely(prev->context.user_cs_base != next->context.user_cs_base || + prev->context.user_cs_limit != next->context.user_cs_limit)) + set_user_cs(next->context.user_cs_base, next->context.user_cs_limit, cpu); #ifdef CONFIG_SMP + else if (unlikely(tlbstate != TLBSTATE_OK)) + set_user_cs(next->context.user_cs_base, next->context.user_cs_limit, cpu); +#endif +#endif + + } else { + +#ifdef CONFIG_PAX_PER_CPU_PGD + pax_open_kernel(); + __clone_user_pgds(get_cpu_pgd(cpu), next->pgd, USER_PGD_PTRS); + __shadow_user_pgds(get_cpu_pgd(cpu) + USER_PGD_PTRS, next->pgd, USER_PGD_PTRS); + pax_close_kernel(); + load_cr3(get_cpu_pgd(cpu)); +#endif + +#ifdef CONFIG_SMP percpu_write(cpu_tlbstate.state, TLBSTATE_OK); BUG_ON(percpu_read(cpu_tlbstate.active_mm) != next); @@ -64,11 +119,28 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, * tlb flush IPI delivery. We must reload CR3 * to make sure to use no freed page tables. */ + +#ifndef CONFIG_PAX_PER_CPU_PGD load_cr3(next->pgd); +#endif + load_LDT_nolock(&next->context); + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) + if (!(__supported_pte_mask & _PAGE_NX)) + cpu_set(cpu, next->context.cpu_user_cs_mask); +#endif + +#if defined(CONFIG_X86_32) && (defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC)) +#ifdef CONFIG_PAX_PAGEEXEC + if (!((next->pax_flags & MF_PAX_PAGEEXEC) && (__supported_pte_mask & _PAGE_NX))) +#endif + set_user_cs(next->context.user_cs_base, next->context.user_cs_limit, cpu); +#endif + } +#endif } -#endif } #define activate_mm(prev, next) \ diff --git a/arch/x86/include/asm/module.h b/arch/x86/include/asm/module.h index 9eae775..c914fea 100644 --- a/arch/x86/include/asm/module.h +++ b/arch/x86/include/asm/module.h @@ -5,6 +5,7 @@ #ifdef CONFIG_X86_64 /* X86_64 does not define MODULE_PROC_FAMILY */ +#define MODULE_PROC_FAMILY "" #elif defined CONFIG_M386 #define MODULE_PROC_FAMILY "386 " #elif defined CONFIG_M486 @@ -59,8 +60,20 @@ #error unknown processor family #endif -#ifdef CONFIG_X86_32 -# define MODULE_ARCH_VERMAGIC MODULE_PROC_FAMILY +#ifdef CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_BTS +#define MODULE_PAX_KERNEXEC "KERNEXEC_BTS " +#elif defined(CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_OR) +#define MODULE_PAX_KERNEXEC "KERNEXEC_OR " +#else +#define MODULE_PAX_KERNEXEC "" #endif +#ifdef CONFIG_PAX_MEMORY_UDEREF +#define MODULE_PAX_UDEREF "UDEREF " +#else +#define MODULE_PAX_UDEREF "" +#endif + +#define MODULE_ARCH_VERMAGIC MODULE_PROC_FAMILY MODULE_PAX_KERNEXEC MODULE_PAX_UDEREF + #endif /* _ASM_X86_MODULE_H */ diff --git a/arch/x86/include/asm/page_64_types.h b/arch/x86/include/asm/page_64_types.h index 7639dbf..e08a58c 100644 --- a/arch/x86/include/asm/page_64_types.h +++ b/arch/x86/include/asm/page_64_types.h @@ -56,7 +56,7 @@ void copy_page(void *to, void *from); /* duplicated to the one in bootmem.h */ extern unsigned long max_pfn; -extern unsigned long phys_base; +extern const unsigned long phys_base; extern unsigned long __phys_addr(unsigned long); #define __phys_reloc_hide(x) (x) diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index a7d2db9..edb023e 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -667,6 +667,18 @@ static inline void set_pgd(pgd_t *pgdp, pgd_t pgd) val); } +static inline void set_pgd_batched(pgd_t *pgdp, pgd_t pgd) +{ + pgdval_t val = native_pgd_val(pgd); + + if (sizeof(pgdval_t) > sizeof(long)) + PVOP_VCALL3(pv_mmu_ops.set_pgd_batched, pgdp, + val, (u64)val >> 32); + else + PVOP_VCALL2(pv_mmu_ops.set_pgd_batched, pgdp, + val); +} + static inline void pgd_clear(pgd_t *pgdp) { set_pgd(pgdp, __pgd(0)); @@ -748,6 +760,21 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx, pv_mmu_ops.set_fixmap(idx, phys, flags); } +#ifdef CONFIG_PAX_KERNEXEC +static inline unsigned long pax_open_kernel(void) +{ + return PVOP_CALL0(unsigned long, pv_mmu_ops.pax_open_kernel); +} + +static inline unsigned long pax_close_kernel(void) +{ + return PVOP_CALL0(unsigned long, pv_mmu_ops.pax_close_kernel); +} +#else +static inline unsigned long pax_open_kernel(void) { return 0; } +static inline unsigned long pax_close_kernel(void) { return 0; } +#endif + #if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS) static inline int arch_spin_is_locked(struct arch_spinlock *lock) @@ -964,7 +991,7 @@ extern void default_banner(void); #define PARA_PATCH(struct, off) ((PARAVIRT_PATCH_##struct + (off)) / 4) #define PARA_SITE(ptype, clobbers, ops) _PVSITE(ptype, clobbers, ops, .long, 4) -#define PARA_INDIRECT(addr) *%cs:addr +#define PARA_INDIRECT(addr) *%ss:addr #endif #define INTERRUPT_RETURN \ @@ -1041,6 +1068,21 @@ extern void default_banner(void); PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_irq_enable_sysexit), \ CLBR_NONE, \ jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_irq_enable_sysexit)) + +#define GET_CR0_INTO_RDI \ + call PARA_INDIRECT(pv_cpu_ops+PV_CPU_read_cr0); \ + mov %rax,%rdi + +#define SET_RDI_INTO_CR0 \ + call PARA_INDIRECT(pv_cpu_ops+PV_CPU_write_cr0) + +#define GET_CR3_INTO_RDI \ + call PARA_INDIRECT(pv_mmu_ops+PV_MMU_read_cr3); \ + mov %rax,%rdi + +#define SET_RDI_INTO_CR3 \ + call PARA_INDIRECT(pv_mmu_ops+PV_MMU_write_cr3) + #endif /* CONFIG_X86_32 */ #endif /* __ASSEMBLY__ */ diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index 8e8b9a4..f07d725 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -84,20 +84,20 @@ struct pv_init_ops { */ unsigned (*patch)(u8 type, u16 clobber, void *insnbuf, unsigned long addr, unsigned len); -}; +} __no_const; struct pv_lazy_ops { /* Set deferred update mode, used for batching operations. */ void (*enter)(void); void (*leave)(void); -}; +} __no_const; struct pv_time_ops { unsigned long long (*sched_clock)(void); unsigned long long (*steal_clock)(int cpu); unsigned long (*get_tsc_khz)(void); -}; +} __no_const; struct pv_cpu_ops { /* hooks for various privileged instructions */ @@ -193,7 +193,7 @@ struct pv_cpu_ops { void (*start_context_switch)(struct task_struct *prev); void (*end_context_switch)(struct task_struct *next); -}; +} __no_const; struct pv_irq_ops { /* @@ -224,7 +224,7 @@ struct pv_apic_ops { unsigned long start_eip, unsigned long start_esp); #endif -}; +} __no_const; struct pv_mmu_ops { unsigned long (*read_cr2)(void); @@ -313,6 +313,7 @@ struct pv_mmu_ops { struct paravirt_callee_save make_pud; void (*set_pgd)(pgd_t *pudp, pgd_t pgdval); + void (*set_pgd_batched)(pgd_t *pudp, pgd_t pgdval); #endif /* PAGETABLE_LEVELS == 4 */ #endif /* PAGETABLE_LEVELS >= 3 */ @@ -324,6 +325,12 @@ struct pv_mmu_ops { an mfn. We can tell which is which from the index. */ void (*set_fixmap)(unsigned /* enum fixed_addresses */ idx, phys_addr_t phys, pgprot_t flags); + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long (*pax_open_kernel)(void); + unsigned long (*pax_close_kernel)(void); +#endif + }; struct arch_spinlock; @@ -334,7 +341,7 @@ struct pv_lock_ops { void (*spin_lock_flags)(struct arch_spinlock *lock, unsigned long flags); int (*spin_trylock)(struct arch_spinlock *lock); void (*spin_unlock)(struct arch_spinlock *lock); -}; +} __no_const; /* This contains all the paravirt structures: we get a convenient * number for each function using the offset which we use to indicate diff --git a/arch/x86/include/asm/pgalloc.h b/arch/x86/include/asm/pgalloc.h index b4389a4..7024269 100644 --- a/arch/x86/include/asm/pgalloc.h +++ b/arch/x86/include/asm/pgalloc.h @@ -63,6 +63,13 @@ static inline void pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte) { paravirt_alloc_pte(mm, __pa(pte) >> PAGE_SHIFT); + set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE)); +} + +static inline void pmd_populate_user(struct mm_struct *mm, + pmd_t *pmd, pte_t *pte) +{ + paravirt_alloc_pte(mm, __pa(pte) >> PAGE_SHIFT); set_pmd(pmd, __pmd(__pa(pte) | _PAGE_TABLE)); } @@ -99,12 +106,22 @@ static inline void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd, #ifdef CONFIG_X86_PAE extern void pud_populate(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd); +static inline void pud_populate_kernel(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd) +{ + pud_populate(mm, pudp, pmd); +} #else /* !CONFIG_X86_PAE */ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) { paravirt_alloc_pmd(mm, __pa(pmd) >> PAGE_SHIFT); set_pud(pud, __pud(_PAGE_TABLE | __pa(pmd))); } + +static inline void pud_populate_kernel(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) +{ + paravirt_alloc_pmd(mm, __pa(pmd) >> PAGE_SHIFT); + set_pud(pud, __pud(_KERNPG_TABLE | __pa(pmd))); +} #endif /* CONFIG_X86_PAE */ #if PAGETABLE_LEVELS > 3 @@ -114,6 +131,12 @@ static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud) set_pgd(pgd, __pgd(_PAGE_TABLE | __pa(pud))); } +static inline void pgd_populate_kernel(struct mm_struct *mm, pgd_t *pgd, pud_t *pud) +{ + paravirt_alloc_pud(mm, __pa(pud) >> PAGE_SHIFT); + set_pgd(pgd, __pgd(_KERNPG_TABLE | __pa(pud))); +} + static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr) { return (pud_t *)get_zeroed_page(GFP_KERNEL|__GFP_REPEAT); diff --git a/arch/x86/include/asm/pgtable-2level.h b/arch/x86/include/asm/pgtable-2level.h index 98391db..8f6984e 100644 --- a/arch/x86/include/asm/pgtable-2level.h +++ b/arch/x86/include/asm/pgtable-2level.h @@ -18,7 +18,9 @@ static inline void native_set_pte(pte_t *ptep , pte_t pte) static inline void native_set_pmd(pmd_t *pmdp, pmd_t pmd) { + pax_open_kernel(); *pmdp = pmd; + pax_close_kernel(); } static inline void native_set_pte_atomic(pte_t *ptep, pte_t pte) diff --git a/arch/x86/include/asm/pgtable-3level.h b/arch/x86/include/asm/pgtable-3level.h index effff47..f9e4035 100644 --- a/arch/x86/include/asm/pgtable-3level.h +++ b/arch/x86/include/asm/pgtable-3level.h @@ -38,12 +38,16 @@ static inline void native_set_pte_atomic(pte_t *ptep, pte_t pte) static inline void native_set_pmd(pmd_t *pmdp, pmd_t pmd) { + pax_open_kernel(); set_64bit((unsigned long long *)(pmdp), native_pmd_val(pmd)); + pax_close_kernel(); } static inline void native_set_pud(pud_t *pudp, pud_t pud) { + pax_open_kernel(); set_64bit((unsigned long long *)(pudp), native_pud_val(pud)); + pax_close_kernel(); } /* diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h index 49afb3f..ed14d07 100644 --- a/arch/x86/include/asm/pgtable.h +++ b/arch/x86/include/asm/pgtable.h @@ -44,6 +44,7 @@ extern struct mm_struct *pgd_page_get_mm(struct page *page); #ifndef __PAGETABLE_PUD_FOLDED #define set_pgd(pgdp, pgd) native_set_pgd(pgdp, pgd) +#define set_pgd_batched(pgdp, pgd) native_set_pgd_batched(pgdp, pgd) #define pgd_clear(pgd) native_pgd_clear(pgd) #endif @@ -81,12 +82,51 @@ extern struct mm_struct *pgd_page_get_mm(struct page *page); #define arch_end_context_switch(prev) do {} while(0) +#define pax_open_kernel() native_pax_open_kernel() +#define pax_close_kernel() native_pax_close_kernel() #endif /* CONFIG_PARAVIRT */ +#define __HAVE_ARCH_PAX_OPEN_KERNEL +#define __HAVE_ARCH_PAX_CLOSE_KERNEL + +#ifdef CONFIG_PAX_KERNEXEC +static inline unsigned long native_pax_open_kernel(void) +{ + unsigned long cr0; + + preempt_disable(); + barrier(); + cr0 = read_cr0() ^ X86_CR0_WP; + BUG_ON(unlikely(cr0 & X86_CR0_WP)); + write_cr0(cr0); + return cr0 ^ X86_CR0_WP; +} + +static inline unsigned long native_pax_close_kernel(void) +{ + unsigned long cr0; + + cr0 = read_cr0() ^ X86_CR0_WP; + BUG_ON(unlikely(!(cr0 & X86_CR0_WP))); + write_cr0(cr0); + barrier(); + preempt_enable_no_resched(); + return cr0 ^ X86_CR0_WP; +} +#else +static inline unsigned long native_pax_open_kernel(void) { return 0; } +static inline unsigned long native_pax_close_kernel(void) { return 0; } +#endif + /* * The following only work if pte_present() is true. * Undefined behaviour if not.. */ +static inline int pte_user(pte_t pte) +{ + return pte_val(pte) & _PAGE_USER; +} + static inline int pte_dirty(pte_t pte) { return pte_flags(pte) & _PAGE_DIRTY; @@ -196,9 +236,29 @@ static inline pte_t pte_wrprotect(pte_t pte) return pte_clear_flags(pte, _PAGE_RW); } +static inline pte_t pte_mkread(pte_t pte) +{ + return __pte(pte_val(pte) | _PAGE_USER); +} + static inline pte_t pte_mkexec(pte_t pte) { - return pte_clear_flags(pte, _PAGE_NX); +#ifdef CONFIG_X86_PAE + if (__supported_pte_mask & _PAGE_NX) + return pte_clear_flags(pte, _PAGE_NX); + else +#endif + return pte_set_flags(pte, _PAGE_USER); +} + +static inline pte_t pte_exprotect(pte_t pte) +{ +#ifdef CONFIG_X86_PAE + if (__supported_pte_mask & _PAGE_NX) + return pte_set_flags(pte, _PAGE_NX); + else +#endif + return pte_clear_flags(pte, _PAGE_USER); } static inline pte_t pte_mkdirty(pte_t pte) @@ -390,6 +450,15 @@ pte_t *populate_extra_pte(unsigned long vaddr); #endif #ifndef __ASSEMBLY__ + +#ifdef CONFIG_PAX_PER_CPU_PGD +extern pgd_t cpu_pgd[NR_CPUS][PTRS_PER_PGD]; +static inline pgd_t *get_cpu_pgd(unsigned int cpu) +{ + return cpu_pgd[cpu]; +} +#endif + #include static inline int pte_none(pte_t pte) @@ -560,7 +629,7 @@ static inline pud_t *pud_offset(pgd_t *pgd, unsigned long address) static inline int pgd_bad(pgd_t pgd) { - return (pgd_flags(pgd) & ~_PAGE_USER) != _KERNPG_TABLE; + return (pgd_flags(pgd) & ~(_PAGE_USER | _PAGE_NX)) != _KERNPG_TABLE; } static inline int pgd_none(pgd_t pgd) @@ -583,7 +652,12 @@ static inline int pgd_none(pgd_t pgd) * pgd_offset() returns a (pgd_t *) * pgd_index() is used get the offset into the pgd page's array of pgd_t's; */ -#define pgd_offset(mm, address) ((mm)->pgd + pgd_index((address))) +#define pgd_offset(mm, address) ((mm)->pgd + pgd_index(address)) + +#ifdef CONFIG_PAX_PER_CPU_PGD +#define pgd_offset_cpu(cpu, address) (get_cpu_pgd(cpu) + pgd_index(address)) +#endif + /* * a shortcut which implies the use of the kernel's pgd, instead * of a process's @@ -594,6 +668,20 @@ static inline int pgd_none(pgd_t pgd) #define KERNEL_PGD_BOUNDARY pgd_index(PAGE_OFFSET) #define KERNEL_PGD_PTRS (PTRS_PER_PGD - KERNEL_PGD_BOUNDARY) +#ifdef CONFIG_X86_32 +#define USER_PGD_PTRS KERNEL_PGD_BOUNDARY +#else +#define TASK_SIZE_MAX_SHIFT CONFIG_TASK_SIZE_MAX_SHIFT +#define USER_PGD_PTRS (_AC(1,UL) << (TASK_SIZE_MAX_SHIFT - PGDIR_SHIFT)) + +#ifdef CONFIG_PAX_MEMORY_UDEREF +#define PAX_USER_SHADOW_BASE (_AC(1,UL) << TASK_SIZE_MAX_SHIFT) +#else +#define PAX_USER_SHADOW_BASE (_AC(0,UL)) +#endif + +#endif + #ifndef __ASSEMBLY__ extern int direct_gbpages; @@ -758,11 +846,23 @@ static inline void pmdp_set_wrprotect(struct mm_struct *mm, * dst and src can be on the same page, but the range must not overlap, * and must not cross a page boundary. */ -static inline void clone_pgd_range(pgd_t *dst, pgd_t *src, int count) +static inline void clone_pgd_range(pgd_t *dst, const pgd_t *src, int count) { - memcpy(dst, src, count * sizeof(pgd_t)); + pax_open_kernel(); + while (count--) + *dst++ = *src++; + pax_close_kernel(); } +#ifdef CONFIG_PAX_PER_CPU_PGD +extern void __clone_user_pgds(pgd_t *dst, const pgd_t *src, int count); +#endif + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +extern void __shadow_user_pgds(pgd_t *dst, const pgd_t *src, int count); +#else +static inline void __shadow_user_pgds(pgd_t *dst, const pgd_t *src, int count) {} +#endif #include #endif /* __ASSEMBLY__ */ diff --git a/arch/x86/include/asm/pgtable_32.h b/arch/x86/include/asm/pgtable_32.h index 0c92113..34a77c6 100644 --- a/arch/x86/include/asm/pgtable_32.h +++ b/arch/x86/include/asm/pgtable_32.h @@ -25,9 +25,6 @@ struct mm_struct; struct vm_area_struct; -extern pgd_t swapper_pg_dir[1024]; -extern pgd_t initial_page_table[1024]; - static inline void pgtable_cache_init(void) { } static inline void check_pgt_cache(void) { } void paging_init(void); @@ -48,6 +45,12 @@ extern void set_pmd_pfn(unsigned long, unsigned long, pgprot_t); # include #endif +extern pgd_t swapper_pg_dir[PTRS_PER_PGD]; +extern pgd_t initial_page_table[PTRS_PER_PGD]; +#ifdef CONFIG_X86_PAE +extern pmd_t swapper_pm_dir[PTRS_PER_PGD][PTRS_PER_PMD]; +#endif + #if defined(CONFIG_HIGHPTE) #define pte_offset_map(dir, address) \ ((pte_t *)kmap_atomic(pmd_page(*(dir))) + \ @@ -62,7 +65,9 @@ extern void set_pmd_pfn(unsigned long, unsigned long, pgprot_t); /* Clear a kernel PTE and flush it from the TLB */ #define kpte_clear_flush(ptep, vaddr) \ do { \ + pax_open_kernel(); \ pte_clear(&init_mm, (vaddr), (ptep)); \ + pax_close_kernel(); \ __flush_tlb_one((vaddr)); \ } while (0) @@ -74,6 +79,9 @@ do { \ #endif /* !__ASSEMBLY__ */ +#define HAVE_ARCH_UNMAPPED_AREA +#define HAVE_ARCH_UNMAPPED_AREA_TOPDOWN + /* * kern_addr_valid() is (1) for FLATMEM and (0) for * SPARSEMEM and DISCONTIGMEM diff --git a/arch/x86/include/asm/pgtable_32_types.h b/arch/x86/include/asm/pgtable_32_types.h index ed5903b..c7fe163 100644 --- a/arch/x86/include/asm/pgtable_32_types.h +++ b/arch/x86/include/asm/pgtable_32_types.h @@ -8,7 +8,7 @@ */ #ifdef CONFIG_X86_PAE # include -# define PMD_SIZE (1UL << PMD_SHIFT) +# define PMD_SIZE (_AC(1, UL) << PMD_SHIFT) # define PMD_MASK (~(PMD_SIZE - 1)) #else # include @@ -46,6 +46,19 @@ extern bool __vmalloc_start_set; /* set once high_memory is set */ # define VMALLOC_END (FIXADDR_START - 2 * PAGE_SIZE) #endif +#ifdef CONFIG_PAX_KERNEXEC +#ifndef __ASSEMBLY__ +extern unsigned char MODULES_EXEC_VADDR[]; +extern unsigned char MODULES_EXEC_END[]; +#endif +#include +#define ktla_ktva(addr) (addr + LOAD_PHYSICAL_ADDR + PAGE_OFFSET) +#define ktva_ktla(addr) (addr - LOAD_PHYSICAL_ADDR - PAGE_OFFSET) +#else +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) +#endif + #define MODULES_VADDR VMALLOC_START #define MODULES_END VMALLOC_END #define MODULES_LEN (MODULES_VADDR - MODULES_END) diff --git a/arch/x86/include/asm/pgtable_64.h b/arch/x86/include/asm/pgtable_64.h index 975f709..9f779c9 100644 --- a/arch/x86/include/asm/pgtable_64.h +++ b/arch/x86/include/asm/pgtable_64.h @@ -16,10 +16,14 @@ extern pud_t level3_kernel_pgt[512]; extern pud_t level3_ident_pgt[512]; +extern pud_t level3_vmalloc_start_pgt[512]; +extern pud_t level3_vmalloc_end_pgt[512]; +extern pud_t level3_vmemmap_pgt[512]; +extern pud_t level2_vmemmap_pgt[512]; extern pmd_t level2_kernel_pgt[512]; extern pmd_t level2_fixmap_pgt[512]; -extern pmd_t level2_ident_pgt[512]; -extern pgd_t init_level4_pgt[]; +extern pmd_t level2_ident_pgt[512*2]; +extern pgd_t init_level4_pgt[512]; #define swapper_pg_dir init_level4_pgt @@ -61,7 +65,9 @@ static inline void native_set_pte_atomic(pte_t *ptep, pte_t pte) static inline void native_set_pmd(pmd_t *pmdp, pmd_t pmd) { + pax_open_kernel(); *pmdp = pmd; + pax_close_kernel(); } static inline void native_pmd_clear(pmd_t *pmd) @@ -97,7 +103,9 @@ static inline pmd_t native_pmdp_get_and_clear(pmd_t *xp) static inline void native_set_pud(pud_t *pudp, pud_t pud) { + pax_open_kernel(); *pudp = pud; + pax_close_kernel(); } static inline void native_pud_clear(pud_t *pud) @@ -107,6 +115,13 @@ static inline void native_pud_clear(pud_t *pud) static inline void native_set_pgd(pgd_t *pgdp, pgd_t pgd) { + pax_open_kernel(); + *pgdp = pgd; + pax_close_kernel(); +} + +static inline void native_set_pgd_batched(pgd_t *pgdp, pgd_t pgd) +{ *pgdp = pgd; } diff --git a/arch/x86/include/asm/pgtable_64_types.h b/arch/x86/include/asm/pgtable_64_types.h index 766ea16..5b96cb3 100644 --- a/arch/x86/include/asm/pgtable_64_types.h +++ b/arch/x86/include/asm/pgtable_64_types.h @@ -59,5 +59,10 @@ typedef struct { pteval_t pte; } pte_t; #define MODULES_VADDR _AC(0xffffffffa0000000, UL) #define MODULES_END _AC(0xffffffffff000000, UL) #define MODULES_LEN (MODULES_END - MODULES_VADDR) +#define MODULES_EXEC_VADDR MODULES_VADDR +#define MODULES_EXEC_END MODULES_END + +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) #endif /* _ASM_X86_PGTABLE_64_DEFS_H */ diff --git a/arch/x86/include/asm/pgtable_types.h b/arch/x86/include/asm/pgtable_types.h index 013286a..8b42f4f 100644 --- a/arch/x86/include/asm/pgtable_types.h +++ b/arch/x86/include/asm/pgtable_types.h @@ -16,13 +16,12 @@ #define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page */ #define _PAGE_BIT_PAT 7 /* on 4KB pages */ #define _PAGE_BIT_GLOBAL 8 /* Global TLB entry PPro+ */ -#define _PAGE_BIT_UNUSED1 9 /* available for programmer */ +#define _PAGE_BIT_SPECIAL 9 /* special mappings, no associated struct page */ #define _PAGE_BIT_IOMAP 10 /* flag used to indicate IO mapping */ #define _PAGE_BIT_HIDDEN 11 /* hidden by kmemcheck */ #define _PAGE_BIT_PAT_LARGE 12 /* On 2MB or 1GB pages */ -#define _PAGE_BIT_SPECIAL _PAGE_BIT_UNUSED1 -#define _PAGE_BIT_CPA_TEST _PAGE_BIT_UNUSED1 -#define _PAGE_BIT_SPLITTING _PAGE_BIT_UNUSED1 /* only valid on a PSE pmd */ +#define _PAGE_BIT_CPA_TEST _PAGE_BIT_SPECIAL +#define _PAGE_BIT_SPLITTING _PAGE_BIT_SPECIAL /* only valid on a PSE pmd */ #define _PAGE_BIT_NX 63 /* No execute: only valid after cpuid check */ /* If _PAGE_BIT_PRESENT is clear, we use these: */ @@ -40,7 +39,6 @@ #define _PAGE_DIRTY (_AT(pteval_t, 1) << _PAGE_BIT_DIRTY) #define _PAGE_PSE (_AT(pteval_t, 1) << _PAGE_BIT_PSE) #define _PAGE_GLOBAL (_AT(pteval_t, 1) << _PAGE_BIT_GLOBAL) -#define _PAGE_UNUSED1 (_AT(pteval_t, 1) << _PAGE_BIT_UNUSED1) #define _PAGE_IOMAP (_AT(pteval_t, 1) << _PAGE_BIT_IOMAP) #define _PAGE_PAT (_AT(pteval_t, 1) << _PAGE_BIT_PAT) #define _PAGE_PAT_LARGE (_AT(pteval_t, 1) << _PAGE_BIT_PAT_LARGE) @@ -57,8 +55,10 @@ #if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) #define _PAGE_NX (_AT(pteval_t, 1) << _PAGE_BIT_NX) -#else +#elif defined(CONFIG_KMEMCHECK) #define _PAGE_NX (_AT(pteval_t, 0)) +#else +#define _PAGE_NX (_AT(pteval_t, 1) << _PAGE_BIT_HIDDEN) #endif #define _PAGE_FILE (_AT(pteval_t, 1) << _PAGE_BIT_FILE) @@ -96,6 +96,9 @@ #define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \ _PAGE_ACCESSED) +#define PAGE_READONLY_NOEXEC PAGE_READONLY +#define PAGE_SHARED_NOEXEC PAGE_SHARED + #define __PAGE_KERNEL_EXEC \ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_GLOBAL) #define __PAGE_KERNEL (__PAGE_KERNEL_EXEC | _PAGE_NX) @@ -106,7 +109,7 @@ #define __PAGE_KERNEL_WC (__PAGE_KERNEL | _PAGE_CACHE_WC) #define __PAGE_KERNEL_NOCACHE (__PAGE_KERNEL | _PAGE_PCD | _PAGE_PWT) #define __PAGE_KERNEL_UC_MINUS (__PAGE_KERNEL | _PAGE_PCD) -#define __PAGE_KERNEL_VSYSCALL (__PAGE_KERNEL_RX | _PAGE_USER) +#define __PAGE_KERNEL_VSYSCALL (__PAGE_KERNEL_RO | _PAGE_USER) #define __PAGE_KERNEL_VVAR (__PAGE_KERNEL_RO | _PAGE_USER) #define __PAGE_KERNEL_VVAR_NOCACHE (__PAGE_KERNEL_VVAR | _PAGE_PCD | _PAGE_PWT) #define __PAGE_KERNEL_LARGE (__PAGE_KERNEL | _PAGE_PSE) @@ -168,8 +171,8 @@ * bits are combined, this will alow user to access the high address mapped * VDSO in the presence of CONFIG_COMPAT_VDSO */ -#define PTE_IDENT_ATTR 0x003 /* PRESENT+RW */ -#define PDE_IDENT_ATTR 0x067 /* PRESENT+RW+USER+DIRTY+ACCESSED */ +#define PTE_IDENT_ATTR 0x063 /* PRESENT+RW+DIRTY+ACCESSED */ +#define PDE_IDENT_ATTR 0x063 /* PRESENT+RW+DIRTY+ACCESSED */ #define PGD_IDENT_ATTR 0x001 /* PRESENT (no other attributes) */ #endif @@ -207,7 +210,17 @@ static inline pgdval_t pgd_flags(pgd_t pgd) { return native_pgd_val(pgd) & PTE_FLAGS_MASK; } +#endif +#if PAGETABLE_LEVELS == 3 +#include +#endif + +#if PAGETABLE_LEVELS == 2 +#include +#endif + +#ifndef __ASSEMBLY__ #if PAGETABLE_LEVELS > 3 typedef struct { pudval_t pud; } pud_t; @@ -221,8 +234,6 @@ static inline pudval_t native_pud_val(pud_t pud) return pud.pud; } #else -#include - static inline pudval_t native_pud_val(pud_t pud) { return native_pgd_val(pud.pgd); @@ -242,8 +253,6 @@ static inline pmdval_t native_pmd_val(pmd_t pmd) return pmd.pmd; } #else -#include - static inline pmdval_t native_pmd_val(pmd_t pmd) { return native_pgd_val(pmd.pud.pgd); @@ -283,7 +292,6 @@ typedef struct page *pgtable_t; extern pteval_t __supported_pte_mask; extern void set_nx(void); -extern int nx_enabled; #define pgprot_writecombine pgprot_writecombine extern pgprot_t pgprot_writecombine(pgprot_t prot); diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index 58545c9..fe6fc38e 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -266,7 +266,7 @@ struct tss_struct { } ____cacheline_aligned; -DECLARE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss); +extern struct tss_struct init_tss[NR_CPUS]; /* * Save the original ist values for checking stack pointers during debugging @@ -860,11 +860,18 @@ static inline void spin_lock_prefetch(const void *x) */ #define TASK_SIZE PAGE_OFFSET #define TASK_SIZE_MAX TASK_SIZE + +#ifdef CONFIG_PAX_SEGMEXEC +#define SEGMEXEC_TASK_SIZE (TASK_SIZE / 2) +#define STACK_TOP ((current->mm->pax_flags & MF_PAX_SEGMEXEC)?SEGMEXEC_TASK_SIZE:TASK_SIZE) +#else #define STACK_TOP TASK_SIZE -#define STACK_TOP_MAX STACK_TOP +#endif + +#define STACK_TOP_MAX TASK_SIZE #define INIT_THREAD { \ - .sp0 = sizeof(init_stack) + (long)&init_stack, \ + .sp0 = sizeof(init_stack) + (long)&init_stack - 8, \ .vm86_info = NULL, \ .sysenter_cs = __KERNEL_CS, \ .io_bitmap_ptr = NULL, \ @@ -878,7 +885,7 @@ static inline void spin_lock_prefetch(const void *x) */ #define INIT_TSS { \ .x86_tss = { \ - .sp0 = sizeof(init_stack) + (long)&init_stack, \ + .sp0 = sizeof(init_stack) + (long)&init_stack - 8, \ .ss0 = __KERNEL_DS, \ .ss1 = __KERNEL_CS, \ .io_bitmap_base = INVALID_IO_BITMAP_OFFSET, \ @@ -889,11 +896,7 @@ static inline void spin_lock_prefetch(const void *x) extern unsigned long thread_saved_pc(struct task_struct *tsk); #define THREAD_SIZE_LONGS (THREAD_SIZE/sizeof(unsigned long)) -#define KSTK_TOP(info) \ -({ \ - unsigned long *__ptr = (unsigned long *)(info); \ - (unsigned long)(&__ptr[THREAD_SIZE_LONGS]); \ -}) +#define KSTK_TOP(info) ((container_of(info, struct task_struct, tinfo))->thread.sp0) /* * The below -8 is to reserve 8 bytes on top of the ring0 stack. @@ -908,7 +911,7 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk); #define task_pt_regs(task) \ ({ \ struct pt_regs *__regs__; \ - __regs__ = (struct pt_regs *)(KSTK_TOP(task_stack_page(task))-8); \ + __regs__ = (struct pt_regs *)((task)->thread.sp0); \ __regs__ - 1; \ }) @@ -918,13 +921,13 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk); /* * User space process size. 47bits minus one guard page. */ -#define TASK_SIZE_MAX ((1UL << 47) - PAGE_SIZE) +#define TASK_SIZE_MAX ((1UL << TASK_SIZE_MAX_SHIFT) - PAGE_SIZE) /* This decides where the kernel will search for a free chunk of vm * space during mmap's. */ #define IA32_PAGE_OFFSET ((current->personality & ADDR_LIMIT_3GB) ? \ - 0xc0000000 : 0xFFFFe000) + 0xc0000000 : 0xFFFFf000) #define TASK_SIZE (test_thread_flag(TIF_IA32) ? \ IA32_PAGE_OFFSET : TASK_SIZE_MAX) @@ -935,11 +938,11 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk); #define STACK_TOP_MAX TASK_SIZE_MAX #define INIT_THREAD { \ - .sp0 = (unsigned long)&init_stack + sizeof(init_stack) \ + .sp0 = (unsigned long)&init_stack + sizeof(init_stack) - 16 \ } #define INIT_TSS { \ - .x86_tss.sp0 = (unsigned long)&init_stack + sizeof(init_stack) \ + .x86_tss.sp0 = (unsigned long)&init_stack + sizeof(init_stack) - 16 \ } /* @@ -961,6 +964,10 @@ extern void start_thread(struct pt_regs *regs, unsigned long new_ip, */ #define TASK_UNMAPPED_BASE (PAGE_ALIGN(TASK_SIZE / 3)) +#ifdef CONFIG_PAX_SEGMEXEC +#define SEGMEXEC_TASK_UNMAPPED_BASE (PAGE_ALIGN(SEGMEXEC_TASK_SIZE / 3)) +#endif + #define KSTK_EIP(task) (task_pt_regs(task)->ip) /* Get/set a process' ability to use the timestamp counter instruction */ diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h index 3566454..4bdfb8c 100644 --- a/arch/x86/include/asm/ptrace.h +++ b/arch/x86/include/asm/ptrace.h @@ -156,28 +156,29 @@ static inline unsigned long regs_return_value(struct pt_regs *regs) } /* - * user_mode_vm(regs) determines whether a register set came from user mode. + * user_mode(regs) determines whether a register set came from user mode. * This is true if V8086 mode was enabled OR if the register set was from * protected mode with RPL-3 CS value. This tricky test checks that with * one comparison. Many places in the kernel can bypass this full check - * if they have already ruled out V8086 mode, so user_mode(regs) can be used. + * if they have already ruled out V8086 mode, so user_mode_novm(regs) can + * be used. */ -static inline int user_mode(struct pt_regs *regs) +static inline int user_mode_novm(struct pt_regs *regs) { #ifdef CONFIG_X86_32 return (regs->cs & SEGMENT_RPL_MASK) == USER_RPL; #else - return !!(regs->cs & 3); + return !!(regs->cs & SEGMENT_RPL_MASK); #endif } -static inline int user_mode_vm(struct pt_regs *regs) +static inline int user_mode(struct pt_regs *regs) { #ifdef CONFIG_X86_32 return ((regs->cs & SEGMENT_RPL_MASK) | (regs->flags & X86_VM_MASK)) >= USER_RPL; #else - return user_mode(regs); + return user_mode_novm(regs); #endif } @@ -193,15 +194,16 @@ static inline int v8086_mode(struct pt_regs *regs) #ifdef CONFIG_X86_64 static inline bool user_64bit_mode(struct pt_regs *regs) { + unsigned long cs = regs->cs & 0xffff; #ifndef CONFIG_PARAVIRT /* * On non-paravirt systems, this is the only long mode CPL 3 * selector. We do not allow long mode selectors in the LDT. */ - return regs->cs == __USER_CS; + return cs == __USER_CS; #else /* Headers are too twisted for this to go in paravirt.h. */ - return regs->cs == __USER_CS || regs->cs == pv_info.extra_user_64bit_cs; + return cs == __USER_CS || cs == pv_info.extra_user_64bit_cs; #endif } #endif diff --git a/arch/x86/include/asm/reboot.h b/arch/x86/include/asm/reboot.h index 92f29706..a79cbbb 100644 --- a/arch/x86/include/asm/reboot.h +++ b/arch/x86/include/asm/reboot.h @@ -6,19 +6,19 @@ struct pt_regs; struct machine_ops { - void (*restart)(char *cmd); - void (*halt)(void); - void (*power_off)(void); + void (* __noreturn restart)(char *cmd); + void (* __noreturn halt)(void); + void (* __noreturn power_off)(void); void (*shutdown)(void); void (*crash_shutdown)(struct pt_regs *); - void (*emergency_restart)(void); -}; + void (* __noreturn emergency_restart)(void); +} __no_const; extern struct machine_ops machine_ops; void native_machine_crash_shutdown(struct pt_regs *regs); void native_machine_shutdown(void); -void machine_real_restart(unsigned int type); +void machine_real_restart(unsigned int type) __noreturn; /* These must match dispatch_table in reboot_32.S */ #define MRR_BIOS 0 #define MRR_APM 1 diff --git a/arch/x86/include/asm/rwsem.h b/arch/x86/include/asm/rwsem.h index 2dbe4a7..ce1db00 100644 --- a/arch/x86/include/asm/rwsem.h +++ b/arch/x86/include/asm/rwsem.h @@ -64,6 +64,14 @@ static inline void __down_read(struct rw_semaphore *sem) { asm volatile("# beginning down_read\n\t" LOCK_PREFIX _ASM_INC "(%1)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX _ASM_DEC "(%1)\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + /* adds 0x00000001 */ " jns 1f\n" " call call_rwsem_down_read_failed\n" @@ -85,6 +93,14 @@ static inline int __down_read_trylock(struct rw_semaphore *sem) "1:\n\t" " mov %1,%2\n\t" " add %3,%2\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "sub %3,%2\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + " jle 2f\n\t" LOCK_PREFIX " cmpxchg %2,%0\n\t" " jnz 1b\n\t" @@ -104,6 +120,14 @@ static inline void __down_write_nested(struct rw_semaphore *sem, int subclass) long tmp; asm volatile("# beginning down_write\n\t" LOCK_PREFIX " xadd %1,(%2)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "mov %1,(%2)\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + /* adds 0xffff0001, returns the old value */ " test %1,%1\n\t" /* was the count 0 before? */ @@ -141,6 +165,14 @@ static inline void __up_read(struct rw_semaphore *sem) long tmp; asm volatile("# beginning __up_read\n\t" LOCK_PREFIX " xadd %1,(%2)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "mov %1,(%2)\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + /* subtracts 1, returns the old value */ " jns 1f\n\t" " call call_rwsem_wake\n" /* expects old value in %edx */ @@ -159,6 +191,14 @@ static inline void __up_write(struct rw_semaphore *sem) long tmp; asm volatile("# beginning __up_write\n\t" LOCK_PREFIX " xadd %1,(%2)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "mov %1,(%2)\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + /* subtracts 0xffff0001, returns the old value */ " jns 1f\n\t" " call call_rwsem_wake\n" /* expects old value in %edx */ @@ -176,6 +216,14 @@ static inline void __downgrade_write(struct rw_semaphore *sem) { asm volatile("# beginning __downgrade_write\n\t" LOCK_PREFIX _ASM_ADD "%2,(%1)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX _ASM_SUB "%2,(%1)\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + /* * transitions 0xZZZZ0001 -> 0xYYYY0001 (i386) * 0xZZZZZZZZ00000001 -> 0xYYYYYYYY00000001 (x86_64) @@ -194,7 +242,15 @@ static inline void __downgrade_write(struct rw_semaphore *sem) */ static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem) { - asm volatile(LOCK_PREFIX _ASM_ADD "%1,%0" + asm volatile(LOCK_PREFIX _ASM_ADD "%1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX _ASM_SUB "%1,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+m" (sem->count) : "er" (delta)); } @@ -204,7 +260,7 @@ static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem) */ static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem) { - return delta + xadd(&sem->count, delta); + return delta + xadd_check_overflow(&sem->count, delta); } #endif /* __KERNEL__ */ diff --git a/arch/x86/include/asm/segment.h b/arch/x86/include/asm/segment.h index 5e64171..f58957e 100644 --- a/arch/x86/include/asm/segment.h +++ b/arch/x86/include/asm/segment.h @@ -64,10 +64,15 @@ * 26 - ESPFIX small SS * 27 - per-cpu [ offset to per-cpu data area ] * 28 - stack_canary-20 [ for stack protector ] - * 29 - unused - * 30 - unused + * 29 - PCI BIOS CS + * 30 - PCI BIOS DS * 31 - TSS for double fault handler */ +#define GDT_ENTRY_KERNEXEC_EFI_CS (1) +#define GDT_ENTRY_KERNEXEC_EFI_DS (2) +#define __KERNEXEC_EFI_CS (GDT_ENTRY_KERNEXEC_EFI_CS*8) +#define __KERNEXEC_EFI_DS (GDT_ENTRY_KERNEXEC_EFI_DS*8) + #define GDT_ENTRY_TLS_MIN 6 #define GDT_ENTRY_TLS_MAX (GDT_ENTRY_TLS_MIN + GDT_ENTRY_TLS_ENTRIES - 1) @@ -79,6 +84,8 @@ #define GDT_ENTRY_KERNEL_CS (GDT_ENTRY_KERNEL_BASE+0) +#define GDT_ENTRY_KERNEXEC_KERNEL_CS (4) + #define GDT_ENTRY_KERNEL_DS (GDT_ENTRY_KERNEL_BASE+1) #define GDT_ENTRY_TSS (GDT_ENTRY_KERNEL_BASE+4) @@ -104,6 +111,12 @@ #define __KERNEL_STACK_CANARY 0 #endif +#define GDT_ENTRY_PCIBIOS_CS (GDT_ENTRY_KERNEL_BASE+17) +#define __PCIBIOS_CS (GDT_ENTRY_PCIBIOS_CS * 8) + +#define GDT_ENTRY_PCIBIOS_DS (GDT_ENTRY_KERNEL_BASE+18) +#define __PCIBIOS_DS (GDT_ENTRY_PCIBIOS_DS * 8) + #define GDT_ENTRY_DOUBLEFAULT_TSS 31 /* @@ -141,7 +154,7 @@ */ /* Matches PNP_CS32 and PNP_CS16 (they must be consecutive) */ -#define SEGMENT_IS_PNP_CODE(x) (((x) & 0xf4) == GDT_ENTRY_PNPBIOS_BASE * 8) +#define SEGMENT_IS_PNP_CODE(x) (((x) & 0xFFFCU) == PNP_CS32 || ((x) & 0xFFFCU) == PNP_CS16) #else @@ -165,6 +178,8 @@ #define __USER32_CS (GDT_ENTRY_DEFAULT_USER32_CS*8+3) #define __USER32_DS __USER_DS +#define GDT_ENTRY_KERNEXEC_KERNEL_CS 7 + #define GDT_ENTRY_TSS 8 /* needs two entries */ #define GDT_ENTRY_LDT 10 /* needs two entries */ #define GDT_ENTRY_TLS_MIN 12 @@ -185,6 +200,7 @@ #endif #define __KERNEL_CS (GDT_ENTRY_KERNEL_CS*8) +#define __KERNEXEC_KERNEL_CS (GDT_ENTRY_KERNEXEC_KERNEL_CS*8) #define __KERNEL_DS (GDT_ENTRY_KERNEL_DS*8) #define __USER_DS (GDT_ENTRY_DEFAULT_USER_DS*8+3) #define __USER_CS (GDT_ENTRY_DEFAULT_USER_CS*8+3) diff --git a/arch/x86/include/asm/smp.h b/arch/x86/include/asm/smp.h index 0434c40..1714bf0 100644 --- a/arch/x86/include/asm/smp.h +++ b/arch/x86/include/asm/smp.h @@ -36,7 +36,7 @@ DECLARE_PER_CPU(cpumask_var_t, cpu_core_map); /* cpus sharing the last level cache: */ DECLARE_PER_CPU(cpumask_var_t, cpu_llc_shared_map); DECLARE_PER_CPU(u16, cpu_llc_id); -DECLARE_PER_CPU(int, cpu_number); +DECLARE_PER_CPU(unsigned int, cpu_number); static inline struct cpumask *cpu_sibling_mask(int cpu) { @@ -77,7 +77,7 @@ struct smp_ops { void (*send_call_func_ipi)(const struct cpumask *mask); void (*send_call_func_single_ipi)(int cpu); -}; +} __no_const; /* Globals due to paravirt */ extern void set_cpu_sibling_map(int cpu); @@ -192,14 +192,8 @@ extern unsigned disabled_cpus __cpuinitdata; extern int safe_smp_processor_id(void); #elif defined(CONFIG_X86_64_SMP) -#define raw_smp_processor_id() (percpu_read(cpu_number)) - -#define stack_smp_processor_id() \ -({ \ - struct thread_info *ti; \ - __asm__("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \ - ti->cpu; \ -}) +#define raw_smp_processor_id() (percpu_read(cpu_number)) +#define stack_smp_processor_id() raw_smp_processor_id() #define safe_smp_processor_id() smp_processor_id() #endif diff --git a/arch/x86/include/asm/spinlock.h b/arch/x86/include/asm/spinlock.h index a82c2bf..2198f61 100644 --- a/arch/x86/include/asm/spinlock.h +++ b/arch/x86/include/asm/spinlock.h @@ -175,6 +175,14 @@ static inline int arch_write_can_lock(arch_rwlock_t *lock) static inline void arch_read_lock(arch_rwlock_t *rw) { asm volatile(LOCK_PREFIX READ_LOCK_SIZE(dec) " (%0)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX READ_LOCK_SIZE(inc) " (%0)\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + "jns 1f\n" "call __read_lock_failed\n\t" "1:\n" @@ -184,6 +192,14 @@ static inline void arch_read_lock(arch_rwlock_t *rw) static inline void arch_write_lock(arch_rwlock_t *rw) { asm volatile(LOCK_PREFIX WRITE_LOCK_SUB(%1) "(%0)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX WRITE_LOCK_ADD(%1) "(%0)\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + "jz 1f\n" "call __write_lock_failed\n\t" "1:\n" @@ -213,13 +229,29 @@ static inline int arch_write_trylock(arch_rwlock_t *lock) static inline void arch_read_unlock(arch_rwlock_t *rw) { - asm volatile(LOCK_PREFIX READ_LOCK_SIZE(inc) " %0" + asm volatile(LOCK_PREFIX READ_LOCK_SIZE(inc) " %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX READ_LOCK_SIZE(dec) " %0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + :"+m" (rw->lock) : : "memory"); } static inline void arch_write_unlock(arch_rwlock_t *rw) { - asm volatile(LOCK_PREFIX WRITE_LOCK_ADD(%1) "%0" + asm volatile(LOCK_PREFIX WRITE_LOCK_ADD(%1) "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX WRITE_LOCK_SUB(%1) "%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+m" (rw->write) : "i" (RW_LOCK_BIAS) : "memory"); } diff --git a/arch/x86/include/asm/stackprotector.h b/arch/x86/include/asm/stackprotector.h index 1575177..cb23f52 100644 --- a/arch/x86/include/asm/stackprotector.h +++ b/arch/x86/include/asm/stackprotector.h @@ -48,7 +48,7 @@ * head_32 for boot CPU and setup_per_cpu_areas() for others. */ #define GDT_STACK_CANARY_INIT \ - [GDT_ENTRY_STACK_CANARY] = GDT_ENTRY_INIT(0x4090, 0, 0x18), + [GDT_ENTRY_STACK_CANARY] = GDT_ENTRY_INIT(0x4090, 0, 0x17), /* * Initialize the stackprotector canary value. @@ -113,7 +113,7 @@ static inline void setup_stack_canary_segment(int cpu) static inline void load_stack_canary_segment(void) { -#ifdef CONFIG_X86_32 +#if defined(CONFIG_X86_32) && !defined(CONFIG_PAX_MEMORY_UDEREF) asm volatile ("mov %0, %%gs" : : "r" (0)); #endif } diff --git a/arch/x86/include/asm/stacktrace.h b/arch/x86/include/asm/stacktrace.h index 70bbe39..4ae2bd4 100644 --- a/arch/x86/include/asm/stacktrace.h +++ b/arch/x86/include/asm/stacktrace.h @@ -11,28 +11,20 @@ extern int kstack_depth_to_print; -struct thread_info; +struct task_struct; struct stacktrace_ops; -typedef unsigned long (*walk_stack_t)(struct thread_info *tinfo, - unsigned long *stack, - unsigned long bp, - const struct stacktrace_ops *ops, - void *data, - unsigned long *end, - int *graph); +typedef unsigned long walk_stack_t(struct task_struct *task, + void *stack_start, + unsigned long *stack, + unsigned long bp, + const struct stacktrace_ops *ops, + void *data, + unsigned long *end, + int *graph); -extern unsigned long -print_context_stack(struct thread_info *tinfo, - unsigned long *stack, unsigned long bp, - const struct stacktrace_ops *ops, void *data, - unsigned long *end, int *graph); - -extern unsigned long -print_context_stack_bp(struct thread_info *tinfo, - unsigned long *stack, unsigned long bp, - const struct stacktrace_ops *ops, void *data, - unsigned long *end, int *graph); +extern walk_stack_t print_context_stack; +extern walk_stack_t print_context_stack_bp; /* Generic stack tracer with callbacks */ @@ -40,7 +32,7 @@ struct stacktrace_ops { void (*address)(void *data, unsigned long address, int reliable); /* On negative return stop dumping */ int (*stack)(void *data, char *name); - walk_stack_t walk_stack; + walk_stack_t *walk_stack; }; void dump_trace(struct task_struct *tsk, struct pt_regs *regs, diff --git a/arch/x86/include/asm/sys_ia32.h b/arch/x86/include/asm/sys_ia32.h index cb23852..2dde194 100644 --- a/arch/x86/include/asm/sys_ia32.h +++ b/arch/x86/include/asm/sys_ia32.h @@ -40,7 +40,7 @@ asmlinkage long sys32_rt_sigprocmask(int, compat_sigset_t __user *, compat_sigset_t __user *, unsigned int); asmlinkage long sys32_alarm(unsigned int); -asmlinkage long sys32_waitpid(compat_pid_t, unsigned int *, int); +asmlinkage long sys32_waitpid(compat_pid_t, unsigned int __user *, int); asmlinkage long sys32_sysfs(int, u32, u32); asmlinkage long sys32_sched_rr_get_interval(compat_pid_t, diff --git a/arch/x86/include/asm/syscalls.h b/arch/x86/include/asm/syscalls.h index f1d8b44..a4de8b7 100644 --- a/arch/x86/include/asm/syscalls.h +++ b/arch/x86/include/asm/syscalls.h @@ -30,7 +30,7 @@ long sys_clone(unsigned long, unsigned long, void __user *, void __user *, struct pt_regs *); /* kernel/ldt.c */ -asmlinkage int sys_modify_ldt(int, void __user *, unsigned long); +asmlinkage int sys_modify_ldt(int, void __user *, unsigned long) __size_overflow(3); /* kernel/signal.c */ long sys_rt_sigreturn(struct pt_regs *); diff --git a/arch/x86/include/asm/system.h b/arch/x86/include/asm/system.h index 2d2f01c..f985723 100644 --- a/arch/x86/include/asm/system.h +++ b/arch/x86/include/asm/system.h @@ -129,7 +129,7 @@ do { \ "call __switch_to\n\t" \ "movq "__percpu_arg([current_task])",%%rsi\n\t" \ __switch_canary \ - "movq %P[thread_info](%%rsi),%%r8\n\t" \ + "movq "__percpu_arg([thread_info])",%%r8\n\t" \ "movq %%rax,%%rdi\n\t" \ "testl %[_tif_fork],%P[ti_flags](%%r8)\n\t" \ "jnz ret_from_fork\n\t" \ @@ -140,7 +140,7 @@ do { \ [threadrsp] "i" (offsetof(struct task_struct, thread.sp)), \ [ti_flags] "i" (offsetof(struct thread_info, flags)), \ [_tif_fork] "i" (_TIF_FORK), \ - [thread_info] "i" (offsetof(struct task_struct, stack)), \ + [thread_info] "m" (current_tinfo), \ [current_task] "m" (current_task) \ __switch_canary_iparam \ : "memory", "cc" __EXTRA_CLOBBER) @@ -200,7 +200,7 @@ static inline unsigned long get_limit(unsigned long segment) { unsigned long __limit; asm("lsll %1,%0" : "=r" (__limit) : "r" (segment)); - return __limit + 1; + return __limit; } static inline void native_clts(void) @@ -397,13 +397,13 @@ void enable_hlt(void); void cpu_idle_wait(void); -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) extern void free_init_pages(char *what, unsigned long begin, unsigned long end); void default_idle(void); bool set_pm_idle_to_default(void); -void stop_this_cpu(void *dummy); +void stop_this_cpu(void *dummy) __noreturn; /* * Force strict CPU ordering. diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index cfd8144..a62ab95 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -10,6 +10,7 @@ #include #include #include +#include /* * low level task data that entry.S needs immediate access to @@ -24,7 +25,6 @@ struct exec_domain; #include struct thread_info { - struct task_struct *task; /* main task structure */ struct exec_domain *exec_domain; /* execution domain */ __u32 flags; /* low level flags */ __u32 status; /* thread synchronous flags */ @@ -34,19 +34,13 @@ struct thread_info { mm_segment_t addr_limit; struct restart_block restart_block; void __user *sysenter_return; -#ifdef CONFIG_X86_32 - unsigned long previous_esp; /* ESP of the previous stack in - case of nested (IRQ) stacks - */ - __u8 supervisor_stack[0]; -#endif + unsigned long lowest_stack; unsigned int sig_on_uaccess_error:1; unsigned int uaccess_err:1; /* uaccess failed */ }; -#define INIT_THREAD_INFO(tsk) \ +#define INIT_THREAD_INFO \ { \ - .task = &tsk, \ .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ @@ -57,7 +51,7 @@ struct thread_info { }, \ } -#define init_thread_info (init_thread_union.thread_info) +#define init_thread_info (init_thread_union.stack) #define init_stack (init_thread_union.stack) #else /* !__ASSEMBLY__ */ @@ -95,6 +89,7 @@ struct thread_info { #define TIF_BLOCKSTEP 25 /* set when we want DEBUGCTLMSR_BTF */ #define TIF_LAZY_MMU_UPDATES 27 /* task is updating the mmu lazily */ #define TIF_SYSCALL_TRACEPOINT 28 /* syscall tracepoint instrumentation */ +#define TIF_GRSEC_SETXID 29 /* update credentials on syscall entry/exit */ #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) @@ -116,16 +111,17 @@ struct thread_info { #define _TIF_BLOCKSTEP (1 << TIF_BLOCKSTEP) #define _TIF_LAZY_MMU_UPDATES (1 << TIF_LAZY_MMU_UPDATES) #define _TIF_SYSCALL_TRACEPOINT (1 << TIF_SYSCALL_TRACEPOINT) +#define _TIF_GRSEC_SETXID (1 << TIF_GRSEC_SETXID) /* work to do in syscall_trace_enter() */ #define _TIF_WORK_SYSCALL_ENTRY \ (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_EMU | _TIF_SYSCALL_AUDIT | \ - _TIF_SECCOMP | _TIF_SINGLESTEP | _TIF_SYSCALL_TRACEPOINT) + _TIF_SECCOMP | _TIF_SINGLESTEP | _TIF_SYSCALL_TRACEPOINT | _TIF_GRSEC_SETXID) /* work to do in syscall_trace_leave() */ #define _TIF_WORK_SYSCALL_EXIT \ (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | _TIF_SINGLESTEP | \ - _TIF_SYSCALL_TRACEPOINT) + _TIF_SYSCALL_TRACEPOINT | _TIF_GRSEC_SETXID) /* work to do on interrupt/exception return */ #define _TIF_WORK_MASK \ @@ -135,7 +131,7 @@ struct thread_info { /* work to do on any return to user space */ #define _TIF_ALLWORK_MASK \ - ((0x0000FFFF & ~_TIF_SECCOMP) | _TIF_SYSCALL_TRACEPOINT) + ((0x0000FFFF & ~_TIF_SECCOMP) | _TIF_SYSCALL_TRACEPOINT | _TIF_GRSEC_SETXID) /* Only used for 64 bit */ #define _TIF_DO_NOTIFY_MASK \ @@ -169,45 +165,40 @@ struct thread_info { ret; \ }) -#ifdef CONFIG_X86_32 - -#define STACK_WARN (THREAD_SIZE/8) -/* - * macros/functions for gaining access to the thread information structure - * - * preempt_count needs to be 1 initially, until the scheduler is functional. - */ -#ifndef __ASSEMBLY__ - - -/* how to get the current stack pointer from C */ -register unsigned long current_stack_pointer asm("esp") __used; - -/* how to get the thread information struct from C */ -static inline struct thread_info *current_thread_info(void) -{ - return (struct thread_info *) - (current_stack_pointer & ~(THREAD_SIZE - 1)); -} - -#else /* !__ASSEMBLY__ */ - +#ifdef __ASSEMBLY__ /* how to get the thread information struct from ASM */ #define GET_THREAD_INFO(reg) \ - movl $-THREAD_SIZE, reg; \ - andl %esp, reg + mov PER_CPU_VAR(current_tinfo), reg /* use this one if reg already contains %esp */ -#define GET_THREAD_INFO_WITH_ESP(reg) \ - andl $-THREAD_SIZE, reg +#define GET_THREAD_INFO_WITH_ESP(reg) GET_THREAD_INFO(reg) +#else +/* how to get the thread information struct from C */ +DECLARE_PER_CPU(struct thread_info *, current_tinfo); + +static __always_inline struct thread_info *current_thread_info(void) +{ + return percpu_read_stable(current_tinfo); +} +#endif + +#ifdef CONFIG_X86_32 + +#define STACK_WARN (THREAD_SIZE/8) +/* + * macros/functions for gaining access to the thread information structure + * + * preempt_count needs to be 1 initially, until the scheduler is functional. + */ +#ifndef __ASSEMBLY__ + +/* how to get the current stack pointer from C */ +register unsigned long current_stack_pointer asm("esp") __used; #endif #else /* X86_32 */ -#include -#define KERNEL_STACK_OFFSET (5*8) - /* * macros/functions for gaining access to the thread information structure * preempt_count needs to be 1 initially, until the scheduler is functional. @@ -215,27 +206,8 @@ static inline struct thread_info *current_thread_info(void) #ifndef __ASSEMBLY__ DECLARE_PER_CPU(unsigned long, kernel_stack); -static inline struct thread_info *current_thread_info(void) -{ - struct thread_info *ti; - ti = (void *)(percpu_read_stable(kernel_stack) + - KERNEL_STACK_OFFSET - THREAD_SIZE); - return ti; -} - -#else /* !__ASSEMBLY__ */ - -/* how to get the thread information struct from ASM */ -#define GET_THREAD_INFO(reg) \ - movq PER_CPU_VAR(kernel_stack),reg ; \ - subq $(THREAD_SIZE-KERNEL_STACK_OFFSET),reg - -/* - * Same if PER_CPU_VAR(kernel_stack) is, perhaps with some offset, already in - * a certain register (to be used in assembler memory operands). - */ -#define THREAD_INFO(reg, off) KERNEL_STACK_OFFSET+(off)-THREAD_SIZE(reg) - +/* how to get the current stack pointer from C */ +register unsigned long current_stack_pointer asm("rsp") __used; #endif #endif /* !X86_32 */ @@ -269,5 +241,16 @@ extern void arch_task_cache_init(void); extern void free_thread_info(struct thread_info *ti); extern int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src); #define arch_task_cache_init arch_task_cache_init + +#define __HAVE_THREAD_FUNCTIONS +#define task_thread_info(task) (&(task)->tinfo) +#define task_stack_page(task) ((task)->stack) +#define setup_thread_stack(p, org) do {} while (0) +#define end_of_stack(p) ((unsigned long *)task_stack_page(p) + 1) + +#define __HAVE_ARCH_TASK_STRUCT_ALLOCATOR +extern struct task_struct *alloc_task_struct_node(int node); +extern void free_task_struct(struct task_struct *); + #endif #endif /* _ASM_X86_THREAD_INFO_H */ diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h index 8be5f54..7ae826d 100644 --- a/arch/x86/include/asm/uaccess.h +++ b/arch/x86/include/asm/uaccess.h @@ -7,12 +7,15 @@ #include #include #include +#include #include #include #define VERIFY_READ 0 #define VERIFY_WRITE 1 +extern void check_object_size(const void *ptr, unsigned long n, bool to); + /* * The fs value determines whether argument validity checking should be * performed or not. If get_fs() == USER_DS, checking is performed, with @@ -28,7 +31,12 @@ #define get_ds() (KERNEL_DS) #define get_fs() (current_thread_info()->addr_limit) +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_MEMORY_UDEREF) +void __set_fs(mm_segment_t x); +void set_fs(mm_segment_t x); +#else #define set_fs(x) (current_thread_info()->addr_limit = (x)) +#endif #define segment_eq(a, b) ((a).seg == (b).seg) @@ -76,7 +84,33 @@ * checks that the pointer is in the user space range - after calling * this function, memory access functions may still return -EFAULT. */ -#define access_ok(type, addr, size) (likely(__range_not_ok(addr, size) == 0)) +#define __access_ok(type, addr, size) (likely(__range_not_ok(addr, size) == 0)) +#define access_ok(type, addr, size) \ +({ \ + long __size = size; \ + unsigned long __addr = (unsigned long)addr; \ + unsigned long __addr_ao = __addr & PAGE_MASK; \ + unsigned long __end_ao = __addr + __size - 1; \ + bool __ret_ao = __range_not_ok(__addr, __size) == 0; \ + if (__ret_ao && unlikely((__end_ao ^ __addr_ao) & PAGE_MASK)) { \ + while(__addr_ao <= __end_ao) { \ + char __c_ao; \ + __addr_ao += PAGE_SIZE; \ + if (__size > PAGE_SIZE) \ + cond_resched(); \ + if (__get_user(__c_ao, (char __user *)__addr)) \ + break; \ + if (type != VERIFY_WRITE) { \ + __addr = __addr_ao; \ + continue; \ + } \ + if (__put_user(__c_ao, (char __user *)__addr)) \ + break; \ + __addr = __addr_ao; \ + } \ + } \ + __ret_ao; \ +}) /* * The exception table consists of pairs of addresses: the first is the @@ -182,12 +216,20 @@ extern int __get_user_bad(void); asm volatile("call __put_user_" #size : "=a" (__ret_pu) \ : "0" ((typeof(*(ptr)))(x)), "c" (ptr) : "ebx") - +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define __copyuser_seg "gs;" +#define __COPYUSER_SET_ES "pushl %%gs; popl %%es\n" +#define __COPYUSER_RESTORE_ES "pushl %%ss; popl %%es\n" +#else +#define __copyuser_seg +#define __COPYUSER_SET_ES +#define __COPYUSER_RESTORE_ES +#endif #ifdef CONFIG_X86_32 #define __put_user_asm_u64(x, addr, err, errret) \ - asm volatile("1: movl %%eax,0(%2)\n" \ - "2: movl %%edx,4(%2)\n" \ + asm volatile("1: "__copyuser_seg"movl %%eax,0(%2)\n" \ + "2: "__copyuser_seg"movl %%edx,4(%2)\n" \ "3:\n" \ ".section .fixup,\"ax\"\n" \ "4: movl %3,%0\n" \ @@ -199,8 +241,8 @@ extern int __get_user_bad(void); : "A" (x), "r" (addr), "i" (errret), "0" (err)) #define __put_user_asm_ex_u64(x, addr) \ - asm volatile("1: movl %%eax,0(%1)\n" \ - "2: movl %%edx,4(%1)\n" \ + asm volatile("1: "__copyuser_seg"movl %%eax,0(%1)\n" \ + "2: "__copyuser_seg"movl %%edx,4(%1)\n" \ "3:\n" \ _ASM_EXTABLE(1b, 2b - 1b) \ _ASM_EXTABLE(2b, 3b - 2b) \ @@ -252,7 +294,7 @@ extern void __put_user_8(void); __typeof__(*(ptr)) __pu_val; \ __chk_user_ptr(ptr); \ might_fault(); \ - __pu_val = x; \ + __pu_val = (x); \ switch (sizeof(*(ptr))) { \ case 1: \ __put_user_x(1, __pu_val, ptr, __ret_pu); \ @@ -373,7 +415,7 @@ do { \ } while (0) #define __get_user_asm(x, addr, err, itype, rtype, ltype, errret) \ - asm volatile("1: mov"itype" %2,%"rtype"1\n" \ + asm volatile("1: "__copyuser_seg"mov"itype" %2,%"rtype"1\n"\ "2:\n" \ ".section .fixup,\"ax\"\n" \ "3: mov %3,%0\n" \ @@ -381,7 +423,7 @@ do { \ " jmp 2b\n" \ ".previous\n" \ _ASM_EXTABLE(1b, 3b) \ - : "=r" (err), ltype(x) \ + : "=r" (err), ltype (x) \ : "m" (__m(addr)), "i" (errret), "0" (err)) #define __get_user_size_ex(x, ptr, size) \ @@ -406,7 +448,7 @@ do { \ } while (0) #define __get_user_asm_ex(x, addr, itype, rtype, ltype) \ - asm volatile("1: mov"itype" %1,%"rtype"0\n" \ + asm volatile("1: "__copyuser_seg"mov"itype" %1,%"rtype"0\n"\ "2:\n" \ _ASM_EXTABLE(1b, 2b - 1b) \ : ltype(x) : "m" (__m(addr))) @@ -423,13 +465,24 @@ do { \ int __gu_err; \ unsigned long __gu_val; \ __get_user_size(__gu_val, (ptr), (size), __gu_err, -EFAULT); \ - (x) = (__force __typeof__(*(ptr)))__gu_val; \ + (x) = (__typeof__(*(ptr)))__gu_val; \ __gu_err; \ }) /* FIXME: this hack is definitely wrong -AK */ struct __large_struct { unsigned long buf[100]; }; -#define __m(x) (*(struct __large_struct __user *)(x)) +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define ____m(x) \ +({ \ + unsigned long ____x = (unsigned long)(x); \ + if (____x < PAX_USER_SHADOW_BASE) \ + ____x += PAX_USER_SHADOW_BASE; \ + (void __user *)____x; \ +}) +#else +#define ____m(x) (x) +#endif +#define __m(x) (*(struct __large_struct __user *)____m(x)) /* * Tell gcc we read from memory instead of writing: this is because @@ -437,7 +490,7 @@ struct __large_struct { unsigned long buf[100]; }; * aliasing issues. */ #define __put_user_asm(x, addr, err, itype, rtype, ltype, errret) \ - asm volatile("1: mov"itype" %"rtype"1,%2\n" \ + asm volatile("1: "__copyuser_seg"mov"itype" %"rtype"1,%2\n"\ "2:\n" \ ".section .fixup,\"ax\"\n" \ "3: mov %3,%0\n" \ @@ -445,10 +498,10 @@ struct __large_struct { unsigned long buf[100]; }; ".previous\n" \ _ASM_EXTABLE(1b, 3b) \ : "=r"(err) \ - : ltype(x), "m" (__m(addr)), "i" (errret), "0" (err)) + : ltype (x), "m" (__m(addr)), "i" (errret), "0" (err)) #define __put_user_asm_ex(x, addr, itype, rtype, ltype) \ - asm volatile("1: mov"itype" %"rtype"0,%1\n" \ + asm volatile("1: "__copyuser_seg"mov"itype" %"rtype"0,%1\n"\ "2:\n" \ _ASM_EXTABLE(1b, 2b - 1b) \ : : ltype(x), "m" (__m(addr))) @@ -487,8 +540,12 @@ struct __large_struct { unsigned long buf[100]; }; * On error, the variable @x is set to zero. */ +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define __get_user(x, ptr) get_user((x), (ptr)) +#else #define __get_user(x, ptr) \ __get_user_nocheck((x), (ptr), sizeof(*(ptr))) +#endif /** * __put_user: - Write a simple value into user space, with less checking. @@ -510,8 +567,12 @@ struct __large_struct { unsigned long buf[100]; }; * Returns zero on success, or -EFAULT on error. */ +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define __put_user(x, ptr) put_user((x), (ptr)) +#else #define __put_user(x, ptr) \ __put_user_nocheck((__typeof__(*(ptr)))(x), (ptr), sizeof(*(ptr))) +#endif #define __get_user_unaligned __get_user #define __put_user_unaligned __put_user @@ -529,7 +590,7 @@ struct __large_struct { unsigned long buf[100]; }; #define get_user_ex(x, ptr) do { \ unsigned long __gue_val; \ __get_user_size_ex((__gue_val), (ptr), (sizeof(*(ptr)))); \ - (x) = (__force __typeof__(*(ptr)))__gue_val; \ + (x) = (__typeof__(*(ptr)))__gue_val; \ } while (0) #ifdef CONFIG_X86_WP_WORKS_OK diff --git a/arch/x86/include/asm/uaccess_32.h b/arch/x86/include/asm/uaccess_32.h index 566e803..7183d0b 100644 --- a/arch/x86/include/asm/uaccess_32.h +++ b/arch/x86/include/asm/uaccess_32.h @@ -11,15 +11,15 @@ #include unsigned long __must_check __copy_to_user_ll - (void __user *to, const void *from, unsigned long n); + (void __user *to, const void *from, unsigned long n) __size_overflow(3); unsigned long __must_check __copy_from_user_ll - (void *to, const void __user *from, unsigned long n); + (void *to, const void __user *from, unsigned long n) __size_overflow(3); unsigned long __must_check __copy_from_user_ll_nozero - (void *to, const void __user *from, unsigned long n); + (void *to, const void __user *from, unsigned long n) __size_overflow(3); unsigned long __must_check __copy_from_user_ll_nocache - (void *to, const void __user *from, unsigned long n); + (void *to, const void __user *from, unsigned long n) __size_overflow(3); unsigned long __must_check __copy_from_user_ll_nocache_nozero - (void *to, const void __user *from, unsigned long n); + (void *to, const void __user *from, unsigned long n) __size_overflow(3); /** * __copy_to_user_inatomic: - Copy a block of data into user space, with less checking. @@ -41,8 +41,13 @@ unsigned long __must_check __copy_from_user_ll_nocache_nozero */ static __always_inline unsigned long __must_check +__copy_to_user_inatomic(void __user *to, const void *from, unsigned long n) __size_overflow(3); +static __always_inline unsigned long __must_check __copy_to_user_inatomic(void __user *to, const void *from, unsigned long n) { + if ((long)n < 0) + return n; + if (__builtin_constant_p(n)) { unsigned long ret; @@ -61,6 +66,8 @@ __copy_to_user_inatomic(void __user *to, const void *from, unsigned long n) return ret; } } + if (!__builtin_constant_p(n)) + check_object_size(from, n, true); return __copy_to_user_ll(to, from, n); } @@ -79,15 +86,23 @@ __copy_to_user_inatomic(void __user *to, const void *from, unsigned long n) * On success, this will be zero. */ static __always_inline unsigned long __must_check +__copy_to_user(void __user *to, const void *from, unsigned long n) __size_overflow(3); +static __always_inline unsigned long __must_check __copy_to_user(void __user *to, const void *from, unsigned long n) { might_fault(); + return __copy_to_user_inatomic(to, from, n); } static __always_inline unsigned long +__copy_from_user_inatomic(void *to, const void __user *from, unsigned long n) __size_overflow(3); +static __always_inline unsigned long __copy_from_user_inatomic(void *to, const void __user *from, unsigned long n) { + if ((long)n < 0) + return n; + /* Avoid zeroing the tail if the copy fails.. * If 'n' is constant and 1, 2, or 4, we do still zero on a failure, * but as the zeroing behaviour is only significant when n is not @@ -134,9 +149,15 @@ __copy_from_user_inatomic(void *to, const void __user *from, unsigned long n) * for explanation of why this is needed. */ static __always_inline unsigned long +__copy_from_user(void *to, const void __user *from, unsigned long n) __size_overflow(3); +static __always_inline unsigned long __copy_from_user(void *to, const void __user *from, unsigned long n) { might_fault(); + + if ((long)n < 0) + return n; + if (__builtin_constant_p(n)) { unsigned long ret; @@ -152,13 +173,21 @@ __copy_from_user(void *to, const void __user *from, unsigned long n) return ret; } } + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); return __copy_from_user_ll(to, from, n); } static __always_inline unsigned long __copy_from_user_nocache(void *to, + const void __user *from, unsigned long n) __size_overflow(3); +static __always_inline unsigned long __copy_from_user_nocache(void *to, const void __user *from, unsigned long n) { might_fault(); + + if ((long)n < 0) + return n; + if (__builtin_constant_p(n)) { unsigned long ret; @@ -179,17 +208,24 @@ static __always_inline unsigned long __copy_from_user_nocache(void *to, static __always_inline unsigned long __copy_from_user_inatomic_nocache(void *to, const void __user *from, + unsigned long n) __size_overflow(3); +static __always_inline unsigned long +__copy_from_user_inatomic_nocache(void *to, const void __user *from, unsigned long n) { - return __copy_from_user_ll_nocache_nozero(to, from, n); + if ((long)n < 0) + return n; + + return __copy_from_user_ll_nocache_nozero(to, from, n); } -unsigned long __must_check copy_to_user(void __user *to, - const void *from, unsigned long n); -unsigned long __must_check _copy_from_user(void *to, - const void __user *from, - unsigned long n); - +extern void copy_to_user_overflow(void) +#ifdef CONFIG_DEBUG_STRICT_USER_COPY_CHECKS + __compiletime_error("copy_to_user() buffer size is not provably correct") +#else + __compiletime_warning("copy_to_user() buffer size is not provably correct") +#endif +; extern void copy_from_user_overflow(void) #ifdef CONFIG_DEBUG_STRICT_USER_COPY_CHECKS @@ -199,17 +235,65 @@ extern void copy_from_user_overflow(void) #endif ; -static inline unsigned long __must_check copy_from_user(void *to, - const void __user *from, - unsigned long n) +/** + * copy_to_user: - Copy a block of data into user space. + * @to: Destination address, in user space. + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * + * Context: User context only. This function may sleep. + * + * Copy data from kernel space to user space. + * + * Returns number of bytes that could not be copied. + * On success, this will be zero. + */ +static inline unsigned long __must_check +copy_to_user(void __user *to, const void *from, unsigned long n) __size_overflow(3); +static inline unsigned long __must_check +copy_to_user(void __user *to, const void *from, unsigned long n) +{ + int sz = __compiletime_object_size(from); + + if (unlikely(sz != -1 && sz < n)) + copy_to_user_overflow(); + else if (access_ok(VERIFY_WRITE, to, n)) + n = __copy_to_user(to, from, n); + return n; +} + +/** + * copy_from_user: - Copy a block of data from user space. + * @to: Destination address, in kernel space. + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * + * Context: User context only. This function may sleep. + * + * Copy data from user space to kernel space. + * + * Returns number of bytes that could not be copied. + * On success, this will be zero. + * + * If some data could not be copied, this function will pad the copied + * data to the requested size using zero bytes. + */ +static inline unsigned long __must_check +copy_from_user(void *to, const void __user *from, unsigned long n) __size_overflow(3); +static inline unsigned long __must_check +copy_from_user(void *to, const void __user *from, unsigned long n) { int sz = __compiletime_object_size(to); - if (likely(sz == -1 || sz >= n)) - n = _copy_from_user(to, from, n); - else + if (unlikely(sz != -1 && sz < n)) copy_from_user_overflow(); - + else if (access_ok(VERIFY_READ, from, n)) + n = __copy_from_user(to, from, n); + else if ((long)n > 0) { + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); + memset(to, 0, n); + } return n; } @@ -235,7 +319,7 @@ long __must_check __strncpy_from_user(char *dst, #define strlen_user(str) strnlen_user(str, LONG_MAX) long strnlen_user(const char __user *str, long n); -unsigned long __must_check clear_user(void __user *mem, unsigned long len); -unsigned long __must_check __clear_user(void __user *mem, unsigned long len); +unsigned long __must_check clear_user(void __user *mem, unsigned long len) __size_overflow(2); +unsigned long __must_check __clear_user(void __user *mem, unsigned long len) __size_overflow(2); #endif /* _ASM_X86_UACCESS_32_H */ diff --git a/arch/x86/include/asm/uaccess_64.h b/arch/x86/include/asm/uaccess_64.h index 1c66d30..e294b5f 100644 --- a/arch/x86/include/asm/uaccess_64.h +++ b/arch/x86/include/asm/uaccess_64.h @@ -10,6 +10,9 @@ #include #include #include +#include + +#define set_fs(x) (current_thread_info()->addr_limit = (x)) /* * Copy To/From Userspace @@ -17,12 +20,14 @@ /* Handles exceptions in both to and from, but doesn't do access_ok */ __must_check unsigned long -copy_user_generic_string(void *to, const void *from, unsigned len); +copy_user_generic_string(void *to, const void *from, unsigned long len) __size_overflow(3); __must_check unsigned long -copy_user_generic_unrolled(void *to, const void *from, unsigned len); +copy_user_generic_unrolled(void *to, const void *from, unsigned long len) __size_overflow(3); static __always_inline __must_check unsigned long -copy_user_generic(void *to, const void *from, unsigned len) +copy_user_generic(void *to, const void *from, unsigned long len) __size_overflow(3); +static __always_inline __must_check unsigned long +copy_user_generic(void *to, const void *from, unsigned long len) { unsigned ret; @@ -32,142 +37,237 @@ copy_user_generic(void *to, const void *from, unsigned len) ASM_OUTPUT2("=a" (ret), "=D" (to), "=S" (from), "=d" (len)), "1" (to), "2" (from), "3" (len) - : "memory", "rcx", "r8", "r9", "r10", "r11"); + : "memory", "rcx", "r8", "r9", "r11"); return ret; } +static __always_inline __must_check unsigned long +__copy_to_user(void __user *to, const void *from, unsigned long len) __size_overflow(3); +static __always_inline __must_check unsigned long +__copy_from_user(void *to, const void __user *from, unsigned long len) __size_overflow(3); __must_check unsigned long -_copy_to_user(void __user *to, const void *from, unsigned len); -__must_check unsigned long -_copy_from_user(void *to, const void __user *from, unsigned len); -__must_check unsigned long -copy_in_user(void __user *to, const void __user *from, unsigned len); +copy_in_user(void __user *to, const void __user *from, unsigned long len) __size_overflow(3); static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, + unsigned long n) __size_overflow(3); +static inline unsigned long __must_check copy_from_user(void *to, + const void __user *from, unsigned long n) { - int sz = __compiletime_object_size(to); - might_fault(); - if (likely(sz == -1 || sz >= n)) - n = _copy_from_user(to, from, n); -#ifdef CONFIG_DEBUG_VM - else - WARN(1, "Buffer overflow detected!\n"); -#endif + + if (access_ok(VERIFY_READ, from, n)) + n = __copy_from_user(to, from, n); + else if (n < INT_MAX) { + if (!__builtin_constant_p(n)) + check_object_size(to, n, false); + memset(to, 0, n); + } return n; } static __always_inline __must_check -int copy_to_user(void __user *dst, const void *src, unsigned size) +int copy_to_user(void __user *dst, const void *src, unsigned long size) __size_overflow(3); +static __always_inline __must_check +int copy_to_user(void __user *dst, const void *src, unsigned long size) { might_fault(); - return _copy_to_user(dst, src, size); + if (access_ok(VERIFY_WRITE, dst, size)) + size = __copy_to_user(dst, src, size); + return size; } static __always_inline __must_check -int __copy_from_user(void *dst, const void __user *src, unsigned size) +unsigned long __copy_from_user(void *dst, const void __user *src, unsigned long size) __size_overflow(3); +static __always_inline __must_check +unsigned long __copy_from_user(void *dst, const void __user *src, unsigned long size) { - int ret = 0; + int sz = __compiletime_object_size(dst); + unsigned ret = 0; might_fault(); - if (!__builtin_constant_p(size)) - return copy_user_generic(dst, (__force void *)src, size); + + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!__access_ok(VERIFY_READ, src, size)) + return size; +#endif + + if (unlikely(sz != -1 && sz < size)) { +#ifdef CONFIG_DEBUG_VM + WARN(1, "Buffer overflow detected!\n"); +#endif + return size; + } + + if (!__builtin_constant_p(size)) { + check_object_size(dst, size, false); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)src < PAX_USER_SHADOW_BASE) + src += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic(dst, (__force_kernel const void *)src, size); + } switch (size) { - case 1:__get_user_asm(*(u8 *)dst, (u8 __user *)src, + case 1:__get_user_asm(*(u8 *)dst, (const u8 __user *)src, ret, "b", "b", "=q", 1); return ret; - case 2:__get_user_asm(*(u16 *)dst, (u16 __user *)src, + case 2:__get_user_asm(*(u16 *)dst, (const u16 __user *)src, ret, "w", "w", "=r", 2); return ret; - case 4:__get_user_asm(*(u32 *)dst, (u32 __user *)src, + case 4:__get_user_asm(*(u32 *)dst, (const u32 __user *)src, ret, "l", "k", "=r", 4); return ret; - case 8:__get_user_asm(*(u64 *)dst, (u64 __user *)src, + case 8:__get_user_asm(*(u64 *)dst, (const u64 __user *)src, ret, "q", "", "=r", 8); return ret; case 10: - __get_user_asm(*(u64 *)dst, (u64 __user *)src, + __get_user_asm(*(u64 *)dst, (const u64 __user *)src, ret, "q", "", "=r", 10); if (unlikely(ret)) return ret; __get_user_asm(*(u16 *)(8 + (char *)dst), - (u16 __user *)(8 + (char __user *)src), + (const u16 __user *)(8 + (const char __user *)src), ret, "w", "w", "=r", 2); return ret; case 16: - __get_user_asm(*(u64 *)dst, (u64 __user *)src, + __get_user_asm(*(u64 *)dst, (const u64 __user *)src, ret, "q", "", "=r", 16); if (unlikely(ret)) return ret; __get_user_asm(*(u64 *)(8 + (char *)dst), - (u64 __user *)(8 + (char __user *)src), + (const u64 __user *)(8 + (const char __user *)src), ret, "q", "", "=r", 8); return ret; default: - return copy_user_generic(dst, (__force void *)src, size); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)src < PAX_USER_SHADOW_BASE) + src += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic(dst, (__force_kernel const void *)src, size); } } static __always_inline __must_check -int __copy_to_user(void __user *dst, const void *src, unsigned size) +unsigned long __copy_to_user(void __user *dst, const void *src, unsigned long size) __size_overflow(3); +static __always_inline __must_check +unsigned long __copy_to_user(void __user *dst, const void *src, unsigned long size) { - int ret = 0; + int sz = __compiletime_object_size(src); + unsigned ret = 0; might_fault(); - if (!__builtin_constant_p(size)) - return copy_user_generic((__force void *)dst, src, size); + + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!__access_ok(VERIFY_WRITE, dst, size)) + return size; +#endif + + if (unlikely(sz != -1 && sz < size)) { +#ifdef CONFIG_DEBUG_VM + WARN(1, "Buffer overflow detected!\n"); +#endif + return size; + } + + if (!__builtin_constant_p(size)) { + check_object_size(src, size, true); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)dst < PAX_USER_SHADOW_BASE) + dst += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic((__force_kernel void *)dst, src, size); + } switch (size) { - case 1:__put_user_asm(*(u8 *)src, (u8 __user *)dst, + case 1:__put_user_asm(*(const u8 *)src, (u8 __user *)dst, ret, "b", "b", "iq", 1); return ret; - case 2:__put_user_asm(*(u16 *)src, (u16 __user *)dst, + case 2:__put_user_asm(*(const u16 *)src, (u16 __user *)dst, ret, "w", "w", "ir", 2); return ret; - case 4:__put_user_asm(*(u32 *)src, (u32 __user *)dst, + case 4:__put_user_asm(*(const u32 *)src, (u32 __user *)dst, ret, "l", "k", "ir", 4); return ret; - case 8:__put_user_asm(*(u64 *)src, (u64 __user *)dst, + case 8:__put_user_asm(*(const u64 *)src, (u64 __user *)dst, ret, "q", "", "er", 8); return ret; case 10: - __put_user_asm(*(u64 *)src, (u64 __user *)dst, + __put_user_asm(*(const u64 *)src, (u64 __user *)dst, ret, "q", "", "er", 10); if (unlikely(ret)) return ret; asm("":::"memory"); - __put_user_asm(4[(u16 *)src], 4 + (u16 __user *)dst, + __put_user_asm(4[(const u16 *)src], 4 + (u16 __user *)dst, ret, "w", "w", "ir", 2); return ret; case 16: - __put_user_asm(*(u64 *)src, (u64 __user *)dst, + __put_user_asm(*(const u64 *)src, (u64 __user *)dst, ret, "q", "", "er", 16); if (unlikely(ret)) return ret; asm("":::"memory"); - __put_user_asm(1[(u64 *)src], 1 + (u64 __user *)dst, + __put_user_asm(1[(const u64 *)src], 1 + (u64 __user *)dst, ret, "q", "", "er", 8); return ret; default: - return copy_user_generic((__force void *)dst, src, size); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)dst < PAX_USER_SHADOW_BASE) + dst += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic((__force_kernel void *)dst, src, size); } } static __always_inline __must_check -int __copy_in_user(void __user *dst, const void __user *src, unsigned size) +unsigned long __copy_in_user(void __user *dst, const void __user *src, unsigned long size) __size_overflow(3); +static __always_inline __must_check +unsigned long __copy_in_user(void __user *dst, const void __user *src, unsigned long size) { - int ret = 0; + unsigned ret = 0; might_fault(); - if (!__builtin_constant_p(size)) - return copy_user_generic((__force void *)dst, - (__force void *)src, size); + + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!__access_ok(VERIFY_READ, src, size)) + return size; + if (!__access_ok(VERIFY_WRITE, dst, size)) + return size; +#endif + + if (!__builtin_constant_p(size)) { + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)src < PAX_USER_SHADOW_BASE) + src += PAX_USER_SHADOW_BASE; + if ((unsigned long)dst < PAX_USER_SHADOW_BASE) + dst += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic((__force_kernel void *)dst, + (__force_kernel const void *)src, size); + } switch (size) { case 1: { u8 tmp; - __get_user_asm(tmp, (u8 __user *)src, + __get_user_asm(tmp, (const u8 __user *)src, ret, "b", "b", "=q", 1); if (likely(!ret)) __put_user_asm(tmp, (u8 __user *)dst, @@ -176,7 +276,7 @@ int __copy_in_user(void __user *dst, const void __user *src, unsigned size) } case 2: { u16 tmp; - __get_user_asm(tmp, (u16 __user *)src, + __get_user_asm(tmp, (const u16 __user *)src, ret, "w", "w", "=r", 2); if (likely(!ret)) __put_user_asm(tmp, (u16 __user *)dst, @@ -186,7 +286,7 @@ int __copy_in_user(void __user *dst, const void __user *src, unsigned size) case 4: { u32 tmp; - __get_user_asm(tmp, (u32 __user *)src, + __get_user_asm(tmp, (const u32 __user *)src, ret, "l", "k", "=r", 4); if (likely(!ret)) __put_user_asm(tmp, (u32 __user *)dst, @@ -195,7 +295,7 @@ int __copy_in_user(void __user *dst, const void __user *src, unsigned size) } case 8: { u64 tmp; - __get_user_asm(tmp, (u64 __user *)src, + __get_user_asm(tmp, (const u64 __user *)src, ret, "q", "", "=r", 8); if (likely(!ret)) __put_user_asm(tmp, (u64 __user *)dst, @@ -203,8 +303,16 @@ int __copy_in_user(void __user *dst, const void __user *src, unsigned size) return ret; } default: - return copy_user_generic((__force void *)dst, - (__force void *)src, size); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)src < PAX_USER_SHADOW_BASE) + src += PAX_USER_SHADOW_BASE; + if ((unsigned long)dst < PAX_USER_SHADOW_BASE) + dst += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic((__force_kernel void *)dst, + (__force_kernel const void *)src, size); } } @@ -215,39 +323,83 @@ __strncpy_from_user(char *dst, const char __user *src, long count); __must_check long strnlen_user(const char __user *str, long n); __must_check long __strnlen_user(const char __user *str, long n); __must_check long strlen_user(const char __user *str); -__must_check unsigned long clear_user(void __user *mem, unsigned long len); -__must_check unsigned long __clear_user(void __user *mem, unsigned long len); +__must_check unsigned long clear_user(void __user *mem, unsigned long len) __size_overflow(2); +__must_check unsigned long __clear_user(void __user *mem, unsigned long len) __size_overflow(2); static __must_check __always_inline int -__copy_from_user_inatomic(void *dst, const void __user *src, unsigned size) +__copy_from_user_inatomic(void *dst, const void __user *src, unsigned long size) __size_overflow(3); +static __must_check __always_inline int +__copy_from_user_inatomic(void *dst, const void __user *src, unsigned long size) { - return copy_user_generic(dst, (__force const void *)src, size); + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!__access_ok(VERIFY_READ, src, size)) + return size; + + if ((unsigned long)src < PAX_USER_SHADOW_BASE) + src += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic(dst, (__force_kernel const void *)src, size); } -static __must_check __always_inline int -__copy_to_user_inatomic(void __user *dst, const void *src, unsigned size) +static __must_check __always_inline unsigned long +__copy_to_user_inatomic(void __user *dst, const void *src, unsigned long size) __size_overflow(3); +static __must_check __always_inline unsigned long +__copy_to_user_inatomic(void __user *dst, const void *src, unsigned long size) { - return copy_user_generic((__force void *)dst, src, size); + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!__access_ok(VERIFY_WRITE, dst, size)) + return size; + + if ((unsigned long)dst < PAX_USER_SHADOW_BASE) + dst += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic((__force_kernel void *)dst, src, size); } -extern long __copy_user_nocache(void *dst, const void __user *src, - unsigned size, int zerorest); +extern unsigned long __copy_user_nocache(void *dst, const void __user *src, + unsigned long size, int zerorest) __size_overflow(3); -static inline int -__copy_from_user_nocache(void *dst, const void __user *src, unsigned size) +static inline unsigned long __copy_from_user_nocache(void *dst, const void __user *src, unsigned long size) __size_overflow(3); +static inline unsigned long __copy_from_user_nocache(void *dst, const void __user *src, unsigned long size) { might_sleep(); + + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!__access_ok(VERIFY_READ, src, size)) + return size; +#endif + return __copy_user_nocache(dst, src, size, 1); } -static inline int -__copy_from_user_inatomic_nocache(void *dst, const void __user *src, - unsigned size) +static inline unsigned long __copy_from_user_inatomic_nocache(void *dst, const void __user *src, + unsigned long size) __size_overflow(3); +static inline unsigned long __copy_from_user_inatomic_nocache(void *dst, const void __user *src, + unsigned long size) { + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!__access_ok(VERIFY_READ, src, size)) + return size; +#endif + return __copy_user_nocache(dst, src, size, 0); } -unsigned long -copy_user_handle_tail(char *to, char *from, unsigned len, unsigned zerorest); +extern unsigned long +copy_user_handle_tail(char __user *to, char __user *from, unsigned long len, unsigned zerorest) __size_overflow(3); #endif /* _ASM_X86_UACCESS_64_H */ diff --git a/arch/x86/include/asm/vdso.h b/arch/x86/include/asm/vdso.h index bb05228..d763d5b 100644 --- a/arch/x86/include/asm/vdso.h +++ b/arch/x86/include/asm/vdso.h @@ -11,7 +11,7 @@ extern const char VDSO32_PRELINK[]; #define VDSO32_SYMBOL(base, name) \ ({ \ extern const char VDSO32_##name[]; \ - (void *)(VDSO32_##name - VDSO32_PRELINK + (unsigned long)(base)); \ + (void __user *)(VDSO32_##name - VDSO32_PRELINK + (unsigned long)(base)); \ }) #endif diff --git a/arch/x86/include/asm/x86_init.h b/arch/x86/include/asm/x86_init.h index a609c39..7a68dc7 100644 --- a/arch/x86/include/asm/x86_init.h +++ b/arch/x86/include/asm/x86_init.h @@ -29,7 +29,7 @@ struct x86_init_mpparse { void (*mpc_oem_bus_info)(struct mpc_bus *m, char *name); void (*find_smp_config)(void); void (*get_smp_config)(unsigned int early); -}; +} __no_const; /** * struct x86_init_resources - platform specific resource related ops @@ -43,7 +43,7 @@ struct x86_init_resources { void (*probe_roms)(void); void (*reserve_resources)(void); char *(*memory_setup)(void); -}; +} __no_const; /** * struct x86_init_irqs - platform specific interrupt setup @@ -56,7 +56,7 @@ struct x86_init_irqs { void (*pre_vector_init)(void); void (*intr_init)(void); void (*trap_init)(void); -}; +} __no_const; /** * struct x86_init_oem - oem platform specific customizing functions @@ -66,7 +66,7 @@ struct x86_init_irqs { struct x86_init_oem { void (*arch_setup)(void); void (*banner)(void); -}; +} __no_const; /** * struct x86_init_mapping - platform specific initial kernel pagetable setup @@ -77,7 +77,7 @@ struct x86_init_oem { */ struct x86_init_mapping { void (*pagetable_reserve)(u64 start, u64 end); -}; +} __no_const; /** * struct x86_init_paging - platform specific paging functions @@ -87,7 +87,7 @@ struct x86_init_mapping { struct x86_init_paging { void (*pagetable_setup_start)(pgd_t *base); void (*pagetable_setup_done)(pgd_t *base); -}; +} __no_const; /** * struct x86_init_timers - platform specific timer setup @@ -102,7 +102,7 @@ struct x86_init_timers { void (*tsc_pre_init)(void); void (*timer_init)(void); void (*wallclock_init)(void); -}; +} __no_const; /** * struct x86_init_iommu - platform specific iommu setup @@ -110,7 +110,7 @@ struct x86_init_timers { */ struct x86_init_iommu { int (*iommu_init)(void); -}; +} __no_const; /** * struct x86_init_pci - platform specific pci init functions @@ -124,7 +124,7 @@ struct x86_init_pci { int (*init)(void); void (*init_irq)(void); void (*fixup_irqs)(void); -}; +} __no_const; /** * struct x86_init_ops - functions for platform specific setup @@ -140,7 +140,7 @@ struct x86_init_ops { struct x86_init_timers timers; struct x86_init_iommu iommu; struct x86_init_pci pci; -}; +} __no_const; /** * struct x86_cpuinit_ops - platform specific cpu hotplug setups @@ -149,7 +149,7 @@ struct x86_init_ops { struct x86_cpuinit_ops { void (*setup_percpu_clockev)(void); void (*fixup_cpu_id)(struct cpuinfo_x86 *c, int node); -}; +} __no_const; /** * struct x86_platform_ops - platform specific runtime functions @@ -171,7 +171,7 @@ struct x86_platform_ops { void (*nmi_init)(void); unsigned char (*get_nmi_reason)(void); int (*i8042_detect)(void); -}; +} __no_const; struct pci_dev; @@ -180,7 +180,7 @@ struct x86_msi_ops { void (*teardown_msi_irq)(unsigned int irq); void (*teardown_msi_irqs)(struct pci_dev *dev); void (*restore_msi_irqs)(struct pci_dev *dev, int irq); -}; +} __no_const; extern struct x86_init_ops x86_init; extern struct x86_cpuinit_ops x86_cpuinit; diff --git a/arch/x86/include/asm/xsave.h b/arch/x86/include/asm/xsave.h index c6ce245..ffbdab7 100644 --- a/arch/x86/include/asm/xsave.h +++ b/arch/x86/include/asm/xsave.h @@ -65,6 +65,11 @@ static inline int xsave_user(struct xsave_struct __user *buf) { int err; +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if ((unsigned long)buf < PAX_USER_SHADOW_BASE) + buf = (struct xsave_struct __user *)((void __user*)buf + PAX_USER_SHADOW_BASE); +#endif + /* * Clear the xsave header first, so that reserved fields are * initialized to zero. @@ -96,10 +101,15 @@ static inline int xsave_user(struct xsave_struct __user *buf) static inline int xrestore_user(struct xsave_struct __user *buf, u64 mask) { int err; - struct xsave_struct *xstate = ((__force struct xsave_struct *)buf); + struct xsave_struct *xstate = ((__force_kernel struct xsave_struct *)buf); u32 lmask = mask; u32 hmask = mask >> 32; +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if ((unsigned long)xstate < PAX_USER_SHADOW_BASE) + xstate = (struct xsave_struct *)((void *)xstate + PAX_USER_SHADOW_BASE); +#endif + __asm__ __volatile__("1: .byte " REX_PREFIX "0x0f,0xae,0x2f\n" "2:\n" ".section .fixup,\"ax\"\n" diff --git a/arch/x86/kernel/acpi/realmode/Makefile b/arch/x86/kernel/acpi/realmode/Makefile index 6a564ac..9b1340c 100644 --- a/arch/x86/kernel/acpi/realmode/Makefile +++ b/arch/x86/kernel/acpi/realmode/Makefile @@ -41,6 +41,9 @@ KBUILD_CFLAGS := $(LINUXINCLUDE) -g -Os -D_SETUP -D_WAKEUP -D__KERNEL__ \ $(call cc-option, -fno-stack-protector) \ $(call cc-option, -mpreferred-stack-boundary=2) KBUILD_CFLAGS += $(call cc-option, -m32) +ifdef CONSTIFY_PLUGIN +KBUILD_CFLAGS += $(CONSTIFY_PLUGIN) -fplugin-arg-constify_plugin-no-constify +endif KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ GCOV_PROFILE := n diff --git a/arch/x86/kernel/acpi/realmode/wakeup.S b/arch/x86/kernel/acpi/realmode/wakeup.S index b4fd836..4358fe3 100644 --- a/arch/x86/kernel/acpi/realmode/wakeup.S +++ b/arch/x86/kernel/acpi/realmode/wakeup.S @@ -108,6 +108,9 @@ wakeup_code: /* Do any other stuff... */ #ifndef CONFIG_64BIT + /* Recheck NX bit overrides (64bit path does this in trampoline */ + call verify_cpu + /* This could also be done in C code... */ movl pmode_cr3, %eax movl %eax, %cr3 @@ -131,6 +134,7 @@ wakeup_code: movl pmode_cr0, %eax movl %eax, %cr0 jmp pmode_return +# include "../../verify_cpu.S" #else pushw $0 pushw trampoline_segment diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index 103b6ab..2004d0a 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -94,8 +94,12 @@ int acpi_suspend_lowlevel(void) header->trampoline_segment = trampoline_address() >> 4; #ifdef CONFIG_SMP stack_start = (unsigned long)temp_stack + sizeof(temp_stack); + + pax_open_kernel(); early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(smp_processor_id()); + pax_close_kernel(); + initial_gs = per_cpu_offset(smp_processor_id()); #endif initial_code = (unsigned long)wakeup_long64; diff --git a/arch/x86/kernel/acpi/wakeup_32.S b/arch/x86/kernel/acpi/wakeup_32.S index 13ab720..95d5442 100644 --- a/arch/x86/kernel/acpi/wakeup_32.S +++ b/arch/x86/kernel/acpi/wakeup_32.S @@ -30,13 +30,11 @@ wakeup_pmode_return: # and restore the stack ... but you need gdt for this to work movl saved_context_esp, %esp - movl %cs:saved_magic, %eax - cmpl $0x12345678, %eax + cmpl $0x12345678, saved_magic jne bogus_magic # jump to place where we left off - movl saved_eip, %eax - jmp *%eax + jmp *(saved_eip) bogus_magic: jmp bogus_magic diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 1f84794..e23f862 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -276,6 +276,13 @@ void __init_or_module apply_alternatives(struct alt_instr *start, */ for (a = start; a < end; a++) { instr = (u8 *)&a->instr_offset + a->instr_offset; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + instr += ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; + if (instr < (u8 *)_text || (u8 *)_einittext <= instr) + instr -= ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; +#endif + replacement = (u8 *)&a->repl_offset + a->repl_offset; BUG_ON(a->replacementlen > a->instrlen); BUG_ON(a->instrlen > sizeof(insnbuf)); @@ -307,10 +314,16 @@ static void alternatives_smp_lock(const s32 *start, const s32 *end, for (poff = start; poff < end; poff++) { u8 *ptr = (u8 *)poff + *poff; +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + ptr += ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; + if (ptr < (u8 *)_text || (u8 *)_einittext <= ptr) + ptr -= ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; +#endif + if (!*poff || ptr < text || ptr >= text_end) continue; /* turn DS segment override prefix into lock prefix */ - if (*ptr == 0x3e) + if (*ktla_ktva(ptr) == 0x3e) text_poke(ptr, ((unsigned char []){0xf0}), 1); }; mutex_unlock(&text_mutex); @@ -328,10 +341,16 @@ static void alternatives_smp_unlock(const s32 *start, const s32 *end, for (poff = start; poff < end; poff++) { u8 *ptr = (u8 *)poff + *poff; +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + ptr += ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; + if (ptr < (u8 *)_text || (u8 *)_einittext <= ptr) + ptr -= ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; +#endif + if (!*poff || ptr < text || ptr >= text_end) continue; /* turn lock prefix into DS segment override prefix */ - if (*ptr == 0xf0) + if (*ktla_ktva(ptr) == 0xf0) text_poke(ptr, ((unsigned char []){0x3E}), 1); }; mutex_unlock(&text_mutex); @@ -500,7 +519,7 @@ void __init_or_module apply_paravirt(struct paravirt_patch_site *start, BUG_ON(p->len > MAX_PATCH_LEN); /* prep the buffer with the original instructions */ - memcpy(insnbuf, p->instr, p->len); + memcpy(insnbuf, ktla_ktva(p->instr), p->len); used = pv_init_ops.patch(p->instrtype, p->clobbers, insnbuf, (unsigned long)p->instr, p->len); @@ -568,7 +587,7 @@ void __init alternative_instructions(void) if (smp_alt_once) free_init_pages("SMP alternatives", (unsigned long)__smp_locks, - (unsigned long)__smp_locks_end); + PAGE_ALIGN((unsigned long)__smp_locks_end)); restart_nmi(); } @@ -585,13 +604,17 @@ void __init alternative_instructions(void) * instructions. And on the local CPU you need to be protected again NMI or MCE * handlers seeing an inconsistent instruction while you patch. */ -void *__init_or_module text_poke_early(void *addr, const void *opcode, +void *__kprobes text_poke_early(void *addr, const void *opcode, size_t len) { unsigned long flags; local_irq_save(flags); - memcpy(addr, opcode, len); + + pax_open_kernel(); + memcpy(ktla_ktva(addr), opcode, len); sync_core(); + pax_close_kernel(); + local_irq_restore(flags); /* Could also do a CLFLUSH here to speed up CPU recovery; but that causes hangs on some VIA CPUs. */ @@ -613,36 +636,22 @@ void *__init_or_module text_poke_early(void *addr, const void *opcode, */ void *__kprobes text_poke(void *addr, const void *opcode, size_t len) { - unsigned long flags; - char *vaddr; + unsigned char *vaddr = ktla_ktva(addr); struct page *pages[2]; - int i; + size_t i; if (!core_kernel_text((unsigned long)addr)) { - pages[0] = vmalloc_to_page(addr); - pages[1] = vmalloc_to_page(addr + PAGE_SIZE); + pages[0] = vmalloc_to_page(vaddr); + pages[1] = vmalloc_to_page(vaddr + PAGE_SIZE); } else { - pages[0] = virt_to_page(addr); + pages[0] = virt_to_page(vaddr); WARN_ON(!PageReserved(pages[0])); - pages[1] = virt_to_page(addr + PAGE_SIZE); + pages[1] = virt_to_page(vaddr + PAGE_SIZE); } BUG_ON(!pages[0]); - local_irq_save(flags); - set_fixmap(FIX_TEXT_POKE0, page_to_phys(pages[0])); - if (pages[1]) - set_fixmap(FIX_TEXT_POKE1, page_to_phys(pages[1])); - vaddr = (char *)fix_to_virt(FIX_TEXT_POKE0); - memcpy(&vaddr[(unsigned long)addr & ~PAGE_MASK], opcode, len); - clear_fixmap(FIX_TEXT_POKE0); - if (pages[1]) - clear_fixmap(FIX_TEXT_POKE1); - local_flush_tlb(); - sync_core(); - /* Could also do a CLFLUSH here to speed up CPU recovery; but - that causes hangs on some VIA CPUs. */ + text_poke_early(addr, opcode, len); for (i = 0; i < len; i++) - BUG_ON(((char *)addr)[i] != ((char *)opcode)[i]); - local_irq_restore(flags); + BUG_ON((vaddr)[i] != ((const unsigned char *)opcode)[i]); return addr; } diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 5b3f88e..61232b4 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -184,7 +184,7 @@ int first_system_vector = 0xfe; /* * Debug level, exported for io_apic.c */ -unsigned int apic_verbosity; +int apic_verbosity; int pic_mode; @@ -1912,7 +1912,7 @@ void smp_error_interrupt(struct pt_regs *regs) apic_write(APIC_ESR, 0); v1 = apic_read(APIC_ESR); ack_APIC_irq(); - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); apic_printk(APIC_DEBUG, KERN_DEBUG "APIC error on CPU%d: %02x(%02x)", smp_processor_id(), v0 , v1); diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index fb07275..e06bb59 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -1096,7 +1096,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin, } EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); -void lock_vector_lock(void) +void lock_vector_lock(void) __acquires(vector_lock) { /* Used to the online set of cpus does not change * during assign_irq_vector. @@ -1104,7 +1104,7 @@ void lock_vector_lock(void) raw_spin_lock(&vector_lock); } -void unlock_vector_lock(void) +void unlock_vector_lock(void) __releases(vector_lock) { raw_spin_unlock(&vector_lock); } @@ -2510,7 +2510,7 @@ static void ack_apic_edge(struct irq_data *data) ack_APIC_irq(); } -atomic_t irq_mis_count; +atomic_unchecked_t irq_mis_count; static void ack_apic_level(struct irq_data *data) { @@ -2576,7 +2576,7 @@ static void ack_apic_level(struct irq_data *data) * at the cpu. */ if (!(v & (1 << (i & 0x1f)))) { - atomic_inc(&irq_mis_count); + atomic_inc_unchecked(&irq_mis_count); eoi_ioapic_irq(irq, cfg); } diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index f76623c..aab694f 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -411,7 +411,7 @@ static DEFINE_MUTEX(apm_mutex); * This is for buggy BIOS's that refer to (real mode) segment 0x40 * even though they are called in protected mode. */ -static struct desc_struct bad_bios_desc = GDT_ENTRY_INIT(0x4092, +static const struct desc_struct bad_bios_desc = GDT_ENTRY_INIT(0x4093, (unsigned long)__va(0x400UL), PAGE_SIZE - 0x400 - 1); static const char driver_version[] = "1.16ac"; /* no spaces */ @@ -589,7 +589,10 @@ static long __apm_bios_call(void *_call) BUG_ON(cpu != 0); gdt = get_cpu_gdt_table(cpu); save_desc_40 = gdt[0x40 / 8]; + + pax_open_kernel(); gdt[0x40 / 8] = bad_bios_desc; + pax_close_kernel(); apm_irq_save(flags); APM_DO_SAVE_SEGS; @@ -598,7 +601,11 @@ static long __apm_bios_call(void *_call) &call->esi); APM_DO_RESTORE_SEGS; apm_irq_restore(flags); + + pax_open_kernel(); gdt[0x40 / 8] = save_desc_40; + pax_close_kernel(); + put_cpu(); return call->eax & 0xff; @@ -665,7 +672,10 @@ static long __apm_bios_call_simple(void *_call) BUG_ON(cpu != 0); gdt = get_cpu_gdt_table(cpu); save_desc_40 = gdt[0x40 / 8]; + + pax_open_kernel(); gdt[0x40 / 8] = bad_bios_desc; + pax_close_kernel(); apm_irq_save(flags); APM_DO_SAVE_SEGS; @@ -673,7 +683,11 @@ static long __apm_bios_call_simple(void *_call) &call->eax); APM_DO_RESTORE_SEGS; apm_irq_restore(flags); + + pax_open_kernel(); gdt[0x40 / 8] = save_desc_40; + pax_close_kernel(); + put_cpu(); return error; } @@ -2347,12 +2361,15 @@ static int __init apm_init(void) * code to that CPU. */ gdt = get_cpu_gdt_table(0); + + pax_open_kernel(); set_desc_base(&gdt[APM_CS >> 3], (unsigned long)__va((unsigned long)apm_info.bios.cseg << 4)); set_desc_base(&gdt[APM_CS_16 >> 3], (unsigned long)__va((unsigned long)apm_info.bios.cseg_16 << 4)); set_desc_base(&gdt[APM_DS >> 3], (unsigned long)__va((unsigned long)apm_info.bios.dseg << 4)); + pax_close_kernel(); proc_create("apm", 0, NULL, &apm_file_ops); diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c index 68de2dc..1f3c720 100644 --- a/arch/x86/kernel/asm-offsets.c +++ b/arch/x86/kernel/asm-offsets.c @@ -33,6 +33,8 @@ void common(void) { OFFSET(TI_status, thread_info, status); OFFSET(TI_addr_limit, thread_info, addr_limit); OFFSET(TI_preempt_count, thread_info, preempt_count); + OFFSET(TI_lowest_stack, thread_info, lowest_stack); + DEFINE(TI_task_thread_sp0, offsetof(struct task_struct, thread.sp0) - offsetof(struct task_struct, tinfo)); BLANK(); OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx); @@ -53,8 +55,26 @@ void common(void) { OFFSET(PV_CPU_irq_enable_sysexit, pv_cpu_ops, irq_enable_sysexit); OFFSET(PV_CPU_read_cr0, pv_cpu_ops, read_cr0); OFFSET(PV_MMU_read_cr2, pv_mmu_ops, read_cr2); + +#ifdef CONFIG_PAX_KERNEXEC + OFFSET(PV_CPU_write_cr0, pv_cpu_ops, write_cr0); #endif +#ifdef CONFIG_PAX_MEMORY_UDEREF + OFFSET(PV_MMU_read_cr3, pv_mmu_ops, read_cr3); + OFFSET(PV_MMU_write_cr3, pv_mmu_ops, write_cr3); +#ifdef CONFIG_X86_64 + OFFSET(PV_MMU_set_pgd_batched, pv_mmu_ops, set_pgd_batched); +#endif +#endif + +#endif + + BLANK(); + DEFINE(PAGE_SIZE_asm, PAGE_SIZE); + DEFINE(PAGE_SHIFT_asm, PAGE_SHIFT); + DEFINE(THREAD_SIZE_asm, THREAD_SIZE); + #ifdef CONFIG_XEN BLANK(); OFFSET(XEN_vcpu_info_mask, vcpu_info, evtchn_upcall_mask); diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c index 834e897..dacddc8 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -70,6 +70,7 @@ int main(void) BLANK(); #undef ENTRY + DEFINE(TSS_size, sizeof(struct tss_struct)); OFFSET(TSS_ist, tss_struct, x86_tss.ist); BLANK(); diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index 25f24dc..4094a7f 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -8,10 +8,6 @@ CFLAGS_REMOVE_common.o = -pg CFLAGS_REMOVE_perf_event.o = -pg endif -# Make sure load_percpu_segment has no stackprotector -nostackp := $(call cc-option, -fno-stack-protector) -CFLAGS_common.o := $(nostackp) - obj-y := intel_cacheinfo.o scattered.o topology.o obj-y += proc.o capflags.o powerflags.o common.o obj-y += vmware.o hypervisor.o sched.o mshyperv.o diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 80ab83d..0a7b34e 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -670,7 +670,7 @@ static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 *c, unsigned int size) { /* AMD errata T13 (order #21922) */ - if ((c->x86 == 6)) { + if (c->x86 == 6) { /* Duron Rev A0 */ if (c->x86_model == 3 && c->x86_mask == 0) size = 64; diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 1a810e4..9fa8201 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -84,60 +84,6 @@ static const struct cpu_dev __cpuinitconst default_cpu = { static const struct cpu_dev *this_cpu __cpuinitdata = &default_cpu; -DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { -#ifdef CONFIG_X86_64 - /* - * We need valid kernel segments for data and code in long mode too - * IRET will check the segment types kkeil 2000/10/28 - * Also sysret mandates a special GDT layout - * - * TLS descriptors are currently at a different place compared to i386. - * Hopefully nobody expects them at a fixed place (Wine?) - */ - [GDT_ENTRY_KERNEL32_CS] = GDT_ENTRY_INIT(0xc09b, 0, 0xfffff), - [GDT_ENTRY_KERNEL_CS] = GDT_ENTRY_INIT(0xa09b, 0, 0xfffff), - [GDT_ENTRY_KERNEL_DS] = GDT_ENTRY_INIT(0xc093, 0, 0xfffff), - [GDT_ENTRY_DEFAULT_USER32_CS] = GDT_ENTRY_INIT(0xc0fb, 0, 0xfffff), - [GDT_ENTRY_DEFAULT_USER_DS] = GDT_ENTRY_INIT(0xc0f3, 0, 0xfffff), - [GDT_ENTRY_DEFAULT_USER_CS] = GDT_ENTRY_INIT(0xa0fb, 0, 0xfffff), -#else - [GDT_ENTRY_KERNEL_CS] = GDT_ENTRY_INIT(0xc09a, 0, 0xfffff), - [GDT_ENTRY_KERNEL_DS] = GDT_ENTRY_INIT(0xc092, 0, 0xfffff), - [GDT_ENTRY_DEFAULT_USER_CS] = GDT_ENTRY_INIT(0xc0fa, 0, 0xfffff), - [GDT_ENTRY_DEFAULT_USER_DS] = GDT_ENTRY_INIT(0xc0f2, 0, 0xfffff), - /* - * Segments used for calling PnP BIOS have byte granularity. - * They code segments and data segments have fixed 64k limits, - * the transfer segment sizes are set at run time. - */ - /* 32-bit code */ - [GDT_ENTRY_PNPBIOS_CS32] = GDT_ENTRY_INIT(0x409a, 0, 0xffff), - /* 16-bit code */ - [GDT_ENTRY_PNPBIOS_CS16] = GDT_ENTRY_INIT(0x009a, 0, 0xffff), - /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_DS] = GDT_ENTRY_INIT(0x0092, 0, 0xffff), - /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_TS1] = GDT_ENTRY_INIT(0x0092, 0, 0), - /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_TS2] = GDT_ENTRY_INIT(0x0092, 0, 0), - /* - * The APM segments have byte granularity and their bases - * are set at run time. All have 64k limits. - */ - /* 32-bit code */ - [GDT_ENTRY_APMBIOS_BASE] = GDT_ENTRY_INIT(0x409a, 0, 0xffff), - /* 16-bit code */ - [GDT_ENTRY_APMBIOS_BASE+1] = GDT_ENTRY_INIT(0x009a, 0, 0xffff), - /* data */ - [GDT_ENTRY_APMBIOS_BASE+2] = GDT_ENTRY_INIT(0x4092, 0, 0xffff), - - [GDT_ENTRY_ESPFIX_SS] = GDT_ENTRY_INIT(0xc092, 0, 0xfffff), - [GDT_ENTRY_PERCPU] = GDT_ENTRY_INIT(0xc092, 0, 0xfffff), - GDT_STACK_CANARY_INIT -#endif -} }; -EXPORT_PER_CPU_SYMBOL_GPL(gdt_page); - static int __init x86_xsave_setup(char *s) { setup_clear_cpu_cap(X86_FEATURE_XSAVE); @@ -372,7 +318,7 @@ void switch_to_new_gdt(int cpu) { struct desc_ptr gdt_descr; - gdt_descr.address = (long)get_cpu_gdt_table(cpu); + gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); gdt_descr.size = GDT_SIZE - 1; load_gdt(&gdt_descr); /* Reload the per-cpu base */ @@ -839,6 +785,10 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) /* Filter out anything that depends on CPUID levels we don't have */ filter_cpuid_features(c, true); +#if defined(CONFIG_X86_32) && (defined(CONFIG_PAX_SEGMEXEC) || defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF)) + setup_clear_cpu_cap(X86_FEATURE_SEP); +#endif + /* If the model name is still unset, do table lookup. */ if (!c->x86_model_id[0]) { const char *p; @@ -1019,10 +969,12 @@ static __init int setup_disablecpuid(char *arg) } __setup("clearcpuid=", setup_disablecpuid); +DEFINE_PER_CPU(struct thread_info *, current_tinfo) = &init_task.tinfo; +EXPORT_PER_CPU_SYMBOL(current_tinfo); + #ifdef CONFIG_X86_64 struct desc_ptr idt_descr = { NR_VECTORS * 16 - 1, (unsigned long) idt_table }; -struct desc_ptr nmi_idt_descr = { NR_VECTORS * 16 - 1, - (unsigned long) nmi_idt_table }; +struct desc_ptr nmi_idt_descr = { NR_VECTORS * 16 - 1, (unsigned long) nmi_idt_table }; DEFINE_PER_CPU_FIRST(union irq_stack_union, irq_stack_union) __aligned(PAGE_SIZE); @@ -1036,7 +988,7 @@ DEFINE_PER_CPU(struct task_struct *, current_task) ____cacheline_aligned = EXPORT_PER_CPU_SYMBOL(current_task); DEFINE_PER_CPU(unsigned long, kernel_stack) = - (unsigned long)&init_thread_union - KERNEL_STACK_OFFSET + THREAD_SIZE; + (unsigned long)&init_thread_union - 16 + THREAD_SIZE; EXPORT_PER_CPU_SYMBOL(kernel_stack); DEFINE_PER_CPU(char *, irq_stack_ptr) = @@ -1126,7 +1078,7 @@ struct pt_regs * __cpuinit idle_regs(struct pt_regs *regs) { memset(regs, 0, sizeof(struct pt_regs)); regs->fs = __KERNEL_PERCPU; - regs->gs = __KERNEL_STACK_CANARY; + savesegment(gs, regs->gs); return regs; } @@ -1181,7 +1133,7 @@ void __cpuinit cpu_init(void) int i; cpu = stack_smp_processor_id(); - t = &per_cpu(init_tss, cpu); + t = init_tss + cpu; oist = &per_cpu(orig_ist, cpu); #ifdef CONFIG_NUMA @@ -1207,7 +1159,7 @@ void __cpuinit cpu_init(void) switch_to_new_gdt(cpu); loadsegment(fs, 0); - load_idt((const struct desc_ptr *)&idt_descr); + load_idt(&idt_descr); memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8); syscall_init(); @@ -1216,7 +1168,6 @@ void __cpuinit cpu_init(void) wrmsrl(MSR_KERNEL_GS_BASE, 0); barrier(); - x86_configure_nx(); if (cpu != 0) enable_x2apic(); @@ -1272,7 +1223,7 @@ void __cpuinit cpu_init(void) { int cpu = smp_processor_id(); struct task_struct *curr = current; - struct tss_struct *t = &per_cpu(init_tss, cpu); + struct tss_struct *t = init_tss + cpu; struct thread_struct *thread = &curr->thread; if (cpumask_test_and_set_cpu(cpu, cpu_initialized_mask)) { diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 3e6ff6c..54b4992 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -174,7 +174,7 @@ static void __cpuinit trap_init_f00f_bug(void) * Update the IDT descriptor and reload the IDT so that * it uses the read-only mapped virtual address. */ - idt_descr.address = fix_to_virt(FIX_F00F_IDT); + idt_descr.address = (struct desc_struct *)fix_to_virt(FIX_F00F_IDT); load_idt(&idt_descr); } #endif diff --git a/arch/x86/kernel/cpu/mcheck/mce-inject.c b/arch/x86/kernel/cpu/mcheck/mce-inject.c index fc4beb3..f20a5a7 100644 --- a/arch/x86/kernel/cpu/mcheck/mce-inject.c +++ b/arch/x86/kernel/cpu/mcheck/mce-inject.c @@ -199,6 +199,8 @@ static void raise_mce(struct mce *m) /* Error injection interface */ static ssize_t mce_write(struct file *filp, const char __user *ubuf, + size_t usize, loff_t *off) __size_overflow(3); +static ssize_t mce_write(struct file *filp, const char __user *ubuf, size_t usize, loff_t *off) { struct mce m; diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index 5a11ae2..a1a1c8a 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -42,6 +42,7 @@ #include #include #include +#include #include "mce-internal.h" @@ -250,7 +251,7 @@ static void print_mce(struct mce *m) !(m->mcgstatus & MCG_STATUS_EIPV) ? " !INEXACT!" : "", m->cs, m->ip); - if (m->cs == __KERNEL_CS) + if (m->cs == __KERNEL_CS || m->cs == __KERNEXEC_KERNEL_CS) print_symbol("{%s}", m->ip); pr_cont("\n"); } @@ -283,10 +284,10 @@ static void print_mce(struct mce *m) #define PANIC_TIMEOUT 5 /* 5 seconds */ -static atomic_t mce_paniced; +static atomic_unchecked_t mce_paniced; static int fake_panic; -static atomic_t mce_fake_paniced; +static atomic_unchecked_t mce_fake_paniced; /* Panic in progress. Enable interrupts and wait for final IPI */ static void wait_for_panic(void) @@ -310,7 +311,7 @@ static void mce_panic(char *msg, struct mce *final, char *exp) /* * Make sure only one CPU runs in machine check panic */ - if (atomic_inc_return(&mce_paniced) > 1) + if (atomic_inc_return_unchecked(&mce_paniced) > 1) wait_for_panic(); barrier(); @@ -318,7 +319,7 @@ static void mce_panic(char *msg, struct mce *final, char *exp) console_verbose(); } else { /* Don't log too much for fake panic */ - if (atomic_inc_return(&mce_fake_paniced) > 1) + if (atomic_inc_return_unchecked(&mce_fake_paniced) > 1) return; } /* First print corrected ones that are still unlogged */ @@ -658,7 +659,7 @@ static int mce_timed_out(u64 *t) * might have been modified by someone else. */ rmb(); - if (atomic_read(&mce_paniced)) + if (atomic_read_unchecked(&mce_paniced)) wait_for_panic(); if (!monarch_timeout) goto out; @@ -1446,7 +1447,7 @@ static void unexpected_machine_check(struct pt_regs *regs, long error_code) } /* Call the installed machine check handler for this CPU setup. */ -void (*machine_check_vector)(struct pt_regs *, long error_code) = +void (*machine_check_vector)(struct pt_regs *, long error_code) __read_only = unexpected_machine_check; /* @@ -1469,7 +1470,9 @@ void __cpuinit mcheck_cpu_init(struct cpuinfo_x86 *c) return; } + pax_open_kernel(); machine_check_vector = do_machine_check; + pax_close_kernel(); __mcheck_cpu_init_generic(); __mcheck_cpu_init_vendor(c); @@ -1483,7 +1486,7 @@ void __cpuinit mcheck_cpu_init(struct cpuinfo_x86 *c) */ static DEFINE_SPINLOCK(mce_chrdev_state_lock); -static int mce_chrdev_open_count; /* #times opened */ +static local_t mce_chrdev_open_count; /* #times opened */ static int mce_chrdev_open_exclu; /* already open exclusive? */ static int mce_chrdev_open(struct inode *inode, struct file *file) @@ -1491,7 +1494,7 @@ static int mce_chrdev_open(struct inode *inode, struct file *file) spin_lock(&mce_chrdev_state_lock); if (mce_chrdev_open_exclu || - (mce_chrdev_open_count && (file->f_flags & O_EXCL))) { + (local_read(&mce_chrdev_open_count) && (file->f_flags & O_EXCL))) { spin_unlock(&mce_chrdev_state_lock); return -EBUSY; @@ -1499,7 +1502,7 @@ static int mce_chrdev_open(struct inode *inode, struct file *file) if (file->f_flags & O_EXCL) mce_chrdev_open_exclu = 1; - mce_chrdev_open_count++; + local_inc(&mce_chrdev_open_count); spin_unlock(&mce_chrdev_state_lock); @@ -1510,7 +1513,7 @@ static int mce_chrdev_release(struct inode *inode, struct file *file) { spin_lock(&mce_chrdev_state_lock); - mce_chrdev_open_count--; + local_dec(&mce_chrdev_open_count); mce_chrdev_open_exclu = 0; spin_unlock(&mce_chrdev_state_lock); @@ -2229,7 +2232,7 @@ struct dentry *mce_get_debugfs_dir(void) static void mce_reset(void) { cpu_missing = 0; - atomic_set(&mce_fake_paniced, 0); + atomic_set_unchecked(&mce_fake_paniced, 0); atomic_set(&mce_executing, 0); atomic_set(&mce_callin, 0); atomic_set(&global_nwo, 0); diff --git a/arch/x86/kernel/cpu/mcheck/p5.c b/arch/x86/kernel/cpu/mcheck/p5.c index 5c0e653..0882b0a 100644 --- a/arch/x86/kernel/cpu/mcheck/p5.c +++ b/arch/x86/kernel/cpu/mcheck/p5.c @@ -12,6 +12,7 @@ #include #include #include +#include /* By default disabled */ int mce_p5_enabled __read_mostly; @@ -50,7 +51,9 @@ void intel_p5_mcheck_init(struct cpuinfo_x86 *c) if (!cpu_has(c, X86_FEATURE_MCE)) return; + pax_open_kernel(); machine_check_vector = pentium_machine_check; + pax_close_kernel(); /* Make sure the vector pointer is visible before we enable MCEs: */ wmb(); diff --git a/arch/x86/kernel/cpu/mcheck/winchip.c b/arch/x86/kernel/cpu/mcheck/winchip.c index 54060f5..c1a7577 100644 --- a/arch/x86/kernel/cpu/mcheck/winchip.c +++ b/arch/x86/kernel/cpu/mcheck/winchip.c @@ -11,6 +11,7 @@ #include #include #include +#include /* Machine check handler for WinChip C6: */ static void winchip_machine_check(struct pt_regs *regs, long error_code) @@ -24,7 +25,9 @@ void winchip_mcheck_init(struct cpuinfo_x86 *c) { u32 lo, hi; + pax_open_kernel(); machine_check_vector = winchip_machine_check; + pax_close_kernel(); /* Make sure the vector pointer is visible before we enable MCEs: */ wmb(); diff --git a/arch/x86/kernel/cpu/mtrr/if.c b/arch/x86/kernel/cpu/mtrr/if.c index 7928963..1b16001 100644 --- a/arch/x86/kernel/cpu/mtrr/if.c +++ b/arch/x86/kernel/cpu/mtrr/if.c @@ -91,6 +91,8 @@ mtrr_file_del(unsigned long base, unsigned long size, * "base=%Lx size=%Lx type=%s" or "disable=%d" */ static ssize_t +mtrr_write(struct file *file, const char __user *buf, size_t len, loff_t * ppos) __size_overflow(3); +static ssize_t mtrr_write(struct file *file, const char __user *buf, size_t len, loff_t * ppos) { int i, err; diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c index 6b96110..0da73eb 100644 --- a/arch/x86/kernel/cpu/mtrr/main.c +++ b/arch/x86/kernel/cpu/mtrr/main.c @@ -62,7 +62,7 @@ static DEFINE_MUTEX(mtrr_mutex); u64 size_or_mask, size_and_mask; static bool mtrr_aps_delayed_init; -static const struct mtrr_ops *mtrr_ops[X86_VENDOR_NUM]; +static const struct mtrr_ops *mtrr_ops[X86_VENDOR_NUM] __read_only; const struct mtrr_ops *mtrr_if; diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.h b/arch/x86/kernel/cpu/mtrr/mtrr.h index df5e41f..816c719 100644 --- a/arch/x86/kernel/cpu/mtrr/mtrr.h +++ b/arch/x86/kernel/cpu/mtrr/mtrr.h @@ -25,7 +25,7 @@ struct mtrr_ops { int (*validate_add_page)(unsigned long base, unsigned long size, unsigned int type); int (*have_wrcomb)(void); -}; +} __do_const; extern int generic_get_free_region(unsigned long base, unsigned long size, int replace_reg); diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 5adce10..99284ec 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1665,7 +1665,7 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) break; perf_callchain_store(entry, frame.return_address); - fp = frame.next_frame; + fp = (const void __force_user *)frame.next_frame; } } diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index 13ad899..f642b9a 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -36,10 +36,8 @@ static void kdump_nmi_callback(int cpu, struct pt_regs *regs) { #ifdef CONFIG_X86_32 struct pt_regs fixed_regs; -#endif -#ifdef CONFIG_X86_32 - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { crash_fixup_ss_esp(&fixed_regs, regs); regs = &fixed_regs; } diff --git a/arch/x86/kernel/doublefault_32.c b/arch/x86/kernel/doublefault_32.c index 37250fe..bf2ec74 100644 --- a/arch/x86/kernel/doublefault_32.c +++ b/arch/x86/kernel/doublefault_32.c @@ -11,7 +11,7 @@ #define DOUBLEFAULT_STACKSIZE (1024) static unsigned long doublefault_stack[DOUBLEFAULT_STACKSIZE]; -#define STACK_START (unsigned long)(doublefault_stack+DOUBLEFAULT_STACKSIZE) +#define STACK_START (unsigned long)(doublefault_stack+DOUBLEFAULT_STACKSIZE-2) #define ptr_ok(x) ((x) > PAGE_OFFSET && (x) < PAGE_OFFSET + MAXMEM) @@ -21,7 +21,7 @@ static void doublefault_fn(void) unsigned long gdt, tss; store_gdt(&gdt_desc); - gdt = gdt_desc.address; + gdt = (unsigned long)gdt_desc.address; printk(KERN_EMERG "PANIC: double fault, gdt at %08lx [%d bytes]\n", gdt, gdt_desc.size); @@ -58,10 +58,10 @@ struct tss_struct doublefault_tss __cacheline_aligned = { /* 0x2 bit is always set */ .flags = X86_EFLAGS_SF | 0x2, .sp = STACK_START, - .es = __USER_DS, + .es = __KERNEL_DS, .cs = __KERNEL_CS, .ss = __KERNEL_DS, - .ds = __USER_DS, + .ds = __KERNEL_DS, .fs = __KERNEL_PERCPU, .__cr3 = __pa_nodebug(swapper_pg_dir), diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index 4025fe4..d8451c6 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -2,6 +2,9 @@ * Copyright (C) 1991, 1992 Linus Torvalds * Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs */ +#ifdef CONFIG_GRKERNSEC_HIDESYM +#define __INCLUDED_BY_HIDESYM 1 +#endif #include #include #include @@ -35,9 +38,8 @@ void printk_address(unsigned long address, int reliable) static void print_ftrace_graph_addr(unsigned long addr, void *data, const struct stacktrace_ops *ops, - struct thread_info *tinfo, int *graph) + struct task_struct *task, int *graph) { - struct task_struct *task = tinfo->task; unsigned long ret_addr; int index = task->curr_ret_stack; @@ -58,7 +60,7 @@ print_ftrace_graph_addr(unsigned long addr, void *data, static inline void print_ftrace_graph_addr(unsigned long addr, void *data, const struct stacktrace_ops *ops, - struct thread_info *tinfo, int *graph) + struct task_struct *task, int *graph) { } #endif @@ -69,10 +71,8 @@ print_ftrace_graph_addr(unsigned long addr, void *data, * severe exception (double fault, nmi, stack fault, debug, mce) hardware stack */ -static inline int valid_stack_ptr(struct thread_info *tinfo, - void *p, unsigned int size, void *end) +static inline int valid_stack_ptr(void *t, void *p, unsigned int size, void *end) { - void *t = tinfo; if (end) { if (p < end && p >= (end-THREAD_SIZE)) return 1; @@ -83,14 +83,14 @@ static inline int valid_stack_ptr(struct thread_info *tinfo, } unsigned long -print_context_stack(struct thread_info *tinfo, +print_context_stack(struct task_struct *task, void *stack_start, unsigned long *stack, unsigned long bp, const struct stacktrace_ops *ops, void *data, unsigned long *end, int *graph) { struct stack_frame *frame = (struct stack_frame *)bp; - while (valid_stack_ptr(tinfo, stack, sizeof(*stack), end)) { + while (valid_stack_ptr(stack_start, stack, sizeof(*stack), end)) { unsigned long addr; addr = *stack; @@ -102,7 +102,7 @@ print_context_stack(struct thread_info *tinfo, } else { ops->address(data, addr, 0); } - print_ftrace_graph_addr(addr, data, ops, tinfo, graph); + print_ftrace_graph_addr(addr, data, ops, task, graph); } stack++; } @@ -111,7 +111,7 @@ print_context_stack(struct thread_info *tinfo, EXPORT_SYMBOL_GPL(print_context_stack); unsigned long -print_context_stack_bp(struct thread_info *tinfo, +print_context_stack_bp(struct task_struct *task, void *stack_start, unsigned long *stack, unsigned long bp, const struct stacktrace_ops *ops, void *data, unsigned long *end, int *graph) @@ -119,7 +119,7 @@ print_context_stack_bp(struct thread_info *tinfo, struct stack_frame *frame = (struct stack_frame *)bp; unsigned long *ret_addr = &frame->return_address; - while (valid_stack_ptr(tinfo, ret_addr, sizeof(*ret_addr), end)) { + while (valid_stack_ptr(stack_start, ret_addr, sizeof(*ret_addr), end)) { unsigned long addr = *ret_addr; if (!__kernel_text_address(addr)) @@ -128,7 +128,7 @@ print_context_stack_bp(struct thread_info *tinfo, ops->address(data, addr, 1); frame = frame->next_frame; ret_addr = &frame->return_address; - print_ftrace_graph_addr(addr, data, ops, tinfo, graph); + print_ftrace_graph_addr(addr, data, ops, task, graph); } return (unsigned long)frame; @@ -186,7 +186,7 @@ void dump_stack(void) bp = stack_frame(current, NULL); printk("Pid: %d, comm: %.20s %s %s %.*s\n", - current->pid, current->comm, print_tainted(), + task_pid_nr(current), current->comm, print_tainted(), init_utsname()->release, (int)strcspn(init_utsname()->version, " "), init_utsname()->version); @@ -222,6 +222,8 @@ unsigned __kprobes long oops_begin(void) } EXPORT_SYMBOL_GPL(oops_begin); +extern void gr_handle_kernel_exploit(void); + void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr) { if (regs && kexec_should_crash(current)) @@ -243,7 +245,10 @@ void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr) panic("Fatal exception in interrupt"); if (panic_on_oops) panic("Fatal exception"); - do_exit(signr); + + gr_handle_kernel_exploit(); + + do_group_exit(signr); } int __kprobes __die(const char *str, struct pt_regs *regs, long err) @@ -270,7 +275,7 @@ int __kprobes __die(const char *str, struct pt_regs *regs, long err) show_registers(regs); #ifdef CONFIG_X86_32 - if (user_mode_vm(regs)) { + if (user_mode(regs)) { sp = regs->sp; ss = regs->ss & 0xffff; } else { @@ -298,7 +303,7 @@ void die(const char *str, struct pt_regs *regs, long err) unsigned long flags = oops_begin(); int sig = SIGSEGV; - if (!user_mode_vm(regs)) + if (!user_mode(regs)) report_bug(regs->ip, regs); if (__die(str, regs, err)) diff --git a/arch/x86/kernel/dumpstack_32.c b/arch/x86/kernel/dumpstack_32.c index c99f9ed..2a15d80 100644 --- a/arch/x86/kernel/dumpstack_32.c +++ b/arch/x86/kernel/dumpstack_32.c @@ -38,15 +38,13 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, bp = stack_frame(task, regs); for (;;) { - struct thread_info *context; + void *stack_start = (void *)((unsigned long)stack & ~(THREAD_SIZE-1)); - context = (struct thread_info *) - ((unsigned long)stack & (~(THREAD_SIZE - 1))); - bp = ops->walk_stack(context, stack, bp, ops, data, NULL, &graph); + bp = ops->walk_stack(task, stack_start, stack, bp, ops, data, NULL, &graph); - stack = (unsigned long *)context->previous_esp; - if (!stack) + if (stack_start == task_stack_page(task)) break; + stack = *(unsigned long **)stack_start; if (ops->stack(data, "IRQ") < 0) break; touch_nmi_watchdog(); @@ -96,21 +94,22 @@ void show_registers(struct pt_regs *regs) * When in-kernel, we also print out the stack and code at the * time of the fault.. */ - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { unsigned int code_prologue = code_bytes * 43 / 64; unsigned int code_len = code_bytes; unsigned char c; u8 *ip; + unsigned long cs_base = get_desc_base(&get_cpu_gdt_table(smp_processor_id())[(0xffff & regs->cs) >> 3]); printk(KERN_EMERG "Stack:\n"); show_stack_log_lvl(NULL, regs, ®s->sp, 0, KERN_EMERG); printk(KERN_EMERG "Code: "); - ip = (u8 *)regs->ip - code_prologue; + ip = (u8 *)regs->ip - code_prologue + cs_base; if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { /* try starting at IP */ - ip = (u8 *)regs->ip; + ip = (u8 *)regs->ip + cs_base; code_len = code_len - code_prologue + 1; } for (i = 0; i < code_len; i++, ip++) { @@ -119,7 +118,7 @@ void show_registers(struct pt_regs *regs) printk(KERN_CONT " Bad EIP value."); break; } - if (ip == (u8 *)regs->ip) + if (ip == (u8 *)regs->ip + cs_base) printk(KERN_CONT "<%02x> ", c); else printk(KERN_CONT "%02x ", c); @@ -132,6 +131,7 @@ int is_valid_bugaddr(unsigned long ip) { unsigned short ud2; + ip = ktla_ktva(ip); if (ip < PAGE_OFFSET) return 0; if (probe_kernel_address((unsigned short *)ip, ud2)) @@ -139,3 +139,15 @@ int is_valid_bugaddr(unsigned long ip) return ud2 == 0x0b0f; } + +#ifdef CONFIG_PAX_MEMORY_STACKLEAK +void pax_check_alloca(unsigned long size) +{ + unsigned long sp = (unsigned long)&sp, stack_left; + + /* all kernel stacks are of the same size */ + stack_left = sp & (THREAD_SIZE - 1); + BUG_ON(stack_left < 256 || size >= stack_left - 256); +} +EXPORT_SYMBOL(pax_check_alloca); +#endif diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c index 17107bd..b2deecf 100644 --- a/arch/x86/kernel/dumpstack_64.c +++ b/arch/x86/kernel/dumpstack_64.c @@ -119,9 +119,9 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *irq_stack_end = (unsigned long *)per_cpu(irq_stack_ptr, cpu); unsigned used = 0; - struct thread_info *tinfo; int graph = 0; unsigned long dummy; + void *stack_start; if (!task) task = current; @@ -142,10 +142,10 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, * current stack address. If the stacks consist of nested * exceptions */ - tinfo = task_thread_info(task); for (;;) { char *id; unsigned long *estack_end; + estack_end = in_exception_stack(cpu, (unsigned long)stack, &used, &id); @@ -153,7 +153,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, if (ops->stack(data, id) < 0) break; - bp = ops->walk_stack(tinfo, stack, bp, ops, + bp = ops->walk_stack(task, estack_end - EXCEPTION_STKSZ, stack, bp, ops, data, estack_end, &graph); ops->stack(data, ""); /* @@ -172,7 +172,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, if (in_irq_stack(stack, irq_stack, irq_stack_end)) { if (ops->stack(data, "IRQ") < 0) break; - bp = ops->walk_stack(tinfo, stack, bp, + bp = ops->walk_stack(task, irq_stack, stack, bp, ops, data, irq_stack_end, &graph); /* * We link to the next stack (which would be @@ -191,7 +191,8 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, /* * This handles the process stack: */ - bp = ops->walk_stack(tinfo, stack, bp, ops, data, NULL, &graph); + stack_start = (void *)((unsigned long)stack & ~(THREAD_SIZE-1)); + bp = ops->walk_stack(task, stack_start, stack, bp, ops, data, NULL, &graph); put_cpu(); } EXPORT_SYMBOL(dump_trace); @@ -305,3 +306,50 @@ int is_valid_bugaddr(unsigned long ip) return ud2 == 0x0b0f; } + +#ifdef CONFIG_PAX_MEMORY_STACKLEAK +void pax_check_alloca(unsigned long size) +{ + unsigned long sp = (unsigned long)&sp, stack_start, stack_end; + unsigned cpu, used; + char *id; + + /* check the process stack first */ + stack_start = (unsigned long)task_stack_page(current); + stack_end = stack_start + THREAD_SIZE; + if (likely(stack_start <= sp && sp < stack_end)) { + unsigned long stack_left = sp & (THREAD_SIZE - 1); + BUG_ON(stack_left < 256 || size >= stack_left - 256); + return; + } + + cpu = get_cpu(); + + /* check the irq stacks */ + stack_end = (unsigned long)per_cpu(irq_stack_ptr, cpu); + stack_start = stack_end - IRQ_STACK_SIZE; + if (stack_start <= sp && sp < stack_end) { + unsigned long stack_left = sp & (IRQ_STACK_SIZE - 1); + put_cpu(); + BUG_ON(stack_left < 256 || size >= stack_left - 256); + return; + } + + /* check the exception stacks */ + used = 0; + stack_end = (unsigned long)in_exception_stack(cpu, sp, &used, &id); + stack_start = stack_end - EXCEPTION_STKSZ; + if (stack_end && stack_start <= sp && sp < stack_end) { + unsigned long stack_left = sp & (EXCEPTION_STKSZ - 1); + put_cpu(); + BUG_ON(stack_left < 256 || size >= stack_left - 256); + return; + } + + put_cpu(); + + /* unknown stack */ + BUG(); +} +EXPORT_SYMBOL(pax_check_alloca); +#endif diff --git a/arch/x86/kernel/early_printk.c b/arch/x86/kernel/early_printk.c index 9b9f18b..9fcaa04 100644 --- a/arch/x86/kernel/early_printk.c +++ b/arch/x86/kernel/early_printk.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 7b784f4..db6b628 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -179,13 +179,146 @@ /*CFI_REL_OFFSET gs, PT_GS*/ .endm .macro SET_KERNEL_GS reg + +#ifdef CONFIG_CC_STACKPROTECTOR movl $(__KERNEL_STACK_CANARY), \reg +#elif defined(CONFIG_PAX_MEMORY_UDEREF) + movl $(__USER_DS), \reg +#else + xorl \reg, \reg +#endif + movl \reg, %gs .endm #endif /* CONFIG_X86_32_LAZY_GS */ -.macro SAVE_ALL +.macro pax_enter_kernel +#ifdef CONFIG_PAX_KERNEXEC + call pax_enter_kernel +#endif +.endm + +.macro pax_exit_kernel +#ifdef CONFIG_PAX_KERNEXEC + call pax_exit_kernel +#endif +.endm + +#ifdef CONFIG_PAX_KERNEXEC +ENTRY(pax_enter_kernel) +#ifdef CONFIG_PARAVIRT + pushl %eax + pushl %ecx + call PARA_INDIRECT(pv_cpu_ops+PV_CPU_read_cr0) + mov %eax, %esi +#else + mov %cr0, %esi +#endif + bts $16, %esi + jnc 1f + mov %cs, %esi + cmp $__KERNEL_CS, %esi + jz 3f + ljmp $__KERNEL_CS, $3f +1: ljmp $__KERNEXEC_KERNEL_CS, $2f +2: +#ifdef CONFIG_PARAVIRT + mov %esi, %eax + call PARA_INDIRECT(pv_cpu_ops+PV_CPU_write_cr0) +#else + mov %esi, %cr0 +#endif +3: +#ifdef CONFIG_PARAVIRT + popl %ecx + popl %eax +#endif + ret +ENDPROC(pax_enter_kernel) + +ENTRY(pax_exit_kernel) +#ifdef CONFIG_PARAVIRT + pushl %eax + pushl %ecx +#endif + mov %cs, %esi + cmp $__KERNEXEC_KERNEL_CS, %esi + jnz 2f +#ifdef CONFIG_PARAVIRT + call PARA_INDIRECT(pv_cpu_ops+PV_CPU_read_cr0); + mov %eax, %esi +#else + mov %cr0, %esi +#endif + btr $16, %esi + ljmp $__KERNEL_CS, $1f +1: +#ifdef CONFIG_PARAVIRT + mov %esi, %eax + call PARA_INDIRECT(pv_cpu_ops+PV_CPU_write_cr0); +#else + mov %esi, %cr0 +#endif +2: +#ifdef CONFIG_PARAVIRT + popl %ecx + popl %eax +#endif + ret +ENDPROC(pax_exit_kernel) +#endif + +.macro pax_erase_kstack +#ifdef CONFIG_PAX_MEMORY_STACKLEAK + call pax_erase_kstack +#endif +.endm + +#ifdef CONFIG_PAX_MEMORY_STACKLEAK +/* + * ebp: thread_info + * ecx, edx: can be clobbered + */ +ENTRY(pax_erase_kstack) + pushl %edi + pushl %eax + + mov TI_lowest_stack(%ebp), %edi + mov $-0xBEEF, %eax + std + +1: mov %edi, %ecx + and $THREAD_SIZE_asm - 1, %ecx + shr $2, %ecx + repne scasl + jecxz 2f + + cmp $2*16, %ecx + jc 2f + + mov $2*16, %ecx + repe scasl + jecxz 2f + jne 1b + +2: cld + mov %esp, %ecx + sub %edi, %ecx + shr $2, %ecx + rep stosl + + mov TI_task_thread_sp0(%ebp), %edi + sub $128, %edi + mov %edi, TI_lowest_stack(%ebp) + + popl %eax + popl %edi + ret +ENDPROC(pax_erase_kstack) +#endif + +.macro __SAVE_ALL _DS cld PUSH_GS pushl_cfi %fs @@ -208,7 +341,7 @@ CFI_REL_OFFSET ecx, 0 pushl_cfi %ebx CFI_REL_OFFSET ebx, 0 - movl $(__USER_DS), %edx + movl $\_DS, %edx movl %edx, %ds movl %edx, %es movl $(__KERNEL_PERCPU), %edx @@ -216,6 +349,15 @@ SET_KERNEL_GS %edx .endm +.macro SAVE_ALL +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + __SAVE_ALL __KERNEL_DS + pax_enter_kernel +#else + __SAVE_ALL __USER_DS +#endif +.endm + .macro RESTORE_INT_REGS popl_cfi %ebx CFI_RESTORE ebx @@ -301,7 +443,7 @@ ENTRY(ret_from_fork) popfl_cfi jmp syscall_exit CFI_ENDPROC -END(ret_from_fork) +ENDPROC(ret_from_fork) /* * Interrupt exit functions should be protected against kprobes @@ -335,7 +477,15 @@ resume_userspace_sig: andl $SEGMENT_RPL_MASK, %eax #endif cmpl $USER_RPL, %eax + +#ifdef CONFIG_PAX_KERNEXEC + jae resume_userspace + + pax_exit_kernel + jmp resume_kernel +#else jb resume_kernel # not returning to v8086 or userspace +#endif ENTRY(resume_userspace) LOCKDEP_SYS_EXIT @@ -347,8 +497,8 @@ ENTRY(resume_userspace) andl $_TIF_WORK_MASK, %ecx # is there any work to be done on # int/exception return? jne work_pending - jmp restore_all -END(ret_from_exception) + jmp restore_all_pax +ENDPROC(ret_from_exception) #ifdef CONFIG_PREEMPT ENTRY(resume_kernel) @@ -363,7 +513,7 @@ need_resched: jz restore_all call preempt_schedule_irq jmp need_resched -END(resume_kernel) +ENDPROC(resume_kernel) #endif CFI_ENDPROC /* @@ -397,23 +547,34 @@ sysenter_past_esp: /*CFI_REL_OFFSET cs, 0*/ /* * Push current_thread_info()->sysenter_return to the stack. - * A tiny bit of offset fixup is necessary - 4*4 means the 4 words - * pushed above; +8 corresponds to copy_thread's esp0 setting. */ - pushl_cfi ((TI_sysenter_return)-THREAD_SIZE+8+4*4)(%esp) + pushl_cfi $0 CFI_REL_OFFSET eip, 0 pushl_cfi %eax SAVE_ALL + GET_THREAD_INFO(%ebp) + movl TI_sysenter_return(%ebp),%ebp + movl %ebp,PT_EIP(%esp) ENABLE_INTERRUPTS(CLBR_NONE) /* * Load the potential sixth argument from user stack. * Careful about security. */ + movl PT_OLDESP(%esp),%ebp + +#ifdef CONFIG_PAX_MEMORY_UDEREF + mov PT_OLDSS(%esp),%ds +1: movl %ds:(%ebp),%ebp + push %ss + pop %ds +#else cmpl $__PAGE_OFFSET-3,%ebp jae syscall_fault 1: movl (%ebp),%ebp +#endif + movl %ebp,PT_EBP(%esp) .section __ex_table,"a" .align 4 @@ -436,12 +597,24 @@ sysenter_do_call: testl $_TIF_ALLWORK_MASK, %ecx jne sysexit_audit sysenter_exit: + +#ifdef CONFIG_PAX_RANDKSTACK + pushl_cfi %eax + movl %esp, %eax + call pax_randomize_kstack + popl_cfi %eax +#endif + + pax_erase_kstack + /* if something modifies registers it must also disable sysexit */ movl PT_EIP(%esp), %edx movl PT_OLDESP(%esp), %ecx xorl %ebp,%ebp TRACE_IRQS_ON 1: mov PT_FS(%esp), %fs +2: mov PT_DS(%esp), %ds +3: mov PT_ES(%esp), %es PTGS_TO_GS ENABLE_INTERRUPTS_SYSEXIT @@ -458,6 +631,9 @@ sysenter_audit: movl %eax,%edx /* 2nd arg: syscall number */ movl $AUDIT_ARCH_I386,%eax /* 1st arg: audit arch */ call __audit_syscall_entry + + pax_erase_kstack + pushl_cfi %ebx movl PT_EAX(%esp),%eax /* reload syscall number */ jmp sysenter_do_call @@ -483,11 +659,17 @@ sysexit_audit: CFI_ENDPROC .pushsection .fixup,"ax" -2: movl $0,PT_FS(%esp) +4: movl $0,PT_FS(%esp) + jmp 1b +5: movl $0,PT_DS(%esp) + jmp 1b +6: movl $0,PT_ES(%esp) jmp 1b .section __ex_table,"a" .align 4 - .long 1b,2b + .long 1b,4b + .long 2b,5b + .long 3b,6b .popsection PTGS_TO_GS_EX ENDPROC(ia32_sysenter_target) @@ -520,6 +702,15 @@ syscall_exit: testl $_TIF_ALLWORK_MASK, %ecx # current->work jne syscall_exit_work +restore_all_pax: + +#ifdef CONFIG_PAX_RANDKSTACK + movl %esp, %eax + call pax_randomize_kstack +#endif + + pax_erase_kstack + restore_all: TRACE_IRQS_IRET restore_all_notrace: @@ -579,14 +770,34 @@ ldt_ss: * compensating for the offset by changing to the ESPFIX segment with * a base address that matches for the difference. */ -#define GDT_ESPFIX_SS PER_CPU_VAR(gdt_page) + (GDT_ENTRY_ESPFIX_SS * 8) +#define GDT_ESPFIX_SS (GDT_ENTRY_ESPFIX_SS * 8)(%ebx) mov %esp, %edx /* load kernel esp */ mov PT_OLDESP(%esp), %eax /* load userspace esp */ mov %dx, %ax /* eax: new kernel esp */ sub %eax, %edx /* offset (low word is 0) */ +#ifdef CONFIG_SMP + movl PER_CPU_VAR(cpu_number), %ebx + shll $PAGE_SHIFT_asm, %ebx + addl $cpu_gdt_table, %ebx +#else + movl $cpu_gdt_table, %ebx +#endif shr $16, %edx - mov %dl, GDT_ESPFIX_SS + 4 /* bits 16..23 */ - mov %dh, GDT_ESPFIX_SS + 7 /* bits 24..31 */ + +#ifdef CONFIG_PAX_KERNEXEC + mov %cr0, %esi + btr $16, %esi + mov %esi, %cr0 +#endif + + mov %dl, 4 + GDT_ESPFIX_SS /* bits 16..23 */ + mov %dh, 7 + GDT_ESPFIX_SS /* bits 24..31 */ + +#ifdef CONFIG_PAX_KERNEXEC + bts $16, %esi + mov %esi, %cr0 +#endif + pushl_cfi $__ESPFIX_SS pushl_cfi %eax /* new kernel esp */ /* Disable interrupts, but do not irqtrace this section: we @@ -615,38 +826,30 @@ work_resched: movl TI_flags(%ebp), %ecx andl $_TIF_WORK_MASK, %ecx # is there any work to be done other # than syscall tracing? - jz restore_all + jz restore_all_pax testb $_TIF_NEED_RESCHED, %cl jnz work_resched work_notifysig: # deal with pending signals and # notify-resume requests + movl %esp, %eax #ifdef CONFIG_VM86 testl $X86_EFLAGS_VM, PT_EFLAGS(%esp) - movl %esp, %eax - jne work_notifysig_v86 # returning to kernel-space or + jz 1f # returning to kernel-space or # vm86-space - TRACE_IRQS_ON - ENABLE_INTERRUPTS(CLBR_NONE) - xorl %edx, %edx - call do_notify_resume - jmp resume_userspace_sig - ALIGN -work_notifysig_v86: pushl_cfi %ecx # save ti_flags for do_notify_resume call save_v86_state # %eax contains pt_regs pointer popl_cfi %ecx movl %eax, %esp -#else - movl %esp, %eax +1: #endif TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_NONE) xorl %edx, %edx call do_notify_resume jmp resume_userspace_sig -END(work_pending) +ENDPROC(work_pending) # perform syscall exit tracing ALIGN @@ -654,11 +857,14 @@ syscall_trace_entry: movl $-ENOSYS,PT_EAX(%esp) movl %esp, %eax call syscall_trace_enter + + pax_erase_kstack + /* What it returned is what we'll actually use. */ cmpl $(NR_syscalls), %eax jnae syscall_call jmp syscall_exit -END(syscall_trace_entry) +ENDPROC(syscall_trace_entry) # perform syscall exit tracing ALIGN @@ -671,20 +877,24 @@ syscall_exit_work: movl %esp, %eax call syscall_trace_leave jmp resume_userspace -END(syscall_exit_work) +ENDPROC(syscall_exit_work) CFI_ENDPROC RING0_INT_FRAME # can't unwind into user space anyway syscall_fault: +#ifdef CONFIG_PAX_MEMORY_UDEREF + push %ss + pop %ds +#endif GET_THREAD_INFO(%ebp) movl $-EFAULT,PT_EAX(%esp) jmp resume_userspace -END(syscall_fault) +ENDPROC(syscall_fault) syscall_badsys: movl $-ENOSYS,PT_EAX(%esp) jmp resume_userspace -END(syscall_badsys) +ENDPROC(syscall_badsys) CFI_ENDPROC /* * End of kprobes section @@ -756,6 +966,36 @@ ENTRY(ptregs_clone) CFI_ENDPROC ENDPROC(ptregs_clone) + ALIGN; +ENTRY(kernel_execve) + CFI_STARTPROC + pushl_cfi %ebp + sub $PT_OLDSS+4,%esp + pushl_cfi %edi + pushl_cfi %ecx + pushl_cfi %eax + lea 3*4(%esp),%edi + mov $PT_OLDSS/4+1,%ecx + xorl %eax,%eax + rep stosl + popl_cfi %eax + popl_cfi %ecx + popl_cfi %edi + movl $X86_EFLAGS_IF,PT_EFLAGS(%esp) + pushl_cfi %esp + call sys_execve + add $4,%esp + CFI_ADJUST_CFA_OFFSET -4 + GET_THREAD_INFO(%ebp) + test %eax,%eax + jz syscall_exit + add $PT_OLDSS+4,%esp + CFI_ADJUST_CFA_OFFSET -PT_OLDSS-4 + popl_cfi %ebp + ret + CFI_ENDPROC +ENDPROC(kernel_execve) + .macro FIXUP_ESPFIX_STACK /* * Switch back for ESPFIX stack to the normal zerobased stack @@ -765,8 +1005,15 @@ ENDPROC(ptregs_clone) * normal stack and adjusts ESP with the matching offset. */ /* fixup the stack */ - mov GDT_ESPFIX_SS + 4, %al /* bits 16..23 */ - mov GDT_ESPFIX_SS + 7, %ah /* bits 24..31 */ +#ifdef CONFIG_SMP + movl PER_CPU_VAR(cpu_number), %ebx + shll $PAGE_SHIFT_asm, %ebx + addl $cpu_gdt_table, %ebx +#else + movl $cpu_gdt_table, %ebx +#endif + mov 4 + GDT_ESPFIX_SS, %al /* bits 16..23 */ + mov 7 + GDT_ESPFIX_SS, %ah /* bits 24..31 */ shl $16, %eax addl %esp, %eax /* the adjusted stack pointer */ pushl_cfi $__KERNEL_DS @@ -819,7 +1066,7 @@ vector=vector+1 .endr 2: jmp common_interrupt .endr -END(irq_entries_start) +ENDPROC(irq_entries_start) .previous END(interrupt) @@ -867,7 +1114,7 @@ ENTRY(coprocessor_error) pushl_cfi $do_coprocessor_error jmp error_code CFI_ENDPROC -END(coprocessor_error) +ENDPROC(coprocessor_error) ENTRY(simd_coprocessor_error) RING0_INT_FRAME @@ -888,7 +1135,7 @@ ENTRY(simd_coprocessor_error) #endif jmp error_code CFI_ENDPROC -END(simd_coprocessor_error) +ENDPROC(simd_coprocessor_error) ENTRY(device_not_available) RING0_INT_FRAME @@ -896,7 +1143,7 @@ ENTRY(device_not_available) pushl_cfi $do_device_not_available jmp error_code CFI_ENDPROC -END(device_not_available) +ENDPROC(device_not_available) #ifdef CONFIG_PARAVIRT ENTRY(native_iret) @@ -905,12 +1152,12 @@ ENTRY(native_iret) .align 4 .long native_iret, iret_exc .previous -END(native_iret) +ENDPROC(native_iret) ENTRY(native_irq_enable_sysexit) sti sysexit -END(native_irq_enable_sysexit) +ENDPROC(native_irq_enable_sysexit) #endif ENTRY(overflow) @@ -919,7 +1166,7 @@ ENTRY(overflow) pushl_cfi $do_overflow jmp error_code CFI_ENDPROC -END(overflow) +ENDPROC(overflow) ENTRY(bounds) RING0_INT_FRAME @@ -927,7 +1174,7 @@ ENTRY(bounds) pushl_cfi $do_bounds jmp error_code CFI_ENDPROC -END(bounds) +ENDPROC(bounds) ENTRY(invalid_op) RING0_INT_FRAME @@ -935,7 +1182,7 @@ ENTRY(invalid_op) pushl_cfi $do_invalid_op jmp error_code CFI_ENDPROC -END(invalid_op) +ENDPROC(invalid_op) ENTRY(coprocessor_segment_overrun) RING0_INT_FRAME @@ -943,35 +1190,35 @@ ENTRY(coprocessor_segment_overrun) pushl_cfi $do_coprocessor_segment_overrun jmp error_code CFI_ENDPROC -END(coprocessor_segment_overrun) +ENDPROC(coprocessor_segment_overrun) ENTRY(invalid_TSS) RING0_EC_FRAME pushl_cfi $do_invalid_TSS jmp error_code CFI_ENDPROC -END(invalid_TSS) +ENDPROC(invalid_TSS) ENTRY(segment_not_present) RING0_EC_FRAME pushl_cfi $do_segment_not_present jmp error_code CFI_ENDPROC -END(segment_not_present) +ENDPROC(segment_not_present) ENTRY(stack_segment) RING0_EC_FRAME pushl_cfi $do_stack_segment jmp error_code CFI_ENDPROC -END(stack_segment) +ENDPROC(stack_segment) ENTRY(alignment_check) RING0_EC_FRAME pushl_cfi $do_alignment_check jmp error_code CFI_ENDPROC -END(alignment_check) +ENDPROC(alignment_check) ENTRY(divide_error) RING0_INT_FRAME @@ -979,7 +1226,7 @@ ENTRY(divide_error) pushl_cfi $do_divide_error jmp error_code CFI_ENDPROC -END(divide_error) +ENDPROC(divide_error) #ifdef CONFIG_X86_MCE ENTRY(machine_check) @@ -988,7 +1235,7 @@ ENTRY(machine_check) pushl_cfi machine_check_vector jmp error_code CFI_ENDPROC -END(machine_check) +ENDPROC(machine_check) #endif ENTRY(spurious_interrupt_bug) @@ -997,7 +1244,7 @@ ENTRY(spurious_interrupt_bug) pushl_cfi $do_spurious_interrupt_bug jmp error_code CFI_ENDPROC -END(spurious_interrupt_bug) +ENDPROC(spurious_interrupt_bug) /* * End of kprobes section */ @@ -1112,7 +1359,7 @@ BUILD_INTERRUPT3(xen_hvm_callback_vector, XEN_HVM_EVTCHN_CALLBACK, ENTRY(mcount) ret -END(mcount) +ENDPROC(mcount) ENTRY(ftrace_caller) cmpl $0, function_trace_stop @@ -1141,7 +1388,7 @@ ftrace_graph_call: .globl ftrace_stub ftrace_stub: ret -END(ftrace_caller) +ENDPROC(ftrace_caller) #else /* ! CONFIG_DYNAMIC_FTRACE */ @@ -1177,7 +1424,7 @@ trace: popl %ecx popl %eax jmp ftrace_stub -END(mcount) +ENDPROC(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_TRACER */ @@ -1198,7 +1445,7 @@ ENTRY(ftrace_graph_caller) popl %ecx popl %eax ret -END(ftrace_graph_caller) +ENDPROC(ftrace_graph_caller) .globl return_to_handler return_to_handler: @@ -1253,15 +1500,18 @@ error_code: movl $-1, PT_ORIG_EAX(%esp) # no syscall to restart REG_TO_PTGS %ecx SET_KERNEL_GS %ecx - movl $(__USER_DS), %ecx + movl $(__KERNEL_DS), %ecx movl %ecx, %ds movl %ecx, %es + + pax_enter_kernel + TRACE_IRQS_OFF movl %esp,%eax # pt_regs pointer call *%edi jmp ret_from_exception CFI_ENDPROC -END(page_fault) +ENDPROC(page_fault) /* * Debug traps and NMI can happen at the one SYSENTER instruction @@ -1303,7 +1553,7 @@ debug_stack_correct: call do_debug jmp ret_from_exception CFI_ENDPROC -END(debug) +ENDPROC(debug) /* * NMI is doubly nasty. It can happen _while_ we're handling @@ -1340,6 +1590,9 @@ nmi_stack_correct: xorl %edx,%edx # zero error code movl %esp,%eax # pt_regs pointer call do_nmi + + pax_exit_kernel + jmp restore_all_notrace CFI_ENDPROC @@ -1376,12 +1629,15 @@ nmi_espfix_stack: FIXUP_ESPFIX_STACK # %eax == %esp xorl %edx,%edx # zero error code call do_nmi + + pax_exit_kernel + RESTORE_REGS lss 12+4(%esp), %esp # back to espfix stack CFI_ADJUST_CFA_OFFSET -24 jmp irq_return CFI_ENDPROC -END(nmi) +ENDPROC(nmi) ENTRY(int3) RING0_INT_FRAME @@ -1393,14 +1649,14 @@ ENTRY(int3) call do_int3 jmp ret_from_exception CFI_ENDPROC -END(int3) +ENDPROC(int3) ENTRY(general_protection) RING0_EC_FRAME pushl_cfi $do_general_protection jmp error_code CFI_ENDPROC -END(general_protection) +ENDPROC(general_protection) #ifdef CONFIG_KVM_GUEST ENTRY(async_page_fault) @@ -1408,7 +1664,7 @@ ENTRY(async_page_fault) pushl_cfi $do_async_page_fault jmp error_code CFI_ENDPROC -END(async_page_fault) +ENDPROC(async_page_fault) #endif /* diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 1333d98..b340ca2 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -56,6 +56,8 @@ #include #include #include +#include +#include /* Avoid __ASSEMBLER__'ifying just for this. */ #include @@ -69,8 +71,9 @@ #ifdef CONFIG_FUNCTION_TRACER #ifdef CONFIG_DYNAMIC_FTRACE ENTRY(mcount) + pax_force_retaddr retq -END(mcount) +ENDPROC(mcount) ENTRY(ftrace_caller) cmpl $0, function_trace_stop @@ -93,8 +96,9 @@ GLOBAL(ftrace_graph_call) #endif GLOBAL(ftrace_stub) + pax_force_retaddr retq -END(ftrace_caller) +ENDPROC(ftrace_caller) #else /* ! CONFIG_DYNAMIC_FTRACE */ ENTRY(mcount) @@ -113,6 +117,7 @@ ENTRY(mcount) #endif GLOBAL(ftrace_stub) + pax_force_retaddr retq trace: @@ -122,12 +127,13 @@ trace: movq 8(%rbp), %rsi subq $MCOUNT_INSN_SIZE, %rdi + pax_force_fptr ftrace_trace_function call *ftrace_trace_function MCOUNT_RESTORE_FRAME jmp ftrace_stub -END(mcount) +ENDPROC(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_TRACER */ @@ -147,8 +153,9 @@ ENTRY(ftrace_graph_caller) MCOUNT_RESTORE_FRAME + pax_force_retaddr retq -END(ftrace_graph_caller) +ENDPROC(ftrace_graph_caller) GLOBAL(return_to_handler) subq $24, %rsp @@ -164,6 +171,7 @@ GLOBAL(return_to_handler) movq 8(%rsp), %rdx movq (%rsp), %rax addq $24, %rsp + pax_force_fptr %rdi jmp *%rdi #endif @@ -179,6 +187,282 @@ ENTRY(native_usergs_sysret64) ENDPROC(native_usergs_sysret64) #endif /* CONFIG_PARAVIRT */ + .macro ljmpq sel, off +#if defined(CONFIG_MPSC) || defined(CONFIG_MCORE2) || defined (CONFIG_MATOM) + .byte 0x48; ljmp *1234f(%rip) + .pushsection .rodata + .align 16 + 1234: .quad \off; .word \sel + .popsection +#else + pushq $\sel + pushq $\off + lretq +#endif + .endm + + .macro pax_enter_kernel + pax_set_fptr_mask +#ifdef CONFIG_PAX_KERNEXEC + call pax_enter_kernel +#endif + .endm + + .macro pax_exit_kernel +#ifdef CONFIG_PAX_KERNEXEC + call pax_exit_kernel +#endif + .endm + +#ifdef CONFIG_PAX_KERNEXEC +ENTRY(pax_enter_kernel) + pushq %rdi + +#ifdef CONFIG_PARAVIRT + PV_SAVE_REGS(CLBR_RDI) +#endif + + GET_CR0_INTO_RDI + bts $16,%rdi + jnc 3f + mov %cs,%edi + cmp $__KERNEL_CS,%edi + jnz 2f +1: + +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI) +#endif + + popq %rdi + pax_force_retaddr + retq + +2: ljmpq __KERNEL_CS,1f +3: ljmpq __KERNEXEC_KERNEL_CS,4f +4: SET_RDI_INTO_CR0 + jmp 1b +ENDPROC(pax_enter_kernel) + +ENTRY(pax_exit_kernel) + pushq %rdi + +#ifdef CONFIG_PARAVIRT + PV_SAVE_REGS(CLBR_RDI) +#endif + + mov %cs,%rdi + cmp $__KERNEXEC_KERNEL_CS,%edi + jz 2f +1: + +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI); +#endif + + popq %rdi + pax_force_retaddr + retq + +2: GET_CR0_INTO_RDI + btr $16,%rdi + ljmpq __KERNEL_CS,3f +3: SET_RDI_INTO_CR0 + jmp 1b +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI); +#endif + + popq %rdi + pax_force_retaddr + retq +ENDPROC(pax_exit_kernel) +#endif + + .macro pax_enter_kernel_user + pax_set_fptr_mask +#ifdef CONFIG_PAX_MEMORY_UDEREF + call pax_enter_kernel_user +#endif + .endm + + .macro pax_exit_kernel_user +#ifdef CONFIG_PAX_MEMORY_UDEREF + call pax_exit_kernel_user +#endif +#ifdef CONFIG_PAX_RANDKSTACK + pushq %rax + call pax_randomize_kstack + popq %rax +#endif + .endm + +#ifdef CONFIG_PAX_MEMORY_UDEREF +ENTRY(pax_enter_kernel_user) + pushq %rdi + pushq %rbx + +#ifdef CONFIG_PARAVIRT + PV_SAVE_REGS(CLBR_RDI) +#endif + + GET_CR3_INTO_RDI + mov %rdi,%rbx + add $__START_KERNEL_map,%rbx + sub phys_base(%rip),%rbx + +#ifdef CONFIG_PARAVIRT + pushq %rdi + cmpl $0, pv_info+PARAVIRT_enabled + jz 1f + i = 0 + .rept USER_PGD_PTRS + mov i*8(%rbx),%rsi + mov $0,%sil + lea i*8(%rbx),%rdi + call PARA_INDIRECT(pv_mmu_ops+PV_MMU_set_pgd_batched) + i = i + 1 + .endr + jmp 2f +1: +#endif + + i = 0 + .rept USER_PGD_PTRS + movb $0,i*8(%rbx) + i = i + 1 + .endr + +#ifdef CONFIG_PARAVIRT +2: popq %rdi +#endif + SET_RDI_INTO_CR3 + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_RDI + bts $16,%rdi + SET_RDI_INTO_CR0 +#endif + +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI) +#endif + + popq %rbx + popq %rdi + pax_force_retaddr + retq +ENDPROC(pax_enter_kernel_user) + +ENTRY(pax_exit_kernel_user) + push %rdi + +#ifdef CONFIG_PARAVIRT + pushq %rbx + PV_SAVE_REGS(CLBR_RDI) +#endif + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_RDI + btr $16,%rdi + SET_RDI_INTO_CR0 +#endif + + GET_CR3_INTO_RDI + add $__START_KERNEL_map,%rdi + sub phys_base(%rip),%rdi + +#ifdef CONFIG_PARAVIRT + cmpl $0, pv_info+PARAVIRT_enabled + jz 1f + mov %rdi,%rbx + i = 0 + .rept USER_PGD_PTRS + mov i*8(%rbx),%rsi + mov $0x67,%sil + lea i*8(%rbx),%rdi + call PARA_INDIRECT(pv_mmu_ops+PV_MMU_set_pgd_batched) + i = i + 1 + .endr + jmp 2f +1: +#endif + + i = 0 + .rept USER_PGD_PTRS + movb $0x67,i*8(%rdi) + i = i + 1 + .endr + +#ifdef CONFIG_PARAVIRT +2: PV_RESTORE_REGS(CLBR_RDI) + popq %rbx +#endif + + popq %rdi + pax_force_retaddr + retq +ENDPROC(pax_exit_kernel_user) +#endif + +.macro pax_erase_kstack +#ifdef CONFIG_PAX_MEMORY_STACKLEAK + call pax_erase_kstack +#endif +.endm + +#ifdef CONFIG_PAX_MEMORY_STACKLEAK +/* + * r11: thread_info + * rcx, rdx: can be clobbered + */ +ENTRY(pax_erase_kstack) + pushq %rdi + pushq %rax + pushq %r11 + + GET_THREAD_INFO(%r11) + mov TI_lowest_stack(%r11), %rdi + mov $-0xBEEF, %rax + std + +1: mov %edi, %ecx + and $THREAD_SIZE_asm - 1, %ecx + shr $3, %ecx + repne scasq + jecxz 2f + + cmp $2*8, %ecx + jc 2f + + mov $2*8, %ecx + repe scasq + jecxz 2f + jne 1b + +2: cld + mov %esp, %ecx + sub %edi, %ecx + + cmp $THREAD_SIZE_asm, %rcx + jb 3f + ud2 +3: + + shr $3, %ecx + rep stosq + + mov TI_task_thread_sp0(%r11), %rdi + sub $256, %rdi + mov %rdi, TI_lowest_stack(%r11) + + popq %r11 + popq %rax + popq %rdi + pax_force_retaddr + ret +ENDPROC(pax_erase_kstack) +#endif .macro TRACE_IRQS_IRETQ offset=ARGOFFSET #ifdef CONFIG_TRACE_IRQFLAGS @@ -232,8 +516,8 @@ ENDPROC(native_usergs_sysret64) .endm .macro UNFAKE_STACK_FRAME - addq $8*6, %rsp - CFI_ADJUST_CFA_OFFSET -(6*8) + addq $8*6 + ARG_SKIP, %rsp + CFI_ADJUST_CFA_OFFSET -(6*8 + ARG_SKIP) .endm /* @@ -320,7 +604,7 @@ ENDPROC(native_usergs_sysret64) movq %rsp, %rsi leaq -RBP(%rsp),%rdi /* arg1 for handler */ - testl $3, CS(%rdi) + testb $3, CS(%rdi) je 1f SWAPGS /* @@ -356,9 +640,10 @@ ENTRY(save_rest) movq_cfi r15, R15+16 movq %r11, 8(%rsp) /* return address */ FIXUP_TOP_OF_STACK %r11, 16 + pax_force_retaddr ret CFI_ENDPROC -END(save_rest) +ENDPROC(save_rest) /* save complete stack frame */ .pushsection .kprobes.text, "ax" @@ -387,9 +672,10 @@ ENTRY(save_paranoid) js 1f /* negative -> in kernel */ SWAPGS xorl %ebx,%ebx -1: ret +1: pax_force_retaddr_bts + ret CFI_ENDPROC -END(save_paranoid) +ENDPROC(save_paranoid) .popsection /* @@ -411,7 +697,7 @@ ENTRY(ret_from_fork) RESTORE_REST - testl $3, CS-ARGOFFSET(%rsp) # from kernel_thread? + testb $3, CS-ARGOFFSET(%rsp) # from kernel_thread? jz retint_restore_args testl $_TIF_IA32, TI_flags(%rcx) # 32-bit compat task needs IRET @@ -421,7 +707,7 @@ ENTRY(ret_from_fork) jmp ret_from_sys_call # go to the SYSRET fastpath CFI_ENDPROC -END(ret_from_fork) +ENDPROC(ret_from_fork) /* * System call entry. Up to 6 arguments in registers are supported. @@ -457,7 +743,7 @@ END(ret_from_fork) ENTRY(system_call) CFI_STARTPROC simple CFI_SIGNAL_FRAME - CFI_DEF_CFA rsp,KERNEL_STACK_OFFSET + CFI_DEF_CFA rsp,0 CFI_REGISTER rip,rcx /*CFI_REGISTER rflags,r11*/ SWAPGS_UNSAFE_STACK @@ -470,21 +756,23 @@ GLOBAL(system_call_after_swapgs) movq %rsp,PER_CPU_VAR(old_rsp) movq PER_CPU_VAR(kernel_stack),%rsp + SAVE_ARGS 8*6,0 + pax_enter_kernel_user /* * No need to follow this irqs off/on section - it's straight * and short: */ ENABLE_INTERRUPTS(CLBR_NONE) - SAVE_ARGS 8,0 movq %rax,ORIG_RAX-ARGOFFSET(%rsp) movq %rcx,RIP-ARGOFFSET(%rsp) CFI_REL_OFFSET rip,RIP-ARGOFFSET - testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + GET_THREAD_INFO(%rcx) + testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%rcx) jnz tracesys system_call_fastpath: cmpq $__NR_syscall_max,%rax ja badsys - movq %r10,%rcx + movq R10-ARGOFFSET(%rsp),%rcx call *sys_call_table(,%rax,8) # XXX: rip relative movq %rax,RAX-ARGOFFSET(%rsp) /* @@ -498,10 +786,13 @@ sysret_check: LOCKDEP_SYS_EXIT DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF - movl TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET),%edx + GET_THREAD_INFO(%rcx) + movl TI_flags(%rcx),%edx andl %edi,%edx jnz sysret_careful CFI_REMEMBER_STATE + pax_exit_kernel_user + pax_erase_kstack /* * sysretq will re-enable interrupts: */ @@ -553,14 +844,18 @@ badsys: * jump back to the normal fast path. */ auditsys: - movq %r10,%r9 /* 6th arg: 4th syscall arg */ + movq R10-ARGOFFSET(%rsp),%r9 /* 6th arg: 4th syscall arg */ movq %rdx,%r8 /* 5th arg: 3rd syscall arg */ movq %rsi,%rcx /* 4th arg: 2nd syscall arg */ movq %rdi,%rdx /* 3rd arg: 1st syscall arg */ movq %rax,%rsi /* 2nd arg: syscall number */ movl $AUDIT_ARCH_X86_64,%edi /* 1st arg: audit arch */ call __audit_syscall_entry + + pax_erase_kstack + LOAD_ARGS 0 /* reload call-clobbered registers */ + pax_set_fptr_mask jmp system_call_fastpath /* @@ -581,7 +876,7 @@ sysret_audit: /* Do syscall tracing */ tracesys: #ifdef CONFIG_AUDITSYSCALL - testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags(%rcx) jz auditsys #endif SAVE_REST @@ -589,16 +884,20 @@ tracesys: FIXUP_TOP_OF_STACK %rdi movq %rsp,%rdi call syscall_trace_enter + + pax_erase_kstack + /* * Reload arg registers from stack in case ptrace changed them. * We don't reload %rax because syscall_trace_enter() returned * the value it wants us to use in the table lookup. */ LOAD_ARGS ARGOFFSET, 1 + pax_set_fptr_mask RESTORE_REST cmpq $__NR_syscall_max,%rax ja int_ret_from_sys_call /* RAX(%rsp) set to -ENOSYS above */ - movq %r10,%rcx /* fixup for C */ + movq R10-ARGOFFSET(%rsp),%rcx /* fixup for C */ call *sys_call_table(,%rax,8) movq %rax,RAX-ARGOFFSET(%rsp) /* Use IRET because user could have changed frame */ @@ -619,6 +918,7 @@ GLOBAL(int_with_check) andl %edi,%edx jnz int_careful andl $~TS_COMPAT,TI_status(%rcx) + pax_erase_kstack jmp retint_swapgs /* Either reschedule or signal or syscall exit tracking needed. */ @@ -665,7 +965,7 @@ int_restore_rest: TRACE_IRQS_OFF jmp int_with_check CFI_ENDPROC -END(system_call) +ENDPROC(system_call) /* * Certain special system calls that need to save a complete full stack frame. @@ -681,7 +981,7 @@ ENTRY(\label) call \func jmp ptregscall_common CFI_ENDPROC -END(\label) +ENDPROC(\label) .endm PTREGSCALL stub_clone, sys_clone, %r8 @@ -699,9 +999,10 @@ ENTRY(ptregscall_common) movq_cfi_restore R12+8, r12 movq_cfi_restore RBP+8, rbp movq_cfi_restore RBX+8, rbx + pax_force_retaddr ret $REST_SKIP /* pop extended registers */ CFI_ENDPROC -END(ptregscall_common) +ENDPROC(ptregscall_common) ENTRY(stub_execve) CFI_STARTPROC @@ -716,7 +1017,7 @@ ENTRY(stub_execve) RESTORE_REST jmp int_ret_from_sys_call CFI_ENDPROC -END(stub_execve) +ENDPROC(stub_execve) /* * sigreturn is special because it needs to restore all registers on return. @@ -734,7 +1035,7 @@ ENTRY(stub_rt_sigreturn) RESTORE_REST jmp int_ret_from_sys_call CFI_ENDPROC -END(stub_rt_sigreturn) +ENDPROC(stub_rt_sigreturn) /* * Build the entry stubs and pointer table with some assembler magic. @@ -769,7 +1070,7 @@ vector=vector+1 2: jmp common_interrupt .endr CFI_ENDPROC -END(irq_entries_start) +ENDPROC(irq_entries_start) .previous END(interrupt) @@ -789,6 +1090,16 @@ END(interrupt) subq $ORIG_RAX-RBP, %rsp CFI_ADJUST_CFA_OFFSET ORIG_RAX-RBP SAVE_ARGS_IRQ +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS(%rdi) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif call \func .endm @@ -820,7 +1131,7 @@ ret_from_intr: exit_intr: GET_THREAD_INFO(%rcx) - testl $3,CS-ARGOFFSET(%rsp) + testb $3,CS-ARGOFFSET(%rsp) je retint_kernel /* Interrupt came from user space */ @@ -842,12 +1153,15 @@ retint_swapgs: /* return to user-space */ * The iretq could re-enable interrupts: */ DISABLE_INTERRUPTS(CLBR_ANY) + pax_exit_kernel_user TRACE_IRQS_IRETQ SWAPGS jmp restore_args retint_restore_args: /* return to kernel space */ DISABLE_INTERRUPTS(CLBR_ANY) + pax_exit_kernel + pax_force_retaddr RIP-ARGOFFSET /* * The iretq could re-enable interrupts: */ @@ -936,7 +1250,7 @@ ENTRY(retint_kernel) #endif CFI_ENDPROC -END(common_interrupt) +ENDPROC(common_interrupt) /* * End of kprobes section */ @@ -953,7 +1267,7 @@ ENTRY(\sym) interrupt \do_sym jmp ret_from_intr CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm #ifdef CONFIG_SMP @@ -1026,12 +1340,22 @@ ENTRY(\sym) CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 call error_entry DEFAULT_FRAME 0 +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif movq %rsp,%rdi /* pt_regs pointer */ xorl %esi,%esi /* no error code */ call \do_sym jmp error_exit /* %ebx: no swapgs flag */ CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm .macro paranoidzeroentry sym do_sym @@ -1043,15 +1367,25 @@ ENTRY(\sym) CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 call save_paranoid TRACE_IRQS_OFF +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif movq %rsp,%rdi /* pt_regs pointer */ xorl %esi,%esi /* no error code */ call \do_sym jmp paranoid_exit /* %ebx: no swapgs flag */ CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm -#define INIT_TSS_IST(x) PER_CPU_VAR(init_tss) + (TSS_ist + ((x) - 1) * 8) +#define INIT_TSS_IST(x) (TSS_ist + ((x) - 1) * 8)(%r12) .macro paranoidzeroentry_ist sym do_sym ist ENTRY(\sym) INTR_FRAME @@ -1061,14 +1395,30 @@ ENTRY(\sym) CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 call save_paranoid TRACE_IRQS_OFF +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif movq %rsp,%rdi /* pt_regs pointer */ xorl %esi,%esi /* no error code */ +#ifdef CONFIG_SMP + imul $TSS_size, PER_CPU_VAR(cpu_number), %r12d + lea init_tss(%r12), %r12 +#else + lea init_tss(%rip), %r12 +#endif subq $EXCEPTION_STKSZ, INIT_TSS_IST(\ist) call \do_sym addq $EXCEPTION_STKSZ, INIT_TSS_IST(\ist) jmp paranoid_exit /* %ebx: no swapgs flag */ CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm .macro errorentry sym do_sym @@ -1079,13 +1429,23 @@ ENTRY(\sym) CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 call error_entry DEFAULT_FRAME 0 +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif movq %rsp,%rdi /* pt_regs pointer */ movq ORIG_RAX(%rsp),%rsi /* get error code */ movq $-1,ORIG_RAX(%rsp) /* no syscall to restart */ call \do_sym jmp error_exit /* %ebx: no swapgs flag */ CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm /* error code is on the stack already */ @@ -1098,13 +1458,23 @@ ENTRY(\sym) call save_paranoid DEFAULT_FRAME 0 TRACE_IRQS_OFF +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif movq %rsp,%rdi /* pt_regs pointer */ movq ORIG_RAX(%rsp),%rsi /* get error code */ movq $-1,ORIG_RAX(%rsp) /* no syscall to restart */ call \do_sym jmp paranoid_exit /* %ebx: no swapgs flag */ CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm zeroentry divide_error do_divide_error @@ -1134,9 +1504,10 @@ gs_change: 2: mfence /* workaround */ SWAPGS popfq_cfi + pax_force_retaddr ret CFI_ENDPROC -END(native_load_gs_index) +ENDPROC(native_load_gs_index) .section __ex_table,"a" .align 8 @@ -1158,13 +1529,14 @@ ENTRY(kernel_thread_helper) * Here we are in the child and the registers are set as they were * at kernel_thread() invocation in the parent. */ + pax_force_fptr %rsi call *%rsi # exit mov %eax, %edi call do_exit ud2 # padding for call trace CFI_ENDPROC -END(kernel_thread_helper) +ENDPROC(kernel_thread_helper) /* * execve(). This function needs to use IRET, not SYSRET, to set up all state properly. @@ -1191,11 +1563,11 @@ ENTRY(kernel_execve) RESTORE_REST testq %rax,%rax je int_ret_from_sys_call - RESTORE_ARGS UNFAKE_STACK_FRAME + pax_force_retaddr ret CFI_ENDPROC -END(kernel_execve) +ENDPROC(kernel_execve) /* Call softirq on interrupt stack. Interrupts are off. */ ENTRY(call_softirq) @@ -1213,9 +1585,10 @@ ENTRY(call_softirq) CFI_DEF_CFA_REGISTER rsp CFI_ADJUST_CFA_OFFSET -8 decl PER_CPU_VAR(irq_count) + pax_force_retaddr ret CFI_ENDPROC -END(call_softirq) +ENDPROC(call_softirq) #ifdef CONFIG_XEN zeroentry xen_hypervisor_callback xen_do_hypervisor_callback @@ -1253,7 +1626,7 @@ ENTRY(xen_do_hypervisor_callback) # do_hypervisor_callback(struct *pt_regs) decl PER_CPU_VAR(irq_count) jmp error_exit CFI_ENDPROC -END(xen_do_hypervisor_callback) +ENDPROC(xen_do_hypervisor_callback) /* * Hypervisor uses this for application faults while it executes. @@ -1312,7 +1685,7 @@ ENTRY(xen_failsafe_callback) SAVE_ALL jmp error_exit CFI_ENDPROC -END(xen_failsafe_callback) +ENDPROC(xen_failsafe_callback) apicinterrupt XEN_HVM_EVTCHN_CALLBACK \ xen_hvm_callback_vector xen_evtchn_do_upcall @@ -1361,16 +1734,31 @@ ENTRY(paranoid_exit) TRACE_IRQS_OFF testl %ebx,%ebx /* swapgs needed? */ jnz paranoid_restore - testl $3,CS(%rsp) + testb $3,CS(%rsp) jnz paranoid_userspace +#ifdef CONFIG_PAX_MEMORY_UDEREF + pax_exit_kernel + TRACE_IRQS_IRETQ 0 + SWAPGS_UNSAFE_STACK + RESTORE_ALL 8 + pax_force_retaddr_bts + jmp irq_return +#endif paranoid_swapgs: +#ifdef CONFIG_PAX_MEMORY_UDEREF + pax_exit_kernel_user +#else + pax_exit_kernel +#endif TRACE_IRQS_IRETQ 0 SWAPGS_UNSAFE_STACK RESTORE_ALL 8 jmp irq_return paranoid_restore: + pax_exit_kernel TRACE_IRQS_IRETQ 0 RESTORE_ALL 8 + pax_force_retaddr_bts jmp irq_return paranoid_userspace: GET_THREAD_INFO(%rcx) @@ -1399,7 +1787,7 @@ paranoid_schedule: TRACE_IRQS_OFF jmp paranoid_userspace CFI_ENDPROC -END(paranoid_exit) +ENDPROC(paranoid_exit) /* * Exception entry point. This expects an error code/orig_rax on the stack. @@ -1426,12 +1814,13 @@ ENTRY(error_entry) movq_cfi r14, R14+8 movq_cfi r15, R15+8 xorl %ebx,%ebx - testl $3,CS+8(%rsp) + testb $3,CS+8(%rsp) je error_kernelspace error_swapgs: SWAPGS error_sti: TRACE_IRQS_OFF + pax_force_retaddr_bts ret /* @@ -1458,7 +1847,7 @@ bstep_iret: movq %rcx,RIP+8(%rsp) jmp error_swapgs CFI_ENDPROC -END(error_entry) +ENDPROC(error_entry) /* ebx: no swapgs flag (1: don't need swapgs, 0: need it) */ @@ -1478,7 +1867,7 @@ ENTRY(error_exit) jnz retint_careful jmp retint_swapgs CFI_ENDPROC -END(error_exit) +ENDPROC(error_exit) /* * Test if a given stack is an NMI stack or not. @@ -1535,9 +1924,11 @@ ENTRY(nmi) * If %cs was not the kernel segment, then the NMI triggered in user * space, which means it is definitely not nested. */ + cmpl $__KERNEXEC_KERNEL_CS, 16(%rsp) + je 1f cmpl $__KERNEL_CS, 16(%rsp) jne first_nmi - +1: /* * Check the special variable on the stack to see if NMIs are * executing. @@ -1659,6 +2050,16 @@ restart_nmi: */ call save_paranoid DEFAULT_FRAME 0 +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif /* paranoidentry do_nmi, 0; without TRACE_IRQS_OFF */ movq %rsp,%rdi movq $-1,%rsi @@ -1666,14 +2067,25 @@ restart_nmi: testl %ebx,%ebx /* swapgs needed? */ jnz nmi_restore nmi_swapgs: +#ifdef CONFIG_PAX_MEMORY_UDEREF + pax_exit_kernel_user +#else + pax_exit_kernel +#endif SWAPGS_UNSAFE_STACK + RESTORE_ALL 8 + /* Clear the NMI executing stack variable */ + movq $0, 10*8(%rsp) + jmp irq_return nmi_restore: + pax_exit_kernel RESTORE_ALL 8 + pax_force_retaddr_bts /* Clear the NMI executing stack variable */ movq $0, 10*8(%rsp) jmp irq_return CFI_ENDPROC -END(nmi) +ENDPROC(nmi) /* * If an NMI hit an iret because of an exception or breakpoint, @@ -1700,7 +2112,7 @@ ENTRY(ignore_sysret) mov $-ENOSYS,%eax sysret CFI_ENDPROC -END(ignore_sysret) +ENDPROC(ignore_sysret) /* * End of kprobes section diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index c9a281f..ce2f317 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -126,7 +126,7 @@ static void *mod_code_ip; /* holds the IP to write to */ static const void *mod_code_newcode; /* holds the text to write to the IP */ static unsigned nmi_wait_count; -static atomic_t nmi_update_count = ATOMIC_INIT(0); +static atomic_unchecked_t nmi_update_count = ATOMIC_INIT(0); int ftrace_arch_read_dyn_info(char *buf, int size) { @@ -134,7 +134,7 @@ int ftrace_arch_read_dyn_info(char *buf, int size) r = snprintf(buf, size, "%u %u", nmi_wait_count, - atomic_read(&nmi_update_count)); + atomic_read_unchecked(&nmi_update_count)); return r; } @@ -177,8 +177,10 @@ void ftrace_nmi_enter(void) if (atomic_inc_return(&nmi_running) & MOD_CODE_WRITE_FLAG) { smp_rmb(); + pax_open_kernel(); ftrace_mod_code(); - atomic_inc(&nmi_update_count); + pax_close_kernel(); + atomic_inc_unchecked(&nmi_update_count); } /* Must have previous changes seen before executions */ smp_mb(); @@ -271,6 +273,8 @@ ftrace_modify_code(unsigned long ip, unsigned const char *old_code, { unsigned char replaced[MCOUNT_INSN_SIZE]; + ip = ktla_ktva(ip); + /* * Note: Due to modules and __init, code can * disappear and change, we need to protect against faulting @@ -327,7 +331,7 @@ int ftrace_update_ftrace_func(ftrace_func_t func) unsigned char old[MCOUNT_INSN_SIZE], *new; int ret; - memcpy(old, &ftrace_call, MCOUNT_INSN_SIZE); + memcpy(old, (void *)ktla_ktva((unsigned long)ftrace_call), MCOUNT_INSN_SIZE); new = ftrace_call_replace(ip, (unsigned long)func); ret = ftrace_modify_code(ip, old, new); @@ -353,6 +357,8 @@ static int ftrace_mod_jmp(unsigned long ip, { unsigned char code[MCOUNT_INSN_SIZE]; + ip = ktla_ktva(ip); + if (probe_kernel_read(code, (void *)ip, MCOUNT_INSN_SIZE)) return -EFAULT; diff --git a/arch/x86/kernel/head32.c b/arch/x86/kernel/head32.c index 51ff186..9e77418 100644 --- a/arch/x86/kernel/head32.c +++ b/arch/x86/kernel/head32.c @@ -19,6 +19,7 @@ #include #include #include +#include static void __init i386_default_early_setup(void) { @@ -31,8 +32,7 @@ static void __init i386_default_early_setup(void) void __init i386_start_kernel(void) { - memblock_reserve(__pa_symbol(&_text), - __pa_symbol(&__bss_stop) - __pa_symbol(&_text)); + memblock_reserve(LOAD_PHYSICAL_ADDR, __pa_symbol(&__bss_stop) - LOAD_PHYSICAL_ADDR); #ifdef CONFIG_BLK_DEV_INITRD /* Reserve INITRD */ diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index ce0be7c..c41476e 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -25,6 +25,12 @@ /* Physical address */ #define pa(X) ((X) - __PAGE_OFFSET) +#ifdef CONFIG_PAX_KERNEXEC +#define ta(X) (X) +#else +#define ta(X) ((X) - __PAGE_OFFSET) +#endif + /* * References to members of the new_cpu_data structure. */ @@ -54,11 +60,7 @@ * and small than max_low_pfn, otherwise will waste some page table entries */ -#if PTRS_PER_PMD > 1 -#define PAGE_TABLE_SIZE(pages) (((pages) / PTRS_PER_PMD) + PTRS_PER_PGD) -#else -#define PAGE_TABLE_SIZE(pages) ((pages) / PTRS_PER_PGD) -#endif +#define PAGE_TABLE_SIZE(pages) ((pages) / PTRS_PER_PTE) /* Number of possible pages in the lowmem region */ LOWMEM_PAGES = (((1<<32) - __PAGE_OFFSET) >> PAGE_SHIFT) @@ -77,6 +79,12 @@ INIT_MAP_SIZE = PAGE_TABLE_SIZE(KERNEL_PAGES) * PAGE_SIZE RESERVE_BRK(pagetables, INIT_MAP_SIZE) /* + * Real beginning of normal "text" segment + */ +ENTRY(stext) +ENTRY(_stext) + +/* * 32-bit kernel entrypoint; only used by the boot CPU. On entry, * %esi points to the real-mode code as a 32-bit pointer. * CS and DS must be 4 GB flat segments, but we don't depend on @@ -84,6 +92,13 @@ RESERVE_BRK(pagetables, INIT_MAP_SIZE) * can. */ __HEAD + +#ifdef CONFIG_PAX_KERNEXEC + jmp startup_32 +/* PaX: fill first page in .text with int3 to catch NULL derefs in kernel mode */ +.fill PAGE_SIZE-5,1,0xcc +#endif + ENTRY(startup_32) movl pa(stack_start),%ecx @@ -105,6 +120,57 @@ ENTRY(startup_32) 2: leal -__PAGE_OFFSET(%ecx),%esp +#ifdef CONFIG_SMP + movl $pa(cpu_gdt_table),%edi + movl $__per_cpu_load,%eax + movw %ax,__KERNEL_PERCPU + 2(%edi) + rorl $16,%eax + movb %al,__KERNEL_PERCPU + 4(%edi) + movb %ah,__KERNEL_PERCPU + 7(%edi) + movl $__per_cpu_end - 1,%eax + subl $__per_cpu_start,%eax + movw %ax,__KERNEL_PERCPU + 0(%edi) +#endif + +#ifdef CONFIG_PAX_MEMORY_UDEREF + movl $NR_CPUS,%ecx + movl $pa(cpu_gdt_table),%edi +1: + movl $((((__PAGE_OFFSET-1) & 0xf0000000) >> 12) | 0x00c09700),GDT_ENTRY_KERNEL_DS * 8 + 4(%edi) + movl $((((__PAGE_OFFSET-1) & 0xf0000000) >> 12) | 0x00c0fb00),GDT_ENTRY_DEFAULT_USER_CS * 8 + 4(%edi) + movl $((((__PAGE_OFFSET-1) & 0xf0000000) >> 12) | 0x00c0f300),GDT_ENTRY_DEFAULT_USER_DS * 8 + 4(%edi) + addl $PAGE_SIZE_asm,%edi + loop 1b +#endif + +#ifdef CONFIG_PAX_KERNEXEC + movl $pa(boot_gdt),%edi + movl $__LOAD_PHYSICAL_ADDR,%eax + movw %ax,__BOOT_CS + 2(%edi) + rorl $16,%eax + movb %al,__BOOT_CS + 4(%edi) + movb %ah,__BOOT_CS + 7(%edi) + rorl $16,%eax + + ljmp $(__BOOT_CS),$1f +1: + + movl $NR_CPUS,%ecx + movl $pa(cpu_gdt_table),%edi + addl $__PAGE_OFFSET,%eax +1: + movw %ax,__KERNEL_CS + 2(%edi) + movw %ax,__KERNEXEC_KERNEL_CS + 2(%edi) + rorl $16,%eax + movb %al,__KERNEL_CS + 4(%edi) + movb %al,__KERNEXEC_KERNEL_CS + 4(%edi) + movb %ah,__KERNEL_CS + 7(%edi) + movb %ah,__KERNEXEC_KERNEL_CS + 7(%edi) + rorl $16,%eax + addl $PAGE_SIZE_asm,%edi + loop 1b +#endif + /* * Clear BSS first so that there are no surprises... */ @@ -195,8 +261,11 @@ ENTRY(startup_32) movl %eax, pa(max_pfn_mapped) /* Do early initialization of the fixmap area */ - movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax - movl %eax,pa(initial_pg_pmd+0x1000*KPMDS-8) +#ifdef CONFIG_COMPAT_VDSO + movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR+_PAGE_USER,pa(initial_pg_pmd+0x1000*KPMDS-8) +#else + movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,pa(initial_pg_pmd+0x1000*KPMDS-8) +#endif #else /* Not PAE */ page_pde_offset = (__PAGE_OFFSET >> 20); @@ -226,8 +295,11 @@ page_pde_offset = (__PAGE_OFFSET >> 20); movl %eax, pa(max_pfn_mapped) /* Do early initialization of the fixmap area */ - movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax - movl %eax,pa(initial_page_table+0xffc) +#ifdef CONFIG_COMPAT_VDSO + movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR+_PAGE_USER,pa(initial_page_table+0xffc) +#else + movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,pa(initial_page_table+0xffc) +#endif #endif #ifdef CONFIG_PARAVIRT @@ -241,9 +313,7 @@ page_pde_offset = (__PAGE_OFFSET >> 20); cmpl $num_subarch_entries, %eax jae bad_subarch - movl pa(subarch_entries)(,%eax,4), %eax - subl $__PAGE_OFFSET, %eax - jmp *%eax + jmp *pa(subarch_entries)(,%eax,4) bad_subarch: WEAK(lguest_entry) @@ -255,10 +325,10 @@ WEAK(xen_entry) __INITDATA subarch_entries: - .long default_entry /* normal x86/PC */ - .long lguest_entry /* lguest hypervisor */ - .long xen_entry /* Xen hypervisor */ - .long default_entry /* Moorestown MID */ + .long ta(default_entry) /* normal x86/PC */ + .long ta(lguest_entry) /* lguest hypervisor */ + .long ta(xen_entry) /* Xen hypervisor */ + .long ta(default_entry) /* Moorestown MID */ num_subarch_entries = (. - subarch_entries) / 4 .previous #else @@ -312,6 +382,7 @@ default_entry: orl %edx,%eax movl %eax,%cr4 +#ifdef CONFIG_X86_PAE testb $X86_CR4_PAE, %al # check if PAE is enabled jz 6f @@ -340,6 +411,9 @@ default_entry: /* Make changes effective */ wrmsr + btsl $_PAGE_BIT_NX-32,pa(__supported_pte_mask+4) +#endif + 6: /* @@ -443,7 +517,7 @@ is386: movl $2,%ecx # set MP 1: movl $(__KERNEL_DS),%eax # reload all the segment registers movl %eax,%ss # after changing gdt. - movl $(__USER_DS),%eax # DS/ES contains default USER segment +# movl $(__KERNEL_DS),%eax # DS/ES contains default KERNEL segment movl %eax,%ds movl %eax,%es @@ -457,15 +531,22 @@ is386: movl $2,%ecx # set MP */ cmpb $0,ready jne 1f - movl $gdt_page,%eax + movl $cpu_gdt_table,%eax movl $stack_canary,%ecx +#ifdef CONFIG_SMP + addl $__per_cpu_load,%ecx +#endif movw %cx, 8 * GDT_ENTRY_STACK_CANARY + 2(%eax) shrl $16, %ecx movb %cl, 8 * GDT_ENTRY_STACK_CANARY + 4(%eax) movb %ch, 8 * GDT_ENTRY_STACK_CANARY + 7(%eax) 1: -#endif movl $(__KERNEL_STACK_CANARY),%eax +#elif defined(CONFIG_PAX_MEMORY_UDEREF) + movl $(__USER_DS),%eax +#else + xorl %eax,%eax +#endif movl %eax,%gs xorl %eax,%eax # Clear LDT @@ -558,22 +639,22 @@ early_page_fault: jmp early_fault early_fault: - cld #ifdef CONFIG_PRINTK + cmpl $1,%ss:early_recursion_flag + je hlt_loop + incl %ss:early_recursion_flag + cld pusha movl $(__KERNEL_DS),%eax movl %eax,%ds movl %eax,%es - cmpl $2,early_recursion_flag - je hlt_loop - incl early_recursion_flag movl %cr2,%eax pushl %eax pushl %edx /* trapno */ pushl $fault_msg call printk +; call dump_stack #endif - call dump_stack hlt_loop: hlt jmp hlt_loop @@ -581,8 +662,11 @@ hlt_loop: /* This is the default interrupt "handler" :-) */ ALIGN ignore_int: - cld #ifdef CONFIG_PRINTK + cmpl $2,%ss:early_recursion_flag + je hlt_loop + incl %ss:early_recursion_flag + cld pushl %eax pushl %ecx pushl %edx @@ -591,9 +675,6 @@ ignore_int: movl $(__KERNEL_DS),%eax movl %eax,%ds movl %eax,%es - cmpl $2,early_recursion_flag - je hlt_loop - incl early_recursion_flag pushl 16(%esp) pushl 24(%esp) pushl 32(%esp) @@ -622,29 +703,43 @@ ENTRY(initial_code) /* * BSS section */ -__PAGE_ALIGNED_BSS - .align PAGE_SIZE #ifdef CONFIG_X86_PAE +.section .initial_pg_pmd,"a",@progbits initial_pg_pmd: .fill 1024*KPMDS,4,0 #else +.section .initial_page_table,"a",@progbits ENTRY(initial_page_table) .fill 1024,4,0 #endif +.section .initial_pg_fixmap,"a",@progbits initial_pg_fixmap: .fill 1024,4,0 +.section .empty_zero_page,"a",@progbits ENTRY(empty_zero_page) .fill 4096,1,0 +.section .swapper_pg_dir,"a",@progbits ENTRY(swapper_pg_dir) +#ifdef CONFIG_X86_PAE + .fill 4,8,0 +#else .fill 1024,4,0 +#endif + +/* + * The IDT has to be page-aligned to simplify the Pentium + * F0 0F bug workaround.. We have a special link segment + * for this. + */ +.section .idt,"a",@progbits +ENTRY(idt_table) + .fill 256,8,0 /* * This starts the data section. */ #ifdef CONFIG_X86_PAE -__PAGE_ALIGNED_DATA - /* Page-aligned for the benefit of paravirt? */ - .align PAGE_SIZE +.section .initial_page_table,"a",@progbits ENTRY(initial_page_table) .long pa(initial_pg_pmd+PGD_IDENT_ATTR),0 /* low identity map */ # if KPMDS == 3 @@ -663,18 +758,27 @@ ENTRY(initial_page_table) # error "Kernel PMDs should be 1, 2 or 3" # endif .align PAGE_SIZE /* needs to be page-sized too */ + +#ifdef CONFIG_PAX_PER_CPU_PGD +ENTRY(cpu_pgd) + .rept NR_CPUS + .fill 4,8,0 + .endr +#endif + #endif .data .balign 4 ENTRY(stack_start) - .long init_thread_union+THREAD_SIZE + .long init_thread_union+THREAD_SIZE-8 +ready: .byte 0 + +.section .rodata,"a",@progbits early_recursion_flag: .long 0 -ready: .byte 0 - int_msg: .asciz "Unknown interrupt or fault at: %p %p %p\n" @@ -707,7 +811,7 @@ fault_msg: .word 0 # 32 bit align gdt_desc.address boot_gdt_descr: .word __BOOT_DS+7 - .long boot_gdt - __PAGE_OFFSET + .long pa(boot_gdt) .word 0 # 32-bit align idt_desc.address idt_descr: @@ -718,7 +822,7 @@ idt_descr: .word 0 # 32 bit align gdt_desc.address ENTRY(early_gdt_descr) .word GDT_ENTRIES*8-1 - .long gdt_page /* Overwritten for secondary CPUs */ + .long cpu_gdt_table /* Overwritten for secondary CPUs */ /* * The boot_gdt must mirror the equivalent in setup.S and is @@ -727,5 +831,65 @@ ENTRY(early_gdt_descr) .align L1_CACHE_BYTES ENTRY(boot_gdt) .fill GDT_ENTRY_BOOT_CS,8,0 - .quad 0x00cf9a000000ffff /* kernel 4GB code at 0x00000000 */ - .quad 0x00cf92000000ffff /* kernel 4GB data at 0x00000000 */ + .quad 0x00cf9b000000ffff /* kernel 4GB code at 0x00000000 */ + .quad 0x00cf93000000ffff /* kernel 4GB data at 0x00000000 */ + + .align PAGE_SIZE_asm +ENTRY(cpu_gdt_table) + .rept NR_CPUS + .quad 0x0000000000000000 /* NULL descriptor */ + .quad 0x0000000000000000 /* 0x0b reserved */ + .quad 0x0000000000000000 /* 0x13 reserved */ + .quad 0x0000000000000000 /* 0x1b reserved */ + +#ifdef CONFIG_PAX_KERNEXEC + .quad 0x00cf9b000000ffff /* 0x20 alternate kernel 4GB code at 0x00000000 */ +#else + .quad 0x0000000000000000 /* 0x20 unused */ +#endif + + .quad 0x0000000000000000 /* 0x28 unused */ + .quad 0x0000000000000000 /* 0x33 TLS entry 1 */ + .quad 0x0000000000000000 /* 0x3b TLS entry 2 */ + .quad 0x0000000000000000 /* 0x43 TLS entry 3 */ + .quad 0x0000000000000000 /* 0x4b reserved */ + .quad 0x0000000000000000 /* 0x53 reserved */ + .quad 0x0000000000000000 /* 0x5b reserved */ + + .quad 0x00cf9b000000ffff /* 0x60 kernel 4GB code at 0x00000000 */ + .quad 0x00cf93000000ffff /* 0x68 kernel 4GB data at 0x00000000 */ + .quad 0x00cffb000000ffff /* 0x73 user 4GB code at 0x00000000 */ + .quad 0x00cff3000000ffff /* 0x7b user 4GB data at 0x00000000 */ + + .quad 0x0000000000000000 /* 0x80 TSS descriptor */ + .quad 0x0000000000000000 /* 0x88 LDT descriptor */ + + /* + * Segments used for calling PnP BIOS have byte granularity. + * The code segments and data segments have fixed 64k limits, + * the transfer segment sizes are set at run time. + */ + .quad 0x00409b000000ffff /* 0x90 32-bit code */ + .quad 0x00009b000000ffff /* 0x98 16-bit code */ + .quad 0x000093000000ffff /* 0xa0 16-bit data */ + .quad 0x0000930000000000 /* 0xa8 16-bit data */ + .quad 0x0000930000000000 /* 0xb0 16-bit data */ + + /* + * The APM segments have byte granularity and their bases + * are set at run time. All have 64k limits. + */ + .quad 0x00409b000000ffff /* 0xb8 APM CS code */ + .quad 0x00009b000000ffff /* 0xc0 APM CS 16 code (16 bit) */ + .quad 0x004093000000ffff /* 0xc8 APM DS data */ + + .quad 0x00c0930000000000 /* 0xd0 - ESPFIX SS */ + .quad 0x0040930000000000 /* 0xd8 - PERCPU */ + .quad 0x0040910000000017 /* 0xe0 - STACK_CANARY */ + .quad 0x0000000000000000 /* 0xe8 - PCIBIOS_CS */ + .quad 0x0000000000000000 /* 0xf0 - PCIBIOS_DS */ + .quad 0x0000000000000000 /* 0xf8 - GDT entry 31: double-fault TSS */ + + /* Be sure this is zeroed to avoid false validations in Xen */ + .fill PAGE_SIZE_asm - GDT_SIZE,1,0 + .endr diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index 40f4eb3..6d24d9d 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -19,6 +19,8 @@ #include #include #include +#include +#include #ifdef CONFIG_PARAVIRT #include @@ -38,6 +40,12 @@ L4_PAGE_OFFSET = pgd_index(__PAGE_OFFSET) L3_PAGE_OFFSET = pud_index(__PAGE_OFFSET) L4_START_KERNEL = pgd_index(__START_KERNEL_map) L3_START_KERNEL = pud_index(__START_KERNEL_map) +L4_VMALLOC_START = pgd_index(VMALLOC_START) +L3_VMALLOC_START = pud_index(VMALLOC_START) +L4_VMALLOC_END = pgd_index(VMALLOC_END) +L3_VMALLOC_END = pud_index(VMALLOC_END) +L4_VMEMMAP_START = pgd_index(VMEMMAP_START) +L3_VMEMMAP_START = pud_index(VMEMMAP_START) .text __HEAD @@ -85,35 +93,23 @@ startup_64: */ addq %rbp, init_level4_pgt + 0(%rip) addq %rbp, init_level4_pgt + (L4_PAGE_OFFSET*8)(%rip) + addq %rbp, init_level4_pgt + (L4_VMALLOC_START*8)(%rip) + addq %rbp, init_level4_pgt + (L4_VMALLOC_END*8)(%rip) + addq %rbp, init_level4_pgt + (L4_VMEMMAP_START*8)(%rip) addq %rbp, init_level4_pgt + (L4_START_KERNEL*8)(%rip) addq %rbp, level3_ident_pgt + 0(%rip) +#ifndef CONFIG_XEN + addq %rbp, level3_ident_pgt + 8(%rip) +#endif - addq %rbp, level3_kernel_pgt + (510*8)(%rip) - addq %rbp, level3_kernel_pgt + (511*8)(%rip) + addq %rbp, level3_vmemmap_pgt + (L3_VMEMMAP_START*8)(%rip) + + addq %rbp, level3_kernel_pgt + (L3_START_KERNEL*8)(%rip) + addq %rbp, level3_kernel_pgt + (L3_START_KERNEL*8+8)(%rip) addq %rbp, level2_fixmap_pgt + (506*8)(%rip) - - /* Add an Identity mapping if I am above 1G */ - leaq _text(%rip), %rdi - andq $PMD_PAGE_MASK, %rdi - - movq %rdi, %rax - shrq $PUD_SHIFT, %rax - andq $(PTRS_PER_PUD - 1), %rax - jz ident_complete - - leaq (level2_spare_pgt - __START_KERNEL_map + _KERNPG_TABLE)(%rbp), %rdx - leaq level3_ident_pgt(%rip), %rbx - movq %rdx, 0(%rbx, %rax, 8) - - movq %rdi, %rax - shrq $PMD_SHIFT, %rax - andq $(PTRS_PER_PMD - 1), %rax - leaq __PAGE_KERNEL_IDENT_LARGE_EXEC(%rdi), %rdx - leaq level2_spare_pgt(%rip), %rbx - movq %rdx, 0(%rbx, %rax, 8) -ident_complete: + addq %rbp, level2_fixmap_pgt + (507*8)(%rip) /* * Fixup the kernel text+data virtual addresses. Note that @@ -160,8 +156,8 @@ ENTRY(secondary_startup_64) * after the boot processor executes this code. */ - /* Enable PAE mode and PGE */ - movl $(X86_CR4_PAE | X86_CR4_PGE), %eax + /* Enable PAE mode and PSE/PGE */ + movl $(X86_CR4_PSE | X86_CR4_PAE | X86_CR4_PGE), %eax movq %rax, %cr4 /* Setup early boot stage 4 level pagetables. */ @@ -183,9 +179,17 @@ ENTRY(secondary_startup_64) movl $MSR_EFER, %ecx rdmsr btsl $_EFER_SCE, %eax /* Enable System Call */ - btl $20,%edi /* No Execute supported? */ + btl $(X86_FEATURE_NX & 31),%edi /* No Execute supported? */ jnc 1f btsl $_EFER_NX, %eax + leaq init_level4_pgt(%rip), %rdi +#ifndef CONFIG_EFI + btsq $_PAGE_BIT_NX, 8*L4_PAGE_OFFSET(%rdi) +#endif + btsq $_PAGE_BIT_NX, 8*L4_VMALLOC_START(%rdi) + btsq $_PAGE_BIT_NX, 8*L4_VMALLOC_END(%rdi) + btsq $_PAGE_BIT_NX, 8*L4_VMEMMAP_START(%rdi) + btsq $_PAGE_BIT_NX, __supported_pte_mask(%rip) 1: wrmsr /* Make changes effective */ /* Setup cr0 */ @@ -247,6 +251,7 @@ ENTRY(secondary_startup_64) * jump. In addition we need to ensure %cs is set so we make this * a far return. */ + pax_set_fptr_mask movq initial_code(%rip),%rax pushq $0 # fake return address to stop unwinder pushq $__KERNEL_CS # set correct cs @@ -269,7 +274,7 @@ ENTRY(secondary_startup_64) bad_address: jmp bad_address - .section ".init.text","ax" + __INIT #ifdef CONFIG_EARLY_PRINTK .globl early_idt_handlers early_idt_handlers: @@ -314,18 +319,23 @@ ENTRY(early_idt_handler) #endif /* EARLY_PRINTK */ 1: hlt jmp 1b + .previous #ifdef CONFIG_EARLY_PRINTK + __INITDATA early_recursion_flag: .long 0 + .previous + .section .rodata,"a",@progbits early_idt_msg: .asciz "PANIC: early exception %02lx rip %lx:%lx error %lx cr2 %lx\n" early_idt_ripmsg: .asciz "RIP %s\n" + .previous #endif /* CONFIG_EARLY_PRINTK */ - .previous + .section .rodata,"a",@progbits #define NEXT_PAGE(name) \ .balign PAGE_SIZE; \ ENTRY(name) @@ -338,7 +348,6 @@ ENTRY(name) i = i + 1 ; \ .endr - .data /* * This default setting generates an ident mapping at address 0x100000 * and a mapping for the kernel that precisely maps virtual address @@ -349,13 +358,41 @@ NEXT_PAGE(init_level4_pgt) .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE .org init_level4_pgt + L4_PAGE_OFFSET*8, 0 .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE + .org init_level4_pgt + L4_VMALLOC_START*8, 0 + .quad level3_vmalloc_start_pgt - __START_KERNEL_map + _KERNPG_TABLE + .org init_level4_pgt + L4_VMALLOC_END*8, 0 + .quad level3_vmalloc_end_pgt - __START_KERNEL_map + _KERNPG_TABLE + .org init_level4_pgt + L4_VMEMMAP_START*8, 0 + .quad level3_vmemmap_pgt - __START_KERNEL_map + _KERNPG_TABLE .org init_level4_pgt + L4_START_KERNEL*8, 0 /* (2^48-(2*1024*1024*1024))/(2^39) = 511 */ .quad level3_kernel_pgt - __START_KERNEL_map + _PAGE_TABLE +#ifdef CONFIG_PAX_PER_CPU_PGD +NEXT_PAGE(cpu_pgd) + .rept NR_CPUS + .fill 512,8,0 + .endr +#endif + NEXT_PAGE(level3_ident_pgt) .quad level2_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE +#ifdef CONFIG_XEN .fill 511,8,0 +#else + .quad level2_ident_pgt + PAGE_SIZE - __START_KERNEL_map + _KERNPG_TABLE + .fill 510,8,0 +#endif + +NEXT_PAGE(level3_vmalloc_start_pgt) + .fill 512,8,0 + +NEXT_PAGE(level3_vmalloc_end_pgt) + .fill 512,8,0 + +NEXT_PAGE(level3_vmemmap_pgt) + .fill L3_VMEMMAP_START,8,0 + .quad level2_vmemmap_pgt - __START_KERNEL_map + _KERNPG_TABLE NEXT_PAGE(level3_kernel_pgt) .fill L3_START_KERNEL,8,0 @@ -363,20 +400,23 @@ NEXT_PAGE(level3_kernel_pgt) .quad level2_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE .quad level2_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE +NEXT_PAGE(level2_vmemmap_pgt) + .fill 512,8,0 + NEXT_PAGE(level2_fixmap_pgt) - .fill 506,8,0 - .quad level1_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE - /* 8MB reserved for vsyscalls + a 2MB hole = 4 + 1 entries */ - .fill 5,8,0 + .fill 507,8,0 + .quad level1_vsyscall_pgt - __START_KERNEL_map + _PAGE_TABLE + /* 6MB reserved for vsyscalls + a 2MB hole = 3 + 1 entries */ + .fill 4,8,0 -NEXT_PAGE(level1_fixmap_pgt) +NEXT_PAGE(level1_vsyscall_pgt) .fill 512,8,0 -NEXT_PAGE(level2_ident_pgt) - /* Since I easily can, map the first 1G. + /* Since I easily can, map the first 2G. * Don't set NX because code runs from these pages. */ - PMDS(0, __PAGE_KERNEL_IDENT_LARGE_EXEC, PTRS_PER_PMD) +NEXT_PAGE(level2_ident_pgt) + PMDS(0, __PAGE_KERNEL_IDENT_LARGE_EXEC, 2*PTRS_PER_PMD) NEXT_PAGE(level2_kernel_pgt) /* @@ -389,37 +429,59 @@ NEXT_PAGE(level2_kernel_pgt) * If you want to increase this then increase MODULES_VADDR * too.) */ - PMDS(0, __PAGE_KERNEL_LARGE_EXEC, - KERNEL_IMAGE_SIZE/PMD_SIZE) - -NEXT_PAGE(level2_spare_pgt) - .fill 512, 8, 0 + PMDS(0, __PAGE_KERNEL_LARGE_EXEC, KERNEL_IMAGE_SIZE/PMD_SIZE) #undef PMDS #undef NEXT_PAGE - .data + .align PAGE_SIZE +ENTRY(cpu_gdt_table) + .rept NR_CPUS + .quad 0x0000000000000000 /* NULL descriptor */ + .quad 0x00cf9b000000ffff /* __KERNEL32_CS */ + .quad 0x00af9b000000ffff /* __KERNEL_CS */ + .quad 0x00cf93000000ffff /* __KERNEL_DS */ + .quad 0x00cffb000000ffff /* __USER32_CS */ + .quad 0x00cff3000000ffff /* __USER_DS, __USER32_DS */ + .quad 0x00affb000000ffff /* __USER_CS */ + +#ifdef CONFIG_PAX_KERNEXEC + .quad 0x00af9b000000ffff /* __KERNEXEC_KERNEL_CS */ +#else + .quad 0x0 /* unused */ +#endif + + .quad 0,0 /* TSS */ + .quad 0,0 /* LDT */ + .quad 0,0,0 /* three TLS descriptors */ + .quad 0x0000f40000000000 /* node/CPU stored in limit */ + /* asm/segment.h:GDT_ENTRIES must match this */ + + /* zero the remaining page */ + .fill PAGE_SIZE / 8 - GDT_ENTRIES,8,0 + .endr + .align 16 .globl early_gdt_descr early_gdt_descr: .word GDT_ENTRIES*8-1 early_gdt_descr_base: - .quad INIT_PER_CPU_VAR(gdt_page) + .quad cpu_gdt_table ENTRY(phys_base) /* This must match the first entry in level2_kernel_pgt */ .quad 0x0000000000000000 #include "../../x86/xen/xen-head.S" - - .section .bss, "aw", @nobits + + .section .rodata,"a",@progbits .align L1_CACHE_BYTES ENTRY(idt_table) - .skip IDT_ENTRIES * 16 + .fill 512,8,0 .align L1_CACHE_BYTES ENTRY(nmi_idt_table) - .skip IDT_ENTRIES * 16 + .fill 512,8,0 __PAGE_ALIGNED_BSS .align PAGE_SIZE diff --git a/arch/x86/kernel/i386_ksyms_32.c b/arch/x86/kernel/i386_ksyms_32.c index 9c3bd4a..e1d9b35 100644 --- a/arch/x86/kernel/i386_ksyms_32.c +++ b/arch/x86/kernel/i386_ksyms_32.c @@ -20,8 +20,12 @@ extern void cmpxchg8b_emu(void); EXPORT_SYMBOL(cmpxchg8b_emu); #endif +EXPORT_SYMBOL_GPL(cpu_gdt_table); + /* Networking helper routines. */ EXPORT_SYMBOL(csum_partial_copy_generic); +EXPORT_SYMBOL(csum_partial_copy_generic_to_user); +EXPORT_SYMBOL(csum_partial_copy_generic_from_user); EXPORT_SYMBOL(__get_user_1); EXPORT_SYMBOL(__get_user_2); @@ -36,3 +40,7 @@ EXPORT_SYMBOL(strstr); EXPORT_SYMBOL(csum_partial); EXPORT_SYMBOL(empty_zero_page); + +#ifdef CONFIG_PAX_KERNEXEC +EXPORT_SYMBOL(__LOAD_PHYSICAL_ADDR); +#endif diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c index f239f30..aab2a58 100644 --- a/arch/x86/kernel/i387.c +++ b/arch/x86/kernel/i387.c @@ -189,6 +189,9 @@ int xfpregs_active(struct task_struct *target, const struct user_regset *regset) int xfpregs_get(struct task_struct *target, const struct user_regset *regset, unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) __size_overflow(4); +int xfpregs_get(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, void *kbuf, void __user *ubuf) { int ret; @@ -208,6 +211,9 @@ int xfpregs_get(struct task_struct *target, const struct user_regset *regset, int xfpregs_set(struct task_struct *target, const struct user_regset *regset, unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) __size_overflow(4); +int xfpregs_set(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, const void *kbuf, const void __user *ubuf) { int ret; @@ -241,6 +247,9 @@ int xfpregs_set(struct task_struct *target, const struct user_regset *regset, int xstateregs_get(struct task_struct *target, const struct user_regset *regset, unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) __size_overflow(4); +int xstateregs_get(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, void *kbuf, void __user *ubuf) { int ret; @@ -270,6 +279,9 @@ int xstateregs_get(struct task_struct *target, const struct user_regset *regset, int xstateregs_set(struct task_struct *target, const struct user_regset *regset, unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) __size_overflow(4); +int xstateregs_set(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, const void *kbuf, const void __user *ubuf) { int ret; @@ -440,6 +452,9 @@ static void convert_to_fxsr(struct task_struct *tsk, int fpregs_get(struct task_struct *target, const struct user_regset *regset, unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) __size_overflow(3,4); +int fpregs_get(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, void *kbuf, void __user *ubuf) { struct user_i387_ia32_struct env; @@ -472,6 +487,9 @@ int fpregs_get(struct task_struct *target, const struct user_regset *regset, int fpregs_set(struct task_struct *target, const struct user_regset *regset, unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) __size_overflow(3,4); +int fpregs_set(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, const void *kbuf, const void __user *ubuf) { struct user_i387_ia32_struct env; @@ -620,6 +638,8 @@ static inline int restore_i387_fsave(struct _fpstate_ia32 __user *buf) } static int restore_i387_fxsave(struct _fpstate_ia32 __user *buf, + unsigned int size) __size_overflow(2); +static int restore_i387_fxsave(struct _fpstate_ia32 __user *buf, unsigned int size) { struct task_struct *tsk = current; diff --git a/arch/x86/kernel/i8259.c b/arch/x86/kernel/i8259.c index 6104852..6114160 100644 --- a/arch/x86/kernel/i8259.c +++ b/arch/x86/kernel/i8259.c @@ -210,7 +210,7 @@ spurious_8259A_irq: "spurious 8259A interrupt: IRQ%d.\n", irq); spurious_irq_mask |= irqmask; } - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); /* * Theoretically we do not have to handle this IRQ, * but in Linux this does not cause problems and is diff --git a/arch/x86/kernel/init_task.c b/arch/x86/kernel/init_task.c index 43e9ccf..44ccf6f 100644 --- a/arch/x86/kernel/init_task.c +++ b/arch/x86/kernel/init_task.c @@ -20,8 +20,7 @@ static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand); * way process stacks are handled. This is done by having a special * "init_task" linker map entry.. */ -union thread_union init_thread_union __init_task_data = - { INIT_THREAD_INFO(init_task) }; +union thread_union init_thread_union __init_task_data; /* * Initial task structure. @@ -38,5 +37,5 @@ EXPORT_SYMBOL(init_task); * section. Since TSS's are completely CPU-local, we want them * on exact cacheline boundaries, to eliminate cacheline ping-pong. */ -DEFINE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss) = INIT_TSS; - +struct tss_struct init_tss[NR_CPUS] ____cacheline_internodealigned_in_smp = { [0 ... NR_CPUS-1] = INIT_TSS }; +EXPORT_SYMBOL(init_tss); diff --git a/arch/x86/kernel/ioport.c b/arch/x86/kernel/ioport.c index 8c96897..be66bfa 100644 --- a/arch/x86/kernel/ioport.c +++ b/arch/x86/kernel/ioport.c @@ -6,6 +6,7 @@ #include #include #include +#include #include #include #include @@ -28,6 +29,12 @@ asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on) if ((from + num <= from) || (from + num > IO_BITMAP_BITS)) return -EINVAL; +#ifdef CONFIG_GRKERNSEC_IO + if (turn_on && grsec_disable_privio) { + gr_handle_ioperm(); + return -EPERM; + } +#endif if (turn_on && !capable(CAP_SYS_RAWIO)) return -EPERM; @@ -54,7 +61,7 @@ asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on) * because the ->io_bitmap_max value must match the bitmap * contents: */ - tss = &per_cpu(init_tss, get_cpu()); + tss = init_tss + get_cpu(); if (turn_on) bitmap_clear(t->io_bitmap_ptr, from, num); @@ -102,6 +109,12 @@ long sys_iopl(unsigned int level, struct pt_regs *regs) return -EINVAL; /* Trying to gain more privileges? */ if (level > old) { +#ifdef CONFIG_GRKERNSEC_IO + if (grsec_disable_privio) { + gr_handle_iopl(); + return -EPERM; + } +#endif if (!capable(CAP_SYS_RAWIO)) return -EPERM; } diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c index 7943e0c..dd32c5c 100644 --- a/arch/x86/kernel/irq.c +++ b/arch/x86/kernel/irq.c @@ -18,7 +18,7 @@ #include #include -atomic_t irq_err_count; +atomic_unchecked_t irq_err_count; /* Function pointer for generic interrupt vector handling */ void (*x86_platform_ipi_callback)(void) = NULL; @@ -121,9 +121,9 @@ int arch_show_interrupts(struct seq_file *p, int prec) seq_printf(p, "%10u ", per_cpu(mce_poll_count, j)); seq_printf(p, " Machine check polls\n"); #endif - seq_printf(p, "%*s: %10u\n", prec, "ERR", atomic_read(&irq_err_count)); + seq_printf(p, "%*s: %10u\n", prec, "ERR", atomic_read_unchecked(&irq_err_count)); #if defined(CONFIG_X86_IO_APIC) - seq_printf(p, "%*s: %10u\n", prec, "MIS", atomic_read(&irq_mis_count)); + seq_printf(p, "%*s: %10u\n", prec, "MIS", atomic_read_unchecked(&irq_mis_count)); #endif return 0; } @@ -164,10 +164,10 @@ u64 arch_irq_stat_cpu(unsigned int cpu) u64 arch_irq_stat(void) { - u64 sum = atomic_read(&irq_err_count); + u64 sum = atomic_read_unchecked(&irq_err_count); #ifdef CONFIG_X86_IO_APIC - sum += atomic_read(&irq_mis_count); + sum += atomic_read_unchecked(&irq_mis_count); #endif return sum; } diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index 40fc861..9b8739b 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -39,7 +39,7 @@ static int check_stack_overflow(void) __asm__ __volatile__("andl %%esp,%0" : "=r" (sp) : "0" (THREAD_SIZE - 1)); - return sp < (sizeof(struct thread_info) + STACK_WARN); + return sp < STACK_WARN; } static void print_stack_overflow(void) @@ -59,8 +59,8 @@ static inline void print_stack_overflow(void) { } * per-CPU IRQ handling contexts (thread information and stack) */ union irq_ctx { - struct thread_info tinfo; - u32 stack[THREAD_SIZE/sizeof(u32)]; + unsigned long previous_esp; + u32 stack[THREAD_SIZE/sizeof(u32)]; } __attribute__((aligned(THREAD_SIZE))); static DEFINE_PER_CPU(union irq_ctx *, hardirq_ctx); @@ -80,10 +80,9 @@ static void call_on_stack(void *func, void *stack) static inline int execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) { - union irq_ctx *curctx, *irqctx; + union irq_ctx *irqctx; u32 *isp, arg1, arg2; - curctx = (union irq_ctx *) current_thread_info(); irqctx = __this_cpu_read(hardirq_ctx); /* @@ -92,21 +91,16 @@ execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) * handler) we can't do that and just have to keep using the * current stack (which is the irq stack already after all) */ - if (unlikely(curctx == irqctx)) + if (unlikely((void *)current_stack_pointer - (void *)irqctx < THREAD_SIZE)) return 0; /* build the stack frame on the IRQ stack */ - isp = (u32 *) ((char *)irqctx + sizeof(*irqctx)); - irqctx->tinfo.task = curctx->tinfo.task; - irqctx->tinfo.previous_esp = current_stack_pointer; + isp = (u32 *) ((char *)irqctx + sizeof(*irqctx) - 8); + irqctx->previous_esp = current_stack_pointer; - /* - * Copy the softirq bits in preempt_count so that the - * softirq checks work in the hardirq context. - */ - irqctx->tinfo.preempt_count = - (irqctx->tinfo.preempt_count & ~SOFTIRQ_MASK) | - (curctx->tinfo.preempt_count & SOFTIRQ_MASK); +#ifdef CONFIG_PAX_MEMORY_UDEREF + __set_fs(MAKE_MM_SEG(0)); +#endif if (unlikely(overflow)) call_on_stack(print_stack_overflow, isp); @@ -118,6 +112,11 @@ execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) : "0" (irq), "1" (desc), "2" (isp), "D" (desc->handle_irq) : "memory", "cc", "ecx"); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + __set_fs(current_thread_info()->addr_limit); +#endif + return 1; } @@ -126,29 +125,11 @@ execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) */ void __cpuinit irq_ctx_init(int cpu) { - union irq_ctx *irqctx; - if (per_cpu(hardirq_ctx, cpu)) return; - irqctx = page_address(alloc_pages_node(cpu_to_node(cpu), - THREAD_FLAGS, - THREAD_ORDER)); - memset(&irqctx->tinfo, 0, sizeof(struct thread_info)); - irqctx->tinfo.cpu = cpu; - irqctx->tinfo.preempt_count = HARDIRQ_OFFSET; - irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); - - per_cpu(hardirq_ctx, cpu) = irqctx; - - irqctx = page_address(alloc_pages_node(cpu_to_node(cpu), - THREAD_FLAGS, - THREAD_ORDER)); - memset(&irqctx->tinfo, 0, sizeof(struct thread_info)); - irqctx->tinfo.cpu = cpu; - irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); - - per_cpu(softirq_ctx, cpu) = irqctx; + per_cpu(hardirq_ctx, cpu) = page_address(alloc_pages_node(cpu_to_node(cpu), THREAD_FLAGS, THREAD_ORDER)); + per_cpu(softirq_ctx, cpu) = page_address(alloc_pages_node(cpu_to_node(cpu), THREAD_FLAGS, THREAD_ORDER)); printk(KERN_DEBUG "CPU %u irqstacks, hard=%p soft=%p\n", cpu, per_cpu(hardirq_ctx, cpu), per_cpu(softirq_ctx, cpu)); @@ -157,7 +138,6 @@ void __cpuinit irq_ctx_init(int cpu) asmlinkage void do_softirq(void) { unsigned long flags; - struct thread_info *curctx; union irq_ctx *irqctx; u32 *isp; @@ -167,15 +147,22 @@ asmlinkage void do_softirq(void) local_irq_save(flags); if (local_softirq_pending()) { - curctx = current_thread_info(); irqctx = __this_cpu_read(softirq_ctx); - irqctx->tinfo.task = curctx->task; - irqctx->tinfo.previous_esp = current_stack_pointer; + irqctx->previous_esp = current_stack_pointer; /* build the stack frame on the softirq stack */ - isp = (u32 *) ((char *)irqctx + sizeof(*irqctx)); + isp = (u32 *) ((char *)irqctx + sizeof(*irqctx) - 8); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + __set_fs(MAKE_MM_SEG(0)); +#endif call_on_stack(__do_softirq, isp); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + __set_fs(current_thread_info()->addr_limit); +#endif + /* * Shouldn't happen, we returned above if in_interrupt(): */ diff --git a/arch/x86/kernel/irq_64.c b/arch/x86/kernel/irq_64.c index d04d3ec..ea4b374 100644 --- a/arch/x86/kernel/irq_64.c +++ b/arch/x86/kernel/irq_64.c @@ -44,7 +44,7 @@ static inline void stack_overflow_check(struct pt_regs *regs) u64 estack_top, estack_bottom; u64 curbase = (u64)task_stack_page(current); - if (user_mode_vm(regs)) + if (user_mode(regs)) return; if (regs->sp >= curbase + sizeof(struct thread_info) + diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c index 90fcf62..e682cdd 100644 --- a/arch/x86/kernel/kdebugfs.c +++ b/arch/x86/kernel/kdebugfs.c @@ -28,6 +28,8 @@ struct setup_data_node { }; static ssize_t setup_data_read(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) __size_overflow(3); +static ssize_t setup_data_read(struct file *file, char __user *user_buf, size_t count, loff_t *ppos) { struct setup_data_node *node = file->private_data; diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c index 2f45c4c..d95504f 100644 --- a/arch/x86/kernel/kgdb.c +++ b/arch/x86/kernel/kgdb.c @@ -126,11 +126,11 @@ char *dbg_get_reg(int regno, void *mem, struct pt_regs *regs) #ifdef CONFIG_X86_32 switch (regno) { case GDB_SS: - if (!user_mode_vm(regs)) + if (!user_mode(regs)) *(unsigned long *)mem = __KERNEL_DS; break; case GDB_SP: - if (!user_mode_vm(regs)) + if (!user_mode(regs)) *(unsigned long *)mem = kernel_stack_pointer(regs); break; case GDB_GS: @@ -475,12 +475,12 @@ int kgdb_arch_handle_exception(int e_vector, int signo, int err_code, case 'k': /* clear the trace bit */ linux_regs->flags &= ~X86_EFLAGS_TF; - atomic_set(&kgdb_cpu_doing_single_step, -1); + atomic_set_unchecked(&kgdb_cpu_doing_single_step, -1); /* set the trace bit if we're stepping */ if (remcomInBuffer[0] == 's') { linux_regs->flags |= X86_EFLAGS_TF; - atomic_set(&kgdb_cpu_doing_single_step, + atomic_set_unchecked(&kgdb_cpu_doing_single_step, raw_smp_processor_id()); } @@ -545,7 +545,7 @@ static int __kgdb_notify(struct die_args *args, unsigned long cmd) switch (cmd) { case DIE_DEBUG: - if (atomic_read(&kgdb_cpu_doing_single_step) != -1) { + if (atomic_read_unchecked(&kgdb_cpu_doing_single_step) != -1) { if (user_mode(regs)) return single_step_cont(regs, args); break; diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c index 7da647d..56fe348 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes.c @@ -118,8 +118,11 @@ static void __kprobes __synthesize_relative_insn(void *from, void *to, u8 op) } __attribute__((packed)) *insn; insn = (struct __arch_relative_insn *)from; + + pax_open_kernel(); insn->raddr = (s32)((long)(to) - ((long)(from) + 5)); insn->op = op; + pax_close_kernel(); } /* Insert a jump instruction at address 'from', which jumps to address 'to'.*/ @@ -156,7 +159,7 @@ static int __kprobes can_boost(kprobe_opcode_t *opcodes) kprobe_opcode_t opcode; kprobe_opcode_t *orig_opcodes = opcodes; - if (search_exception_tables((unsigned long)opcodes)) + if (search_exception_tables(ktva_ktla((unsigned long)opcodes))) return 0; /* Page fault may occur on this address. */ retry: @@ -317,7 +320,9 @@ static int __kprobes __copy_instruction(u8 *dest, u8 *src, int recover) } } insn_get_length(&insn); + pax_open_kernel(); memcpy(dest, insn.kaddr, insn.length); + pax_close_kernel(); #ifdef CONFIG_X86_64 if (insn_rip_relative(&insn)) { @@ -341,7 +346,9 @@ static int __kprobes __copy_instruction(u8 *dest, u8 *src, int recover) (u8 *) dest; BUG_ON((s64) (s32) newdisp != newdisp); /* Sanity check. */ disp = (u8 *) dest + insn_offset_displacement(&insn); + pax_open_kernel(); *(s32 *) disp = (s32) newdisp; + pax_close_kernel(); } #endif return insn.length; @@ -355,12 +362,12 @@ static void __kprobes arch_copy_kprobe(struct kprobe *p) */ __copy_instruction(p->ainsn.insn, p->addr, 0); - if (can_boost(p->addr)) + if (can_boost(ktla_ktva(p->addr))) p->ainsn.boostable = 0; else p->ainsn.boostable = -1; - p->opcode = *p->addr; + p->opcode = *(ktla_ktva(p->addr)); } int __kprobes arch_prepare_kprobe(struct kprobe *p) @@ -477,7 +484,7 @@ static void __kprobes setup_singlestep(struct kprobe *p, struct pt_regs *regs, * nor set current_kprobe, because it doesn't use single * stepping. */ - regs->ip = (unsigned long)p->ainsn.insn; + regs->ip = ktva_ktla((unsigned long)p->ainsn.insn); preempt_enable_no_resched(); return; } @@ -496,7 +503,7 @@ static void __kprobes setup_singlestep(struct kprobe *p, struct pt_regs *regs, if (p->opcode == BREAKPOINT_INSTRUCTION) regs->ip = (unsigned long)p->addr; else - regs->ip = (unsigned long)p->ainsn.insn; + regs->ip = ktva_ktla((unsigned long)p->ainsn.insn); } /* @@ -575,7 +582,7 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) setup_singlestep(p, regs, kcb, 0); return 1; } - } else if (*addr != BREAKPOINT_INSTRUCTION) { + } else if (*(kprobe_opcode_t *)ktla_ktva((unsigned long)addr) != BREAKPOINT_INSTRUCTION) { /* * The breakpoint instruction was removed right * after we hit it. Another cpu has removed @@ -683,6 +690,9 @@ static void __used __kprobes kretprobe_trampoline_holder(void) " movq %rax, 152(%rsp)\n" RESTORE_REGS_STRING " popfq\n" +#ifdef KERNEXEC_PLUGIN + " btsq $63,(%rsp)\n" +#endif #else " pushf\n" SAVE_REGS_STRING @@ -820,7 +830,7 @@ static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs, struct kprobe_ctlblk *kcb) { unsigned long *tos = stack_addr(regs); - unsigned long copy_ip = (unsigned long)p->ainsn.insn; + unsigned long copy_ip = ktva_ktla((unsigned long)p->ainsn.insn); unsigned long orig_ip = (unsigned long)p->addr; kprobe_opcode_t *insn = p->ainsn.insn; @@ -1002,7 +1012,7 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self, struct die_args *args = data; int ret = NOTIFY_DONE; - if (args->regs && user_mode_vm(args->regs)) + if (args->regs && user_mode(args->regs)) return ret; switch (val) { @@ -1384,7 +1394,7 @@ int __kprobes arch_prepare_optimized_kprobe(struct optimized_kprobe *op) * Verify if the address gap is in 2GB range, because this uses * a relative jump. */ - rel = (long)op->optinsn.insn - (long)op->kp.addr + RELATIVEJUMP_SIZE; + rel = (long)op->optinsn.insn - ktla_ktva((long)op->kp.addr) + RELATIVEJUMP_SIZE; if (abs(rel) > 0x7fffffff) return -ERANGE; @@ -1405,11 +1415,11 @@ int __kprobes arch_prepare_optimized_kprobe(struct optimized_kprobe *op) synthesize_set_arg1(buf + TMPL_MOVE_IDX, (unsigned long)op); /* Set probe function call */ - synthesize_relcall(buf + TMPL_CALL_IDX, optimized_callback); + synthesize_relcall(buf + TMPL_CALL_IDX, ktla_ktva(optimized_callback)); /* Set returning jmp instruction at the tail of out-of-line buffer */ synthesize_reljump(buf + TMPL_END_IDX + op->optinsn.size, - (u8 *)op->kp.addr + op->optinsn.size); + (u8 *)ktla_ktva(op->kp.addr) + op->optinsn.size); flush_icache_range((unsigned long) buf, (unsigned long) buf + TMPL_END_IDX + @@ -1431,7 +1441,7 @@ static void __kprobes setup_optimize_kprobe(struct text_poke_param *tprm, ((long)op->kp.addr + RELATIVEJUMP_SIZE)); /* Backup instructions which will be replaced by jump address */ - memcpy(op->optinsn.copied_insn, op->kp.addr + INT3_SIZE, + memcpy(op->optinsn.copied_insn, ktla_ktva(op->kp.addr) + INT3_SIZE, RELATIVE_ADDR_SIZE); insn_buf[0] = RELATIVEJUMP_OPCODE; diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c index ea69726..a305f16 100644 --- a/arch/x86/kernel/ldt.c +++ b/arch/x86/kernel/ldt.c @@ -67,13 +67,13 @@ static int alloc_ldt(mm_context_t *pc, int mincount, int reload) if (reload) { #ifdef CONFIG_SMP preempt_disable(); - load_LDT(pc); + load_LDT_nolock(pc); if (!cpumask_equal(mm_cpumask(current->mm), cpumask_of(smp_processor_id()))) smp_call_function(flush_ldt, current->mm, 1); preempt_enable(); #else - load_LDT(pc); + load_LDT_nolock(pc); #endif } if (oldsize) { @@ -95,7 +95,7 @@ static inline int copy_ldt(mm_context_t *new, mm_context_t *old) return err; for (i = 0; i < old->size; i++) - write_ldt_entry(new->ldt, i, old->ldt + i * LDT_ENTRY_SIZE); + write_ldt_entry(new->ldt, i, old->ldt + i); return 0; } @@ -116,6 +116,24 @@ int init_new_context(struct task_struct *tsk, struct mm_struct *mm) retval = copy_ldt(&mm->context, &old_mm->context); mutex_unlock(&old_mm->context.lock); } + + if (tsk == current) { + mm->context.vdso = 0; + +#ifdef CONFIG_X86_32 +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) + mm->context.user_cs_base = 0UL; + mm->context.user_cs_limit = ~0UL; + +#if defined(CONFIG_PAX_PAGEEXEC) && defined(CONFIG_SMP) + cpus_clear(mm->context.cpu_user_cs_mask); +#endif + +#endif +#endif + + } + return retval; } @@ -141,6 +159,7 @@ void destroy_context(struct mm_struct *mm) } } +static int read_ldt(void __user *ptr, unsigned long bytecount) __size_overflow(2); static int read_ldt(void __user *ptr, unsigned long bytecount) { int err; @@ -175,6 +194,7 @@ error_return: return err; } +static int read_default_ldt(void __user *ptr, unsigned long bytecount) __size_overflow(2); static int read_default_ldt(void __user *ptr, unsigned long bytecount) { /* CHECKME: Can we use _one_ random number ? */ @@ -230,6 +250,13 @@ static int write_ldt(void __user *ptr, unsigned long bytecount, int oldmode) } } +#ifdef CONFIG_PAX_SEGMEXEC + if ((mm->pax_flags & MF_PAX_SEGMEXEC) && (ldt_info.contents & MODIFY_LDT_CONTENTS_CODE)) { + error = -EINVAL; + goto out_unlock; + } +#endif + fill_ldt(&ldt, &ldt_info); if (oldmode) ldt.avl = 0; diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index a3fa43b..8966f4c 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -27,7 +27,7 @@ #include #include -static void set_idt(void *newidt, __u16 limit) +static void set_idt(struct desc_struct *newidt, __u16 limit) { struct desc_ptr curidt; @@ -39,7 +39,7 @@ static void set_idt(void *newidt, __u16 limit) } -static void set_gdt(void *newgdt, __u16 limit) +static void set_gdt(struct desc_struct *newgdt, __u16 limit) { struct desc_ptr curgdt; @@ -217,7 +217,7 @@ void machine_kexec(struct kimage *image) } control_page = page_address(image->control_code_page); - memcpy(control_page, relocate_kernel, KEXEC_CONTROL_CODE_MAX_SIZE); + memcpy(control_page, (void *)ktla_ktva((unsigned long)relocate_kernel), KEXEC_CONTROL_CODE_MAX_SIZE); relocate_kernel_ptr = control_page; page_list[PA_CONTROL_PAGE] = __pa(control_page); diff --git a/arch/x86/kernel/microcode_intel.c b/arch/x86/kernel/microcode_intel.c index 3ca42d0..79d24cd 100644 --- a/arch/x86/kernel/microcode_intel.c +++ b/arch/x86/kernel/microcode_intel.c @@ -434,15 +434,16 @@ static enum ucode_state request_microcode_fw(int cpu, struct device *device) return ret; } +static int get_ucode_user(void *to, const void *from, size_t n) __size_overflow(3); static int get_ucode_user(void *to, const void *from, size_t n) { - return copy_from_user(to, from, n); + return copy_from_user(to, (const void __force_user *)from, n); } static enum ucode_state request_microcode_user(int cpu, const void __user *buf, size_t size) { - return generic_load_microcode(cpu, (void *)buf, size, &get_ucode_user); + return generic_load_microcode(cpu, (__force_kernel void *)buf, size, &get_ucode_user); } static void microcode_fini_cpu(int cpu) diff --git a/arch/x86/kernel/module.c b/arch/x86/kernel/module.c index 925179f..1f0d561 100644 --- a/arch/x86/kernel/module.c +++ b/arch/x86/kernel/module.c @@ -36,15 +36,61 @@ #define DEBUGP(fmt...) #endif -void *module_alloc(unsigned long size) +static inline void *__module_alloc(unsigned long size, pgprot_t prot) __size_overflow(1); +static inline void *__module_alloc(unsigned long size, pgprot_t prot) { - if (PAGE_ALIGN(size) > MODULES_LEN) + if (size == 0 || PAGE_ALIGN(size) > MODULES_LEN) return NULL; return __vmalloc_node_range(size, 1, MODULES_VADDR, MODULES_END, - GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC, + GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO, prot, -1, __builtin_return_address(0)); } +void *module_alloc(unsigned long size) +{ + +#ifdef CONFIG_PAX_KERNEXEC + return __module_alloc(size, PAGE_KERNEL); +#else + return __module_alloc(size, PAGE_KERNEL_EXEC); +#endif + +} + +#ifdef CONFIG_PAX_KERNEXEC +#ifdef CONFIG_X86_32 +void *module_alloc_exec(unsigned long size) +{ + struct vm_struct *area; + + if (size == 0) + return NULL; + + area = __get_vm_area(size, VM_ALLOC, (unsigned long)&MODULES_EXEC_VADDR, (unsigned long)&MODULES_EXEC_END); + return area ? area->addr : NULL; +} +EXPORT_SYMBOL(module_alloc_exec); + +void module_free_exec(struct module *mod, void *module_region) +{ + vunmap(module_region); +} +EXPORT_SYMBOL(module_free_exec); +#else +void module_free_exec(struct module *mod, void *module_region) +{ + module_free(mod, module_region); +} +EXPORT_SYMBOL(module_free_exec); + +void *module_alloc_exec(unsigned long size) +{ + return __module_alloc(size, PAGE_KERNEL_RX); +} +EXPORT_SYMBOL(module_alloc_exec); +#endif +#endif + #ifdef CONFIG_X86_32 int apply_relocate(Elf32_Shdr *sechdrs, const char *strtab, @@ -55,14 +101,16 @@ int apply_relocate(Elf32_Shdr *sechdrs, unsigned int i; Elf32_Rel *rel = (void *)sechdrs[relsec].sh_addr; Elf32_Sym *sym; - uint32_t *location; + uint32_t *plocation, location; DEBUGP("Applying relocate section %u to %u\n", relsec, sechdrs[relsec].sh_info); for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rel); i++) { /* This is where to make the change */ - location = (void *)sechdrs[sechdrs[relsec].sh_info].sh_addr - + rel[i].r_offset; + plocation = (void *)sechdrs[sechdrs[relsec].sh_info].sh_addr + rel[i].r_offset; + location = (uint32_t)plocation; + if (sechdrs[sechdrs[relsec].sh_info].sh_flags & SHF_EXECINSTR) + plocation = ktla_ktva((void *)plocation); /* This is the symbol it is referring to. Note that all undefined symbols have been resolved. */ sym = (Elf32_Sym *)sechdrs[symindex].sh_addr @@ -71,11 +119,15 @@ int apply_relocate(Elf32_Shdr *sechdrs, switch (ELF32_R_TYPE(rel[i].r_info)) { case R_386_32: /* We add the value into the location given */ - *location += sym->st_value; + pax_open_kernel(); + *plocation += sym->st_value; + pax_close_kernel(); break; case R_386_PC32: /* Add the value, subtract its postition */ - *location += sym->st_value - (uint32_t)location; + pax_open_kernel(); + *plocation += sym->st_value - location; + pax_close_kernel(); break; default: printk(KERN_ERR "module %s: Unknown relocation: %u\n", @@ -120,21 +172,30 @@ int apply_relocate_add(Elf64_Shdr *sechdrs, case R_X86_64_NONE: break; case R_X86_64_64: + pax_open_kernel(); *(u64 *)loc = val; + pax_close_kernel(); break; case R_X86_64_32: + pax_open_kernel(); *(u32 *)loc = val; + pax_close_kernel(); if (val != *(u32 *)loc) goto overflow; break; case R_X86_64_32S: + pax_open_kernel(); *(s32 *)loc = val; + pax_close_kernel(); if ((s64)val != *(s32 *)loc) goto overflow; break; case R_X86_64_PC32: val -= (u64)loc; + pax_open_kernel(); *(u32 *)loc = val; + pax_close_kernel(); + #if 0 if ((s64)val != *(s32 *)loc) goto overflow; diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c index 47acaf3..ec48ab6 100644 --- a/arch/x86/kernel/nmi.c +++ b/arch/x86/kernel/nmi.c @@ -505,6 +505,17 @@ static inline void nmi_nesting_postprocess(void) dotraplinkage notrace __kprobes void do_nmi(struct pt_regs *regs, long error_code) { + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + if (!user_mode(regs)) { + unsigned long cs = regs->cs & 0xFFFF; + unsigned long ip = ktva_ktla(regs->ip); + + if ((cs == __KERNEL_CS || cs == __KERNEXEC_KERNEL_CS) && ip <= (unsigned long)_etext) + regs->ip = ip; + } +#endif + nmi_nesting_preprocess(regs); nmi_enter(); diff --git a/arch/x86/kernel/paravirt-spinlocks.c b/arch/x86/kernel/paravirt-spinlocks.c index 676b8c7..870ba04 100644 --- a/arch/x86/kernel/paravirt-spinlocks.c +++ b/arch/x86/kernel/paravirt-spinlocks.c @@ -13,7 +13,7 @@ default_spin_lock_flags(arch_spinlock_t *lock, unsigned long flags) arch_spin_lock(lock); } -struct pv_lock_ops pv_lock_ops = { +struct pv_lock_ops pv_lock_ops __read_only = { #ifdef CONFIG_SMP .spin_is_locked = __ticket_spin_is_locked, .spin_is_contended = __ticket_spin_is_contended, diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index d90272e..6bb013b 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -53,6 +53,9 @@ u64 _paravirt_ident_64(u64 x) { return x; } +#if defined(CONFIG_X86_32) && defined(CONFIG_X86_PAE) +PV_CALLEE_SAVE_REGS_THUNK(_paravirt_ident_64); +#endif void __init default_banner(void) { @@ -145,15 +148,19 @@ unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf, if (opfunc == NULL) /* If there's no function, patch it with a ud2a (BUG) */ ret = paravirt_patch_insns(insnbuf, len, ud2a, ud2a+sizeof(ud2a)); - else if (opfunc == _paravirt_nop) + else if (opfunc == (void *)_paravirt_nop) /* If the operation is a nop, then nop the callsite */ ret = paravirt_patch_nop(); /* identity functions just return their single argument */ - else if (opfunc == _paravirt_ident_32) + else if (opfunc == (void *)_paravirt_ident_32) ret = paravirt_patch_ident_32(insnbuf, len); - else if (opfunc == _paravirt_ident_64) + else if (opfunc == (void *)_paravirt_ident_64) ret = paravirt_patch_ident_64(insnbuf, len); +#if defined(CONFIG_X86_32) && defined(CONFIG_X86_PAE) + else if (opfunc == (void *)__raw_callee_save__paravirt_ident_64) + ret = paravirt_patch_ident_64(insnbuf, len); +#endif else if (type == PARAVIRT_PATCH(pv_cpu_ops.iret) || type == PARAVIRT_PATCH(pv_cpu_ops.irq_enable_sysexit) || @@ -178,7 +185,7 @@ unsigned paravirt_patch_insns(void *insnbuf, unsigned len, if (insn_len > len || start == NULL) insn_len = len; else - memcpy(insnbuf, start, insn_len); + memcpy(insnbuf, ktla_ktva(start), insn_len); return insn_len; } @@ -302,7 +309,7 @@ void arch_flush_lazy_mmu_mode(void) preempt_enable(); } -struct pv_info pv_info = { +struct pv_info pv_info __read_only = { .name = "bare hardware", .paravirt_enabled = 0, .kernel_rpl = 0, @@ -313,16 +320,16 @@ struct pv_info pv_info = { #endif }; -struct pv_init_ops pv_init_ops = { +struct pv_init_ops pv_init_ops __read_only = { .patch = native_patch, }; -struct pv_time_ops pv_time_ops = { +struct pv_time_ops pv_time_ops __read_only = { .sched_clock = native_sched_clock, .steal_clock = native_steal_clock, }; -struct pv_irq_ops pv_irq_ops = { +struct pv_irq_ops pv_irq_ops __read_only = { .save_fl = __PV_IS_CALLEE_SAVE(native_save_fl), .restore_fl = __PV_IS_CALLEE_SAVE(native_restore_fl), .irq_disable = __PV_IS_CALLEE_SAVE(native_irq_disable), @@ -334,7 +341,7 @@ struct pv_irq_ops pv_irq_ops = { #endif }; -struct pv_cpu_ops pv_cpu_ops = { +struct pv_cpu_ops pv_cpu_ops __read_only = { .cpuid = native_cpuid, .get_debugreg = native_get_debugreg, .set_debugreg = native_set_debugreg, @@ -395,21 +402,26 @@ struct pv_cpu_ops pv_cpu_ops = { .end_context_switch = paravirt_nop, }; -struct pv_apic_ops pv_apic_ops = { +struct pv_apic_ops pv_apic_ops __read_only = { #ifdef CONFIG_X86_LOCAL_APIC .startup_ipi_hook = paravirt_nop, #endif }; -#if defined(CONFIG_X86_32) && !defined(CONFIG_X86_PAE) +#ifdef CONFIG_X86_32 +#ifdef CONFIG_X86_PAE +/* 64-bit pagetable entries */ +#define PTE_IDENT PV_CALLEE_SAVE(_paravirt_ident_64) +#else /* 32-bit pagetable entries */ #define PTE_IDENT __PV_IS_CALLEE_SAVE(_paravirt_ident_32) +#endif #else /* 64-bit pagetable entries */ #define PTE_IDENT __PV_IS_CALLEE_SAVE(_paravirt_ident_64) #endif -struct pv_mmu_ops pv_mmu_ops = { +struct pv_mmu_ops pv_mmu_ops __read_only = { .read_cr2 = native_read_cr2, .write_cr2 = native_write_cr2, @@ -459,6 +471,7 @@ struct pv_mmu_ops pv_mmu_ops = { .make_pud = PTE_IDENT, .set_pgd = native_set_pgd, + .set_pgd_batched = native_set_pgd_batched, #endif #endif /* PAGETABLE_LEVELS >= 3 */ @@ -478,6 +491,12 @@ struct pv_mmu_ops pv_mmu_ops = { }, .set_fixmap = native_set_fixmap, + +#ifdef CONFIG_PAX_KERNEXEC + .pax_open_kernel = native_pax_open_kernel, + .pax_close_kernel = native_pax_close_kernel, +#endif + }; EXPORT_SYMBOL_GPL(pv_time_ops); diff --git a/arch/x86/kernel/pci-iommu_table.c b/arch/x86/kernel/pci-iommu_table.c index 35ccf75..7a15747 100644 --- a/arch/x86/kernel/pci-iommu_table.c +++ b/arch/x86/kernel/pci-iommu_table.c @@ -2,7 +2,7 @@ #include #include #include - +#include #define DEBUG 1 diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 15763af..da59ada 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -48,16 +48,33 @@ void free_thread_xstate(struct task_struct *tsk) void free_thread_info(struct thread_info *ti) { - free_thread_xstate(ti->task); free_pages((unsigned long)ti, THREAD_ORDER); } +static struct kmem_cache *task_struct_cachep; + void arch_task_cache_init(void) { - task_xstate_cachep = - kmem_cache_create("task_xstate", xstate_size, + /* create a slab on which task_structs can be allocated */ + task_struct_cachep = + kmem_cache_create("task_struct", sizeof(struct task_struct), + ARCH_MIN_TASKALIGN, SLAB_PANIC | SLAB_NOTRACK, NULL); + + task_xstate_cachep = + kmem_cache_create("task_xstate", xstate_size, __alignof__(union thread_xstate), - SLAB_PANIC | SLAB_NOTRACK, NULL); + SLAB_PANIC | SLAB_NOTRACK | SLAB_USERCOPY, NULL); +} + +struct task_struct *alloc_task_struct_node(int node) +{ + return kmem_cache_alloc_node(task_struct_cachep, GFP_KERNEL, node); +} + +void free_task_struct(struct task_struct *task) +{ + free_thread_xstate(task); + kmem_cache_free(task_struct_cachep, task); } /* @@ -70,7 +87,7 @@ void exit_thread(void) unsigned long *bp = t->io_bitmap_ptr; if (bp) { - struct tss_struct *tss = &per_cpu(init_tss, get_cpu()); + struct tss_struct *tss = init_tss + get_cpu(); t->io_bitmap_ptr = NULL; clear_thread_flag(TIF_IO_BITMAP); @@ -106,7 +123,7 @@ void show_regs_common(void) printk(KERN_CONT "\n"); printk(KERN_DEFAULT "Pid: %d, comm: %.20s %s %s %.*s", - current->pid, current->comm, print_tainted(), + task_pid_nr(current), current->comm, print_tainted(), init_utsname()->release, (int)strcspn(init_utsname()->version, " "), init_utsname()->version); @@ -120,6 +137,9 @@ void flush_thread(void) { struct task_struct *tsk = current; +#if defined(CONFIG_X86_32) && !defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_PAX_MEMORY_UDEREF) + loadsegment(gs, 0); +#endif flush_ptrace_hw_breakpoint(tsk); memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array)); /* @@ -282,10 +302,10 @@ int kernel_thread(int (*fn)(void *), void *arg, unsigned long flags) regs.di = (unsigned long) arg; #ifdef CONFIG_X86_32 - regs.ds = __USER_DS; - regs.es = __USER_DS; + regs.ds = __KERNEL_DS; + regs.es = __KERNEL_DS; regs.fs = __KERNEL_PERCPU; - regs.gs = __KERNEL_STACK_CANARY; + savesegment(gs, regs.gs); #else regs.ss = __KERNEL_DS; #endif @@ -411,7 +431,7 @@ bool set_pm_idle_to_default(void) return ret; } -void stop_this_cpu(void *dummy) +__noreturn void stop_this_cpu(void *dummy) { local_irq_disable(); /* @@ -653,16 +673,37 @@ static int __init idle_setup(char *str) } early_param("idle", idle_setup); -unsigned long arch_align_stack(unsigned long sp) +#ifdef CONFIG_PAX_RANDKSTACK +void pax_randomize_kstack(struct pt_regs *regs) { - if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) - sp -= get_random_int() % 8192; - return sp & ~0xf; -} + struct thread_struct *thread = ¤t->thread; + unsigned long time; -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long range_end = mm->brk + 0x02000000; - return randomize_range(mm->brk, range_end, 0) ? : mm->brk; -} + if (!randomize_va_space) + return; + + if (v8086_mode(regs)) + return; + rdtscl(time); + + /* P4 seems to return a 0 LSB, ignore it */ +#ifdef CONFIG_MPENTIUM4 + time &= 0x3EUL; + time <<= 2; +#elif defined(CONFIG_X86_64) + time &= 0xFUL; + time <<= 4; +#else + time &= 0x1FUL; + time <<= 3; +#endif + + thread->sp0 ^= time; + load_sp0(init_tss + smp_processor_id(), thread); + +#ifdef CONFIG_X86_64 + percpu_write(kernel_stack, thread->sp0); +#endif +} +#endif diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index c08d1ff..6ae1c81 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -67,6 +67,7 @@ asmlinkage void ret_from_fork(void) __asm__("ret_from_fork"); unsigned long thread_saved_pc(struct task_struct *tsk) { return ((unsigned long *)tsk->thread.sp)[3]; +//XXX return tsk->thread.eip; } #ifndef CONFIG_SMP @@ -132,15 +133,14 @@ void __show_regs(struct pt_regs *regs, int all) unsigned long sp; unsigned short ss, gs; - if (user_mode_vm(regs)) { + if (user_mode(regs)) { sp = regs->sp; ss = regs->ss & 0xffff; - gs = get_user_gs(regs); } else { sp = kernel_stack_pointer(regs); savesegment(ss, ss); - savesegment(gs, gs); } + gs = get_user_gs(regs); show_regs_common(); @@ -202,13 +202,14 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, struct task_struct *tsk; int err; - childregs = task_pt_regs(p); + childregs = task_stack_page(p) + THREAD_SIZE - sizeof(struct pt_regs) - 8; *childregs = *regs; childregs->ax = 0; childregs->sp = sp; p->thread.sp = (unsigned long) childregs; p->thread.sp0 = (unsigned long) (childregs+1); + p->tinfo.lowest_stack = (unsigned long)task_stack_page(p); p->thread.ip = (unsigned long) ret_from_fork; @@ -299,7 +300,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) struct thread_struct *prev = &prev_p->thread, *next = &next_p->thread; int cpu = smp_processor_id(); - struct tss_struct *tss = &per_cpu(init_tss, cpu); + struct tss_struct *tss = init_tss + cpu; fpu_switch_t fpu; /* never put a printk in __switch_to... printk() calls wake_up*() indirectly */ @@ -323,6 +324,10 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) */ lazy_save_gs(prev->gs); +#ifdef CONFIG_PAX_MEMORY_UDEREF + __set_fs(task_thread_info(next_p)->addr_limit); +#endif + /* * Load the per-thread Thread-Local Storage descriptor. */ @@ -353,6 +358,9 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) */ arch_end_context_switch(next_p); + percpu_write(current_task, next_p); + percpu_write(current_tinfo, &next_p->tinfo); + /* * Restore %gs if needed (which is common) */ @@ -361,8 +369,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) switch_fpu_finish(next_p, fpu); - percpu_write(current_task, next_p); - return prev_p; } @@ -392,4 +398,3 @@ unsigned long get_wchan(struct task_struct *p) } while (count++ < 16); return 0; } - diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index cfa5c90..4facd28 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -89,7 +89,7 @@ static void __exit_idle(void) void exit_idle(void) { /* idle loop has pid 0 */ - if (current->pid) + if (task_pid_nr(current)) return; __exit_idle(); } @@ -270,8 +270,7 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, struct pt_regs *childregs; struct task_struct *me = current; - childregs = ((struct pt_regs *) - (THREAD_SIZE + task_stack_page(p))) - 1; + childregs = task_stack_page(p) + THREAD_SIZE - sizeof(struct pt_regs) - 16; *childregs = *regs; childregs->ax = 0; @@ -283,6 +282,7 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, p->thread.sp = (unsigned long) childregs; p->thread.sp0 = (unsigned long) (childregs+1); p->thread.usersp = me->thread.usersp; + p->tinfo.lowest_stack = (unsigned long)task_stack_page(p); set_tsk_thread_flag(p, TIF_FORK); @@ -385,7 +385,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) struct thread_struct *prev = &prev_p->thread; struct thread_struct *next = &next_p->thread; int cpu = smp_processor_id(); - struct tss_struct *tss = &per_cpu(init_tss, cpu); + struct tss_struct *tss = init_tss + cpu; unsigned fsindex, gsindex; fpu_switch_t fpu; @@ -467,10 +467,9 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) prev->usersp = percpu_read(old_rsp); percpu_write(old_rsp, next->usersp); percpu_write(current_task, next_p); + percpu_write(current_tinfo, &next_p->tinfo); - percpu_write(kernel_stack, - (unsigned long)task_stack_page(next_p) + - THREAD_SIZE - KERNEL_STACK_OFFSET); + percpu_write(kernel_stack, next->sp0); /* * Now maybe reload the debug registers and handle I/O bitmaps @@ -525,12 +524,11 @@ unsigned long get_wchan(struct task_struct *p) if (!p || p == current || p->state == TASK_RUNNING) return 0; stack = (unsigned long)task_stack_page(p); - if (p->thread.sp < stack || p->thread.sp >= stack+THREAD_SIZE) + if (p->thread.sp < stack || p->thread.sp > stack+THREAD_SIZE-16-sizeof(u64)) return 0; fp = *(u64 *)(p->thread.sp); do { - if (fp < (unsigned long)stack || - fp >= (unsigned long)stack+THREAD_SIZE) + if (fp < stack || fp > stack+THREAD_SIZE-16-sizeof(u64)) return 0; ip = *(u64 *)(fp+8); if (!in_sched_functions(ip)) diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 5026738..574f70a 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -792,6 +792,10 @@ static int ioperm_active(struct task_struct *target, static int ioperm_get(struct task_struct *target, const struct user_regset *regset, unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) __size_overflow(3,4); +static int ioperm_get(struct task_struct *target, + const struct user_regset *regset, + unsigned int pos, unsigned int count, void *kbuf, void __user *ubuf) { if (!target->thread.io_bitmap_ptr) @@ -823,7 +827,7 @@ long arch_ptrace(struct task_struct *child, long request, unsigned long addr, unsigned long data) { int ret; - unsigned long __user *datap = (unsigned long __user *)data; + unsigned long __user *datap = (__force unsigned long __user *)data; switch (request) { /* read the word at location addr in the USER area. */ @@ -908,14 +912,14 @@ long arch_ptrace(struct task_struct *child, long request, if ((int) addr < 0) return -EIO; ret = do_get_thread_area(child, addr, - (struct user_desc __user *)data); + (__force struct user_desc __user *) data); break; case PTRACE_SET_THREAD_AREA: if ((int) addr < 0) return -EIO; ret = do_set_thread_area(child, addr, - (struct user_desc __user *)data, 0); + (__force struct user_desc __user *) data, 0); break; #endif @@ -1332,7 +1336,7 @@ static void fill_sigtrap_info(struct task_struct *tsk, memset(info, 0, sizeof(*info)); info->si_signo = SIGTRAP; info->si_code = si_code; - info->si_addr = user_mode_vm(regs) ? (void __user *)regs->ip : NULL; + info->si_addr = user_mode(regs) ? (__force void __user *)regs->ip : NULL; } void user_single_step_siginfo(struct task_struct *tsk, @@ -1361,6 +1365,10 @@ void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, # define IS_IA32 0 #endif +#ifdef CONFIG_GRKERNSEC_SETXID +extern void gr_delayed_cred_worker(void); +#endif + /* * We must return the syscall number to actually look up in the table. * This can be -1L to skip running any syscall at all. @@ -1369,6 +1377,11 @@ long syscall_trace_enter(struct pt_regs *regs) { long ret = 0; +#ifdef CONFIG_GRKERNSEC_SETXID + if (unlikely(test_and_clear_thread_flag(TIF_GRSEC_SETXID))) + gr_delayed_cred_worker(); +#endif + /* * If we stepped into a sysenter/syscall insn, it trapped in * kernel mode; do_debug() cleared TF and set TIF_SINGLESTEP. @@ -1412,6 +1425,11 @@ void syscall_trace_leave(struct pt_regs *regs) { bool step; +#ifdef CONFIG_GRKERNSEC_SETXID + if (unlikely(test_and_clear_thread_flag(TIF_GRSEC_SETXID))) + gr_delayed_cred_worker(); +#endif + audit_syscall_exit(regs); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c index 42eb330..139955c 100644 --- a/arch/x86/kernel/pvclock.c +++ b/arch/x86/kernel/pvclock.c @@ -81,11 +81,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src) return pv_tsc_khz; } -static atomic64_t last_value = ATOMIC64_INIT(0); +static atomic64_unchecked_t last_value = ATOMIC64_INIT(0); void pvclock_resume(void) { - atomic64_set(&last_value, 0); + atomic64_set_unchecked(&last_value, 0); } cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src) @@ -121,11 +121,11 @@ cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src) * updating at the same time, and one of them could be slightly behind, * making the assumption that last_value always go forward fail to hold. */ - last = atomic64_read(&last_value); + last = atomic64_read_unchecked(&last_value); do { if (ret < last) return last; - last = atomic64_cmpxchg(&last_value, last, ret); + last = atomic64_cmpxchg_unchecked(&last_value, last, ret); } while (unlikely(last != ret)); return ret; diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index d840e69..98e9581 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -35,7 +35,7 @@ void (*pm_power_off)(void); EXPORT_SYMBOL(pm_power_off); static const struct desc_ptr no_idt = {}; -static int reboot_mode; +static unsigned short reboot_mode; enum reboot_type reboot_type = BOOT_ACPI; int reboot_force; @@ -335,13 +335,17 @@ core_initcall(reboot_init); extern const unsigned char machine_real_restart_asm[]; extern const u64 machine_real_restart_gdt[3]; -void machine_real_restart(unsigned int type) +__noreturn void machine_real_restart(unsigned int type) { void *restart_va; unsigned long restart_pa; - void (*restart_lowmem)(unsigned int); + void (* __noreturn restart_lowmem)(unsigned int); u64 *lowmem_gdt; +#if defined(CONFIG_X86_32) && (defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF)) + struct desc_struct *gdt; +#endif + local_irq_disable(); /* Write zero to CMOS register number 0x0f, which the BIOS POST @@ -367,14 +371,14 @@ void machine_real_restart(unsigned int type) boot)". This seems like a fairly standard thing that gets set by REBOOT.COM programs, and the previous reset routine did this too. */ - *((unsigned short *)0x472) = reboot_mode; + *(unsigned short *)(__va(0x472)) = reboot_mode; /* Patch the GDT in the low memory trampoline */ lowmem_gdt = TRAMPOLINE_SYM(machine_real_restart_gdt); restart_va = TRAMPOLINE_SYM(machine_real_restart_asm); restart_pa = virt_to_phys(restart_va); - restart_lowmem = (void (*)(unsigned int))restart_pa; + restart_lowmem = (void *)restart_pa; /* GDT[0]: GDT self-pointer */ lowmem_gdt[0] = @@ -385,7 +389,33 @@ void machine_real_restart(unsigned int type) GDT_ENTRY(0x009b, restart_pa, 0xffff); /* Jump to the identity-mapped low memory code */ + +#if defined(CONFIG_X86_32) && (defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF)) + gdt = get_cpu_gdt_table(smp_processor_id()); + pax_open_kernel(); +#ifdef CONFIG_PAX_MEMORY_UDEREF + gdt[GDT_ENTRY_KERNEL_DS].type = 3; + gdt[GDT_ENTRY_KERNEL_DS].limit = 0xf; + asm("mov %0, %%ds; mov %0, %%es; mov %0, %%ss" : : "r" (__KERNEL_DS) : "memory"); +#endif +#ifdef CONFIG_PAX_KERNEXEC + gdt[GDT_ENTRY_KERNEL_CS].base0 = 0; + gdt[GDT_ENTRY_KERNEL_CS].base1 = 0; + gdt[GDT_ENTRY_KERNEL_CS].base2 = 0; + gdt[GDT_ENTRY_KERNEL_CS].limit0 = 0xffff; + gdt[GDT_ENTRY_KERNEL_CS].limit = 0xf; + gdt[GDT_ENTRY_KERNEL_CS].g = 1; +#endif + pax_close_kernel(); +#endif + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + asm volatile("push %0; push %1; lret\n" : : "i" (__KERNEL_CS), "rm" (restart_lowmem), "a" (type)); + unreachable(); +#else restart_lowmem(type); +#endif + } #ifdef CONFIG_APM_MODULE EXPORT_SYMBOL(machine_real_restart); @@ -556,7 +586,7 @@ void __attribute__((weak)) mach_reboot_fixups(void) * try to force a triple fault and then cycle between hitting the keyboard * controller and doing that */ -static void native_machine_emergency_restart(void) +__noreturn static void native_machine_emergency_restart(void) { int i; int attempt = 0; @@ -680,13 +710,13 @@ void native_machine_shutdown(void) #endif } -static void __machine_emergency_restart(int emergency) +static __noreturn void __machine_emergency_restart(int emergency) { reboot_emergency = emergency; machine_ops.emergency_restart(); } -static void native_machine_restart(char *__unused) +static __noreturn void native_machine_restart(char *__unused) { printk("machine restart\n"); @@ -695,7 +725,7 @@ static void native_machine_restart(char *__unused) __machine_emergency_restart(0); } -static void native_machine_halt(void) +static __noreturn void native_machine_halt(void) { /* stop other cpus and apics */ machine_shutdown(); @@ -706,7 +736,7 @@ static void native_machine_halt(void) stop_this_cpu(NULL); } -static void native_machine_power_off(void) +__noreturn static void native_machine_power_off(void) { if (pm_power_off) { if (!reboot_force) @@ -715,6 +745,7 @@ static void native_machine_power_off(void) } /* a fallback in case there is no PM info available */ tboot_shutdown(TB_SHUTDOWN_HALT); + unreachable(); } struct machine_ops machine_ops = { diff --git a/arch/x86/kernel/relocate_kernel_64.S b/arch/x86/kernel/relocate_kernel_64.S index 7a6f3b3..bed145d7 100644 --- a/arch/x86/kernel/relocate_kernel_64.S +++ b/arch/x86/kernel/relocate_kernel_64.S @@ -11,6 +11,7 @@ #include #include #include +#include /* * Must be relocatable PIC code callable as a C function @@ -160,13 +161,14 @@ identity_mapped: xorq %rbp, %rbp xorq %r8, %r8 xorq %r9, %r9 - xorq %r10, %r9 + xorq %r10, %r10 xorq %r11, %r11 xorq %r12, %r12 xorq %r13, %r13 xorq %r14, %r14 xorq %r15, %r15 + pax_force_retaddr 0, 1 ret 1: diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index d7d5099..28555d0 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -448,7 +448,7 @@ static void __init parse_setup_data(void) switch (data->type) { case SETUP_E820_EXT: - parse_e820_ext(data); + parse_e820_ext((struct setup_data __force_kernel *)data); break; case SETUP_DTB: add_dtb(pa_data); @@ -649,7 +649,7 @@ static void __init trim_bios_range(void) * area (640->1Mb) as ram even though it is not. * take them out. */ - e820_remove_range(BIOS_BEGIN, BIOS_END - BIOS_BEGIN, E820_RAM, 1); + e820_remove_range(ISA_START_ADDRESS, ISA_END_ADDRESS - ISA_START_ADDRESS, E820_RAM, 1); sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); } @@ -767,14 +767,14 @@ void __init setup_arch(char **cmdline_p) if (!boot_params.hdr.root_flags) root_mountflags &= ~MS_RDONLY; - init_mm.start_code = (unsigned long) _text; - init_mm.end_code = (unsigned long) _etext; + init_mm.start_code = ktla_ktva((unsigned long) _text); + init_mm.end_code = ktla_ktva((unsigned long) _etext); init_mm.end_data = (unsigned long) _edata; init_mm.brk = _brk_end; - code_resource.start = virt_to_phys(_text); - code_resource.end = virt_to_phys(_etext)-1; - data_resource.start = virt_to_phys(_etext); + code_resource.start = virt_to_phys(ktla_ktva(_text)); + code_resource.end = virt_to_phys(ktla_ktva(_etext))-1; + data_resource.start = virt_to_phys(_sdata); data_resource.end = virt_to_phys(_edata)-1; bss_resource.start = virt_to_phys(&__bss_start); bss_resource.end = virt_to_phys(&__bss_stop)-1; diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 5a98aa2..848d2be 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -21,19 +21,17 @@ #include #include -DEFINE_PER_CPU(int, cpu_number); +#ifdef CONFIG_SMP +DEFINE_PER_CPU(unsigned int, cpu_number); EXPORT_PER_CPU_SYMBOL(cpu_number); +#endif -#ifdef CONFIG_X86_64 #define BOOT_PERCPU_OFFSET ((unsigned long)__per_cpu_load) -#else -#define BOOT_PERCPU_OFFSET 0 -#endif DEFINE_PER_CPU(unsigned long, this_cpu_off) = BOOT_PERCPU_OFFSET; EXPORT_PER_CPU_SYMBOL(this_cpu_off); -unsigned long __per_cpu_offset[NR_CPUS] __read_mostly = { +unsigned long __per_cpu_offset[NR_CPUS] __read_only = { [0 ... NR_CPUS-1] = BOOT_PERCPU_OFFSET, }; EXPORT_SYMBOL(__per_cpu_offset); @@ -96,6 +94,8 @@ static bool __init pcpu_need_numa(void) * Pointer to the allocated area on success, NULL on failure. */ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, + unsigned long align) __size_overflow(2); +static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, unsigned long align) { const unsigned long goal = __pa(MAX_DMA_ADDRESS); @@ -124,6 +124,8 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, /* * Helpers for first chunk memory allocation */ +static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, size_t align) __size_overflow(2); + static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, size_t align) { return pcpu_alloc_bootmem(cpu, size, align); @@ -155,10 +157,10 @@ static inline void setup_percpu_segment(int cpu) { #ifdef CONFIG_X86_32 struct desc_struct gdt; + unsigned long base = per_cpu_offset(cpu); - pack_descriptor(&gdt, per_cpu_offset(cpu), 0xFFFFF, - 0x2 | DESCTYPE_S, 0x8); - gdt.s = 1; + pack_descriptor(&gdt, base, (VMALLOC_END - base - 1) >> PAGE_SHIFT, + 0x83 | DESCTYPE_S, 0xC); write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_PERCPU, &gdt, DESCTYPE_S); #endif @@ -219,6 +221,11 @@ void __init setup_per_cpu_areas(void) /* alrighty, percpu areas up and running */ delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; for_each_possible_cpu(cpu) { +#ifdef CONFIG_CC_STACKPROTECTOR +#ifdef CONFIG_X86_32 + unsigned long canary = per_cpu(stack_canary.canary, cpu); +#endif +#endif per_cpu_offset(cpu) = delta + pcpu_unit_offsets[cpu]; per_cpu(this_cpu_off, cpu) = per_cpu_offset(cpu); per_cpu(cpu_number, cpu) = cpu; @@ -259,6 +266,12 @@ void __init setup_per_cpu_areas(void) */ set_cpu_numa_node(cpu, early_cpu_to_node(cpu)); #endif +#ifdef CONFIG_CC_STACKPROTECTOR +#ifdef CONFIG_X86_32 + if (!cpu) + per_cpu(stack_canary.canary, cpu) = canary; +#endif +#endif /* * Up to this point, the boot CPU has been using .init.data * area. Reload any changed state for the boot CPU. diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c index 46a01bd..2e88e6d 100644 --- a/arch/x86/kernel/signal.c +++ b/arch/x86/kernel/signal.c @@ -198,7 +198,7 @@ static unsigned long align_sigframe(unsigned long sp) * Align the stack pointer according to the i386 ABI, * i.e. so that on function entry ((sp + 4) & 15) == 0. */ - sp = ((sp + 4) & -16ul) - 4; + sp = ((sp - 12) & -16ul) - 4; #else /* !CONFIG_X86_32 */ sp = round_down(sp, 16) - 8; #endif @@ -249,11 +249,11 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size, * Return an always-bogus address instead so we will die with SIGSEGV. */ if (onsigstack && !likely(on_sig_stack(sp))) - return (void __user *)-1L; + return (__force void __user *)-1L; /* save i387 state */ if (used_math() && save_i387_xstate(*fpstate) < 0) - return (void __user *)-1L; + return (__force void __user *)-1L; return (void __user *)sp; } @@ -308,9 +308,9 @@ __setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, } if (current->mm->context.vdso) - restorer = VDSO32_SYMBOL(current->mm->context.vdso, sigreturn); + restorer = (__force void __user *)VDSO32_SYMBOL(current->mm->context.vdso, sigreturn); else - restorer = &frame->retcode; + restorer = (void __user *)&frame->retcode; if (ka->sa.sa_flags & SA_RESTORER) restorer = ka->sa.sa_restorer; @@ -324,7 +324,7 @@ __setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, * reasons and because gdb uses it as a signature to notice * signal handler stack frames. */ - err |= __put_user(*((u64 *)&retcode), (u64 *)frame->retcode); + err |= __put_user(*((u64 *)&retcode), (u64 __user *)frame->retcode); if (err) return -EFAULT; @@ -378,7 +378,10 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); /* Set up to return from userspace. */ - restorer = VDSO32_SYMBOL(current->mm->context.vdso, rt_sigreturn); + if (current->mm->context.vdso) + restorer = (__force void __user *)VDSO32_SYMBOL(current->mm->context.vdso, rt_sigreturn); + else + restorer = (void __user *)&frame->retcode; if (ka->sa.sa_flags & SA_RESTORER) restorer = ka->sa.sa_restorer; put_user_ex(restorer, &frame->pretcode); @@ -390,7 +393,7 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, * reasons and because gdb uses it as a signature to notice * signal handler stack frames. */ - put_user_ex(*((u64 *)&rt_retcode), (u64 *)frame->retcode); + put_user_ex(*((u64 *)&rt_retcode), (u64 __user *)frame->retcode); } put_user_catch(err); if (err) @@ -765,7 +768,7 @@ static void do_signal(struct pt_regs *regs) * X86_32: vm86 regs switched out by assembly code before reaching * here, so testing against kernel CS suffices. */ - if (!user_mode(regs)) + if (!user_mode_novm(regs)) return; signr = get_signal_to_deliver(&info, &ka, regs, NULL); diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 66d250c..f1b10bd 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -715,17 +715,20 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) set_idle_for_cpu(cpu, c_idle.idle); do_rest: per_cpu(current_task, cpu) = c_idle.idle; + per_cpu(current_tinfo, cpu) = &c_idle.idle->tinfo; #ifdef CONFIG_X86_32 /* Stack for startup_32 can be just as for start_secondary onwards */ irq_ctx_init(cpu); #else clear_tsk_thread_flag(c_idle.idle, TIF_FORK); initial_gs = per_cpu_offset(cpu); - per_cpu(kernel_stack, cpu) = - (unsigned long)task_stack_page(c_idle.idle) - - KERNEL_STACK_OFFSET + THREAD_SIZE; + per_cpu(kernel_stack, cpu) = (unsigned long)task_stack_page(c_idle.idle) - 16 + THREAD_SIZE; #endif + + pax_open_kernel(); early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); + pax_close_kernel(); + initial_code = (unsigned long)start_secondary; stack_start = c_idle.idle->thread.sp; @@ -868,6 +871,12 @@ int __cpuinit native_cpu_up(unsigned int cpu) per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; +#ifdef CONFIG_PAX_PER_CPU_PGD + clone_pgd_range(get_cpu_pgd(cpu) + KERNEL_PGD_BOUNDARY, + swapper_pg_dir + KERNEL_PGD_BOUNDARY, + KERNEL_PGD_PTRS); +#endif + err = do_boot_cpu(apicid, cpu); if (err) { pr_debug("do_boot_cpu failed %d\n", err); diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c index c346d11..d43b163 100644 --- a/arch/x86/kernel/step.c +++ b/arch/x86/kernel/step.c @@ -27,10 +27,10 @@ unsigned long convert_ip_to_linear(struct task_struct *child, struct pt_regs *re struct desc_struct *desc; unsigned long base; - seg &= ~7UL; + seg >>= 3; mutex_lock(&child->mm->context.lock); - if (unlikely((seg >> 3) >= child->mm->context.size)) + if (unlikely(seg >= child->mm->context.size)) addr = -1L; /* bogus selector, access would fault */ else { desc = child->mm->context.ldt + seg; @@ -42,7 +42,8 @@ unsigned long convert_ip_to_linear(struct task_struct *child, struct pt_regs *re addr += base; } mutex_unlock(&child->mm->context.lock); - } + } else if (seg == __KERNEL_CS || seg == __KERNEXEC_KERNEL_CS) + addr = ktla_ktva(addr); return addr; } @@ -53,6 +54,9 @@ static int is_setting_trap_flag(struct task_struct *child, struct pt_regs *regs) unsigned char opcode[15]; unsigned long addr = convert_ip_to_linear(child, regs); + if (addr == -EINVAL) + return 0; + copied = access_process_vm(child, addr, opcode, sizeof(opcode), 0); for (i = 0; i < copied; i++) { switch (opcode[i]) { diff --git a/arch/x86/kernel/sys_i386_32.c b/arch/x86/kernel/sys_i386_32.c index 0b0cb5f..db6b9ed 100644 --- a/arch/x86/kernel/sys_i386_32.c +++ b/arch/x86/kernel/sys_i386_32.c @@ -24,17 +24,224 @@ #include -/* - * Do a system call from kernel instead of calling sys_execve so we - * end up with proper pt_regs. - */ -int kernel_execve(const char *filename, - const char *const argv[], - const char *const envp[]) +int i386_mmap_check(unsigned long addr, unsigned long len, unsigned long flags) { - long __res; - asm volatile ("int $0x80" - : "=a" (__res) - : "0" (__NR_execve), "b" (filename), "c" (argv), "d" (envp) : "memory"); - return __res; + unsigned long pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (current->mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + if (len > pax_task_size || addr > pax_task_size - len) + return -EINVAL; + + return 0; +} + +unsigned long +arch_get_unmapped_area(struct file *filp, unsigned long addr, + unsigned long len, unsigned long pgoff, unsigned long flags) +{ + struct mm_struct *mm = current->mm; + struct vm_area_struct *vma; + unsigned long start_addr, pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + pax_task_size -= PAGE_SIZE; + + if (len > pax_task_size) + return -ENOMEM; + + if (flags & MAP_FIXED) + return addr; + +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + + if (addr) { + addr = PAGE_ALIGN(addr); + if (pax_task_size - len >= addr) { + vma = find_vma(mm, addr); + if (check_heap_stack_gap(vma, addr, len)) + return addr; + } + } + if (len > mm->cached_hole_size) { + start_addr = addr = mm->free_area_cache; + } else { + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; + } + +#ifdef CONFIG_PAX_PAGEEXEC + if (!(__supported_pte_mask & _PAGE_NX) && (mm->pax_flags & MF_PAX_PAGEEXEC) && (flags & MAP_EXECUTABLE) && start_addr >= mm->mmap_base) { + start_addr = 0x00110000UL; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + start_addr += mm->delta_mmap & 0x03FFF000UL; +#endif + + if (mm->start_brk <= start_addr && start_addr < mm->mmap_base) + start_addr = addr = mm->mmap_base; + else + addr = start_addr; + } +#endif + +full_search: + for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { + /* At this point: (!vma || addr < vma->vm_end). */ + if (pax_task_size - len < addr) { + /* + * Start a new search - just in case we missed + * some holes. + */ + if (start_addr != mm->mmap_base) { + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; + goto full_search; + } + return -ENOMEM; + } + if (check_heap_stack_gap(vma, addr, len)) + break; + if (addr + mm->cached_hole_size < vma->vm_start) + mm->cached_hole_size = vma->vm_start - addr; + addr = vma->vm_end; + if (mm->start_brk <= addr && addr < mm->mmap_base) { + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; + goto full_search; + } + } + + /* + * Remember the place where we stopped the search: + */ + mm->free_area_cache = addr + len; + return addr; +} + +unsigned long +arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, + const unsigned long len, const unsigned long pgoff, + const unsigned long flags) +{ + struct vm_area_struct *vma; + struct mm_struct *mm = current->mm; + unsigned long base = mm->mmap_base, addr = addr0, pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + pax_task_size -= PAGE_SIZE; + + /* requested length too big for entire address space */ + if (len > pax_task_size) + return -ENOMEM; + + if (flags & MAP_FIXED) + return addr; + +#ifdef CONFIG_PAX_PAGEEXEC + if (!(__supported_pte_mask & _PAGE_NX) && (mm->pax_flags & MF_PAX_PAGEEXEC) && (flags & MAP_EXECUTABLE)) + goto bottomup; +#endif + +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + + /* requesting a specific address */ + if (addr) { + addr = PAGE_ALIGN(addr); + if (pax_task_size - len >= addr) { + vma = find_vma(mm, addr); + if (check_heap_stack_gap(vma, addr, len)) + return addr; + } + } + + /* check if free_area_cache is useful for us */ + if (len <= mm->cached_hole_size) { + mm->cached_hole_size = 0; + mm->free_area_cache = mm->mmap_base; + } + + /* either no address requested or can't fit in requested address hole */ + addr = mm->free_area_cache; + + /* make sure it can fit in the remaining address space */ + if (addr > len) { + vma = find_vma(mm, addr-len); + if (check_heap_stack_gap(vma, addr - len, len)) + /* remember the address as a hint for next time */ + return (mm->free_area_cache = addr-len); + } + + if (mm->mmap_base < len) + goto bottomup; + + addr = mm->mmap_base-len; + + do { + /* + * Lookup failure means no vma is above this address, + * else if new region fits below vma->vm_start, + * return with success: + */ + vma = find_vma(mm, addr); + if (check_heap_stack_gap(vma, addr, len)) + /* remember the address as a hint for next time */ + return (mm->free_area_cache = addr); + + /* remember the largest hole we saw so far */ + if (addr + mm->cached_hole_size < vma->vm_start) + mm->cached_hole_size = vma->vm_start - addr; + + /* try just below the current vma->vm_start */ + addr = skip_heap_stack_gap(vma, len); + } while (!IS_ERR_VALUE(addr)); + +bottomup: + /* + * A failed mmap() very likely causes application failure, + * so fall back to the bottom-up function here. This scenario + * can happen with large stack limits and large mmap() + * allocations. + */ + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + mm->mmap_base = SEGMEXEC_TASK_UNMAPPED_BASE; + else +#endif + + mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + + mm->free_area_cache = mm->mmap_base; + mm->cached_hole_size = ~0UL; + addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags); + /* + * Restore the topdown base: + */ + mm->mmap_base = base; + mm->free_area_cache = base; + mm->cached_hole_size = ~0UL; + + return addr; } diff --git a/arch/x86/kernel/sys_x86_64.c b/arch/x86/kernel/sys_x86_64.c index 0514890..3dbebce 100644 --- a/arch/x86/kernel/sys_x86_64.c +++ b/arch/x86/kernel/sys_x86_64.c @@ -95,8 +95,8 @@ out: return error; } -static void find_start_end(unsigned long flags, unsigned long *begin, - unsigned long *end) +static void find_start_end(struct mm_struct *mm, unsigned long flags, + unsigned long *begin, unsigned long *end) { if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) { unsigned long new_begin; @@ -115,7 +115,7 @@ static void find_start_end(unsigned long flags, unsigned long *begin, *begin = new_begin; } } else { - *begin = TASK_UNMAPPED_BASE; + *begin = mm->mmap_base; *end = TASK_SIZE; } } @@ -132,16 +132,19 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, if (flags & MAP_FIXED) return addr; - find_start_end(flags, &begin, &end); + find_start_end(mm, flags, &begin, &end); if (len > end) return -ENOMEM; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { addr = PAGE_ALIGN(addr); vma = find_vma(mm, addr); - if (end - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (end - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } if (((flags & MAP_32BIT) || test_thread_flag(TIF_IA32)) @@ -172,7 +175,7 @@ full_search: } return -ENOMEM; } - if (!vma || addr + len <= vma->vm_start) { + if (check_heap_stack_gap(vma, addr, len)) { /* * Remember the place where we stopped the search: */ @@ -195,7 +198,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, { struct vm_area_struct *vma; struct mm_struct *mm = current->mm; - unsigned long addr = addr0; + unsigned long base = mm->mmap_base, addr = addr0; /* requested length too big for entire address space */ if (len > TASK_SIZE) @@ -208,13 +211,18 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) goto bottomup; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + /* requesting a specific address */ if (addr) { addr = PAGE_ALIGN(addr); - vma = find_vma(mm, addr); - if (TASK_SIZE - len >= addr && - (!vma || addr + len <= vma->vm_start)) - return addr; + if (TASK_SIZE - len >= addr) { + vma = find_vma(mm, addr); + if (check_heap_stack_gap(vma, addr, len)) + return addr; + } } /* check if free_area_cache is useful for us */ @@ -232,7 +240,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, ALIGN_TOPDOWN); vma = find_vma(mm, tmp_addr); - if (!vma || tmp_addr + len <= vma->vm_start) + if (check_heap_stack_gap(vma, tmp_addr, len)) /* remember the address as a hint for next time */ return mm->free_area_cache = tmp_addr; } @@ -251,7 +259,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, * return with success: */ vma = find_vma(mm, addr); - if (!vma || addr+len <= vma->vm_start) + if (check_heap_stack_gap(vma, addr, len)) /* remember the address as a hint for next time */ return mm->free_area_cache = addr; @@ -260,8 +268,8 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, mm->cached_hole_size = vma->vm_start - addr; /* try just below the current vma->vm_start */ - addr = vma->vm_start-len; - } while (len < vma->vm_start); + addr = skip_heap_stack_gap(vma, len); + } while (!IS_ERR_VALUE(addr)); bottomup: /* @@ -270,13 +278,21 @@ bottomup: * can happen with large stack limits and large mmap() * allocations. */ + mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + + mm->free_area_cache = mm->mmap_base; mm->cached_hole_size = ~0UL; - mm->free_area_cache = TASK_UNMAPPED_BASE; addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags); /* * Restore the topdown base: */ - mm->free_area_cache = mm->mmap_base; + mm->mmap_base = base; + mm->free_area_cache = base; mm->cached_hole_size = ~0UL; return addr; diff --git a/arch/x86/kernel/tboot.c b/arch/x86/kernel/tboot.c index e2410e2..4fe3fbc 100644 --- a/arch/x86/kernel/tboot.c +++ b/arch/x86/kernel/tboot.c @@ -219,7 +219,7 @@ static int tboot_setup_sleep(void) void tboot_shutdown(u32 shutdown_type) { - void (*shutdown)(void); + void (* __noreturn shutdown)(void); if (!tboot_enabled()) return; @@ -241,7 +241,7 @@ void tboot_shutdown(u32 shutdown_type) switch_to_tboot_pt(); - shutdown = (void(*)(void))(unsigned long)tboot->shutdown_entry; + shutdown = (void *)tboot->shutdown_entry; shutdown(); /* should not reach here */ @@ -298,7 +298,7 @@ void tboot_sleep(u8 sleep_state, u32 pm1a_control, u32 pm1b_control) tboot_shutdown(acpi_shutdown_map[sleep_state]); } -static atomic_t ap_wfs_count; +static atomic_unchecked_t ap_wfs_count; static int tboot_wait_for_aps(int num_aps) { @@ -322,9 +322,9 @@ static int __cpuinit tboot_cpu_callback(struct notifier_block *nfb, { switch (action) { case CPU_DYING: - atomic_inc(&ap_wfs_count); + atomic_inc_unchecked(&ap_wfs_count); if (num_online_cpus() == 1) - if (tboot_wait_for_aps(atomic_read(&ap_wfs_count))) + if (tboot_wait_for_aps(atomic_read_unchecked(&ap_wfs_count))) return NOTIFY_BAD; break; } @@ -343,7 +343,7 @@ static __init int tboot_late_init(void) tboot_create_trampoline(); - atomic_set(&ap_wfs_count, 0); + atomic_set_unchecked(&ap_wfs_count, 0); register_hotcpu_notifier(&tboot_cpu_notifier); return 0; } diff --git a/arch/x86/kernel/time.c b/arch/x86/kernel/time.c index dd5fbf4..b7f2232 100644 --- a/arch/x86/kernel/time.c +++ b/arch/x86/kernel/time.c @@ -31,9 +31,9 @@ unsigned long profile_pc(struct pt_regs *regs) { unsigned long pc = instruction_pointer(regs); - if (!user_mode_vm(regs) && in_lock_functions(pc)) { + if (!user_mode(regs) && in_lock_functions(pc)) { #ifdef CONFIG_FRAME_POINTER - return *(unsigned long *)(regs->bp + sizeof(long)); + return ktla_ktva(*(unsigned long *)(regs->bp + sizeof(long))); #else unsigned long *sp = (unsigned long *)kernel_stack_pointer(regs); @@ -42,11 +42,17 @@ unsigned long profile_pc(struct pt_regs *regs) * or above a saved flags. Eflags has bits 22-31 zero, * kernel addresses don't. */ + +#ifdef CONFIG_PAX_KERNEXEC + return ktla_ktva(sp[0]); +#else if (sp[0] >> 22) return sp[0]; if (sp[1] >> 22) return sp[1]; #endif + +#endif } return pc; } diff --git a/arch/x86/kernel/tls.c b/arch/x86/kernel/tls.c index bcfec2d..8f88b4a 100644 --- a/arch/x86/kernel/tls.c +++ b/arch/x86/kernel/tls.c @@ -85,6 +85,11 @@ int do_set_thread_area(struct task_struct *p, int idx, if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX) return -EINVAL; +#ifdef CONFIG_PAX_SEGMEXEC + if ((p->mm->pax_flags & MF_PAX_SEGMEXEC) && (info.contents & MODIFY_LDT_CONTENTS_CODE)) + return -EINVAL; +#endif + set_tls_desc(p, idx, &info, 1); return 0; diff --git a/arch/x86/kernel/tls.h b/arch/x86/kernel/tls.h index 2f083a2..7d3fecc 100644 --- a/arch/x86/kernel/tls.h +++ b/arch/x86/kernel/tls.h @@ -16,6 +16,6 @@ extern user_regset_active_fn regset_tls_active; extern user_regset_get_fn regset_tls_get; -extern user_regset_set_fn regset_tls_set; +extern user_regset_set_fn regset_tls_set __size_overflow(4); #endif /* _ARCH_X86_KERNEL_TLS_H */ diff --git a/arch/x86/kernel/trampoline_32.S b/arch/x86/kernel/trampoline_32.S index 451c0a7..e57f551 100644 --- a/arch/x86/kernel/trampoline_32.S +++ b/arch/x86/kernel/trampoline_32.S @@ -32,6 +32,12 @@ #include #include +#ifdef CONFIG_PAX_KERNEXEC +#define ta(X) (X) +#else +#define ta(X) ((X) - __PAGE_OFFSET) +#endif + #ifdef CONFIG_SMP .section ".x86_trampoline","a" @@ -62,7 +68,7 @@ r_base = . inc %ax # protected mode (PE) bit lmsw %ax # into protected mode # flush prefetch and jump to startup_32_smp in arch/i386/kernel/head.S - ljmpl $__BOOT_CS, $(startup_32_smp-__PAGE_OFFSET) + ljmpl $__BOOT_CS, $ta(startup_32_smp) # These need to be in the same 64K segment as the above; # hence we don't use the boot_gdt_descr defined in head.S diff --git a/arch/x86/kernel/trampoline_64.S b/arch/x86/kernel/trampoline_64.S index 09ff517..df19fbff 100644 --- a/arch/x86/kernel/trampoline_64.S +++ b/arch/x86/kernel/trampoline_64.S @@ -90,7 +90,7 @@ startup_32: movl $__KERNEL_DS, %eax # Initialize the %ds segment register movl %eax, %ds - movl $X86_CR4_PAE, %eax + movl $(X86_CR4_PSE | X86_CR4_PAE | X86_CR4_PGE), %eax movl %eax, %cr4 # Enable PAE mode # Setup trampoline 4 level pagetables @@ -138,7 +138,7 @@ tidt: # so the kernel can live anywhere .balign 4 tgdt: - .short tgdt_end - tgdt # gdt limit + .short tgdt_end - tgdt - 1 # gdt limit .long tgdt - r_base .short 0 .quad 0x00cf9b000000ffff # __KERNEL32_CS diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 4bbe04d..41d0943 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -70,12 +70,6 @@ asmlinkage int system_call(void); /* Do we ignore FPU interrupts ? */ char ignore_fpu_irq; - -/* - * The IDT has to be page-aligned to simplify the Pentium - * F0 0F bug workaround. - */ -gate_desc idt_table[NR_VECTORS] __page_aligned_data = { { { { 0, 0 } } }, }; #endif DECLARE_BITMAP(used_vectors, NR_VECTORS); @@ -108,13 +102,13 @@ static inline void preempt_conditional_cli(struct pt_regs *regs) } static void __kprobes -do_trap(int trapnr, int signr, char *str, struct pt_regs *regs, +do_trap(int trapnr, int signr, const char *str, struct pt_regs *regs, long error_code, siginfo_t *info) { struct task_struct *tsk = current; #ifdef CONFIG_X86_32 - if (regs->flags & X86_VM_MASK) { + if (v8086_mode(regs)) { /* * traps 0, 1, 3, 4, and 5 should be forwarded to vm86. * On nmi (interrupt 2), do_trap should not be called. @@ -125,7 +119,7 @@ do_trap(int trapnr, int signr, char *str, struct pt_regs *regs, } #endif - if (!user_mode(regs)) + if (!user_mode_novm(regs)) goto kernel_trap; #ifdef CONFIG_X86_32 @@ -148,7 +142,7 @@ trap_signal: printk_ratelimit()) { printk(KERN_INFO "%s[%d] trap %s ip:%lx sp:%lx error:%lx", - tsk->comm, tsk->pid, str, + tsk->comm, task_pid_nr(tsk), str, regs->ip, regs->sp, error_code); print_vma_addr(" in ", regs->ip); printk("\n"); @@ -165,8 +159,20 @@ kernel_trap: if (!fixup_exception(regs)) { tsk->thread.error_code = error_code; tsk->thread.trap_no = trapnr; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + if (trapnr == 12 && ((regs->cs & 0xFFFF) == __KERNEL_CS || (regs->cs & 0xFFFF) == __KERNEXEC_KERNEL_CS)) + str = "PAX: suspicious stack segment fault"; +#endif + die(str, regs, error_code); } + +#ifdef CONFIG_PAX_REFCOUNT + if (trapnr == 4) + pax_report_refcount_overflow(regs); +#endif + return; #ifdef CONFIG_X86_32 @@ -255,14 +261,30 @@ do_general_protection(struct pt_regs *regs, long error_code) conditional_sti(regs); #ifdef CONFIG_X86_32 - if (regs->flags & X86_VM_MASK) + if (v8086_mode(regs)) goto gp_in_vm86; #endif tsk = current; - if (!user_mode(regs)) + if (!user_mode_novm(regs)) goto gp_in_kernel; +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) + if (!(__supported_pte_mask & _PAGE_NX) && tsk->mm && (tsk->mm->pax_flags & MF_PAX_PAGEEXEC)) { + struct mm_struct *mm = tsk->mm; + unsigned long limit; + + down_write(&mm->mmap_sem); + limit = mm->context.user_cs_limit; + if (limit < TASK_SIZE) { + track_exec_limit(mm, limit, TASK_SIZE, VM_EXEC); + up_write(&mm->mmap_sem); + return; + } + up_write(&mm->mmap_sem); + } +#endif + tsk->thread.error_code = error_code; tsk->thread.trap_no = 13; @@ -295,6 +317,13 @@ gp_in_kernel: if (notify_die(DIE_GPF, "general protection fault", regs, error_code, 13, SIGSEGV) == NOTIFY_STOP) return; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + if ((regs->cs & 0xFFFF) == __KERNEL_CS || (regs->cs & 0xFFFF) == __KERNEXEC_KERNEL_CS) + die("PAX: suspicious general protection fault", regs, error_code); + else +#endif + die("general protection fault", regs, error_code); } @@ -421,7 +450,7 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) /* It's safe to allow irq's after DR6 has been saved */ preempt_conditional_sti(regs); - if (regs->flags & X86_VM_MASK) { + if (v8086_mode(regs)) { handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, 1); preempt_conditional_cli(regs); @@ -436,7 +465,7 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) * We already checked v86 mode above, so we can check for kernel mode * by just checking the CPL of CS. */ - if ((dr6 & DR_STEP) && !user_mode(regs)) { + if ((dr6 & DR_STEP) && !user_mode_novm(regs)) { tsk->thread.debugreg6 &= ~DR_STEP; set_tsk_thread_flag(tsk, TIF_SINGLESTEP); regs->flags &= ~X86_EFLAGS_TF; @@ -466,7 +495,7 @@ void math_error(struct pt_regs *regs, int error_code, int trapnr) return; conditional_sti(regs); - if (!user_mode_vm(regs)) + if (!user_mode(regs)) { if (!fixup_exception(regs)) { task->thread.error_code = error_code; diff --git a/arch/x86/kernel/verify_cpu.S b/arch/x86/kernel/verify_cpu.S index b9242ba..50c5edd 100644 --- a/arch/x86/kernel/verify_cpu.S +++ b/arch/x86/kernel/verify_cpu.S @@ -20,6 +20,7 @@ * arch/x86/boot/compressed/head_64.S: Boot cpu verification * arch/x86/kernel/trampoline_64.S: secondary processor verification * arch/x86/kernel/head_32.S: processor startup + * arch/x86/kernel/acpi/realmode/wakeup.S: 32bit processor resume * * verify_cpu, returns the status of longmode and SSE in register %eax. * 0: Success 1: Failure diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c index 328cb37..f37fee1 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -41,6 +41,7 @@ #include #include #include +#include #include #include @@ -109,6 +110,9 @@ static int copy_vm86_regs_to_user(struct vm86_regs __user *user, /* convert vm86_regs to kernel_vm86_regs */ static int copy_vm86_regs_from_user(struct kernel_vm86_regs *regs, const struct vm86_regs __user *user, + unsigned extra) __size_overflow(3); +static int copy_vm86_regs_from_user(struct kernel_vm86_regs *regs, + const struct vm86_regs __user *user, unsigned extra) { int ret = 0; @@ -148,7 +152,7 @@ struct pt_regs *save_v86_state(struct kernel_vm86_regs *regs) do_exit(SIGSEGV); } - tss = &per_cpu(init_tss, get_cpu()); + tss = init_tss + get_cpu(); current->thread.sp0 = current->thread.saved_sp0; current->thread.sysenter_cs = __KERNEL_CS; load_sp0(tss, ¤t->thread); @@ -210,6 +214,13 @@ int sys_vm86old(struct vm86_struct __user *v86, struct pt_regs *regs) struct task_struct *tsk; int tmp, ret = -EPERM; +#ifdef CONFIG_GRKERNSEC_VM86 + if (!capable(CAP_SYS_RAWIO)) { + gr_handle_vm86(); + goto out; + } +#endif + tsk = current; if (tsk->thread.saved_sp0) goto out; @@ -240,6 +251,14 @@ int sys_vm86(unsigned long cmd, unsigned long arg, struct pt_regs *regs) int tmp, ret; struct vm86plus_struct __user *v86; +#ifdef CONFIG_GRKERNSEC_VM86 + if (!capable(CAP_SYS_RAWIO)) { + gr_handle_vm86(); + ret = -EPERM; + goto out; + } +#endif + tsk = current; switch (cmd) { case VM86_REQUEST_IRQ: @@ -326,7 +345,7 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk tsk->thread.saved_fs = info->regs32->fs; tsk->thread.saved_gs = get_user_gs(info->regs32); - tss = &per_cpu(init_tss, get_cpu()); + tss = init_tss + get_cpu(); tsk->thread.sp0 = (unsigned long) &info->VM86_TSS_ESP0; if (cpu_has_sep) tsk->thread.sysenter_cs = 0; @@ -533,7 +552,7 @@ static void do_int(struct kernel_vm86_regs *regs, int i, goto cannot_handle; if (i == 0x21 && is_revectored(AH(regs), &KVM86->int21_revectored)) goto cannot_handle; - intr_ptr = (unsigned long __user *) (i << 2); + intr_ptr = (__force unsigned long __user *) (i << 2); if (get_user(segoffs, intr_ptr)) goto cannot_handle; if ((segoffs >> 16) == BIOSSEG) diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 0f703f1..9e15f64 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -26,6 +26,13 @@ #include #include #include +#include + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) +#define __KERNEL_TEXT_OFFSET (LOAD_OFFSET + ____LOAD_PHYSICAL_ADDR) +#else +#define __KERNEL_TEXT_OFFSET 0 +#endif #undef i386 /* in case the preprocessor is a 32bit one */ @@ -69,30 +76,43 @@ jiffies_64 = jiffies; PHDRS { text PT_LOAD FLAGS(5); /* R_E */ +#ifdef CONFIG_X86_32 + module PT_LOAD FLAGS(5); /* R_E */ +#endif +#ifdef CONFIG_XEN + rodata PT_LOAD FLAGS(5); /* R_E */ +#else + rodata PT_LOAD FLAGS(4); /* R__ */ +#endif data PT_LOAD FLAGS(6); /* RW_ */ -#ifdef CONFIG_X86_64 + init.begin PT_LOAD FLAGS(6); /* RW_ */ #ifdef CONFIG_SMP percpu PT_LOAD FLAGS(6); /* RW_ */ #endif + text.init PT_LOAD FLAGS(5); /* R_E */ + text.exit PT_LOAD FLAGS(5); /* R_E */ init PT_LOAD FLAGS(7); /* RWE */ -#endif note PT_NOTE FLAGS(0); /* ___ */ } SECTIONS { #ifdef CONFIG_X86_32 - . = LOAD_OFFSET + LOAD_PHYSICAL_ADDR; - phys_startup_32 = startup_32 - LOAD_OFFSET; + . = LOAD_OFFSET + ____LOAD_PHYSICAL_ADDR; #else - . = __START_KERNEL; - phys_startup_64 = startup_64 - LOAD_OFFSET; + . = __START_KERNEL; #endif /* Text and read-only data */ - .text : AT(ADDR(.text) - LOAD_OFFSET) { - _text = .; + .text (. - __KERNEL_TEXT_OFFSET): AT(ADDR(.text) - LOAD_OFFSET + __KERNEL_TEXT_OFFSET) { /* bootstrapping code */ +#ifdef CONFIG_X86_32 + phys_startup_32 = startup_32 - LOAD_OFFSET + __KERNEL_TEXT_OFFSET; +#else + phys_startup_64 = startup_64 - LOAD_OFFSET + __KERNEL_TEXT_OFFSET; +#endif + __LOAD_PHYSICAL_ADDR = . - LOAD_OFFSET + __KERNEL_TEXT_OFFSET; + _text = .; HEAD_TEXT #ifdef CONFIG_X86_32 . = ALIGN(PAGE_SIZE); @@ -108,13 +128,47 @@ SECTIONS IRQENTRY_TEXT *(.fixup) *(.gnu.warning) - /* End of text section */ - _etext = .; } :text = 0x9090 - NOTES :text :note + . += __KERNEL_TEXT_OFFSET; - EXCEPTION_TABLE(16) :text = 0x9090 +#ifdef CONFIG_X86_32 + . = ALIGN(PAGE_SIZE); + .module.text : AT(ADDR(.module.text) - LOAD_OFFSET) { + +#if defined(CONFIG_PAX_KERNEXEC) && defined(CONFIG_MODULES) + MODULES_EXEC_VADDR = .; + BYTE(0) + . += (CONFIG_PAX_KERNEXEC_MODULE_TEXT * 1024 * 1024); + . = ALIGN(HPAGE_SIZE); + MODULES_EXEC_END = . - 1; +#endif + + } :module +#endif + + .text.end : AT(ADDR(.text.end) - LOAD_OFFSET) { + /* End of text section */ + _etext = . - __KERNEL_TEXT_OFFSET; + } + +#ifdef CONFIG_X86_32 + . = ALIGN(PAGE_SIZE); + .rodata.page_aligned : AT(ADDR(.rodata.page_aligned) - LOAD_OFFSET) { + *(.idt) + . = ALIGN(PAGE_SIZE); + *(.empty_zero_page) + *(.initial_pg_fixmap) + *(.initial_pg_pmd) + *(.initial_page_table) + *(.swapper_pg_dir) + } :rodata +#endif + + . = ALIGN(PAGE_SIZE); + NOTES :rodata :note + + EXCEPTION_TABLE(16) :rodata #if defined(CONFIG_DEBUG_RODATA) /* .text should occupy whole number of pages */ @@ -126,16 +180,20 @@ SECTIONS /* Data */ .data : AT(ADDR(.data) - LOAD_OFFSET) { + +#ifdef CONFIG_PAX_KERNEXEC + . = ALIGN(HPAGE_SIZE); +#else + . = ALIGN(PAGE_SIZE); +#endif + /* Start of data section */ _sdata = .; /* init_task */ INIT_TASK_DATA(THREAD_SIZE) -#ifdef CONFIG_X86_32 - /* 32 bit has nosave before _edata */ NOSAVE_DATA -#endif PAGE_ALIGNED_DATA(PAGE_SIZE) @@ -176,12 +234,19 @@ SECTIONS #endif /* CONFIG_X86_64 */ /* Init code and data - will be freed after init */ - . = ALIGN(PAGE_SIZE); .init.begin : AT(ADDR(.init.begin) - LOAD_OFFSET) { + BYTE(0) + +#ifdef CONFIG_PAX_KERNEXEC + . = ALIGN(HPAGE_SIZE); +#else + . = ALIGN(PAGE_SIZE); +#endif + __init_begin = .; /* paired with __init_end */ - } + } :init.begin -#if defined(CONFIG_X86_64) && defined(CONFIG_SMP) +#ifdef CONFIG_SMP /* * percpu offsets are zero-based on SMP. PERCPU_VADDR() changes the * output PHDR, so the next output section - .init.text - should @@ -190,12 +255,27 @@ SECTIONS PERCPU_VADDR(INTERNODE_CACHE_BYTES, 0, :percpu) #endif - INIT_TEXT_SECTION(PAGE_SIZE) -#ifdef CONFIG_X86_64 - :init -#endif + . = ALIGN(PAGE_SIZE); + init_begin = .; + .init.text (. - __KERNEL_TEXT_OFFSET): AT(init_begin - LOAD_OFFSET) { + VMLINUX_SYMBOL(_sinittext) = .; + INIT_TEXT + VMLINUX_SYMBOL(_einittext) = .; + . = ALIGN(PAGE_SIZE); + } :text.init - INIT_DATA_SECTION(16) + /* + * .exit.text is discard at runtime, not link time, to deal with + * references from .altinstructions and .eh_frame + */ + .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET + __KERNEL_TEXT_OFFSET) { + EXIT_TEXT + . = ALIGN(16); + } :text.exit + . = init_begin + SIZEOF(.init.text) + SIZEOF(.exit.text); + + . = ALIGN(PAGE_SIZE); + INIT_DATA_SECTION(16) :init /* * Code and data for a variety of lowlevel trampolines, to be @@ -269,19 +349,12 @@ SECTIONS } . = ALIGN(8); - /* - * .exit.text is discard at runtime, not link time, to deal with - * references from .altinstructions and .eh_frame - */ - .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET) { - EXIT_TEXT - } .exit.data : AT(ADDR(.exit.data) - LOAD_OFFSET) { EXIT_DATA } -#if !defined(CONFIG_X86_64) || !defined(CONFIG_SMP) +#ifndef CONFIG_SMP PERCPU_SECTION(INTERNODE_CACHE_BYTES) #endif @@ -300,16 +373,10 @@ SECTIONS .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { __smp_locks = .; *(.smp_locks) - . = ALIGN(PAGE_SIZE); __smp_locks_end = .; + . = ALIGN(PAGE_SIZE); } -#ifdef CONFIG_X86_64 - .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { - NOSAVE_DATA - } -#endif - /* BSS */ . = ALIGN(PAGE_SIZE); .bss : AT(ADDR(.bss) - LOAD_OFFSET) { @@ -325,6 +392,7 @@ SECTIONS __brk_base = .; . += 64 * 1024; /* 64k alignment slop space */ *(.brk_reservation) /* areas brk users have reserved */ + . = ALIGN(HPAGE_SIZE); __brk_limit = .; } @@ -351,13 +419,12 @@ SECTIONS * for the boot processor. */ #define INIT_PER_CPU(x) init_per_cpu__##x = x + __per_cpu_load -INIT_PER_CPU(gdt_page); INIT_PER_CPU(irq_stack_union); /* * Build-time check on the image size: */ -. = ASSERT((_end - _text <= KERNEL_IMAGE_SIZE), +. = ASSERT((_end - _text - __KERNEL_TEXT_OFFSET <= KERNEL_IMAGE_SIZE), "kernel image bigger than KERNEL_IMAGE_SIZE"); #ifdef CONFIG_SMP diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c index b07ba93..a212969 100644 --- a/arch/x86/kernel/vsyscall_64.c +++ b/arch/x86/kernel/vsyscall_64.c @@ -57,15 +57,13 @@ DEFINE_VVAR(struct vsyscall_gtod_data, vsyscall_gtod_data) = .lock = __SEQLOCK_UNLOCKED(__vsyscall_gtod_data.lock), }; -static enum { EMULATE, NATIVE, NONE } vsyscall_mode = EMULATE; +static enum { EMULATE, NONE } vsyscall_mode = EMULATE; static int __init vsyscall_setup(char *str) { if (str) { if (!strcmp("emulate", str)) vsyscall_mode = EMULATE; - else if (!strcmp("native", str)) - vsyscall_mode = NATIVE; else if (!strcmp("none", str)) vsyscall_mode = NONE; else @@ -207,7 +205,7 @@ bool emulate_vsyscall(struct pt_regs *regs, unsigned long address) tsk = current; if (seccomp_mode(&tsk->seccomp)) - do_exit(SIGKILL); + do_group_exit(SIGKILL); /* * With a real vsyscall, page faults cause SIGSEGV. We want to @@ -279,8 +277,7 @@ bool emulate_vsyscall(struct pt_regs *regs, unsigned long address) return true; sigsegv: - force_sig(SIGSEGV, current); - return true; + do_group_exit(SIGKILL); } /* @@ -333,10 +330,7 @@ void __init map_vsyscall(void) extern char __vvar_page; unsigned long physaddr_vvar_page = __pa_symbol(&__vvar_page); - __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_vsyscall, - vsyscall_mode == NATIVE - ? PAGE_KERNEL_VSYSCALL - : PAGE_KERNEL_VVAR); + __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_vsyscall, PAGE_KERNEL_VVAR); BUILD_BUG_ON((unsigned long)__fix_to_virt(VSYSCALL_FIRST_PAGE) != (unsigned long)VSYSCALL_START); diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c index 9796c2f..f686fbf 100644 --- a/arch/x86/kernel/x8664_ksyms_64.c +++ b/arch/x86/kernel/x8664_ksyms_64.c @@ -29,8 +29,6 @@ EXPORT_SYMBOL(__put_user_8); EXPORT_SYMBOL(copy_user_generic_string); EXPORT_SYMBOL(copy_user_generic_unrolled); EXPORT_SYMBOL(__copy_user_nocache); -EXPORT_SYMBOL(_copy_from_user); -EXPORT_SYMBOL(_copy_to_user); EXPORT_SYMBOL(copy_page); EXPORT_SYMBOL(clear_page); diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index 7110911..e8cdee5 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -130,7 +130,7 @@ int check_for_xstate(struct i387_fxsave_struct __user *buf, fx_sw_user->xstate_size > fx_sw_user->extended_size) return -EINVAL; - err = __get_user(magic2, (__u32 *) (((void *)fpstate) + + err = __get_user(magic2, (__u32 __user *) (((void __user *)fpstate) + fx_sw_user->extended_size - FP_XSTATE_MAGIC2_SIZE)); if (err) @@ -266,7 +266,7 @@ fx_only: * the other extended state. */ xrstor_state(init_xstate_buf, pcntxt_mask & ~XSTATE_FPSSE); - return fxrstor_checking((__force struct i387_fxsave_struct *)buf); + return fxrstor_checking((struct i387_fxsave_struct __force_kernel *)buf); } /* @@ -295,7 +295,7 @@ int restore_i387_xstate(void __user *buf) if (use_xsave()) err = restore_user_xstate(buf); else - err = fxrstor_checking((__force struct i387_fxsave_struct *) + err = fxrstor_checking((struct i387_fxsave_struct __force_kernel *) buf); if (unlikely(err)) { /* diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c index 89b02bf..0f6511d 100644 --- a/arch/x86/kvm/cpuid.c +++ b/arch/x86/kvm/cpuid.c @@ -124,15 +124,20 @@ int kvm_vcpu_ioctl_set_cpuid2(struct kvm_vcpu *vcpu, struct kvm_cpuid2 *cpuid, struct kvm_cpuid_entry2 __user *entries) { - int r; + int r, i; r = -E2BIG; if (cpuid->nent > KVM_MAX_CPUID_ENTRIES) goto out; r = -EFAULT; - if (copy_from_user(&vcpu->arch.cpuid_entries, entries, - cpuid->nent * sizeof(struct kvm_cpuid_entry2))) + if (!access_ok(VERIFY_READ, entries, cpuid->nent * sizeof(struct kvm_cpuid_entry2))) goto out; + for (i = 0; i < cpuid->nent; ++i) { + struct kvm_cpuid_entry2 cpuid_entry; + if (__copy_from_user(&cpuid_entry, entries + i, sizeof(cpuid_entry))) + goto out; + vcpu->arch.cpuid_entries[i] = cpuid_entry; + } vcpu->arch.cpuid_nent = cpuid->nent; kvm_apic_set_version(vcpu); kvm_x86_ops->cpuid_update(vcpu); @@ -147,15 +152,19 @@ int kvm_vcpu_ioctl_get_cpuid2(struct kvm_vcpu *vcpu, struct kvm_cpuid2 *cpuid, struct kvm_cpuid_entry2 __user *entries) { - int r; + int r, i; r = -E2BIG; if (cpuid->nent < vcpu->arch.cpuid_nent) goto out; r = -EFAULT; - if (copy_to_user(entries, &vcpu->arch.cpuid_entries, - vcpu->arch.cpuid_nent * sizeof(struct kvm_cpuid_entry2))) + if (!access_ok(VERIFY_WRITE, entries, vcpu->arch.cpuid_nent * sizeof(struct kvm_cpuid_entry2))) goto out; + for (i = 0; i < vcpu->arch.cpuid_nent; ++i) { + struct kvm_cpuid_entry2 cpuid_entry = vcpu->arch.cpuid_entries[i]; + if (__copy_to_user(entries + i, &cpuid_entry, sizeof(cpuid_entry))) + goto out; + } return 0; out: diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index 0982507..7f6d72f 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -250,6 +250,7 @@ struct gprefix { #define ____emulate_2op(ctxt, _op, _x, _y, _suffix, _dsttype) \ do { \ + unsigned long _tmp; \ __asm__ __volatile__ ( \ _PRE_EFLAGS("0", "4", "2") \ _op _suffix " %"_x"3,%1; " \ @@ -264,8 +265,6 @@ struct gprefix { /* Raw emulation: instruction has two explicit operands. */ #define __emulate_2op_nobyte(ctxt,_op,_wx,_wy,_lx,_ly,_qx,_qy) \ do { \ - unsigned long _tmp; \ - \ switch ((ctxt)->dst.bytes) { \ case 2: \ ____emulate_2op(ctxt,_op,_wx,_wy,"w",u16); \ @@ -281,7 +280,6 @@ struct gprefix { #define __emulate_2op(ctxt,_op,_bx,_by,_wx,_wy,_lx,_ly,_qx,_qy) \ do { \ - unsigned long _tmp; \ switch ((ctxt)->dst.bytes) { \ case 1: \ ____emulate_2op(ctxt,_op,_bx,_by,"b",u8); \ diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index cfdc6e0..ab92e84 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -54,7 +54,7 @@ #define APIC_BUS_CYCLE_NS 1 /* #define apic_debug(fmt,arg...) printk(KERN_WARNING fmt,##arg) */ -#define apic_debug(fmt, arg...) +#define apic_debug(fmt, arg...) do {} while (0) #define APIC_LVT_NUM 6 /* 14 is the version for Xeon and Pentium 8.4.8*/ diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h index 1561028..0ed7f14 100644 --- a/arch/x86/kvm/paging_tmpl.h +++ b/arch/x86/kvm/paging_tmpl.h @@ -197,7 +197,7 @@ retry_walk: if (unlikely(kvm_is_error_hva(host_addr))) goto error; - ptep_user = (pt_element_t __user *)((void *)host_addr + offset); + ptep_user = (pt_element_t __force_user *)((void *)host_addr + offset); if (unlikely(__copy_from_user(&pte, ptep_user, sizeof(pte)))) goto error; diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index e385214..f8df033 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -3420,7 +3420,11 @@ static void reload_tss(struct kvm_vcpu *vcpu) int cpu = raw_smp_processor_id(); struct svm_cpu_data *sd = per_cpu(svm_data, cpu); + + pax_open_kernel(); sd->tss_desc->type = 9; /* available 32/64-bit TSS */ + pax_close_kernel(); + load_TR_desc(); } @@ -3798,6 +3802,10 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu) #endif #endif +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_MEMORY_UDEREF) + __set_fs(current_thread_info()->addr_limit); +#endif + reload_tss(vcpu); local_irq_disable(); diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index a7a6f60..04b745a 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -1306,7 +1306,11 @@ static void reload_tss(void) struct desc_struct *descs; descs = (void *)gdt->address; + + pax_open_kernel(); descs[GDT_ENTRY_TSS].type = 9; /* available TSS */ + pax_close_kernel(); + load_TR_desc(); } @@ -2637,8 +2641,11 @@ static __init int hardware_setup(void) if (!cpu_has_vmx_flexpriority()) flexpriority_enabled = 0; - if (!cpu_has_vmx_tpr_shadow()) - kvm_x86_ops->update_cr8_intercept = NULL; + if (!cpu_has_vmx_tpr_shadow()) { + pax_open_kernel(); + *(void **)&kvm_x86_ops->update_cr8_intercept = NULL; + pax_close_kernel(); + } if (enable_ept && !cpu_has_vmx_ept_2m_page()) kvm_disable_largepages(); @@ -3654,7 +3661,7 @@ static void vmx_set_constant_host_state(void) vmcs_writel(HOST_IDTR_BASE, dt.address); /* 22.2.4 */ asm("mov $.Lkvm_vmx_return, %0" : "=r"(tmpl)); - vmcs_writel(HOST_RIP, tmpl); /* 22.2.5 */ + vmcs_writel(HOST_RIP, ktla_ktva(tmpl)); /* 22.2.5 */ rdmsr(MSR_IA32_SYSENTER_CS, low32, high32); vmcs_write32(HOST_IA32_SYSENTER_CS, low32); @@ -6192,6 +6199,12 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) "jmp .Lkvm_vmx_return \n\t" ".Llaunched: " __ex(ASM_VMX_VMRESUME) "\n\t" ".Lkvm_vmx_return: " + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + "ljmp %[cs],$.Lkvm_vmx_return2\n\t" + ".Lkvm_vmx_return2: " +#endif + /* Save guest registers, load host registers, keep flags */ "mov %0, %c[wordsize](%%"R"sp) \n\t" "pop %0 \n\t" @@ -6240,6 +6253,11 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) #endif [cr2]"i"(offsetof(struct vcpu_vmx, vcpu.arch.cr2)), [wordsize]"i"(sizeof(ulong)) + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + ,[cs]"i"(__KERNEL_CS) +#endif + : "cc", "memory" , R"ax", R"bx", R"di", R"si" #ifdef CONFIG_X86_64 @@ -6268,7 +6286,16 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) } } - asm("mov %0, %%ds; mov %0, %%es" : : "r"(__USER_DS)); + asm("mov %0, %%ds; mov %0, %%es; mov %0, %%ss" : : "r"(__KERNEL_DS)); + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + loadsegment(fs, __KERNEL_PERCPU); +#endif + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_MEMORY_UDEREF) + __set_fs(current_thread_info()->addr_limit); +#endif + vmx->loaded_vmcs->launched = 1; vmx->exit_reason = vmcs_read32(VM_EXIT_REASON); diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 8d1c6c6..6e6d611 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -873,6 +873,7 @@ static int do_set_msr(struct kvm_vcpu *vcpu, unsigned index, u64 *data) return kvm_set_msr(vcpu, index, *data); } +static void kvm_write_wall_clock(struct kvm *kvm, gpa_t wall_clock) __size_overflow(2); static void kvm_write_wall_clock(struct kvm *kvm, gpa_t wall_clock) { int version; @@ -1307,12 +1308,13 @@ static int set_msr_mce(struct kvm_vcpu *vcpu, u32 msr, u64 data) return 0; } +static int xen_hvm_config(struct kvm_vcpu *vcpu, u64 data) __size_overflow(2); static int xen_hvm_config(struct kvm_vcpu *vcpu, u64 data) { struct kvm *kvm = vcpu->kvm; int lm = is_long_mode(vcpu); - u8 *blob_addr = lm ? (u8 *)(long)kvm->arch.xen_hvm_config.blob_addr_64 - : (u8 *)(long)kvm->arch.xen_hvm_config.blob_addr_32; + u8 __user *blob_addr = lm ? (u8 __user *)(long)kvm->arch.xen_hvm_config.blob_addr_64 + : (u8 __user *)(long)kvm->arch.xen_hvm_config.blob_addr_32; u8 blob_size = lm ? kvm->arch.xen_hvm_config.blob_size_64 : kvm->arch.xen_hvm_config.blob_size_32; u32 page_num = data & ~PAGE_MASK; @@ -2145,6 +2147,8 @@ long kvm_arch_dev_ioctl(struct file *filp, if (n < msr_list.nmsrs) goto out; r = -EFAULT; + if (num_msrs_to_save > ARRAY_SIZE(msrs_to_save)) + goto out; if (copy_to_user(user_msr_list->indices, &msrs_to_save, num_msrs_to_save * sizeof(u32))) goto out; @@ -2266,7 +2270,7 @@ static int kvm_vcpu_ioctl_set_lapic(struct kvm_vcpu *vcpu, static int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq) { - if (irq->irq < 0 || irq->irq >= 256) + if (irq->irq >= 256) return -EINVAL; if (irqchip_in_kernel(vcpu->kvm)) return -ENXIO; @@ -3499,6 +3503,9 @@ gpa_t kvm_mmu_gva_to_gpa_system(struct kvm_vcpu *vcpu, gva_t gva, static int kvm_read_guest_virt_helper(gva_t addr, void *val, unsigned int bytes, struct kvm_vcpu *vcpu, u32 access, + struct x86_exception *exception) __size_overflow(1,3); +static int kvm_read_guest_virt_helper(gva_t addr, void *val, unsigned int bytes, + struct kvm_vcpu *vcpu, u32 access, struct x86_exception *exception) { void *data = val; @@ -3530,6 +3537,9 @@ out: /* used for instruction fetching */ static int kvm_fetch_guest_virt(struct x86_emulate_ctxt *ctxt, gva_t addr, void *val, unsigned int bytes, + struct x86_exception *exception) __size_overflow(2,4); +static int kvm_fetch_guest_virt(struct x86_emulate_ctxt *ctxt, + gva_t addr, void *val, unsigned int bytes, struct x86_exception *exception) { struct kvm_vcpu *vcpu = emul_to_vcpu(ctxt); @@ -3554,6 +3564,9 @@ EXPORT_SYMBOL_GPL(kvm_read_guest_virt); static int kvm_read_guest_virt_system(struct x86_emulate_ctxt *ctxt, gva_t addr, void *val, unsigned int bytes, + struct x86_exception *exception) __size_overflow(2,4); +static int kvm_read_guest_virt_system(struct x86_emulate_ctxt *ctxt, + gva_t addr, void *val, unsigned int bytes, struct x86_exception *exception) { struct kvm_vcpu *vcpu = emul_to_vcpu(ctxt); @@ -3667,12 +3680,16 @@ static int read_prepare(struct kvm_vcpu *vcpu, void *val, int bytes) } static int read_emulate(struct kvm_vcpu *vcpu, gpa_t gpa, + void *val, int bytes) __size_overflow(2); +static int read_emulate(struct kvm_vcpu *vcpu, gpa_t gpa, void *val, int bytes) { return !kvm_read_guest(vcpu->kvm, gpa, val, bytes); } static int write_emulate(struct kvm_vcpu *vcpu, gpa_t gpa, + void *val, int bytes) __size_overflow(2); +static int write_emulate(struct kvm_vcpu *vcpu, gpa_t gpa, void *val, int bytes) { return emulator_write_phys(vcpu, gpa, val, bytes); @@ -3823,6 +3840,12 @@ static int emulator_cmpxchg_emulated(struct x86_emulate_ctxt *ctxt, const void *old, const void *new, unsigned int bytes, + struct x86_exception *exception) __size_overflow(5); +static int emulator_cmpxchg_emulated(struct x86_emulate_ctxt *ctxt, + unsigned long addr, + const void *old, + const void *new, + unsigned int bytes, struct x86_exception *exception) { struct kvm_vcpu *vcpu = emul_to_vcpu(ctxt); @@ -4782,7 +4805,7 @@ static void kvm_set_mmio_spte_mask(void) kvm_mmu_set_mmio_spte_mask(mask); } -int kvm_arch_init(void *opaque) +int kvm_arch_init(const void *opaque) { int r; struct kvm_x86_ops *ops = (struct kvm_x86_ops *)opaque; diff --git a/arch/x86/kvm/x86.h b/arch/x86/kvm/x86.h index cb80c29..aeee86c 100644 --- a/arch/x86/kvm/x86.h +++ b/arch/x86/kvm/x86.h @@ -116,11 +116,11 @@ void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data); int kvm_read_guest_virt(struct x86_emulate_ctxt *ctxt, gva_t addr, void *val, unsigned int bytes, - struct x86_exception *exception); + struct x86_exception *exception) __size_overflow(2,4); int kvm_write_guest_virt_system(struct x86_emulate_ctxt *ctxt, gva_t addr, void *val, unsigned int bytes, - struct x86_exception *exception); + struct x86_exception *exception) __size_overflow(2,4); extern u64 host_xcr0; diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 642d880..44e0f3f 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -1200,9 +1200,10 @@ static __init int early_put_chars(u32 vtermno, const char *buf, int count) * Rebooting also tells the Host we're finished, but the RESTART flag tells the * Launcher to reboot us. */ -static void lguest_restart(char *reason) +static __noreturn void lguest_restart(char *reason) { hcall(LHCALL_SHUTDOWN, __pa(reason), LGUEST_SHUTDOWN_RESTART, 0, 0); + BUG(); } /*G:050 diff --git a/arch/x86/lib/atomic64_32.c b/arch/x86/lib/atomic64_32.c index 042f682..c92afb6 100644 --- a/arch/x86/lib/atomic64_32.c +++ b/arch/x86/lib/atomic64_32.c @@ -8,18 +8,30 @@ long long atomic64_read_cx8(long long, const atomic64_t *v); EXPORT_SYMBOL(atomic64_read_cx8); +long long atomic64_read_unchecked_cx8(long long, const atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_read_unchecked_cx8); long long atomic64_set_cx8(long long, const atomic64_t *v); EXPORT_SYMBOL(atomic64_set_cx8); +long long atomic64_set_unchecked_cx8(long long, const atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_set_unchecked_cx8); long long atomic64_xchg_cx8(long long, unsigned high); EXPORT_SYMBOL(atomic64_xchg_cx8); long long atomic64_add_return_cx8(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_add_return_cx8); +long long atomic64_add_return_unchecked_cx8(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_add_return_unchecked_cx8); long long atomic64_sub_return_cx8(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_sub_return_cx8); +long long atomic64_sub_return_unchecked_cx8(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_sub_return_unchecked_cx8); long long atomic64_inc_return_cx8(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_inc_return_cx8); +long long atomic64_inc_return_unchecked_cx8(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_inc_return_unchecked_cx8); long long atomic64_dec_return_cx8(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_dec_return_cx8); +long long atomic64_dec_return_unchecked_cx8(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_dec_return_unchecked_cx8); long long atomic64_dec_if_positive_cx8(atomic64_t *v); EXPORT_SYMBOL(atomic64_dec_if_positive_cx8); int atomic64_inc_not_zero_cx8(atomic64_t *v); @@ -30,26 +42,46 @@ EXPORT_SYMBOL(atomic64_add_unless_cx8); #ifndef CONFIG_X86_CMPXCHG64 long long atomic64_read_386(long long, const atomic64_t *v); EXPORT_SYMBOL(atomic64_read_386); +long long atomic64_read_unchecked_386(long long, const atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_read_unchecked_386); long long atomic64_set_386(long long, const atomic64_t *v); EXPORT_SYMBOL(atomic64_set_386); +long long atomic64_set_unchecked_386(long long, const atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_set_unchecked_386); long long atomic64_xchg_386(long long, unsigned high); EXPORT_SYMBOL(atomic64_xchg_386); long long atomic64_add_return_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_add_return_386); +long long atomic64_add_return_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_add_return_unchecked_386); long long atomic64_sub_return_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_sub_return_386); +long long atomic64_sub_return_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_sub_return_unchecked_386); long long atomic64_inc_return_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_inc_return_386); +long long atomic64_inc_return_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_inc_return_unchecked_386); long long atomic64_dec_return_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_dec_return_386); +long long atomic64_dec_return_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_dec_return_unchecked_386); long long atomic64_add_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_add_386); +long long atomic64_add_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_add_unchecked_386); long long atomic64_sub_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_sub_386); +long long atomic64_sub_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_sub_unchecked_386); long long atomic64_inc_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_inc_386); +long long atomic64_inc_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_inc_unchecked_386); long long atomic64_dec_386(long long a, atomic64_t *v); EXPORT_SYMBOL(atomic64_dec_386); +long long atomic64_dec_unchecked_386(long long a, atomic64_unchecked_t *v); +EXPORT_SYMBOL(atomic64_dec_unchecked_386); long long atomic64_dec_if_positive_386(atomic64_t *v); EXPORT_SYMBOL(atomic64_dec_if_positive_386); int atomic64_inc_not_zero_386(atomic64_t *v); diff --git a/arch/x86/lib/atomic64_386_32.S b/arch/x86/lib/atomic64_386_32.S index e8e7e0d..56fd1b0 100644 --- a/arch/x86/lib/atomic64_386_32.S +++ b/arch/x86/lib/atomic64_386_32.S @@ -48,6 +48,10 @@ BEGIN(read) movl (v), %eax movl 4(v), %edx RET_ENDP +BEGIN(read_unchecked) + movl (v), %eax + movl 4(v), %edx +RET_ENDP #undef v #define v %esi @@ -55,6 +59,10 @@ BEGIN(set) movl %ebx, (v) movl %ecx, 4(v) RET_ENDP +BEGIN(set_unchecked) + movl %ebx, (v) + movl %ecx, 4(v) +RET_ENDP #undef v #define v %esi @@ -70,6 +78,20 @@ RET_ENDP BEGIN(add) addl %eax, (v) adcl %edx, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT + jno 0f + subl %eax, (v) + sbbl %edx, 4(v) + int $4 +0: + _ASM_EXTABLE(0b, 0b) +#endif + +RET_ENDP +BEGIN(add_unchecked) + addl %eax, (v) + adcl %edx, 4(v) RET_ENDP #undef v @@ -77,6 +99,24 @@ RET_ENDP BEGIN(add_return) addl (v), %eax adcl 4(v), %edx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 2f) +#endif + + movl %eax, (v) + movl %edx, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT +2: +#endif + +RET_ENDP +BEGIN(add_return_unchecked) + addl (v), %eax + adcl 4(v), %edx movl %eax, (v) movl %edx, 4(v) RET_ENDP @@ -86,6 +126,20 @@ RET_ENDP BEGIN(sub) subl %eax, (v) sbbl %edx, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT + jno 0f + addl %eax, (v) + adcl %edx, 4(v) + int $4 +0: + _ASM_EXTABLE(0b, 0b) +#endif + +RET_ENDP +BEGIN(sub_unchecked) + subl %eax, (v) + sbbl %edx, 4(v) RET_ENDP #undef v @@ -96,6 +150,27 @@ BEGIN(sub_return) sbbl $0, %edx addl (v), %eax adcl 4(v), %edx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 2f) +#endif + + movl %eax, (v) + movl %edx, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT +2: +#endif + +RET_ENDP +BEGIN(sub_return_unchecked) + negl %edx + negl %eax + sbbl $0, %edx + addl (v), %eax + adcl 4(v), %edx movl %eax, (v) movl %edx, 4(v) RET_ENDP @@ -105,6 +180,20 @@ RET_ENDP BEGIN(inc) addl $1, (v) adcl $0, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT + jno 0f + subl $1, (v) + sbbl $0, 4(v) + int $4 +0: + _ASM_EXTABLE(0b, 0b) +#endif + +RET_ENDP +BEGIN(inc_unchecked) + addl $1, (v) + adcl $0, 4(v) RET_ENDP #undef v @@ -114,6 +203,26 @@ BEGIN(inc_return) movl 4(v), %edx addl $1, %eax adcl $0, %edx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 2f) +#endif + + movl %eax, (v) + movl %edx, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT +2: +#endif + +RET_ENDP +BEGIN(inc_return_unchecked) + movl (v), %eax + movl 4(v), %edx + addl $1, %eax + adcl $0, %edx movl %eax, (v) movl %edx, 4(v) RET_ENDP @@ -123,6 +232,20 @@ RET_ENDP BEGIN(dec) subl $1, (v) sbbl $0, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT + jno 0f + addl $1, (v) + adcl $0, 4(v) + int $4 +0: + _ASM_EXTABLE(0b, 0b) +#endif + +RET_ENDP +BEGIN(dec_unchecked) + subl $1, (v) + sbbl $0, 4(v) RET_ENDP #undef v @@ -132,6 +255,26 @@ BEGIN(dec_return) movl 4(v), %edx subl $1, %eax sbbl $0, %edx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 2f) +#endif + + movl %eax, (v) + movl %edx, 4(v) + +#ifdef CONFIG_PAX_REFCOUNT +2: +#endif + +RET_ENDP +BEGIN(dec_return_unchecked) + movl (v), %eax + movl 4(v), %edx + subl $1, %eax + sbbl $0, %edx movl %eax, (v) movl %edx, 4(v) RET_ENDP @@ -143,6 +286,13 @@ BEGIN(add_unless) adcl %edx, %edi addl (v), %eax adcl 4(v), %edx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 2f) +#endif + cmpl %eax, %esi je 3f 1: @@ -168,6 +318,13 @@ BEGIN(inc_not_zero) 1: addl $1, %eax adcl $0, %edx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 2f) +#endif + movl %eax, (v) movl %edx, 4(v) movl $1, %eax @@ -186,6 +343,13 @@ BEGIN(dec_if_positive) movl 4(v), %edx subl $1, %eax sbbl $0, %edx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 1f) +#endif + js 1f movl %eax, (v) movl %edx, 4(v) diff --git a/arch/x86/lib/atomic64_cx8_32.S b/arch/x86/lib/atomic64_cx8_32.S index 391a083..3a2cf39 100644 --- a/arch/x86/lib/atomic64_cx8_32.S +++ b/arch/x86/lib/atomic64_cx8_32.S @@ -35,10 +35,20 @@ ENTRY(atomic64_read_cx8) CFI_STARTPROC read64 %ecx + pax_force_retaddr ret CFI_ENDPROC ENDPROC(atomic64_read_cx8) +ENTRY(atomic64_read_unchecked_cx8) + CFI_STARTPROC + + read64 %ecx + pax_force_retaddr + ret + CFI_ENDPROC +ENDPROC(atomic64_read_unchecked_cx8) + ENTRY(atomic64_set_cx8) CFI_STARTPROC @@ -48,10 +58,25 @@ ENTRY(atomic64_set_cx8) cmpxchg8b (%esi) jne 1b + pax_force_retaddr ret CFI_ENDPROC ENDPROC(atomic64_set_cx8) +ENTRY(atomic64_set_unchecked_cx8) + CFI_STARTPROC + +1: +/* we don't need LOCK_PREFIX since aligned 64-bit writes + * are atomic on 586 and newer */ + cmpxchg8b (%esi) + jne 1b + + pax_force_retaddr + ret + CFI_ENDPROC +ENDPROC(atomic64_set_unchecked_cx8) + ENTRY(atomic64_xchg_cx8) CFI_STARTPROC @@ -62,12 +87,13 @@ ENTRY(atomic64_xchg_cx8) cmpxchg8b (%esi) jne 1b + pax_force_retaddr ret CFI_ENDPROC ENDPROC(atomic64_xchg_cx8) -.macro addsub_return func ins insc -ENTRY(atomic64_\func\()_return_cx8) +.macro addsub_return func ins insc unchecked="" +ENTRY(atomic64_\func\()_return\unchecked\()_cx8) CFI_STARTPROC SAVE ebp SAVE ebx @@ -84,27 +110,44 @@ ENTRY(atomic64_\func\()_return_cx8) movl %edx, %ecx \ins\()l %esi, %ebx \insc\()l %edi, %ecx + +.ifb \unchecked +#ifdef CONFIG_PAX_REFCOUNT + into +2: + _ASM_EXTABLE(2b, 3f) +#endif +.endif + LOCK_PREFIX cmpxchg8b (%ebp) jne 1b - -10: movl %ebx, %eax movl %ecx, %edx + +.ifb \unchecked +#ifdef CONFIG_PAX_REFCOUNT +3: +#endif +.endif + RESTORE edi RESTORE esi RESTORE ebx RESTORE ebp + pax_force_retaddr ret CFI_ENDPROC -ENDPROC(atomic64_\func\()_return_cx8) +ENDPROC(atomic64_\func\()_return\unchecked\()_cx8) .endm addsub_return add add adc addsub_return sub sub sbb +addsub_return add add adc _unchecked +addsub_return sub sub sbb _unchecked -.macro incdec_return func ins insc -ENTRY(atomic64_\func\()_return_cx8) +.macro incdec_return func ins insc unchecked="" +ENTRY(atomic64_\func\()_return\unchecked\()_cx8) CFI_STARTPROC SAVE ebx @@ -114,21 +157,39 @@ ENTRY(atomic64_\func\()_return_cx8) movl %edx, %ecx \ins\()l $1, %ebx \insc\()l $0, %ecx + +.ifb \unchecked +#ifdef CONFIG_PAX_REFCOUNT + into +2: + _ASM_EXTABLE(2b, 3f) +#endif +.endif + LOCK_PREFIX cmpxchg8b (%esi) jne 1b -10: movl %ebx, %eax movl %ecx, %edx + +.ifb \unchecked +#ifdef CONFIG_PAX_REFCOUNT +3: +#endif +.endif + RESTORE ebx + pax_force_retaddr ret CFI_ENDPROC -ENDPROC(atomic64_\func\()_return_cx8) +ENDPROC(atomic64_\func\()_return\unchecked\()_cx8) .endm incdec_return inc add adc incdec_return dec sub sbb +incdec_return inc add adc _unchecked +incdec_return dec sub sbb _unchecked ENTRY(atomic64_dec_if_positive_cx8) CFI_STARTPROC @@ -140,6 +201,13 @@ ENTRY(atomic64_dec_if_positive_cx8) movl %edx, %ecx subl $1, %ebx sbb $0, %ecx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 2f) +#endif + js 2f LOCK_PREFIX cmpxchg8b (%esi) @@ -149,6 +217,7 @@ ENTRY(atomic64_dec_if_positive_cx8) movl %ebx, %eax movl %ecx, %edx RESTORE ebx + pax_force_retaddr ret CFI_ENDPROC ENDPROC(atomic64_dec_if_positive_cx8) @@ -174,6 +243,13 @@ ENTRY(atomic64_add_unless_cx8) movl %edx, %ecx addl %esi, %ebx adcl %edi, %ecx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 3f) +#endif + LOCK_PREFIX cmpxchg8b (%ebp) jne 1b @@ -184,6 +260,7 @@ ENTRY(atomic64_add_unless_cx8) CFI_ADJUST_CFA_OFFSET -8 RESTORE ebx RESTORE ebp + pax_force_retaddr ret 4: cmpl %edx, 4(%esp) @@ -206,6 +283,13 @@ ENTRY(atomic64_inc_not_zero_cx8) movl %edx, %ecx addl $1, %ebx adcl $0, %ecx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 3f) +#endif + LOCK_PREFIX cmpxchg8b (%esi) jne 1b @@ -213,6 +297,7 @@ ENTRY(atomic64_inc_not_zero_cx8) movl $1, %eax 3: RESTORE ebx + pax_force_retaddr ret 4: testl %edx, %edx diff --git a/arch/x86/lib/checksum_32.S b/arch/x86/lib/checksum_32.S index 78d16a5..fbcf666 100644 --- a/arch/x86/lib/checksum_32.S +++ b/arch/x86/lib/checksum_32.S @@ -28,7 +28,8 @@ #include #include #include - +#include + /* * computes a partial checksum, e.g. for TCP/UDP fragments */ @@ -296,9 +297,24 @@ unsigned int csum_partial_copy_generic (const char *src, char *dst, #define ARGBASE 16 #define FP 12 - -ENTRY(csum_partial_copy_generic) + +ENTRY(csum_partial_copy_generic_to_user) CFI_STARTPROC + +#ifdef CONFIG_PAX_MEMORY_UDEREF + pushl_cfi %gs + popl_cfi %es + jmp csum_partial_copy_generic +#endif + +ENTRY(csum_partial_copy_generic_from_user) + +#ifdef CONFIG_PAX_MEMORY_UDEREF + pushl_cfi %gs + popl_cfi %ds +#endif + +ENTRY(csum_partial_copy_generic) subl $4,%esp CFI_ADJUST_CFA_OFFSET 4 pushl_cfi %edi @@ -320,7 +336,7 @@ ENTRY(csum_partial_copy_generic) jmp 4f SRC(1: movw (%esi), %bx ) addl $2, %esi -DST( movw %bx, (%edi) ) +DST( movw %bx, %es:(%edi) ) addl $2, %edi addw %bx, %ax adcl $0, %eax @@ -332,30 +348,30 @@ DST( movw %bx, (%edi) ) SRC(1: movl (%esi), %ebx ) SRC( movl 4(%esi), %edx ) adcl %ebx, %eax -DST( movl %ebx, (%edi) ) +DST( movl %ebx, %es:(%edi) ) adcl %edx, %eax -DST( movl %edx, 4(%edi) ) +DST( movl %edx, %es:4(%edi) ) SRC( movl 8(%esi), %ebx ) SRC( movl 12(%esi), %edx ) adcl %ebx, %eax -DST( movl %ebx, 8(%edi) ) +DST( movl %ebx, %es:8(%edi) ) adcl %edx, %eax -DST( movl %edx, 12(%edi) ) +DST( movl %edx, %es:12(%edi) ) SRC( movl 16(%esi), %ebx ) SRC( movl 20(%esi), %edx ) adcl %ebx, %eax -DST( movl %ebx, 16(%edi) ) +DST( movl %ebx, %es:16(%edi) ) adcl %edx, %eax -DST( movl %edx, 20(%edi) ) +DST( movl %edx, %es:20(%edi) ) SRC( movl 24(%esi), %ebx ) SRC( movl 28(%esi), %edx ) adcl %ebx, %eax -DST( movl %ebx, 24(%edi) ) +DST( movl %ebx, %es:24(%edi) ) adcl %edx, %eax -DST( movl %edx, 28(%edi) ) +DST( movl %edx, %es:28(%edi) ) lea 32(%esi), %esi lea 32(%edi), %edi @@ -369,7 +385,7 @@ DST( movl %edx, 28(%edi) ) shrl $2, %edx # This clears CF SRC(3: movl (%esi), %ebx ) adcl %ebx, %eax -DST( movl %ebx, (%edi) ) +DST( movl %ebx, %es:(%edi) ) lea 4(%esi), %esi lea 4(%edi), %edi dec %edx @@ -381,12 +397,12 @@ DST( movl %ebx, (%edi) ) jb 5f SRC( movw (%esi), %cx ) leal 2(%esi), %esi -DST( movw %cx, (%edi) ) +DST( movw %cx, %es:(%edi) ) leal 2(%edi), %edi je 6f shll $16,%ecx SRC(5: movb (%esi), %cl ) -DST( movb %cl, (%edi) ) +DST( movb %cl, %es:(%edi) ) 6: addl %ecx, %eax adcl $0, %eax 7: @@ -397,7 +413,7 @@ DST( movb %cl, (%edi) ) 6001: movl ARGBASE+20(%esp), %ebx # src_err_ptr - movl $-EFAULT, (%ebx) + movl $-EFAULT, %ss:(%ebx) # zero the complete destination - computing the rest # is too much work @@ -410,11 +426,15 @@ DST( movb %cl, (%edi) ) 6002: movl ARGBASE+24(%esp), %ebx # dst_err_ptr - movl $-EFAULT,(%ebx) + movl $-EFAULT,%ss:(%ebx) jmp 5000b .previous + pushl_cfi %ss + popl_cfi %ds + pushl_cfi %ss + popl_cfi %es popl_cfi %ebx CFI_RESTORE ebx popl_cfi %esi @@ -424,26 +444,43 @@ DST( movb %cl, (%edi) ) popl_cfi %ecx # equivalent to addl $4,%esp ret CFI_ENDPROC -ENDPROC(csum_partial_copy_generic) +ENDPROC(csum_partial_copy_generic_to_user) #else /* Version for PentiumII/PPro */ #define ROUND1(x) \ + nop; nop; nop; \ SRC(movl x(%esi), %ebx ) ; \ addl %ebx, %eax ; \ - DST(movl %ebx, x(%edi) ) ; + DST(movl %ebx, %es:x(%edi)) ; #define ROUND(x) \ + nop; nop; nop; \ SRC(movl x(%esi), %ebx ) ; \ adcl %ebx, %eax ; \ - DST(movl %ebx, x(%edi) ) ; + DST(movl %ebx, %es:x(%edi)) ; #define ARGBASE 12 - -ENTRY(csum_partial_copy_generic) + +ENTRY(csum_partial_copy_generic_to_user) CFI_STARTPROC + +#ifdef CONFIG_PAX_MEMORY_UDEREF + pushl_cfi %gs + popl_cfi %es + jmp csum_partial_copy_generic +#endif + +ENTRY(csum_partial_copy_generic_from_user) + +#ifdef CONFIG_PAX_MEMORY_UDEREF + pushl_cfi %gs + popl_cfi %ds +#endif + +ENTRY(csum_partial_copy_generic) pushl_cfi %ebx CFI_REL_OFFSET ebx, 0 pushl_cfi %edi @@ -464,7 +501,7 @@ ENTRY(csum_partial_copy_generic) subl %ebx, %edi lea -1(%esi),%edx andl $-32,%edx - lea 3f(%ebx,%ebx), %ebx + lea 3f(%ebx,%ebx,2), %ebx testl %esi, %esi jmp *%ebx 1: addl $64,%esi @@ -485,19 +522,19 @@ ENTRY(csum_partial_copy_generic) jb 5f SRC( movw (%esi), %dx ) leal 2(%esi), %esi -DST( movw %dx, (%edi) ) +DST( movw %dx, %es:(%edi) ) leal 2(%edi), %edi je 6f shll $16,%edx 5: SRC( movb (%esi), %dl ) -DST( movb %dl, (%edi) ) +DST( movb %dl, %es:(%edi) ) 6: addl %edx, %eax adcl $0, %eax 7: .section .fixup, "ax" 6001: movl ARGBASE+20(%esp), %ebx # src_err_ptr - movl $-EFAULT, (%ebx) + movl $-EFAULT, %ss:(%ebx) # zero the complete destination (computing the rest is too much work) movl ARGBASE+8(%esp),%edi # dst movl ARGBASE+12(%esp),%ecx # len @@ -505,10 +542,17 @@ DST( movb %dl, (%edi) ) rep; stosb jmp 7b 6002: movl ARGBASE+24(%esp), %ebx # dst_err_ptr - movl $-EFAULT, (%ebx) + movl $-EFAULT, %ss:(%ebx) jmp 7b .previous +#ifdef CONFIG_PAX_MEMORY_UDEREF + pushl_cfi %ss + popl_cfi %ds + pushl_cfi %ss + popl_cfi %es +#endif + popl_cfi %esi CFI_RESTORE esi popl_cfi %edi @@ -517,7 +561,7 @@ DST( movb %dl, (%edi) ) CFI_RESTORE ebx ret CFI_ENDPROC -ENDPROC(csum_partial_copy_generic) +ENDPROC(csum_partial_copy_generic_to_user) #undef ROUND #undef ROUND1 diff --git a/arch/x86/lib/clear_page_64.S b/arch/x86/lib/clear_page_64.S index f2145cf..cea889d 100644 --- a/arch/x86/lib/clear_page_64.S +++ b/arch/x86/lib/clear_page_64.S @@ -11,6 +11,7 @@ ENTRY(clear_page_c) movl $4096/8,%ecx xorl %eax,%eax rep stosq + pax_force_retaddr ret CFI_ENDPROC ENDPROC(clear_page_c) @@ -20,6 +21,7 @@ ENTRY(clear_page_c_e) movl $4096,%ecx xorl %eax,%eax rep stosb + pax_force_retaddr ret CFI_ENDPROC ENDPROC(clear_page_c_e) @@ -43,6 +45,7 @@ ENTRY(clear_page) leaq 64(%rdi),%rdi jnz .Lloop nop + pax_force_retaddr ret CFI_ENDPROC .Lclear_page_end: @@ -58,7 +61,7 @@ ENDPROC(clear_page) #include - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 1: .byte 0xeb /* jmp */ .byte (clear_page_c - clear_page) - (2f - 1b) /* offset */ 2: .byte 0xeb /* jmp */ diff --git a/arch/x86/lib/cmpxchg16b_emu.S b/arch/x86/lib/cmpxchg16b_emu.S index 1e572c5..2a162cd 100644 --- a/arch/x86/lib/cmpxchg16b_emu.S +++ b/arch/x86/lib/cmpxchg16b_emu.S @@ -53,11 +53,13 @@ this_cpu_cmpxchg16b_emu: popf mov $1, %al + pax_force_retaddr ret not_same: popf xor %al,%al + pax_force_retaddr ret CFI_ENDPROC diff --git a/arch/x86/lib/copy_page_64.S b/arch/x86/lib/copy_page_64.S index 01c805b..dccb07f 100644 --- a/arch/x86/lib/copy_page_64.S +++ b/arch/x86/lib/copy_page_64.S @@ -9,6 +9,7 @@ copy_page_c: CFI_STARTPROC movl $4096/8,%ecx rep movsq + pax_force_retaddr ret CFI_ENDPROC ENDPROC(copy_page_c) @@ -39,7 +40,7 @@ ENTRY(copy_page) movq 16 (%rsi), %rdx movq 24 (%rsi), %r8 movq 32 (%rsi), %r9 - movq 40 (%rsi), %r10 + movq 40 (%rsi), %r13 movq 48 (%rsi), %r11 movq 56 (%rsi), %r12 @@ -50,7 +51,7 @@ ENTRY(copy_page) movq %rdx, 16 (%rdi) movq %r8, 24 (%rdi) movq %r9, 32 (%rdi) - movq %r10, 40 (%rdi) + movq %r13, 40 (%rdi) movq %r11, 48 (%rdi) movq %r12, 56 (%rdi) @@ -69,7 +70,7 @@ ENTRY(copy_page) movq 16 (%rsi), %rdx movq 24 (%rsi), %r8 movq 32 (%rsi), %r9 - movq 40 (%rsi), %r10 + movq 40 (%rsi), %r13 movq 48 (%rsi), %r11 movq 56 (%rsi), %r12 @@ -78,7 +79,7 @@ ENTRY(copy_page) movq %rdx, 16 (%rdi) movq %r8, 24 (%rdi) movq %r9, 32 (%rdi) - movq %r10, 40 (%rdi) + movq %r13, 40 (%rdi) movq %r11, 48 (%rdi) movq %r12, 56 (%rdi) @@ -95,6 +96,7 @@ ENTRY(copy_page) CFI_RESTORE r13 addq $3*8,%rsp CFI_ADJUST_CFA_OFFSET -3*8 + pax_force_retaddr ret .Lcopy_page_end: CFI_ENDPROC @@ -105,7 +107,7 @@ ENDPROC(copy_page) #include - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 1: .byte 0xeb /* jmp */ .byte (copy_page_c - copy_page) - (2f - 1b) /* offset */ 2: diff --git a/arch/x86/lib/copy_user_64.S b/arch/x86/lib/copy_user_64.S index 0248402..821c786 100644 --- a/arch/x86/lib/copy_user_64.S +++ b/arch/x86/lib/copy_user_64.S @@ -16,6 +16,7 @@ #include #include #include +#include /* * By placing feature2 after feature1 in altinstructions section, we logically @@ -29,7 +30,7 @@ .byte 0xe9 /* 32bit jump */ .long \orig-1f /* by default jump to orig */ 1: - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 2: .byte 0xe9 /* near jump with 32bit immediate */ .long \alt1-1b /* offset */ /* or alternatively to alt1 */ 3: .byte 0xe9 /* near jump with 32bit immediate */ @@ -71,47 +72,20 @@ #endif .endm -/* Standard copy_to_user with segment limit checking */ -ENTRY(_copy_to_user) - CFI_STARTPROC - GET_THREAD_INFO(%rax) - movq %rdi,%rcx - addq %rdx,%rcx - jc bad_to_user - cmpq TI_addr_limit(%rax),%rcx - ja bad_to_user - ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,X86_FEATURE_ERMS, \ - copy_user_generic_unrolled,copy_user_generic_string, \ - copy_user_enhanced_fast_string - CFI_ENDPROC -ENDPROC(_copy_to_user) - -/* Standard copy_from_user with segment limit checking */ -ENTRY(_copy_from_user) - CFI_STARTPROC - GET_THREAD_INFO(%rax) - movq %rsi,%rcx - addq %rdx,%rcx - jc bad_from_user - cmpq TI_addr_limit(%rax),%rcx - ja bad_from_user - ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,X86_FEATURE_ERMS, \ - copy_user_generic_unrolled,copy_user_generic_string, \ - copy_user_enhanced_fast_string - CFI_ENDPROC -ENDPROC(_copy_from_user) - .section .fixup,"ax" /* must zero dest */ ENTRY(bad_from_user) bad_from_user: CFI_STARTPROC + testl %edx,%edx + js bad_to_user movl %edx,%ecx xorl %eax,%eax rep stosb bad_to_user: movl %edx,%eax + pax_force_retaddr ret CFI_ENDPROC ENDPROC(bad_from_user) @@ -141,19 +115,19 @@ ENTRY(copy_user_generic_unrolled) jz 17f 1: movq (%rsi),%r8 2: movq 1*8(%rsi),%r9 -3: movq 2*8(%rsi),%r10 +3: movq 2*8(%rsi),%rax 4: movq 3*8(%rsi),%r11 5: movq %r8,(%rdi) 6: movq %r9,1*8(%rdi) -7: movq %r10,2*8(%rdi) +7: movq %rax,2*8(%rdi) 8: movq %r11,3*8(%rdi) 9: movq 4*8(%rsi),%r8 10: movq 5*8(%rsi),%r9 -11: movq 6*8(%rsi),%r10 +11: movq 6*8(%rsi),%rax 12: movq 7*8(%rsi),%r11 13: movq %r8,4*8(%rdi) 14: movq %r9,5*8(%rdi) -15: movq %r10,6*8(%rdi) +15: movq %rax,6*8(%rdi) 16: movq %r11,7*8(%rdi) leaq 64(%rsi),%rsi leaq 64(%rdi),%rdi @@ -179,6 +153,7 @@ ENTRY(copy_user_generic_unrolled) decl %ecx jnz 21b 23: xor %eax,%eax + pax_force_retaddr ret .section .fixup,"ax" @@ -251,6 +226,7 @@ ENTRY(copy_user_generic_string) 3: rep movsb 4: xorl %eax,%eax + pax_force_retaddr ret .section .fixup,"ax" @@ -287,6 +263,7 @@ ENTRY(copy_user_enhanced_fast_string) 1: rep movsb 2: xorl %eax,%eax + pax_force_retaddr ret .section .fixup,"ax" diff --git a/arch/x86/lib/copy_user_nocache_64.S b/arch/x86/lib/copy_user_nocache_64.S index cb0c112..e3a6895 100644 --- a/arch/x86/lib/copy_user_nocache_64.S +++ b/arch/x86/lib/copy_user_nocache_64.S @@ -8,12 +8,14 @@ #include #include +#include #define FIX_ALIGNMENT 1 #include #include #include +#include .macro ALIGN_DESTINATION #ifdef FIX_ALIGNMENT @@ -50,6 +52,15 @@ */ ENTRY(__copy_user_nocache) CFI_STARTPROC + +#ifdef CONFIG_PAX_MEMORY_UDEREF + mov $PAX_USER_SHADOW_BASE,%rcx + cmp %rcx,%rsi + jae 1f + add %rcx,%rsi +1: +#endif + cmpl $8,%edx jb 20f /* less then 8 bytes, go to byte copy loop */ ALIGN_DESTINATION @@ -59,19 +70,19 @@ ENTRY(__copy_user_nocache) jz 17f 1: movq (%rsi),%r8 2: movq 1*8(%rsi),%r9 -3: movq 2*8(%rsi),%r10 +3: movq 2*8(%rsi),%rax 4: movq 3*8(%rsi),%r11 5: movnti %r8,(%rdi) 6: movnti %r9,1*8(%rdi) -7: movnti %r10,2*8(%rdi) +7: movnti %rax,2*8(%rdi) 8: movnti %r11,3*8(%rdi) 9: movq 4*8(%rsi),%r8 10: movq 5*8(%rsi),%r9 -11: movq 6*8(%rsi),%r10 +11: movq 6*8(%rsi),%rax 12: movq 7*8(%rsi),%r11 13: movnti %r8,4*8(%rdi) 14: movnti %r9,5*8(%rdi) -15: movnti %r10,6*8(%rdi) +15: movnti %rax,6*8(%rdi) 16: movnti %r11,7*8(%rdi) leaq 64(%rsi),%rsi leaq 64(%rdi),%rdi @@ -98,6 +109,7 @@ ENTRY(__copy_user_nocache) jnz 21b 23: xorl %eax,%eax sfence + pax_force_retaddr ret .section .fixup,"ax" diff --git a/arch/x86/lib/csum-copy_64.S b/arch/x86/lib/csum-copy_64.S index fb903b7..c92b7f7 100644 --- a/arch/x86/lib/csum-copy_64.S +++ b/arch/x86/lib/csum-copy_64.S @@ -8,6 +8,7 @@ #include #include #include +#include /* * Checksum copy with exception handling. @@ -228,6 +229,7 @@ ENTRY(csum_partial_copy_generic) CFI_RESTORE rbp addq $7*8, %rsp CFI_ADJUST_CFA_OFFSET -7*8 + pax_force_retaddr 0, 1 ret CFI_RESTORE_STATE diff --git a/arch/x86/lib/csum-wrappers_64.c b/arch/x86/lib/csum-wrappers_64.c index 459b58a..9570bc7 100644 --- a/arch/x86/lib/csum-wrappers_64.c +++ b/arch/x86/lib/csum-wrappers_64.c @@ -52,7 +52,13 @@ csum_partial_copy_from_user(const void __user *src, void *dst, len -= 2; } } - isum = csum_partial_copy_generic((__force const void *)src, + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)src < PAX_USER_SHADOW_BASE) + src += PAX_USER_SHADOW_BASE; +#endif + + isum = csum_partial_copy_generic((const void __force_kernel *)src, dst, len, isum, errp, NULL); if (unlikely(*errp)) goto out_err; @@ -105,7 +111,13 @@ csum_partial_copy_to_user(const void *src, void __user *dst, } *errp = 0; - return csum_partial_copy_generic(src, (void __force *)dst, + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)dst < PAX_USER_SHADOW_BASE) + dst += PAX_USER_SHADOW_BASE; +#endif + + return csum_partial_copy_generic(src, (void __force_kernel *)dst, len, isum, NULL, errp); } EXPORT_SYMBOL(csum_partial_copy_to_user); diff --git a/arch/x86/lib/getuser.S b/arch/x86/lib/getuser.S index 51f1504..ddac4c1 100644 --- a/arch/x86/lib/getuser.S +++ b/arch/x86/lib/getuser.S @@ -33,15 +33,38 @@ #include #include #include +#include +#include +#include + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define __copyuser_seg gs; +#else +#define __copyuser_seg +#endif .text ENTRY(__get_user_1) CFI_STARTPROC + +#if !defined(CONFIG_X86_32) || !defined(CONFIG_PAX_MEMORY_UDEREF) GET_THREAD_INFO(%_ASM_DX) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user -1: movzb (%_ASM_AX),%edx + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov $PAX_USER_SHADOW_BASE,%_ASM_DX + cmp %_ASM_DX,%_ASM_AX + jae 1234f + add %_ASM_DX,%_ASM_AX +1234: +#endif + +#endif + +1: __copyuser_seg movzb (%_ASM_AX),%edx xor %eax,%eax + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__get_user_1) @@ -49,12 +72,26 @@ ENDPROC(__get_user_1) ENTRY(__get_user_2) CFI_STARTPROC add $1,%_ASM_AX + +#if !defined(CONFIG_X86_32) || !defined(CONFIG_PAX_MEMORY_UDEREF) jc bad_get_user GET_THREAD_INFO(%_ASM_DX) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user -2: movzwl -1(%_ASM_AX),%edx + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov $PAX_USER_SHADOW_BASE,%_ASM_DX + cmp %_ASM_DX,%_ASM_AX + jae 1234f + add %_ASM_DX,%_ASM_AX +1234: +#endif + +#endif + +2: __copyuser_seg movzwl -1(%_ASM_AX),%edx xor %eax,%eax + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__get_user_2) @@ -62,12 +99,26 @@ ENDPROC(__get_user_2) ENTRY(__get_user_4) CFI_STARTPROC add $3,%_ASM_AX + +#if !defined(CONFIG_X86_32) || !defined(CONFIG_PAX_MEMORY_UDEREF) jc bad_get_user GET_THREAD_INFO(%_ASM_DX) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user -3: mov -3(%_ASM_AX),%edx + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov $PAX_USER_SHADOW_BASE,%_ASM_DX + cmp %_ASM_DX,%_ASM_AX + jae 1234f + add %_ASM_DX,%_ASM_AX +1234: +#endif + +#endif + +3: __copyuser_seg mov -3(%_ASM_AX),%edx xor %eax,%eax + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__get_user_4) @@ -80,8 +131,18 @@ ENTRY(__get_user_8) GET_THREAD_INFO(%_ASM_DX) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user + +#ifdef CONFIG_PAX_MEMORY_UDEREF + mov $PAX_USER_SHADOW_BASE,%_ASM_DX + cmp %_ASM_DX,%_ASM_AX + jae 1234f + add %_ASM_DX,%_ASM_AX +1234: +#endif + 4: movq -7(%_ASM_AX),%_ASM_DX xor %eax,%eax + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__get_user_8) @@ -91,6 +152,7 @@ bad_get_user: CFI_STARTPROC xor %edx,%edx mov $(-EFAULT),%_ASM_AX + pax_force_retaddr ret CFI_ENDPROC END(bad_get_user) diff --git a/arch/x86/lib/insn.c b/arch/x86/lib/insn.c index 5a1f9f3..ba9f577 100644 --- a/arch/x86/lib/insn.c +++ b/arch/x86/lib/insn.c @@ -21,6 +21,11 @@ #include #include #include +#ifdef __KERNEL__ +#include +#else +#define ktla_ktva(addr) addr +#endif /* Verify next sizeof(t) bytes can be on the same instruction */ #define validate_next(t, insn, n) \ @@ -49,8 +54,8 @@ void insn_init(struct insn *insn, const void *kaddr, int x86_64) { memset(insn, 0, sizeof(*insn)); - insn->kaddr = kaddr; - insn->next_byte = kaddr; + insn->kaddr = ktla_ktva(kaddr); + insn->next_byte = ktla_ktva(kaddr); insn->x86_64 = x86_64 ? 1 : 0; insn->opnd_bytes = 4; if (x86_64) diff --git a/arch/x86/lib/iomap_copy_64.S b/arch/x86/lib/iomap_copy_64.S index 05a95e7..326f2fa 100644 --- a/arch/x86/lib/iomap_copy_64.S +++ b/arch/x86/lib/iomap_copy_64.S @@ -17,6 +17,7 @@ #include #include +#include /* * override generic version in lib/iomap_copy.c @@ -25,6 +26,7 @@ ENTRY(__iowrite32_copy) CFI_STARTPROC movl %edx,%ecx rep movsd + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__iowrite32_copy) diff --git a/arch/x86/lib/memcpy_64.S b/arch/x86/lib/memcpy_64.S index efbf2a0..8893637 100644 --- a/arch/x86/lib/memcpy_64.S +++ b/arch/x86/lib/memcpy_64.S @@ -34,6 +34,7 @@ rep movsq movl %edx, %ecx rep movsb + pax_force_retaddr ret .Lmemcpy_e: .previous @@ -51,6 +52,7 @@ movl %edx, %ecx rep movsb + pax_force_retaddr ret .Lmemcpy_e_e: .previous @@ -81,13 +83,13 @@ ENTRY(memcpy) */ movq 0*8(%rsi), %r8 movq 1*8(%rsi), %r9 - movq 2*8(%rsi), %r10 + movq 2*8(%rsi), %rcx movq 3*8(%rsi), %r11 leaq 4*8(%rsi), %rsi movq %r8, 0*8(%rdi) movq %r9, 1*8(%rdi) - movq %r10, 2*8(%rdi) + movq %rcx, 2*8(%rdi) movq %r11, 3*8(%rdi) leaq 4*8(%rdi), %rdi jae .Lcopy_forward_loop @@ -110,12 +112,12 @@ ENTRY(memcpy) subq $0x20, %rdx movq -1*8(%rsi), %r8 movq -2*8(%rsi), %r9 - movq -3*8(%rsi), %r10 + movq -3*8(%rsi), %rcx movq -4*8(%rsi), %r11 leaq -4*8(%rsi), %rsi movq %r8, -1*8(%rdi) movq %r9, -2*8(%rdi) - movq %r10, -3*8(%rdi) + movq %rcx, -3*8(%rdi) movq %r11, -4*8(%rdi) leaq -4*8(%rdi), %rdi jae .Lcopy_backward_loop @@ -135,12 +137,13 @@ ENTRY(memcpy) */ movq 0*8(%rsi), %r8 movq 1*8(%rsi), %r9 - movq -2*8(%rsi, %rdx), %r10 + movq -2*8(%rsi, %rdx), %rcx movq -1*8(%rsi, %rdx), %r11 movq %r8, 0*8(%rdi) movq %r9, 1*8(%rdi) - movq %r10, -2*8(%rdi, %rdx) + movq %rcx, -2*8(%rdi, %rdx) movq %r11, -1*8(%rdi, %rdx) + pax_force_retaddr retq .p2align 4 .Lless_16bytes: @@ -153,6 +156,7 @@ ENTRY(memcpy) movq -1*8(%rsi, %rdx), %r9 movq %r8, 0*8(%rdi) movq %r9, -1*8(%rdi, %rdx) + pax_force_retaddr retq .p2align 4 .Lless_8bytes: @@ -166,6 +170,7 @@ ENTRY(memcpy) movl -4(%rsi, %rdx), %r8d movl %ecx, (%rdi) movl %r8d, -4(%rdi, %rdx) + pax_force_retaddr retq .p2align 4 .Lless_3bytes: @@ -183,6 +188,7 @@ ENTRY(memcpy) jnz .Lloop_1 .Lend: + pax_force_retaddr retq CFI_ENDPROC ENDPROC(memcpy) diff --git a/arch/x86/lib/memmove_64.S b/arch/x86/lib/memmove_64.S index ee16461..c39c199 100644 --- a/arch/x86/lib/memmove_64.S +++ b/arch/x86/lib/memmove_64.S @@ -61,13 +61,13 @@ ENTRY(memmove) 5: sub $0x20, %rdx movq 0*8(%rsi), %r11 - movq 1*8(%rsi), %r10 + movq 1*8(%rsi), %rcx movq 2*8(%rsi), %r9 movq 3*8(%rsi), %r8 leaq 4*8(%rsi), %rsi movq %r11, 0*8(%rdi) - movq %r10, 1*8(%rdi) + movq %rcx, 1*8(%rdi) movq %r9, 2*8(%rdi) movq %r8, 3*8(%rdi) leaq 4*8(%rdi), %rdi @@ -81,10 +81,10 @@ ENTRY(memmove) 4: movq %rdx, %rcx movq -8(%rsi, %rdx), %r11 - lea -8(%rdi, %rdx), %r10 + lea -8(%rdi, %rdx), %r9 shrq $3, %rcx rep movsq - movq %r11, (%r10) + movq %r11, (%r9) jmp 13f .Lmemmove_end_forward: @@ -95,14 +95,14 @@ ENTRY(memmove) 7: movq %rdx, %rcx movq (%rsi), %r11 - movq %rdi, %r10 + movq %rdi, %r9 leaq -8(%rsi, %rdx), %rsi leaq -8(%rdi, %rdx), %rdi shrq $3, %rcx std rep movsq cld - movq %r11, (%r10) + movq %r11, (%r9) jmp 13f /* @@ -127,13 +127,13 @@ ENTRY(memmove) 8: subq $0x20, %rdx movq -1*8(%rsi), %r11 - movq -2*8(%rsi), %r10 + movq -2*8(%rsi), %rcx movq -3*8(%rsi), %r9 movq -4*8(%rsi), %r8 leaq -4*8(%rsi), %rsi movq %r11, -1*8(%rdi) - movq %r10, -2*8(%rdi) + movq %rcx, -2*8(%rdi) movq %r9, -3*8(%rdi) movq %r8, -4*8(%rdi) leaq -4*8(%rdi), %rdi @@ -151,11 +151,11 @@ ENTRY(memmove) * Move data from 16 bytes to 31 bytes. */ movq 0*8(%rsi), %r11 - movq 1*8(%rsi), %r10 + movq 1*8(%rsi), %rcx movq -2*8(%rsi, %rdx), %r9 movq -1*8(%rsi, %rdx), %r8 movq %r11, 0*8(%rdi) - movq %r10, 1*8(%rdi) + movq %rcx, 1*8(%rdi) movq %r9, -2*8(%rdi, %rdx) movq %r8, -1*8(%rdi, %rdx) jmp 13f @@ -167,9 +167,9 @@ ENTRY(memmove) * Move data from 8 bytes to 15 bytes. */ movq 0*8(%rsi), %r11 - movq -1*8(%rsi, %rdx), %r10 + movq -1*8(%rsi, %rdx), %r9 movq %r11, 0*8(%rdi) - movq %r10, -1*8(%rdi, %rdx) + movq %r9, -1*8(%rdi, %rdx) jmp 13f 10: cmpq $4, %rdx @@ -178,9 +178,9 @@ ENTRY(memmove) * Move data from 4 bytes to 7 bytes. */ movl (%rsi), %r11d - movl -4(%rsi, %rdx), %r10d + movl -4(%rsi, %rdx), %r9d movl %r11d, (%rdi) - movl %r10d, -4(%rdi, %rdx) + movl %r9d, -4(%rdi, %rdx) jmp 13f 11: cmp $2, %rdx @@ -189,9 +189,9 @@ ENTRY(memmove) * Move data from 2 bytes to 3 bytes. */ movw (%rsi), %r11w - movw -2(%rsi, %rdx), %r10w + movw -2(%rsi, %rdx), %r9w movw %r11w, (%rdi) - movw %r10w, -2(%rdi, %rdx) + movw %r9w, -2(%rdi, %rdx) jmp 13f 12: cmp $1, %rdx @@ -202,6 +202,7 @@ ENTRY(memmove) movb (%rsi), %r11b movb %r11b, (%rdi) 13: + pax_force_retaddr retq CFI_ENDPROC @@ -210,6 +211,7 @@ ENTRY(memmove) /* Forward moving data. */ movq %rdx, %rcx rep movsb + pax_force_retaddr retq .Lmemmove_end_forward_efs: .previous diff --git a/arch/x86/lib/memset_64.S b/arch/x86/lib/memset_64.S index 79bd454..dff325a 100644 --- a/arch/x86/lib/memset_64.S +++ b/arch/x86/lib/memset_64.S @@ -31,6 +31,7 @@ movl %r8d,%ecx rep stosb movq %r9,%rax + pax_force_retaddr ret .Lmemset_e: .previous @@ -53,6 +54,7 @@ movl %edx,%ecx rep stosb movq %r9,%rax + pax_force_retaddr ret .Lmemset_e_e: .previous @@ -60,13 +62,13 @@ ENTRY(memset) ENTRY(__memset) CFI_STARTPROC - movq %rdi,%r10 movq %rdx,%r11 /* expand byte value */ movzbl %sil,%ecx movabs $0x0101010101010101,%rax mul %rcx /* with rax, clobbers rdx */ + movq %rdi,%rdx /* align dst */ movl %edi,%r9d @@ -120,7 +122,8 @@ ENTRY(__memset) jnz .Lloop_1 .Lende: - movq %r10,%rax + movq %rdx,%rax + pax_force_retaddr ret CFI_RESTORE_STATE diff --git a/arch/x86/lib/mmx_32.c b/arch/x86/lib/mmx_32.c index c9f2d9b..e7fd2c0 100644 --- a/arch/x86/lib/mmx_32.c +++ b/arch/x86/lib/mmx_32.c @@ -29,6 +29,7 @@ void *_mmx_memcpy(void *to, const void *from, size_t len) { void *p; int i; + unsigned long cr0; if (unlikely(in_interrupt())) return __memcpy(to, from, len); @@ -39,44 +40,72 @@ void *_mmx_memcpy(void *to, const void *from, size_t len) kernel_fpu_begin(); __asm__ __volatile__ ( - "1: prefetch (%0)\n" /* This set is 28 bytes */ - " prefetch 64(%0)\n" - " prefetch 128(%0)\n" - " prefetch 192(%0)\n" - " prefetch 256(%0)\n" + "1: prefetch (%1)\n" /* This set is 28 bytes */ + " prefetch 64(%1)\n" + " prefetch 128(%1)\n" + " prefetch 192(%1)\n" + " prefetch 256(%1)\n" "2: \n" ".section .fixup, \"ax\"\n" - "3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ + "3: \n" + +#ifdef CONFIG_PAX_KERNEXEC + " movl %%cr0, %0\n" + " movl %0, %%eax\n" + " andl $0xFFFEFFFF, %%eax\n" + " movl %%eax, %%cr0\n" +#endif + + " movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ + +#ifdef CONFIG_PAX_KERNEXEC + " movl %0, %%cr0\n" +#endif + " jmp 2b\n" ".previous\n" _ASM_EXTABLE(1b, 3b) - : : "r" (from)); + : "=&r" (cr0) : "r" (from) : "ax"); for ( ; i > 5; i--) { __asm__ __volatile__ ( - "1: prefetch 320(%0)\n" - "2: movq (%0), %%mm0\n" - " movq 8(%0), %%mm1\n" - " movq 16(%0), %%mm2\n" - " movq 24(%0), %%mm3\n" - " movq %%mm0, (%1)\n" - " movq %%mm1, 8(%1)\n" - " movq %%mm2, 16(%1)\n" - " movq %%mm3, 24(%1)\n" - " movq 32(%0), %%mm0\n" - " movq 40(%0), %%mm1\n" - " movq 48(%0), %%mm2\n" - " movq 56(%0), %%mm3\n" - " movq %%mm0, 32(%1)\n" - " movq %%mm1, 40(%1)\n" - " movq %%mm2, 48(%1)\n" - " movq %%mm3, 56(%1)\n" + "1: prefetch 320(%1)\n" + "2: movq (%1), %%mm0\n" + " movq 8(%1), %%mm1\n" + " movq 16(%1), %%mm2\n" + " movq 24(%1), %%mm3\n" + " movq %%mm0, (%2)\n" + " movq %%mm1, 8(%2)\n" + " movq %%mm2, 16(%2)\n" + " movq %%mm3, 24(%2)\n" + " movq 32(%1), %%mm0\n" + " movq 40(%1), %%mm1\n" + " movq 48(%1), %%mm2\n" + " movq 56(%1), %%mm3\n" + " movq %%mm0, 32(%2)\n" + " movq %%mm1, 40(%2)\n" + " movq %%mm2, 48(%2)\n" + " movq %%mm3, 56(%2)\n" ".section .fixup, \"ax\"\n" - "3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */ + "3:\n" + +#ifdef CONFIG_PAX_KERNEXEC + " movl %%cr0, %0\n" + " movl %0, %%eax\n" + " andl $0xFFFEFFFF, %%eax\n" + " movl %%eax, %%cr0\n" +#endif + + " movw $0x05EB, 1b\n" /* jmp on 5 bytes */ + +#ifdef CONFIG_PAX_KERNEXEC + " movl %0, %%cr0\n" +#endif + " jmp 2b\n" ".previous\n" _ASM_EXTABLE(1b, 3b) - : : "r" (from), "r" (to) : "memory"); + : "=&r" (cr0) : "r" (from), "r" (to) : "memory", "ax"); from += 64; to += 64; @@ -158,6 +187,7 @@ static void fast_clear_page(void *page) static void fast_copy_page(void *to, void *from) { int i; + unsigned long cr0; kernel_fpu_begin(); @@ -166,42 +196,70 @@ static void fast_copy_page(void *to, void *from) * but that is for later. -AV */ __asm__ __volatile__( - "1: prefetch (%0)\n" - " prefetch 64(%0)\n" - " prefetch 128(%0)\n" - " prefetch 192(%0)\n" - " prefetch 256(%0)\n" + "1: prefetch (%1)\n" + " prefetch 64(%1)\n" + " prefetch 128(%1)\n" + " prefetch 192(%1)\n" + " prefetch 256(%1)\n" "2: \n" ".section .fixup, \"ax\"\n" - "3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ + "3: \n" + +#ifdef CONFIG_PAX_KERNEXEC + " movl %%cr0, %0\n" + " movl %0, %%eax\n" + " andl $0xFFFEFFFF, %%eax\n" + " movl %%eax, %%cr0\n" +#endif + + " movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ + +#ifdef CONFIG_PAX_KERNEXEC + " movl %0, %%cr0\n" +#endif + " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b, 3b) : : "r" (from)); + _ASM_EXTABLE(1b, 3b) : "=&r" (cr0) : "r" (from) : "ax"); for (i = 0; i < (4096-320)/64; i++) { __asm__ __volatile__ ( - "1: prefetch 320(%0)\n" - "2: movq (%0), %%mm0\n" - " movntq %%mm0, (%1)\n" - " movq 8(%0), %%mm1\n" - " movntq %%mm1, 8(%1)\n" - " movq 16(%0), %%mm2\n" - " movntq %%mm2, 16(%1)\n" - " movq 24(%0), %%mm3\n" - " movntq %%mm3, 24(%1)\n" - " movq 32(%0), %%mm4\n" - " movntq %%mm4, 32(%1)\n" - " movq 40(%0), %%mm5\n" - " movntq %%mm5, 40(%1)\n" - " movq 48(%0), %%mm6\n" - " movntq %%mm6, 48(%1)\n" - " movq 56(%0), %%mm7\n" - " movntq %%mm7, 56(%1)\n" + "1: prefetch 320(%1)\n" + "2: movq (%1), %%mm0\n" + " movntq %%mm0, (%2)\n" + " movq 8(%1), %%mm1\n" + " movntq %%mm1, 8(%2)\n" + " movq 16(%1), %%mm2\n" + " movntq %%mm2, 16(%2)\n" + " movq 24(%1), %%mm3\n" + " movntq %%mm3, 24(%2)\n" + " movq 32(%1), %%mm4\n" + " movntq %%mm4, 32(%2)\n" + " movq 40(%1), %%mm5\n" + " movntq %%mm5, 40(%2)\n" + " movq 48(%1), %%mm6\n" + " movntq %%mm6, 48(%2)\n" + " movq 56(%1), %%mm7\n" + " movntq %%mm7, 56(%2)\n" ".section .fixup, \"ax\"\n" - "3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */ + "3:\n" + +#ifdef CONFIG_PAX_KERNEXEC + " movl %%cr0, %0\n" + " movl %0, %%eax\n" + " andl $0xFFFEFFFF, %%eax\n" + " movl %%eax, %%cr0\n" +#endif + + " movw $0x05EB, 1b\n" /* jmp on 5 bytes */ + +#ifdef CONFIG_PAX_KERNEXEC + " movl %0, %%cr0\n" +#endif + " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b, 3b) : : "r" (from), "r" (to) : "memory"); + _ASM_EXTABLE(1b, 3b) : "=&r" (cr0) : "r" (from), "r" (to) : "memory", "ax"); from += 64; to += 64; @@ -280,47 +338,76 @@ static void fast_clear_page(void *page) static void fast_copy_page(void *to, void *from) { int i; + unsigned long cr0; kernel_fpu_begin(); __asm__ __volatile__ ( - "1: prefetch (%0)\n" - " prefetch 64(%0)\n" - " prefetch 128(%0)\n" - " prefetch 192(%0)\n" - " prefetch 256(%0)\n" + "1: prefetch (%1)\n" + " prefetch 64(%1)\n" + " prefetch 128(%1)\n" + " prefetch 192(%1)\n" + " prefetch 256(%1)\n" "2: \n" ".section .fixup, \"ax\"\n" - "3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ + "3: \n" + +#ifdef CONFIG_PAX_KERNEXEC + " movl %%cr0, %0\n" + " movl %0, %%eax\n" + " andl $0xFFFEFFFF, %%eax\n" + " movl %%eax, %%cr0\n" +#endif + + " movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ + +#ifdef CONFIG_PAX_KERNEXEC + " movl %0, %%cr0\n" +#endif + " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b, 3b) : : "r" (from)); + _ASM_EXTABLE(1b, 3b) : "=&r" (cr0) : "r" (from) : "ax"); for (i = 0; i < 4096/64; i++) { __asm__ __volatile__ ( - "1: prefetch 320(%0)\n" - "2: movq (%0), %%mm0\n" - " movq 8(%0), %%mm1\n" - " movq 16(%0), %%mm2\n" - " movq 24(%0), %%mm3\n" - " movq %%mm0, (%1)\n" - " movq %%mm1, 8(%1)\n" - " movq %%mm2, 16(%1)\n" - " movq %%mm3, 24(%1)\n" - " movq 32(%0), %%mm0\n" - " movq 40(%0), %%mm1\n" - " movq 48(%0), %%mm2\n" - " movq 56(%0), %%mm3\n" - " movq %%mm0, 32(%1)\n" - " movq %%mm1, 40(%1)\n" - " movq %%mm2, 48(%1)\n" - " movq %%mm3, 56(%1)\n" + "1: prefetch 320(%1)\n" + "2: movq (%1), %%mm0\n" + " movq 8(%1), %%mm1\n" + " movq 16(%1), %%mm2\n" + " movq 24(%1), %%mm3\n" + " movq %%mm0, (%2)\n" + " movq %%mm1, 8(%2)\n" + " movq %%mm2, 16(%2)\n" + " movq %%mm3, 24(%2)\n" + " movq 32(%1), %%mm0\n" + " movq 40(%1), %%mm1\n" + " movq 48(%1), %%mm2\n" + " movq 56(%1), %%mm3\n" + " movq %%mm0, 32(%2)\n" + " movq %%mm1, 40(%2)\n" + " movq %%mm2, 48(%2)\n" + " movq %%mm3, 56(%2)\n" ".section .fixup, \"ax\"\n" - "3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */ + "3:\n" + +#ifdef CONFIG_PAX_KERNEXEC + " movl %%cr0, %0\n" + " movl %0, %%eax\n" + " andl $0xFFFEFFFF, %%eax\n" + " movl %%eax, %%cr0\n" +#endif + + " movw $0x05EB, 1b\n" /* jmp on 5 bytes */ + +#ifdef CONFIG_PAX_KERNEXEC + " movl %0, %%cr0\n" +#endif + " jmp 2b\n" ".previous\n" _ASM_EXTABLE(1b, 3b) - : : "r" (from), "r" (to) : "memory"); + : "=&r" (cr0) : "r" (from), "r" (to) : "memory", "ax"); from += 64; to += 64; diff --git a/arch/x86/lib/msr-reg.S b/arch/x86/lib/msr-reg.S index 69fa106..adda88b 100644 --- a/arch/x86/lib/msr-reg.S +++ b/arch/x86/lib/msr-reg.S @@ -3,6 +3,7 @@ #include #include #include +#include #ifdef CONFIG_X86_64 /* @@ -16,7 +17,7 @@ ENTRY(native_\op\()_safe_regs) CFI_STARTPROC pushq_cfi %rbx pushq_cfi %rbp - movq %rdi, %r10 /* Save pointer */ + movq %rdi, %r9 /* Save pointer */ xorl %r11d, %r11d /* Return value */ movl (%rdi), %eax movl 4(%rdi), %ecx @@ -27,16 +28,17 @@ ENTRY(native_\op\()_safe_regs) movl 28(%rdi), %edi CFI_REMEMBER_STATE 1: \op -2: movl %eax, (%r10) +2: movl %eax, (%r9) movl %r11d, %eax /* Return value */ - movl %ecx, 4(%r10) - movl %edx, 8(%r10) - movl %ebx, 12(%r10) - movl %ebp, 20(%r10) - movl %esi, 24(%r10) - movl %edi, 28(%r10) + movl %ecx, 4(%r9) + movl %edx, 8(%r9) + movl %ebx, 12(%r9) + movl %ebp, 20(%r9) + movl %esi, 24(%r9) + movl %edi, 28(%r9) popq_cfi %rbp popq_cfi %rbx + pax_force_retaddr ret 3: CFI_RESTORE_STATE diff --git a/arch/x86/lib/putuser.S b/arch/x86/lib/putuser.S index 36b0d15..d381858 100644 --- a/arch/x86/lib/putuser.S +++ b/arch/x86/lib/putuser.S @@ -15,7 +15,9 @@ #include #include #include - +#include +#include +#include /* * __put_user_X @@ -29,52 +31,119 @@ * as they get called from within inline assembly. */ -#define ENTER CFI_STARTPROC ; \ - GET_THREAD_INFO(%_ASM_BX) -#define EXIT ret ; \ +#define ENTER CFI_STARTPROC +#define EXIT pax_force_retaddr; ret ; \ CFI_ENDPROC +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define _DEST %_ASM_CX,%_ASM_BX +#else +#define _DEST %_ASM_CX +#endif + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define __copyuser_seg gs; +#else +#define __copyuser_seg +#endif + .text ENTRY(__put_user_1) ENTER + +#if !defined(CONFIG_X86_32) || !defined(CONFIG_PAX_MEMORY_UDEREF) + GET_THREAD_INFO(%_ASM_BX) cmp TI_addr_limit(%_ASM_BX),%_ASM_CX jae bad_put_user -1: movb %al,(%_ASM_CX) + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov $PAX_USER_SHADOW_BASE,%_ASM_BX + cmp %_ASM_BX,%_ASM_CX + jb 1234f + xor %ebx,%ebx +1234: +#endif + +#endif + +1: __copyuser_seg movb %al,(_DEST) xor %eax,%eax EXIT ENDPROC(__put_user_1) ENTRY(__put_user_2) ENTER + +#if !defined(CONFIG_X86_32) || !defined(CONFIG_PAX_MEMORY_UDEREF) + GET_THREAD_INFO(%_ASM_BX) mov TI_addr_limit(%_ASM_BX),%_ASM_BX sub $1,%_ASM_BX cmp %_ASM_BX,%_ASM_CX jae bad_put_user -2: movw %ax,(%_ASM_CX) + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov $PAX_USER_SHADOW_BASE,%_ASM_BX + cmp %_ASM_BX,%_ASM_CX + jb 1234f + xor %ebx,%ebx +1234: +#endif + +#endif + +2: __copyuser_seg movw %ax,(_DEST) xor %eax,%eax EXIT ENDPROC(__put_user_2) ENTRY(__put_user_4) ENTER + +#if !defined(CONFIG_X86_32) || !defined(CONFIG_PAX_MEMORY_UDEREF) + GET_THREAD_INFO(%_ASM_BX) mov TI_addr_limit(%_ASM_BX),%_ASM_BX sub $3,%_ASM_BX cmp %_ASM_BX,%_ASM_CX jae bad_put_user -3: movl %eax,(%_ASM_CX) + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov $PAX_USER_SHADOW_BASE,%_ASM_BX + cmp %_ASM_BX,%_ASM_CX + jb 1234f + xor %ebx,%ebx +1234: +#endif + +#endif + +3: __copyuser_seg movl %eax,(_DEST) xor %eax,%eax EXIT ENDPROC(__put_user_4) ENTRY(__put_user_8) ENTER + +#if !defined(CONFIG_X86_32) || !defined(CONFIG_PAX_MEMORY_UDEREF) + GET_THREAD_INFO(%_ASM_BX) mov TI_addr_limit(%_ASM_BX),%_ASM_BX sub $7,%_ASM_BX cmp %_ASM_BX,%_ASM_CX jae bad_put_user -4: mov %_ASM_AX,(%_ASM_CX) + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov $PAX_USER_SHADOW_BASE,%_ASM_BX + cmp %_ASM_BX,%_ASM_CX + jb 1234f + xor %ebx,%ebx +1234: +#endif + +#endif + +4: __copyuser_seg mov %_ASM_AX,(_DEST) #ifdef CONFIG_X86_32 -5: movl %edx,4(%_ASM_CX) +5: __copyuser_seg movl %edx,4(_DEST) #endif xor %eax,%eax EXIT diff --git a/arch/x86/lib/rwlock.S b/arch/x86/lib/rwlock.S index 1cad221..de671ee 100644 --- a/arch/x86/lib/rwlock.S +++ b/arch/x86/lib/rwlock.S @@ -16,13 +16,34 @@ ENTRY(__write_lock_failed) FRAME 0: LOCK_PREFIX WRITE_LOCK_ADD($RW_LOCK_BIAS) (%__lock_ptr) + +#ifdef CONFIG_PAX_REFCOUNT + jno 1234f + LOCK_PREFIX + WRITE_LOCK_SUB($RW_LOCK_BIAS) (%__lock_ptr) + int $4 +1234: + _ASM_EXTABLE(1234b, 1234b) +#endif + 1: rep; nop cmpl $WRITE_LOCK_CMP, (%__lock_ptr) jne 1b LOCK_PREFIX WRITE_LOCK_SUB($RW_LOCK_BIAS) (%__lock_ptr) + +#ifdef CONFIG_PAX_REFCOUNT + jno 1234f + LOCK_PREFIX + WRITE_LOCK_ADD($RW_LOCK_BIAS) (%__lock_ptr) + int $4 +1234: + _ASM_EXTABLE(1234b, 1234b) +#endif + jnz 0b ENDFRAME + pax_force_retaddr ret CFI_ENDPROC END(__write_lock_failed) @@ -32,13 +53,34 @@ ENTRY(__read_lock_failed) FRAME 0: LOCK_PREFIX READ_LOCK_SIZE(inc) (%__lock_ptr) + +#ifdef CONFIG_PAX_REFCOUNT + jno 1234f + LOCK_PREFIX + READ_LOCK_SIZE(dec) (%__lock_ptr) + int $4 +1234: + _ASM_EXTABLE(1234b, 1234b) +#endif + 1: rep; nop READ_LOCK_SIZE(cmp) $1, (%__lock_ptr) js 1b LOCK_PREFIX READ_LOCK_SIZE(dec) (%__lock_ptr) + +#ifdef CONFIG_PAX_REFCOUNT + jno 1234f + LOCK_PREFIX + READ_LOCK_SIZE(inc) (%__lock_ptr) + int $4 +1234: + _ASM_EXTABLE(1234b, 1234b) +#endif + js 0b ENDFRAME + pax_force_retaddr ret CFI_ENDPROC END(__read_lock_failed) diff --git a/arch/x86/lib/rwsem.S b/arch/x86/lib/rwsem.S index 5dff5f0..cadebf4 100644 --- a/arch/x86/lib/rwsem.S +++ b/arch/x86/lib/rwsem.S @@ -94,6 +94,7 @@ ENTRY(call_rwsem_down_read_failed) __ASM_SIZE(pop,_cfi) %__ASM_REG(dx) CFI_RESTORE __ASM_REG(dx) restore_common_regs + pax_force_retaddr ret CFI_ENDPROC ENDPROC(call_rwsem_down_read_failed) @@ -104,6 +105,7 @@ ENTRY(call_rwsem_down_write_failed) movq %rax,%rdi call rwsem_down_write_failed restore_common_regs + pax_force_retaddr ret CFI_ENDPROC ENDPROC(call_rwsem_down_write_failed) @@ -117,7 +119,8 @@ ENTRY(call_rwsem_wake) movq %rax,%rdi call rwsem_wake restore_common_regs -1: ret +1: pax_force_retaddr + ret CFI_ENDPROC ENDPROC(call_rwsem_wake) @@ -131,6 +134,7 @@ ENTRY(call_rwsem_downgrade_wake) __ASM_SIZE(pop,_cfi) %__ASM_REG(dx) CFI_RESTORE __ASM_REG(dx) restore_common_regs + pax_force_retaddr ret CFI_ENDPROC ENDPROC(call_rwsem_downgrade_wake) diff --git a/arch/x86/lib/thunk_64.S b/arch/x86/lib/thunk_64.S index a63efd6..ccecad8 100644 --- a/arch/x86/lib/thunk_64.S +++ b/arch/x86/lib/thunk_64.S @@ -8,6 +8,7 @@ #include #include #include +#include /* rdi: arg1 ... normal C conventions. rax is saved/restored. */ .macro THUNK name, func, put_ret_addr_in_rdi=0 @@ -41,5 +42,6 @@ SAVE_ARGS restore: RESTORE_ARGS + pax_force_retaddr ret CFI_ENDPROC diff --git a/arch/x86/lib/usercopy_32.c b/arch/x86/lib/usercopy_32.c index e218d5d..a99a1eb 100644 --- a/arch/x86/lib/usercopy_32.c +++ b/arch/x86/lib/usercopy_32.c @@ -43,7 +43,7 @@ do { \ __asm__ __volatile__( \ " testl %1,%1\n" \ " jz 2f\n" \ - "0: lodsb\n" \ + "0: "__copyuser_seg"lodsb\n" \ " stosb\n" \ " testb %%al,%%al\n" \ " jz 1f\n" \ @@ -128,10 +128,12 @@ do { \ int __d0; \ might_fault(); \ __asm__ __volatile__( \ + __COPYUSER_SET_ES \ "0: rep; stosl\n" \ " movl %2,%0\n" \ "1: rep; stosb\n" \ "2:\n" \ + __COPYUSER_RESTORE_ES \ ".section .fixup,\"ax\"\n" \ "3: lea 0(%2,%0,4),%0\n" \ " jmp 2b\n" \ @@ -200,6 +202,7 @@ long strnlen_user(const char __user *s, long n) might_fault(); __asm__ __volatile__( + __COPYUSER_SET_ES " testl %0, %0\n" " jz 3f\n" " andl %0,%%ecx\n" @@ -208,6 +211,7 @@ long strnlen_user(const char __user *s, long n) " subl %%ecx,%0\n" " addl %0,%%eax\n" "1:\n" + __COPYUSER_RESTORE_ES ".section .fixup,\"ax\"\n" "2: xorl %%eax,%%eax\n" " jmp 1b\n" @@ -227,7 +231,7 @@ EXPORT_SYMBOL(strnlen_user); #ifdef CONFIG_X86_INTEL_USERCOPY static unsigned long -__copy_user_intel(void __user *to, const void *from, unsigned long size) +__generic_copy_to_user_intel(void __user *to, const void *from, unsigned long size) { int d0, d1; __asm__ __volatile__( @@ -239,36 +243,36 @@ __copy_user_intel(void __user *to, const void *from, unsigned long size) " .align 2,0x90\n" "3: movl 0(%4), %%eax\n" "4: movl 4(%4), %%edx\n" - "5: movl %%eax, 0(%3)\n" - "6: movl %%edx, 4(%3)\n" + "5: "__copyuser_seg" movl %%eax, 0(%3)\n" + "6: "__copyuser_seg" movl %%edx, 4(%3)\n" "7: movl 8(%4), %%eax\n" "8: movl 12(%4),%%edx\n" - "9: movl %%eax, 8(%3)\n" - "10: movl %%edx, 12(%3)\n" + "9: "__copyuser_seg" movl %%eax, 8(%3)\n" + "10: "__copyuser_seg" movl %%edx, 12(%3)\n" "11: movl 16(%4), %%eax\n" "12: movl 20(%4), %%edx\n" - "13: movl %%eax, 16(%3)\n" - "14: movl %%edx, 20(%3)\n" + "13: "__copyuser_seg" movl %%eax, 16(%3)\n" + "14: "__copyuser_seg" movl %%edx, 20(%3)\n" "15: movl 24(%4), %%eax\n" "16: movl 28(%4), %%edx\n" - "17: movl %%eax, 24(%3)\n" - "18: movl %%edx, 28(%3)\n" + "17: "__copyuser_seg" movl %%eax, 24(%3)\n" + "18: "__copyuser_seg" movl %%edx, 28(%3)\n" "19: movl 32(%4), %%eax\n" "20: movl 36(%4), %%edx\n" - "21: movl %%eax, 32(%3)\n" - "22: movl %%edx, 36(%3)\n" + "21: "__copyuser_seg" movl %%eax, 32(%3)\n" + "22: "__copyuser_seg" movl %%edx, 36(%3)\n" "23: movl 40(%4), %%eax\n" "24: movl 44(%4), %%edx\n" - "25: movl %%eax, 40(%3)\n" - "26: movl %%edx, 44(%3)\n" + "25: "__copyuser_seg" movl %%eax, 40(%3)\n" + "26: "__copyuser_seg" movl %%edx, 44(%3)\n" "27: movl 48(%4), %%eax\n" "28: movl 52(%4), %%edx\n" - "29: movl %%eax, 48(%3)\n" - "30: movl %%edx, 52(%3)\n" + "29: "__copyuser_seg" movl %%eax, 48(%3)\n" + "30: "__copyuser_seg" movl %%edx, 52(%3)\n" "31: movl 56(%4), %%eax\n" "32: movl 60(%4), %%edx\n" - "33: movl %%eax, 56(%3)\n" - "34: movl %%edx, 60(%3)\n" + "33: "__copyuser_seg" movl %%eax, 56(%3)\n" + "34: "__copyuser_seg" movl %%edx, 60(%3)\n" " addl $-64, %0\n" " addl $64, %4\n" " addl $64, %3\n" @@ -278,10 +282,12 @@ __copy_user_intel(void __user *to, const void *from, unsigned long size) " shrl $2, %0\n" " andl $3, %%eax\n" " cld\n" + __COPYUSER_SET_ES "99: rep; movsl\n" "36: movl %%eax, %0\n" "37: rep; movsb\n" "100:\n" + __COPYUSER_RESTORE_ES ".section .fixup,\"ax\"\n" "101: lea 0(%%eax,%0,4),%0\n" " jmp 100b\n" @@ -334,46 +340,155 @@ __copy_user_intel(void __user *to, const void *from, unsigned long size) } static unsigned long +__generic_copy_from_user_intel(void *to, const void __user *from, unsigned long size) +{ + int d0, d1; + __asm__ __volatile__( + " .align 2,0x90\n" + "1: "__copyuser_seg" movl 32(%4), %%eax\n" + " cmpl $67, %0\n" + " jbe 3f\n" + "2: "__copyuser_seg" movl 64(%4), %%eax\n" + " .align 2,0x90\n" + "3: "__copyuser_seg" movl 0(%4), %%eax\n" + "4: "__copyuser_seg" movl 4(%4), %%edx\n" + "5: movl %%eax, 0(%3)\n" + "6: movl %%edx, 4(%3)\n" + "7: "__copyuser_seg" movl 8(%4), %%eax\n" + "8: "__copyuser_seg" movl 12(%4),%%edx\n" + "9: movl %%eax, 8(%3)\n" + "10: movl %%edx, 12(%3)\n" + "11: "__copyuser_seg" movl 16(%4), %%eax\n" + "12: "__copyuser_seg" movl 20(%4), %%edx\n" + "13: movl %%eax, 16(%3)\n" + "14: movl %%edx, 20(%3)\n" + "15: "__copyuser_seg" movl 24(%4), %%eax\n" + "16: "__copyuser_seg" movl 28(%4), %%edx\n" + "17: movl %%eax, 24(%3)\n" + "18: movl %%edx, 28(%3)\n" + "19: "__copyuser_seg" movl 32(%4), %%eax\n" + "20: "__copyuser_seg" movl 36(%4), %%edx\n" + "21: movl %%eax, 32(%3)\n" + "22: movl %%edx, 36(%3)\n" + "23: "__copyuser_seg" movl 40(%4), %%eax\n" + "24: "__copyuser_seg" movl 44(%4), %%edx\n" + "25: movl %%eax, 40(%3)\n" + "26: movl %%edx, 44(%3)\n" + "27: "__copyuser_seg" movl 48(%4), %%eax\n" + "28: "__copyuser_seg" movl 52(%4), %%edx\n" + "29: movl %%eax, 48(%3)\n" + "30: movl %%edx, 52(%3)\n" + "31: "__copyuser_seg" movl 56(%4), %%eax\n" + "32: "__copyuser_seg" movl 60(%4), %%edx\n" + "33: movl %%eax, 56(%3)\n" + "34: movl %%edx, 60(%3)\n" + " addl $-64, %0\n" + " addl $64, %4\n" + " addl $64, %3\n" + " cmpl $63, %0\n" + " ja 1b\n" + "35: movl %0, %%eax\n" + " shrl $2, %0\n" + " andl $3, %%eax\n" + " cld\n" + "99: rep; "__copyuser_seg" movsl\n" + "36: movl %%eax, %0\n" + "37: rep; "__copyuser_seg" movsb\n" + "100:\n" + ".section .fixup,\"ax\"\n" + "101: lea 0(%%eax,%0,4),%0\n" + " jmp 100b\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .align 4\n" + " .long 1b,100b\n" + " .long 2b,100b\n" + " .long 3b,100b\n" + " .long 4b,100b\n" + " .long 5b,100b\n" + " .long 6b,100b\n" + " .long 7b,100b\n" + " .long 8b,100b\n" + " .long 9b,100b\n" + " .long 10b,100b\n" + " .long 11b,100b\n" + " .long 12b,100b\n" + " .long 13b,100b\n" + " .long 14b,100b\n" + " .long 15b,100b\n" + " .long 16b,100b\n" + " .long 17b,100b\n" + " .long 18b,100b\n" + " .long 19b,100b\n" + " .long 20b,100b\n" + " .long 21b,100b\n" + " .long 22b,100b\n" + " .long 23b,100b\n" + " .long 24b,100b\n" + " .long 25b,100b\n" + " .long 26b,100b\n" + " .long 27b,100b\n" + " .long 28b,100b\n" + " .long 29b,100b\n" + " .long 30b,100b\n" + " .long 31b,100b\n" + " .long 32b,100b\n" + " .long 33b,100b\n" + " .long 34b,100b\n" + " .long 35b,100b\n" + " .long 36b,100b\n" + " .long 37b,100b\n" + " .long 99b,101b\n" + ".previous" + : "=&c"(size), "=&D" (d0), "=&S" (d1) + : "1"(to), "2"(from), "0"(size) + : "eax", "edx", "memory"); + return size; +} + +static unsigned long +__copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) __size_overflow(3); +static unsigned long __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) { int d0, d1; __asm__ __volatile__( " .align 2,0x90\n" - "0: movl 32(%4), %%eax\n" + "0: "__copyuser_seg" movl 32(%4), %%eax\n" " cmpl $67, %0\n" " jbe 2f\n" - "1: movl 64(%4), %%eax\n" + "1: "__copyuser_seg" movl 64(%4), %%eax\n" " .align 2,0x90\n" - "2: movl 0(%4), %%eax\n" - "21: movl 4(%4), %%edx\n" + "2: "__copyuser_seg" movl 0(%4), %%eax\n" + "21: "__copyuser_seg" movl 4(%4), %%edx\n" " movl %%eax, 0(%3)\n" " movl %%edx, 4(%3)\n" - "3: movl 8(%4), %%eax\n" - "31: movl 12(%4),%%edx\n" + "3: "__copyuser_seg" movl 8(%4), %%eax\n" + "31: "__copyuser_seg" movl 12(%4),%%edx\n" " movl %%eax, 8(%3)\n" " movl %%edx, 12(%3)\n" - "4: movl 16(%4), %%eax\n" - "41: movl 20(%4), %%edx\n" + "4: "__copyuser_seg" movl 16(%4), %%eax\n" + "41: "__copyuser_seg" movl 20(%4), %%edx\n" " movl %%eax, 16(%3)\n" " movl %%edx, 20(%3)\n" - "10: movl 24(%4), %%eax\n" - "51: movl 28(%4), %%edx\n" + "10: "__copyuser_seg" movl 24(%4), %%eax\n" + "51: "__copyuser_seg" movl 28(%4), %%edx\n" " movl %%eax, 24(%3)\n" " movl %%edx, 28(%3)\n" - "11: movl 32(%4), %%eax\n" - "61: movl 36(%4), %%edx\n" + "11: "__copyuser_seg" movl 32(%4), %%eax\n" + "61: "__copyuser_seg" movl 36(%4), %%edx\n" " movl %%eax, 32(%3)\n" " movl %%edx, 36(%3)\n" - "12: movl 40(%4), %%eax\n" - "71: movl 44(%4), %%edx\n" + "12: "__copyuser_seg" movl 40(%4), %%eax\n" + "71: "__copyuser_seg" movl 44(%4), %%edx\n" " movl %%eax, 40(%3)\n" " movl %%edx, 44(%3)\n" - "13: movl 48(%4), %%eax\n" - "81: movl 52(%4), %%edx\n" + "13: "__copyuser_seg" movl 48(%4), %%eax\n" + "81: "__copyuser_seg" movl 52(%4), %%edx\n" " movl %%eax, 48(%3)\n" " movl %%edx, 52(%3)\n" - "14: movl 56(%4), %%eax\n" - "91: movl 60(%4), %%edx\n" + "14: "__copyuser_seg" movl 56(%4), %%eax\n" + "91: "__copyuser_seg" movl 60(%4), %%edx\n" " movl %%eax, 56(%3)\n" " movl %%edx, 60(%3)\n" " addl $-64, %0\n" @@ -385,9 +500,9 @@ __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) " shrl $2, %0\n" " andl $3, %%eax\n" " cld\n" - "6: rep; movsl\n" + "6: rep; "__copyuser_seg" movsl\n" " movl %%eax,%0\n" - "7: rep; movsb\n" + "7: rep; "__copyuser_seg" movsb\n" "8:\n" ".section .fixup,\"ax\"\n" "9: lea 0(%%eax,%0,4),%0\n" @@ -434,47 +549,49 @@ __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) */ static unsigned long __copy_user_zeroing_intel_nocache(void *to, + const void __user *from, unsigned long size) __size_overflow(3); +static unsigned long __copy_user_zeroing_intel_nocache(void *to, const void __user *from, unsigned long size) { int d0, d1; __asm__ __volatile__( " .align 2,0x90\n" - "0: movl 32(%4), %%eax\n" + "0: "__copyuser_seg" movl 32(%4), %%eax\n" " cmpl $67, %0\n" " jbe 2f\n" - "1: movl 64(%4), %%eax\n" + "1: "__copyuser_seg" movl 64(%4), %%eax\n" " .align 2,0x90\n" - "2: movl 0(%4), %%eax\n" - "21: movl 4(%4), %%edx\n" + "2: "__copyuser_seg" movl 0(%4), %%eax\n" + "21: "__copyuser_seg" movl 4(%4), %%edx\n" " movnti %%eax, 0(%3)\n" " movnti %%edx, 4(%3)\n" - "3: movl 8(%4), %%eax\n" - "31: movl 12(%4),%%edx\n" + "3: "__copyuser_seg" movl 8(%4), %%eax\n" + "31: "__copyuser_seg" movl 12(%4),%%edx\n" " movnti %%eax, 8(%3)\n" " movnti %%edx, 12(%3)\n" - "4: movl 16(%4), %%eax\n" - "41: movl 20(%4), %%edx\n" + "4: "__copyuser_seg" movl 16(%4), %%eax\n" + "41: "__copyuser_seg" movl 20(%4), %%edx\n" " movnti %%eax, 16(%3)\n" " movnti %%edx, 20(%3)\n" - "10: movl 24(%4), %%eax\n" - "51: movl 28(%4), %%edx\n" + "10: "__copyuser_seg" movl 24(%4), %%eax\n" + "51: "__copyuser_seg" movl 28(%4), %%edx\n" " movnti %%eax, 24(%3)\n" " movnti %%edx, 28(%3)\n" - "11: movl 32(%4), %%eax\n" - "61: movl 36(%4), %%edx\n" + "11: "__copyuser_seg" movl 32(%4), %%eax\n" + "61: "__copyuser_seg" movl 36(%4), %%edx\n" " movnti %%eax, 32(%3)\n" " movnti %%edx, 36(%3)\n" - "12: movl 40(%4), %%eax\n" - "71: movl 44(%4), %%edx\n" + "12: "__copyuser_seg" movl 40(%4), %%eax\n" + "71: "__copyuser_seg" movl 44(%4), %%edx\n" " movnti %%eax, 40(%3)\n" " movnti %%edx, 44(%3)\n" - "13: movl 48(%4), %%eax\n" - "81: movl 52(%4), %%edx\n" + "13: "__copyuser_seg" movl 48(%4), %%eax\n" + "81: "__copyuser_seg" movl 52(%4), %%edx\n" " movnti %%eax, 48(%3)\n" " movnti %%edx, 52(%3)\n" - "14: movl 56(%4), %%eax\n" - "91: movl 60(%4), %%edx\n" + "14: "__copyuser_seg" movl 56(%4), %%eax\n" + "91: "__copyuser_seg" movl 60(%4), %%edx\n" " movnti %%eax, 56(%3)\n" " movnti %%edx, 60(%3)\n" " addl $-64, %0\n" @@ -487,9 +604,9 @@ static unsigned long __copy_user_zeroing_intel_nocache(void *to, " shrl $2, %0\n" " andl $3, %%eax\n" " cld\n" - "6: rep; movsl\n" + "6: rep; "__copyuser_seg" movsl\n" " movl %%eax,%0\n" - "7: rep; movsb\n" + "7: rep; "__copyuser_seg" movsb\n" "8:\n" ".section .fixup,\"ax\"\n" "9: lea 0(%%eax,%0,4),%0\n" @@ -531,47 +648,49 @@ static unsigned long __copy_user_zeroing_intel_nocache(void *to, } static unsigned long __copy_user_intel_nocache(void *to, + const void __user *from, unsigned long size) __size_overflow(3); +static unsigned long __copy_user_intel_nocache(void *to, const void __user *from, unsigned long size) { int d0, d1; __asm__ __volatile__( " .align 2,0x90\n" - "0: movl 32(%4), %%eax\n" + "0: "__copyuser_seg" movl 32(%4), %%eax\n" " cmpl $67, %0\n" " jbe 2f\n" - "1: movl 64(%4), %%eax\n" + "1: "__copyuser_seg" movl 64(%4), %%eax\n" " .align 2,0x90\n" - "2: movl 0(%4), %%eax\n" - "21: movl 4(%4), %%edx\n" + "2: "__copyuser_seg" movl 0(%4), %%eax\n" + "21: "__copyuser_seg" movl 4(%4), %%edx\n" " movnti %%eax, 0(%3)\n" " movnti %%edx, 4(%3)\n" - "3: movl 8(%4), %%eax\n" - "31: movl 12(%4),%%edx\n" + "3: "__copyuser_seg" movl 8(%4), %%eax\n" + "31: "__copyuser_seg" movl 12(%4),%%edx\n" " movnti %%eax, 8(%3)\n" " movnti %%edx, 12(%3)\n" - "4: movl 16(%4), %%eax\n" - "41: movl 20(%4), %%edx\n" + "4: "__copyuser_seg" movl 16(%4), %%eax\n" + "41: "__copyuser_seg" movl 20(%4), %%edx\n" " movnti %%eax, 16(%3)\n" " movnti %%edx, 20(%3)\n" - "10: movl 24(%4), %%eax\n" - "51: movl 28(%4), %%edx\n" + "10: "__copyuser_seg" movl 24(%4), %%eax\n" + "51: "__copyuser_seg" movl 28(%4), %%edx\n" " movnti %%eax, 24(%3)\n" " movnti %%edx, 28(%3)\n" - "11: movl 32(%4), %%eax\n" - "61: movl 36(%4), %%edx\n" + "11: "__copyuser_seg" movl 32(%4), %%eax\n" + "61: "__copyuser_seg" movl 36(%4), %%edx\n" " movnti %%eax, 32(%3)\n" " movnti %%edx, 36(%3)\n" - "12: movl 40(%4), %%eax\n" - "71: movl 44(%4), %%edx\n" + "12: "__copyuser_seg" movl 40(%4), %%eax\n" + "71: "__copyuser_seg" movl 44(%4), %%edx\n" " movnti %%eax, 40(%3)\n" " movnti %%edx, 44(%3)\n" - "13: movl 48(%4), %%eax\n" - "81: movl 52(%4), %%edx\n" + "13: "__copyuser_seg" movl 48(%4), %%eax\n" + "81: "__copyuser_seg" movl 52(%4), %%edx\n" " movnti %%eax, 48(%3)\n" " movnti %%edx, 52(%3)\n" - "14: movl 56(%4), %%eax\n" - "91: movl 60(%4), %%edx\n" + "14: "__copyuser_seg" movl 56(%4), %%eax\n" + "91: "__copyuser_seg" movl 60(%4), %%edx\n" " movnti %%eax, 56(%3)\n" " movnti %%edx, 60(%3)\n" " addl $-64, %0\n" @@ -584,9 +703,9 @@ static unsigned long __copy_user_intel_nocache(void *to, " shrl $2, %0\n" " andl $3, %%eax\n" " cld\n" - "6: rep; movsl\n" + "6: rep; "__copyuser_seg" movsl\n" " movl %%eax,%0\n" - "7: rep; movsb\n" + "7: rep; "__copyuser_seg" movsb\n" "8:\n" ".section .fixup,\"ax\"\n" "9: lea 0(%%eax,%0,4),%0\n" @@ -629,32 +748,36 @@ static unsigned long __copy_user_intel_nocache(void *to, */ unsigned long __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size); -unsigned long __copy_user_intel(void __user *to, const void *from, +unsigned long __generic_copy_to_user_intel(void __user *to, const void *from, + unsigned long size); +unsigned long __generic_copy_from_user_intel(void *to, const void __user *from, unsigned long size); unsigned long __copy_user_zeroing_intel_nocache(void *to, const void __user *from, unsigned long size); #endif /* CONFIG_X86_INTEL_USERCOPY */ /* Generic arbitrary sized copy. */ -#define __copy_user(to, from, size) \ +#define __copy_user(to, from, size, prefix, set, restore) \ do { \ int __d0, __d1, __d2; \ __asm__ __volatile__( \ + set \ " cmp $7,%0\n" \ " jbe 1f\n" \ " movl %1,%0\n" \ " negl %0\n" \ " andl $7,%0\n" \ " subl %0,%3\n" \ - "4: rep; movsb\n" \ + "4: rep; "prefix"movsb\n" \ " movl %3,%0\n" \ " shrl $2,%0\n" \ " andl $3,%3\n" \ " .align 2,0x90\n" \ - "0: rep; movsl\n" \ + "0: rep; "prefix"movsl\n" \ " movl %3,%0\n" \ - "1: rep; movsb\n" \ + "1: rep; "prefix"movsb\n" \ "2:\n" \ + restore \ ".section .fixup,\"ax\"\n" \ "5: addl %3,%0\n" \ " jmp 2b\n" \ @@ -682,14 +805,14 @@ do { \ " negl %0\n" \ " andl $7,%0\n" \ " subl %0,%3\n" \ - "4: rep; movsb\n" \ + "4: rep; "__copyuser_seg"movsb\n" \ " movl %3,%0\n" \ " shrl $2,%0\n" \ " andl $3,%3\n" \ " .align 2,0x90\n" \ - "0: rep; movsl\n" \ + "0: rep; "__copyuser_seg"movsl\n" \ " movl %3,%0\n" \ - "1: rep; movsb\n" \ + "1: rep; "__copyuser_seg"movsb\n" \ "2:\n" \ ".section .fixup,\"ax\"\n" \ "5: addl %3,%0\n" \ @@ -775,9 +898,9 @@ survive: } #endif if (movsl_is_ok(to, from, n)) - __copy_user(to, from, n); + __copy_user(to, from, n, "", __COPYUSER_SET_ES, __COPYUSER_RESTORE_ES); else - n = __copy_user_intel(to, from, n); + n = __generic_copy_to_user_intel(to, from, n); return n; } EXPORT_SYMBOL(__copy_to_user_ll); @@ -797,10 +920,9 @@ unsigned long __copy_from_user_ll_nozero(void *to, const void __user *from, unsigned long n) { if (movsl_is_ok(to, from, n)) - __copy_user(to, from, n); + __copy_user(to, from, n, __copyuser_seg, "", ""); else - n = __copy_user_intel((void __user *)to, - (const void *)from, n); + n = __generic_copy_from_user_intel(to, from, n); return n; } EXPORT_SYMBOL(__copy_from_user_ll_nozero); @@ -827,65 +949,50 @@ unsigned long __copy_from_user_ll_nocache_nozero(void *to, const void __user *fr if (n > 64 && cpu_has_xmm2) n = __copy_user_intel_nocache(to, from, n); else - __copy_user(to, from, n); + __copy_user(to, from, n, __copyuser_seg, "", ""); #else - __copy_user(to, from, n); + __copy_user(to, from, n, __copyuser_seg, "", ""); #endif return n; } EXPORT_SYMBOL(__copy_from_user_ll_nocache_nozero); -/** - * copy_to_user: - Copy a block of data into user space. - * @to: Destination address, in user space. - * @from: Source address, in kernel space. - * @n: Number of bytes to copy. - * - * Context: User context only. This function may sleep. - * - * Copy data from kernel space to user space. - * - * Returns number of bytes that could not be copied. - * On success, this will be zero. - */ -unsigned long -copy_to_user(void __user *to, const void *from, unsigned long n) -{ - if (access_ok(VERIFY_WRITE, to, n)) - n = __copy_to_user(to, from, n); - return n; -} -EXPORT_SYMBOL(copy_to_user); - -/** - * copy_from_user: - Copy a block of data from user space. - * @to: Destination address, in kernel space. - * @from: Source address, in user space. - * @n: Number of bytes to copy. - * - * Context: User context only. This function may sleep. - * - * Copy data from user space to kernel space. - * - * Returns number of bytes that could not be copied. - * On success, this will be zero. - * - * If some data could not be copied, this function will pad the copied - * data to the requested size using zero bytes. - */ -unsigned long -_copy_from_user(void *to, const void __user *from, unsigned long n) -{ - if (access_ok(VERIFY_READ, from, n)) - n = __copy_from_user(to, from, n); - else - memset(to, 0, n); - return n; -} -EXPORT_SYMBOL(_copy_from_user); - void copy_from_user_overflow(void) { WARN(1, "Buffer overflow detected!\n"); } EXPORT_SYMBOL(copy_from_user_overflow); + +void copy_to_user_overflow(void) +{ + WARN(1, "Buffer overflow detected!\n"); +} +EXPORT_SYMBOL(copy_to_user_overflow); + +#ifdef CONFIG_PAX_MEMORY_UDEREF +void __set_fs(mm_segment_t x) +{ + switch (x.seg) { + case 0: + loadsegment(gs, 0); + break; + case TASK_SIZE_MAX: + loadsegment(gs, __USER_DS); + break; + case -1UL: + loadsegment(gs, __KERNEL_DS); + break; + default: + BUG(); + } + return; +} +EXPORT_SYMBOL(__set_fs); + +void set_fs(mm_segment_t x) +{ + current_thread_info()->addr_limit = x; + __set_fs(x); +} +EXPORT_SYMBOL(set_fs); +#endif diff --git a/arch/x86/lib/usercopy_64.c b/arch/x86/lib/usercopy_64.c index b7c2849..8633ad8 100644 --- a/arch/x86/lib/usercopy_64.c +++ b/arch/x86/lib/usercopy_64.c @@ -42,6 +42,12 @@ long __strncpy_from_user(char *dst, const char __user *src, long count) { long res; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)src < PAX_USER_SHADOW_BASE) + src += PAX_USER_SHADOW_BASE; +#endif + __do_strncpy_from_user(dst, src, count, res); return res; } @@ -65,6 +71,12 @@ unsigned long __clear_user(void __user *addr, unsigned long size) { long __d0; might_fault(); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)addr < PAX_USER_SHADOW_BASE) + addr += PAX_USER_SHADOW_BASE; +#endif + /* no memory constraint because it doesn't change any memory gcc knows about */ asm volatile( @@ -149,12 +161,20 @@ long strlen_user(const char __user *s) } EXPORT_SYMBOL(strlen_user); -unsigned long copy_in_user(void __user *to, const void __user *from, unsigned len) +unsigned long copy_in_user(void __user *to, const void __user *from, unsigned long len) { - if (access_ok(VERIFY_WRITE, to, len) && access_ok(VERIFY_READ, from, len)) { - return copy_user_generic((__force void *)to, (__force void *)from, len); - } - return len; + if (access_ok(VERIFY_WRITE, to, len) && access_ok(VERIFY_READ, from, len)) { + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if ((unsigned long)to < PAX_USER_SHADOW_BASE) + to += PAX_USER_SHADOW_BASE; + if ((unsigned long)from < PAX_USER_SHADOW_BASE) + from += PAX_USER_SHADOW_BASE; +#endif + + return copy_user_generic((void __force_kernel *)to, (void __force_kernel *)from, len); + } + return len; } EXPORT_SYMBOL(copy_in_user); @@ -164,7 +184,7 @@ EXPORT_SYMBOL(copy_in_user); * it is not necessary to optimize tail handling. */ unsigned long -copy_user_handle_tail(char *to, char *from, unsigned len, unsigned zerorest) +copy_user_handle_tail(char __user *to, char __user *from, unsigned long len, unsigned zerorest) { char c; unsigned zero_len; diff --git a/arch/x86/mm/extable.c b/arch/x86/mm/extable.c index 1fb85db..8b3540b 100644 --- a/arch/x86/mm/extable.c +++ b/arch/x86/mm/extable.c @@ -8,7 +8,7 @@ int fixup_exception(struct pt_regs *regs) const struct exception_table_entry *fixup; #ifdef CONFIG_PNPBIOS - if (unlikely(SEGMENT_IS_PNP_CODE(regs->cs))) { + if (unlikely(!v8086_mode(regs) && SEGMENT_IS_PNP_CODE(regs->cs))) { extern u32 pnp_bios_fault_eip, pnp_bios_fault_esp; extern u32 pnp_bios_is_utter_crap; pnp_bios_is_utter_crap = 1; diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index f0b4caf..d92fd42 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -13,11 +13,18 @@ #include /* perf_sw_event */ #include /* hstate_index_to_shift */ #include /* prefetchw */ +#include +#include #include /* dotraplinkage, ... */ #include /* pgd_*(), ... */ #include /* kmemcheck_*(), ... */ #include /* VSYSCALL_START */ +#include + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +#include +#endif /* * Page fault error code bits: @@ -55,7 +62,7 @@ static inline int __kprobes notify_page_fault(struct pt_regs *regs) int ret = 0; /* kprobe_running() needs smp_processor_id() */ - if (kprobes_built_in() && !user_mode_vm(regs)) { + if (kprobes_built_in() && !user_mode(regs)) { preempt_disable(); if (kprobe_running() && kprobe_fault_handler(regs, 14)) ret = 1; @@ -116,7 +123,10 @@ check_prefetch_opcode(struct pt_regs *regs, unsigned char *instr, return !instr_lo || (instr_lo>>1) == 1; case 0x00: /* Prefetch instruction is 0x0F0D or 0x0F18 */ - if (probe_kernel_address(instr, opcode)) + if (user_mode(regs)) { + if (__copy_from_user_inatomic(&opcode, (unsigned char __force_user *)(instr), 1)) + return 0; + } else if (probe_kernel_address(instr, opcode)) return 0; *prefetch = (instr_lo == 0xF) && @@ -150,7 +160,10 @@ is_prefetch(struct pt_regs *regs, unsigned long error_code, unsigned long addr) while (instr < max_instr) { unsigned char opcode; - if (probe_kernel_address(instr, opcode)) + if (user_mode(regs)) { + if (__copy_from_user_inatomic(&opcode, (unsigned char __force_user *)(instr), 1)) + break; + } else if (probe_kernel_address(instr, opcode)) break; instr++; @@ -181,6 +194,34 @@ force_sig_info_fault(int si_signo, int si_code, unsigned long address, force_sig_info(si_signo, &info, tsk); } +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) +static bool pax_is_fetch_fault(struct pt_regs *regs, unsigned long error_code, unsigned long address); +#endif + +#ifdef CONFIG_PAX_EMUTRAMP +static int pax_handle_fetch_fault(struct pt_regs *regs); +#endif + +#ifdef CONFIG_PAX_PAGEEXEC +static inline pmd_t * pax_get_pmd(struct mm_struct *mm, unsigned long address) +{ + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + + pgd = pgd_offset(mm, address); + if (!pgd_present(*pgd)) + return NULL; + pud = pud_offset(pgd, address); + if (!pud_present(*pud)) + return NULL; + pmd = pmd_offset(pud, address); + if (!pmd_present(*pmd)) + return NULL; + return pmd; +} +#endif + DEFINE_SPINLOCK(pgd_lock); LIST_HEAD(pgd_list); @@ -231,10 +272,22 @@ void vmalloc_sync_all(void) for (address = VMALLOC_START & PMD_MASK; address >= TASK_SIZE && address < FIXADDR_TOP; address += PMD_SIZE) { + +#ifdef CONFIG_PAX_PER_CPU_PGD + unsigned long cpu; +#else struct page *page; +#endif spin_lock(&pgd_lock); + +#ifdef CONFIG_PAX_PER_CPU_PGD + for (cpu = 0; cpu < nr_cpu_ids; ++cpu) { + pgd_t *pgd = get_cpu_pgd(cpu); + pmd_t *ret; +#else list_for_each_entry(page, &pgd_list, lru) { + pgd_t *pgd = page_address(page); spinlock_t *pgt_lock; pmd_t *ret; @@ -242,8 +295,13 @@ void vmalloc_sync_all(void) pgt_lock = &pgd_page_get_mm(page)->page_table_lock; spin_lock(pgt_lock); - ret = vmalloc_sync_one(page_address(page), address); +#endif + + ret = vmalloc_sync_one(pgd, address); + +#ifndef CONFIG_PAX_PER_CPU_PGD spin_unlock(pgt_lock); +#endif if (!ret) break; @@ -277,6 +335,11 @@ static noinline __kprobes int vmalloc_fault(unsigned long address) * an interrupt in the middle of a task switch.. */ pgd_paddr = read_cr3(); + +#ifdef CONFIG_PAX_PER_CPU_PGD + BUG_ON(__pa(get_cpu_pgd(smp_processor_id())) != (pgd_paddr & PHYSICAL_PAGE_MASK)); +#endif + pmd_k = vmalloc_sync_one(__va(pgd_paddr), address); if (!pmd_k) return -1; @@ -372,7 +435,14 @@ static noinline __kprobes int vmalloc_fault(unsigned long address) * happen within a race in page table update. In the later * case just flush: */ + +#ifdef CONFIG_PAX_PER_CPU_PGD + BUG_ON(__pa(get_cpu_pgd(smp_processor_id())) != (read_cr3() & PHYSICAL_PAGE_MASK)); + pgd = pgd_offset_cpu(smp_processor_id(), address); +#else pgd = pgd_offset(current->active_mm, address); +#endif + pgd_ref = pgd_offset_k(address); if (pgd_none(*pgd_ref)) return -1; @@ -540,7 +610,7 @@ static int is_errata93(struct pt_regs *regs, unsigned long address) static int is_errata100(struct pt_regs *regs, unsigned long address) { #ifdef CONFIG_X86_64 - if ((regs->cs == __USER32_CS || (regs->cs & (1<<2))) && (address >> 32)) + if ((regs->cs == __USER32_CS || (regs->cs & SEGMENT_LDT)) && (address >> 32)) return 1; #endif return 0; @@ -567,7 +637,7 @@ static int is_f00f_bug(struct pt_regs *regs, unsigned long address) } static const char nx_warning[] = KERN_CRIT -"kernel tried to execute NX-protected page - exploit attempt? (uid: %d)\n"; +"kernel tried to execute NX-protected page - exploit attempt? (uid: %d, task: %s, pid: %d)\n"; static void show_fault_oops(struct pt_regs *regs, unsigned long error_code, @@ -576,15 +646,26 @@ show_fault_oops(struct pt_regs *regs, unsigned long error_code, if (!oops_may_print()) return; - if (error_code & PF_INSTR) { + if ((__supported_pte_mask & _PAGE_NX) && (error_code & PF_INSTR)) { unsigned int level; pte_t *pte = lookup_address(address, &level); if (pte && pte_present(*pte) && !pte_exec(*pte)) - printk(nx_warning, current_uid()); + printk(nx_warning, current_uid(), current->comm, task_pid_nr(current)); } +#ifdef CONFIG_PAX_KERNEXEC + if (init_mm.start_code <= address && address < init_mm.end_code) { + if (current->signal->curr_ip) + printk(KERN_ERR "PAX: From %pI4: %s:%d, uid/euid: %u/%u, attempted to modify kernel code\n", + ¤t->signal->curr_ip, current->comm, task_pid_nr(current), current_uid(), current_euid()); + else + printk(KERN_ERR "PAX: %s:%d, uid/euid: %u/%u, attempted to modify kernel code\n", + current->comm, task_pid_nr(current), current_uid(), current_euid()); + } +#endif + printk(KERN_ALERT "BUG: unable to handle kernel "); if (address < PAGE_SIZE) printk(KERN_CONT "NULL pointer dereference"); @@ -748,6 +829,21 @@ __bad_area_nosemaphore(struct pt_regs *regs, unsigned long error_code, } #endif +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) + if (pax_is_fetch_fault(regs, error_code, address)) { + +#ifdef CONFIG_PAX_EMUTRAMP + switch (pax_handle_fetch_fault(regs)) { + case 2: + return; + } +#endif + + pax_report_fault(regs, (void *)regs->ip, (void *)regs->sp); + do_group_exit(SIGKILL); + } +#endif + if (unlikely(show_unhandled_signals)) show_signal_msg(regs, error_code, address, tsk); @@ -844,7 +940,7 @@ do_sigbus(struct pt_regs *regs, unsigned long error_code, unsigned long address, if (fault & (VM_FAULT_HWPOISON|VM_FAULT_HWPOISON_LARGE)) { printk(KERN_ERR "MCE: Killing %s:%d due to hardware memory corruption fault at %lx\n", - tsk->comm, tsk->pid, address); + tsk->comm, task_pid_nr(tsk), address); code = BUS_MCEERR_AR; } #endif @@ -900,6 +996,99 @@ static int spurious_fault_check(unsigned long error_code, pte_t *pte) return 1; } +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) +static int pax_handle_pageexec_fault(struct pt_regs *regs, struct mm_struct *mm, unsigned long address, unsigned long error_code) +{ + pte_t *pte; + pmd_t *pmd; + spinlock_t *ptl; + unsigned char pte_mask; + + if ((__supported_pte_mask & _PAGE_NX) || (error_code & (PF_PROT|PF_USER)) != (PF_PROT|PF_USER) || v8086_mode(regs) || + !(mm->pax_flags & MF_PAX_PAGEEXEC)) + return 0; + + /* PaX: it's our fault, let's handle it if we can */ + + /* PaX: take a look at read faults before acquiring any locks */ + if (unlikely(!(error_code & PF_WRITE) && (regs->ip == address))) { + /* instruction fetch attempt from a protected page in user mode */ + up_read(&mm->mmap_sem); + +#ifdef CONFIG_PAX_EMUTRAMP + switch (pax_handle_fetch_fault(regs)) { + case 2: + return 1; + } +#endif + + pax_report_fault(regs, (void *)regs->ip, (void *)regs->sp); + do_group_exit(SIGKILL); + } + + pmd = pax_get_pmd(mm, address); + if (unlikely(!pmd)) + return 0; + + pte = pte_offset_map_lock(mm, pmd, address, &ptl); + if (unlikely(!(pte_val(*pte) & _PAGE_PRESENT) || pte_user(*pte))) { + pte_unmap_unlock(pte, ptl); + return 0; + } + + if (unlikely((error_code & PF_WRITE) && !pte_write(*pte))) { + /* write attempt to a protected page in user mode */ + pte_unmap_unlock(pte, ptl); + return 0; + } + +#ifdef CONFIG_SMP + if (likely(address > get_limit(regs->cs) && cpu_isset(smp_processor_id(), mm->context.cpu_user_cs_mask))) +#else + if (likely(address > get_limit(regs->cs))) +#endif + { + set_pte(pte, pte_mkread(*pte)); + __flush_tlb_one(address); + pte_unmap_unlock(pte, ptl); + up_read(&mm->mmap_sem); + return 1; + } + + pte_mask = _PAGE_ACCESSED | _PAGE_USER | ((error_code & PF_WRITE) << (_PAGE_BIT_DIRTY-1)); + + /* + * PaX: fill DTLB with user rights and retry + */ + __asm__ __volatile__ ( + "orb %2,(%1)\n" +#if defined(CONFIG_M586) || defined(CONFIG_M586TSC) +/* + * PaX: let this uncommented 'invlpg' remind us on the behaviour of Intel's + * (and AMD's) TLBs. namely, they do not cache PTEs that would raise *any* + * page fault when examined during a TLB load attempt. this is true not only + * for PTEs holding a non-present entry but also present entries that will + * raise a page fault (such as those set up by PaX, or the copy-on-write + * mechanism). in effect it means that we do *not* need to flush the TLBs + * for our target pages since their PTEs are simply not in the TLBs at all. + + * the best thing in omitting it is that we gain around 15-20% speed in the + * fast path of the page fault handler and can get rid of tracing since we + * can no longer flush unintended entries. + */ + "invlpg (%0)\n" +#endif + __copyuser_seg"testb $0,(%0)\n" + "xorb %3,(%1)\n" + : + : "r" (address), "r" (pte), "q" (pte_mask), "i" (_PAGE_USER) + : "memory", "cc"); + pte_unmap_unlock(pte, ptl); + up_read(&mm->mmap_sem); + return 1; +} +#endif + /* * Handle a spurious fault caused by a stale TLB entry. * @@ -972,6 +1161,9 @@ int show_unhandled_signals = 1; static inline int access_error(unsigned long error_code, struct vm_area_struct *vma) { + if ((__supported_pte_mask & _PAGE_NX) && (error_code & PF_INSTR) && !(vma->vm_flags & VM_EXEC)) + return 1; + if (error_code & PF_WRITE) { /* write, present and write, not present: */ if (unlikely(!(vma->vm_flags & VM_WRITE))) @@ -1005,18 +1197,32 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code) { struct vm_area_struct *vma; struct task_struct *tsk; - unsigned long address; struct mm_struct *mm; int fault; int write = error_code & PF_WRITE; unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE | (write ? FAULT_FLAG_WRITE : 0); - tsk = current; - mm = tsk->mm; - /* Get the faulting address: */ - address = read_cr2(); + unsigned long address = read_cr2(); + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (!user_mode(regs) && address < 2 * PAX_USER_SHADOW_BASE) { + if (!search_exception_tables(regs->ip)) { + bad_area_nosemaphore(regs, error_code, address); + return; + } + if (address < PAX_USER_SHADOW_BASE) { + printk(KERN_ERR "PAX: please report this to pageexec@freemail.hu\n"); + printk(KERN_ERR "PAX: faulting IP: %pS\n", (void *)regs->ip); + show_trace_log_lvl(NULL, NULL, (void *)regs->sp, regs->bp, KERN_ERR); + } else + address -= PAX_USER_SHADOW_BASE; + } +#endif + + tsk = current; + mm = tsk->mm; /* * Detect and handle instructions that would cause a page fault for @@ -1077,7 +1283,7 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code) * User-mode registers count as a user access even for any * potential system fault or CPU buglet: */ - if (user_mode_vm(regs)) { + if (user_mode(regs)) { local_irq_enable(); error_code |= PF_USER; } else { @@ -1132,6 +1338,11 @@ retry: might_sleep(); } +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) + if (pax_handle_pageexec_fault(regs, mm, address, error_code)) + return; +#endif + vma = find_vma(mm, address); if (unlikely(!vma)) { bad_area(regs, error_code, address); @@ -1143,18 +1354,24 @@ retry: bad_area(regs, error_code, address); return; } - if (error_code & PF_USER) { - /* - * Accessing the stack below %sp is always a bug. - * The large cushion allows instructions like enter - * and pusha to work. ("enter $65535, $31" pushes - * 32 pointers and then decrements %sp by 65535.) - */ - if (unlikely(address + 65536 + 32 * sizeof(unsigned long) < regs->sp)) { - bad_area(regs, error_code, address); - return; - } + /* + * Accessing the stack below %sp is always a bug. + * The large cushion allows instructions like enter + * and pusha to work. ("enter $65535, $31" pushes + * 32 pointers and then decrements %sp by 65535.) + */ + if (unlikely(address + 65536 + 32 * sizeof(unsigned long) < task_pt_regs(tsk)->sp)) { + bad_area(regs, error_code, address); + return; } + +#ifdef CONFIG_PAX_SEGMEXEC + if (unlikely((mm->pax_flags & MF_PAX_SEGMEXEC) && vma->vm_end - SEGMEXEC_TASK_SIZE - 1 < address - SEGMEXEC_TASK_SIZE - 1)) { + bad_area(regs, error_code, address); + return; + } +#endif + if (unlikely(expand_stack(vma, address))) { bad_area(regs, error_code, address); return; @@ -1209,3 +1426,292 @@ good_area: up_read(&mm->mmap_sem); } + +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) +static bool pax_is_fetch_fault(struct pt_regs *regs, unsigned long error_code, unsigned long address) +{ + struct mm_struct *mm = current->mm; + unsigned long ip = regs->ip; + + if (v8086_mode(regs)) + ip = ((regs->cs & 0xffff) << 4) + (ip & 0xffff); + +#ifdef CONFIG_PAX_PAGEEXEC + if (mm->pax_flags & MF_PAX_PAGEEXEC) { + if ((__supported_pte_mask & _PAGE_NX) && (error_code & PF_INSTR)) + return true; + if (!(error_code & (PF_PROT | PF_WRITE)) && ip == address) + return true; + return false; + } +#endif + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) { + if (!(error_code & (PF_PROT | PF_WRITE)) && (ip + SEGMEXEC_TASK_SIZE == address)) + return true; + return false; + } +#endif + + return false; +} +#endif + +#ifdef CONFIG_PAX_EMUTRAMP +static int pax_handle_fetch_fault_32(struct pt_regs *regs) +{ + int err; + + do { /* PaX: libffi trampoline emulation */ + unsigned char mov, jmp; + unsigned int addr1, addr2; + +#ifdef CONFIG_X86_64 + if ((regs->ip + 9) >> 32) + break; +#endif + + err = get_user(mov, (unsigned char __user *)regs->ip); + err |= get_user(addr1, (unsigned int __user *)(regs->ip + 1)); + err |= get_user(jmp, (unsigned char __user *)(regs->ip + 5)); + err |= get_user(addr2, (unsigned int __user *)(regs->ip + 6)); + + if (err) + break; + + if (mov == 0xB8 && jmp == 0xE9) { + regs->ax = addr1; + regs->ip = (unsigned int)(regs->ip + addr2 + 10); + return 2; + } + } while (0); + + do { /* PaX: gcc trampoline emulation #1 */ + unsigned char mov1, mov2; + unsigned short jmp; + unsigned int addr1, addr2; + +#ifdef CONFIG_X86_64 + if ((regs->ip + 11) >> 32) + break; +#endif + + err = get_user(mov1, (unsigned char __user *)regs->ip); + err |= get_user(addr1, (unsigned int __user *)(regs->ip + 1)); + err |= get_user(mov2, (unsigned char __user *)(regs->ip + 5)); + err |= get_user(addr2, (unsigned int __user *)(regs->ip + 6)); + err |= get_user(jmp, (unsigned short __user *)(regs->ip + 10)); + + if (err) + break; + + if (mov1 == 0xB9 && mov2 == 0xB8 && jmp == 0xE0FF) { + regs->cx = addr1; + regs->ax = addr2; + regs->ip = addr2; + return 2; + } + } while (0); + + do { /* PaX: gcc trampoline emulation #2 */ + unsigned char mov, jmp; + unsigned int addr1, addr2; + +#ifdef CONFIG_X86_64 + if ((regs->ip + 9) >> 32) + break; +#endif + + err = get_user(mov, (unsigned char __user *)regs->ip); + err |= get_user(addr1, (unsigned int __user *)(regs->ip + 1)); + err |= get_user(jmp, (unsigned char __user *)(regs->ip + 5)); + err |= get_user(addr2, (unsigned int __user *)(regs->ip + 6)); + + if (err) + break; + + if (mov == 0xB9 && jmp == 0xE9) { + regs->cx = addr1; + regs->ip = (unsigned int)(regs->ip + addr2 + 10); + return 2; + } + } while (0); + + return 1; /* PaX in action */ +} + +#ifdef CONFIG_X86_64 +static int pax_handle_fetch_fault_64(struct pt_regs *regs) +{ + int err; + + do { /* PaX: libffi trampoline emulation */ + unsigned short mov1, mov2, jmp1; + unsigned char stcclc, jmp2; + unsigned long addr1, addr2; + + err = get_user(mov1, (unsigned short __user *)regs->ip); + err |= get_user(addr1, (unsigned long __user *)(regs->ip + 2)); + err |= get_user(mov2, (unsigned short __user *)(regs->ip + 10)); + err |= get_user(addr2, (unsigned long __user *)(regs->ip + 12)); + err |= get_user(stcclc, (unsigned char __user *)(regs->ip + 20)); + err |= get_user(jmp1, (unsigned short __user *)(regs->ip + 21)); + err |= get_user(jmp2, (unsigned char __user *)(regs->ip + 23)); + + if (err) + break; + + if (mov1 == 0xBB49 && mov2 == 0xBA49 && (stcclc == 0xF8 || stcclc == 0xF9) && jmp1 == 0xFF49 && jmp2 == 0xE3) { + regs->r11 = addr1; + regs->r10 = addr2; + if (stcclc == 0xF8) + regs->flags &= ~X86_EFLAGS_CF; + else + regs->flags |= X86_EFLAGS_CF; + regs->ip = addr1; + return 2; + } + } while (0); + + do { /* PaX: gcc trampoline emulation #1 */ + unsigned short mov1, mov2, jmp1; + unsigned char jmp2; + unsigned int addr1; + unsigned long addr2; + + err = get_user(mov1, (unsigned short __user *)regs->ip); + err |= get_user(addr1, (unsigned int __user *)(regs->ip + 2)); + err |= get_user(mov2, (unsigned short __user *)(regs->ip + 6)); + err |= get_user(addr2, (unsigned long __user *)(regs->ip + 8)); + err |= get_user(jmp1, (unsigned short __user *)(regs->ip + 16)); + err |= get_user(jmp2, (unsigned char __user *)(regs->ip + 18)); + + if (err) + break; + + if (mov1 == 0xBB41 && mov2 == 0xBA49 && jmp1 == 0xFF49 && jmp2 == 0xE3) { + regs->r11 = addr1; + regs->r10 = addr2; + regs->ip = addr1; + return 2; + } + } while (0); + + do { /* PaX: gcc trampoline emulation #2 */ + unsigned short mov1, mov2, jmp1; + unsigned char jmp2; + unsigned long addr1, addr2; + + err = get_user(mov1, (unsigned short __user *)regs->ip); + err |= get_user(addr1, (unsigned long __user *)(regs->ip + 2)); + err |= get_user(mov2, (unsigned short __user *)(regs->ip + 10)); + err |= get_user(addr2, (unsigned long __user *)(regs->ip + 12)); + err |= get_user(jmp1, (unsigned short __user *)(regs->ip + 20)); + err |= get_user(jmp2, (unsigned char __user *)(regs->ip + 22)); + + if (err) + break; + + if (mov1 == 0xBB49 && mov2 == 0xBA49 && jmp1 == 0xFF49 && jmp2 == 0xE3) { + regs->r11 = addr1; + regs->r10 = addr2; + regs->ip = addr1; + return 2; + } + } while (0); + + return 1; /* PaX in action */ +} +#endif + +/* + * PaX: decide what to do with offenders (regs->ip = fault address) + * + * returns 1 when task should be killed + * 2 when gcc trampoline was detected + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + if (v8086_mode(regs)) + return 1; + + if (!(current->mm->pax_flags & MF_PAX_EMUTRAMP)) + return 1; + +#ifdef CONFIG_X86_32 + return pax_handle_fetch_fault_32(regs); +#else + if (regs->cs == __USER32_CS || (regs->cs & SEGMENT_LDT)) + return pax_handle_fetch_fault_32(regs); + else + return pax_handle_fetch_fault_64(regs); +#endif +} +#endif + +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) +void pax_report_insns(struct pt_regs *regs, void *pc, void *sp) +{ + long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 20; i++) { + unsigned char c; + if (get_user(c, (unsigned char __force_user *)pc+i)) + printk(KERN_CONT "?? "); + else + printk(KERN_CONT "%02x ", c); + } + printk("\n"); + + printk(KERN_ERR "PAX: bytes at SP-%lu: ", (unsigned long)sizeof(long)); + for (i = -1; i < 80 / (long)sizeof(long); i++) { + unsigned long c; + if (get_user(c, (unsigned long __force_user *)sp+i)) { +#ifdef CONFIG_X86_32 + printk(KERN_CONT "???????? "); +#else + if ((regs->cs == __USER32_CS || (regs->cs & SEGMENT_LDT))) + printk(KERN_CONT "???????? ???????? "); + else + printk(KERN_CONT "???????????????? "); +#endif + } else { +#ifdef CONFIG_X86_64 + if ((regs->cs == __USER32_CS || (regs->cs & SEGMENT_LDT))) { + printk(KERN_CONT "%08x ", (unsigned int)c); + printk(KERN_CONT "%08x ", (unsigned int)(c >> 32)); + } else +#endif + printk(KERN_CONT "%0*lx ", 2 * (int)sizeof(long), c); + } + } + printk("\n"); +} +#endif + +/** + * probe_kernel_write(): safely attempt to write to a location + * @dst: address to write to + * @src: pointer to the data that shall be written + * @size: size of the data chunk + * + * Safely write to address @dst from the buffer at @src. If a kernel fault + * happens, handle that and return -EFAULT. + */ +long notrace probe_kernel_write(void *dst, const void *src, size_t size) +{ + long ret; + mm_segment_t old_fs = get_fs(); + + set_fs(KERNEL_DS); + pagefault_disable(); + pax_open_kernel(); + ret = __copy_to_user_inatomic((void __force_user *)dst, src, size); + pax_close_kernel(); + pagefault_enable(); + set_fs(old_fs); + + return ret ? -EFAULT : 0; +} diff --git a/arch/x86/mm/gup.c b/arch/x86/mm/gup.c index dd74e46..7d26398 100644 --- a/arch/x86/mm/gup.c +++ b/arch/x86/mm/gup.c @@ -255,7 +255,7 @@ int __get_user_pages_fast(unsigned long start, int nr_pages, int write, addr = start; len = (unsigned long) nr_pages << PAGE_SHIFT; end = start + len; - if (unlikely(!access_ok(write ? VERIFY_WRITE : VERIFY_READ, + if (unlikely(!__access_ok(write ? VERIFY_WRITE : VERIFY_READ, (void __user *)start, len))) return 0; diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c index f4f29b1..5cac4fb 100644 --- a/arch/x86/mm/highmem_32.c +++ b/arch/x86/mm/highmem_32.c @@ -44,7 +44,11 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot) idx = type + KM_TYPE_NR*smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); BUG_ON(!pte_none(*(kmap_pte-idx))); + + pax_open_kernel(); set_pte(kmap_pte-idx, mk_pte(page, prot)); + pax_close_kernel(); + arch_flush_lazy_mmu_mode(); return (void *)vaddr; diff --git a/arch/x86/mm/hugetlbpage.c b/arch/x86/mm/hugetlbpage.c index 8ecbb4b..a269cab 100644 --- a/arch/x86/mm/hugetlbpage.c +++ b/arch/x86/mm/hugetlbpage.c @@ -266,13 +266,20 @@ static unsigned long hugetlb_get_unmapped_area_bottomup(struct file *file, struct hstate *h = hstate_file(file); struct mm_struct *mm = current->mm; struct vm_area_struct *vma; - unsigned long start_addr; + unsigned long start_addr, pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + pax_task_size -= PAGE_SIZE; if (len > mm->cached_hole_size) { - start_addr = mm->free_area_cache; + start_addr = mm->free_area_cache; } else { - start_addr = TASK_UNMAPPED_BASE; - mm->cached_hole_size = 0; + start_addr = mm->mmap_base; + mm->cached_hole_size = 0; } full_search: @@ -280,26 +287,27 @@ full_search: for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { /* At this point: (!vma || addr < vma->vm_end). */ - if (TASK_SIZE - len < addr) { + if (pax_task_size - len < addr) { /* * Start a new search - just in case we missed * some holes. */ - if (start_addr != TASK_UNMAPPED_BASE) { - start_addr = TASK_UNMAPPED_BASE; + if (start_addr != mm->mmap_base) { + start_addr = mm->mmap_base; mm->cached_hole_size = 0; goto full_search; } return -ENOMEM; } - if (!vma || addr + len <= vma->vm_start) { - mm->free_area_cache = addr + len; - return addr; - } + if (check_heap_stack_gap(vma, addr, len)) + break; if (addr + mm->cached_hole_size < vma->vm_start) mm->cached_hole_size = vma->vm_start - addr; addr = ALIGN(vma->vm_end, huge_page_size(h)); } + + mm->free_area_cache = addr + len; + return addr; } static unsigned long hugetlb_get_unmapped_area_topdown(struct file *file, @@ -308,10 +316,9 @@ static unsigned long hugetlb_get_unmapped_area_topdown(struct file *file, { struct hstate *h = hstate_file(file); struct mm_struct *mm = current->mm; - struct vm_area_struct *vma, *prev_vma; - unsigned long base = mm->mmap_base, addr = addr0; + struct vm_area_struct *vma; + unsigned long base = mm->mmap_base, addr; unsigned long largest_hole = mm->cached_hole_size; - int first_time = 1; /* don't allow allocations above current base */ if (mm->free_area_cache > base) @@ -321,14 +328,15 @@ static unsigned long hugetlb_get_unmapped_area_topdown(struct file *file, largest_hole = 0; mm->free_area_cache = base; } -try_again: + /* make sure it can fit in the remaining address space */ if (mm->free_area_cache < len) goto fail; /* either no address requested or can't fit in requested address hole */ - addr = (mm->free_area_cache - len) & huge_page_mask(h); + addr = (mm->free_area_cache - len); do { + addr &= huge_page_mask(h); /* * Lookup failure means no vma is above this address, * i.e. return with success: @@ -341,46 +349,47 @@ try_again: * new region fits between prev_vma->vm_end and * vma->vm_start, use it: */ - prev_vma = vma->vm_prev; - if (addr + len <= vma->vm_start && - (!prev_vma || (addr >= prev_vma->vm_end))) { + if (check_heap_stack_gap(vma, addr, len)) { /* remember the address as a hint for next time */ - mm->cached_hole_size = largest_hole; - return (mm->free_area_cache = addr); - } else { - /* pull free_area_cache down to the first hole */ - if (mm->free_area_cache == vma->vm_end) { - mm->free_area_cache = vma->vm_start; - mm->cached_hole_size = largest_hole; - } + mm->cached_hole_size = largest_hole; + return (mm->free_area_cache = addr); + } + /* pull free_area_cache down to the first hole */ + if (mm->free_area_cache == vma->vm_end) { + mm->free_area_cache = vma->vm_start; + mm->cached_hole_size = largest_hole; } /* remember the largest hole we saw so far */ if (addr + largest_hole < vma->vm_start) - largest_hole = vma->vm_start - addr; + largest_hole = vma->vm_start - addr; /* try just below the current vma->vm_start */ - addr = (vma->vm_start - len) & huge_page_mask(h); - } while (len <= vma->vm_start); + addr = skip_heap_stack_gap(vma, len); + } while (!IS_ERR_VALUE(addr)); fail: /* - * if hint left us with no space for the requested - * mapping then try again: - */ - if (first_time) { - mm->free_area_cache = base; - largest_hole = 0; - first_time = 0; - goto try_again; - } - /* * A failed mmap() very likely causes application failure, * so fall back to the bottom-up function here. This scenario * can happen with large stack limits and large mmap() * allocations. */ - mm->free_area_cache = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + mm->mmap_base = SEGMEXEC_TASK_UNMAPPED_BASE; + else +#endif + + mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + + mm->free_area_cache = mm->mmap_base; mm->cached_hole_size = ~0UL; addr = hugetlb_get_unmapped_area_bottomup(file, addr0, len, pgoff, flags); @@ -388,6 +397,7 @@ fail: /* * Restore the topdown base: */ + mm->mmap_base = base; mm->free_area_cache = base; mm->cached_hole_size = ~0UL; @@ -401,10 +411,19 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, struct hstate *h = hstate_file(file); struct mm_struct *mm = current->mm; struct vm_area_struct *vma; + unsigned long pax_task_size = TASK_SIZE; if (len & ~huge_page_mask(h)) return -EINVAL; - if (len > TASK_SIZE) + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + pax_task_size -= PAGE_SIZE; + + if (len > pax_task_size) return -ENOMEM; if (flags & MAP_FIXED) { @@ -416,8 +435,7 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, if (addr) { addr = ALIGN(addr, huge_page_size(h)); vma = find_vma(mm, addr); - if (TASK_SIZE - len >= addr && - (!vma || addr + len <= vma->vm_start)) + if (pax_task_size - len >= addr && check_heap_stack_gap(vma, addr, len)) return addr; } if (mm->get_unmapped_area == arch_get_unmapped_area) diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index 6cabf65..77e9c1c 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -17,6 +17,7 @@ #include #include #include /* for MAX_DMA_PFN */ +#include unsigned long __initdata pgt_buf_start; unsigned long __meminitdata pgt_buf_end; @@ -33,7 +34,7 @@ int direct_gbpages static void __init find_early_table_space(unsigned long end, int use_pse, int use_gbpages) { - unsigned long puds, pmds, ptes, tables, start = 0, good_end = end; + unsigned long puds, pmds, ptes, tables, start = 0x100000, good_end = end; phys_addr_t base; puds = (end + PUD_SIZE - 1) >> PUD_SHIFT; @@ -314,8 +315,29 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, */ int devmem_is_allowed(unsigned long pagenr) { +#ifdef CONFIG_GRKERNSEC_KMEM + /* allow BDA */ + if (!pagenr) + return 1; + /* allow EBDA */ + if ((0x9f000 >> PAGE_SHIFT) == pagenr) + return 1; +#else + if (!pagenr) + return 1; +#ifdef CONFIG_VM86 + if (pagenr < (ISA_START_ADDRESS >> PAGE_SHIFT)) + return 1; +#endif +#endif + + if ((ISA_START_ADDRESS >> PAGE_SHIFT) <= pagenr && pagenr < (ISA_END_ADDRESS >> PAGE_SHIFT)) + return 1; +#ifdef CONFIG_GRKERNSEC_KMEM + /* throw out everything else below 1MB */ if (pagenr <= 256) - return 1; + return 0; +#endif if (iomem_is_exclusive(pagenr << PAGE_SHIFT)) return 0; if (!page_is_ram(pagenr)) @@ -374,6 +396,86 @@ void free_init_pages(char *what, unsigned long begin, unsigned long end) void free_initmem(void) { + +#ifdef CONFIG_PAX_KERNEXEC +#ifdef CONFIG_X86_32 + /* PaX: limit KERNEL_CS to actual size */ + unsigned long addr, limit; + struct desc_struct d; + int cpu; + + limit = paravirt_enabled() ? ktva_ktla(0xffffffff) : (unsigned long)&_etext; + limit = (limit - 1UL) >> PAGE_SHIFT; + + memset(__LOAD_PHYSICAL_ADDR + PAGE_OFFSET, POISON_FREE_INITMEM, PAGE_SIZE); + for (cpu = 0; cpu < nr_cpu_ids; cpu++) { + pack_descriptor(&d, get_desc_base(&get_cpu_gdt_table(cpu)[GDT_ENTRY_KERNEL_CS]), limit, 0x9B, 0xC); + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_KERNEL_CS, &d, DESCTYPE_S); + } + + /* PaX: make KERNEL_CS read-only */ + addr = PFN_ALIGN(ktla_ktva((unsigned long)&_text)); + if (!paravirt_enabled()) + set_memory_ro(addr, (PFN_ALIGN(_sdata) - addr) >> PAGE_SHIFT); +/* + for (addr = ktla_ktva((unsigned long)&_text); addr < (unsigned long)&_sdata; addr += PMD_SIZE) { + pgd = pgd_offset_k(addr); + pud = pud_offset(pgd, addr); + pmd = pmd_offset(pud, addr); + set_pmd(pmd, __pmd(pmd_val(*pmd) & ~_PAGE_RW)); + } +*/ +#ifdef CONFIG_X86_PAE + set_memory_nx(PFN_ALIGN(__init_begin), (PFN_ALIGN(__init_end) - PFN_ALIGN(__init_begin)) >> PAGE_SHIFT); +/* + for (addr = (unsigned long)&__init_begin; addr < (unsigned long)&__init_end; addr += PMD_SIZE) { + pgd = pgd_offset_k(addr); + pud = pud_offset(pgd, addr); + pmd = pmd_offset(pud, addr); + set_pmd(pmd, __pmd(pmd_val(*pmd) | (_PAGE_NX & __supported_pte_mask))); + } +*/ +#endif + +#ifdef CONFIG_MODULES + set_memory_4k((unsigned long)MODULES_EXEC_VADDR, (MODULES_EXEC_END - MODULES_EXEC_VADDR) >> PAGE_SHIFT); +#endif + +#else + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + unsigned long addr, end; + + /* PaX: make kernel code/rodata read-only, rest non-executable */ + for (addr = __START_KERNEL_map; addr < __START_KERNEL_map + KERNEL_IMAGE_SIZE; addr += PMD_SIZE) { + pgd = pgd_offset_k(addr); + pud = pud_offset(pgd, addr); + pmd = pmd_offset(pud, addr); + if (!pmd_present(*pmd)) + continue; + if ((unsigned long)_text <= addr && addr < (unsigned long)_sdata) + set_pmd(pmd, __pmd(pmd_val(*pmd) & ~_PAGE_RW)); + else + set_pmd(pmd, __pmd(pmd_val(*pmd) | (_PAGE_NX & __supported_pte_mask))); + } + + addr = (unsigned long)__va(__pa(__START_KERNEL_map)); + end = addr + KERNEL_IMAGE_SIZE; + for (; addr < end; addr += PMD_SIZE) { + pgd = pgd_offset_k(addr); + pud = pud_offset(pgd, addr); + pmd = pmd_offset(pud, addr); + if (!pmd_present(*pmd)) + continue; + if ((unsigned long)__va(__pa(_text)) <= addr && addr < (unsigned long)__va(__pa(_sdata))) + set_pmd(pmd, __pmd(pmd_val(*pmd) & ~_PAGE_RW)); + } +#endif + + flush_tlb_all(); +#endif + free_init_pages("unused kernel memory", (unsigned long)(&__init_begin), (unsigned long)(&__init_end)); diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index 8663f6c..829ae76 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -74,36 +74,6 @@ static __init void *alloc_low_page(void) } /* - * Creates a middle page table and puts a pointer to it in the - * given global directory entry. This only returns the gd entry - * in non-PAE compilation mode, since the middle layer is folded. - */ -static pmd_t * __init one_md_table_init(pgd_t *pgd) -{ - pud_t *pud; - pmd_t *pmd_table; - -#ifdef CONFIG_X86_PAE - if (!(pgd_val(*pgd) & _PAGE_PRESENT)) { - if (after_bootmem) - pmd_table = (pmd_t *)alloc_bootmem_pages(PAGE_SIZE); - else - pmd_table = (pmd_t *)alloc_low_page(); - paravirt_alloc_pmd(&init_mm, __pa(pmd_table) >> PAGE_SHIFT); - set_pgd(pgd, __pgd(__pa(pmd_table) | _PAGE_PRESENT)); - pud = pud_offset(pgd, 0); - BUG_ON(pmd_table != pmd_offset(pud, 0)); - - return pmd_table; - } -#endif - pud = pud_offset(pgd, 0); - pmd_table = pmd_offset(pud, 0); - - return pmd_table; -} - -/* * Create a page table and place a pointer to it in a middle page * directory entry: */ @@ -123,13 +93,28 @@ static pte_t * __init one_page_table_init(pmd_t *pmd) page_table = (pte_t *)alloc_low_page(); paravirt_alloc_pte(&init_mm, __pa(page_table) >> PAGE_SHIFT); +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) + set_pmd(pmd, __pmd(__pa(page_table) | _KERNPG_TABLE)); +#else set_pmd(pmd, __pmd(__pa(page_table) | _PAGE_TABLE)); +#endif BUG_ON(page_table != pte_offset_kernel(pmd, 0)); } return pte_offset_kernel(pmd, 0); } +static pmd_t * __init one_md_table_init(pgd_t *pgd) +{ + pud_t *pud; + pmd_t *pmd_table; + + pud = pud_offset(pgd, 0); + pmd_table = pmd_offset(pud, 0); + + return pmd_table; +} + pmd_t * __init populate_extra_pmd(unsigned long vaddr) { int pgd_idx = pgd_index(vaddr); @@ -203,6 +188,7 @@ page_table_range_init(unsigned long start, unsigned long end, pgd_t *pgd_base) int pgd_idx, pmd_idx; unsigned long vaddr; pgd_t *pgd; + pud_t *pud; pmd_t *pmd; pte_t *pte = NULL; @@ -212,8 +198,13 @@ page_table_range_init(unsigned long start, unsigned long end, pgd_t *pgd_base) pgd = pgd_base + pgd_idx; for ( ; (pgd_idx < PTRS_PER_PGD) && (vaddr != end); pgd++, pgd_idx++) { - pmd = one_md_table_init(pgd); - pmd = pmd + pmd_index(vaddr); + pud = pud_offset(pgd, vaddr); + pmd = pmd_offset(pud, vaddr); + +#ifdef CONFIG_X86_PAE + paravirt_alloc_pmd(&init_mm, __pa(pmd) >> PAGE_SHIFT); +#endif + for (; (pmd_idx < PTRS_PER_PMD) && (vaddr != end); pmd++, pmd_idx++) { pte = page_table_kmap_check(one_page_table_init(pmd), @@ -225,11 +216,20 @@ page_table_range_init(unsigned long start, unsigned long end, pgd_t *pgd_base) } } -static inline int is_kernel_text(unsigned long addr) +static inline int is_kernel_text(unsigned long start, unsigned long end) { - if (addr >= (unsigned long)_text && addr <= (unsigned long)__init_end) - return 1; - return 0; + if ((start > ktla_ktva((unsigned long)_etext) || + end <= ktla_ktva((unsigned long)_stext)) && + (start > ktla_ktva((unsigned long)_einittext) || + end <= ktla_ktva((unsigned long)_sinittext)) && + +#ifdef CONFIG_ACPI_SLEEP + (start > (unsigned long)__va(acpi_wakeup_address) + 0x4000 || end <= (unsigned long)__va(acpi_wakeup_address)) && +#endif + + (start > (unsigned long)__va(0xfffff) || end <= (unsigned long)__va(0xc0000))) + return 0; + return 1; } /* @@ -246,9 +246,10 @@ kernel_physical_mapping_init(unsigned long start, unsigned long last_map_addr = end; unsigned long start_pfn, end_pfn; pgd_t *pgd_base = swapper_pg_dir; - int pgd_idx, pmd_idx, pte_ofs; + unsigned int pgd_idx, pmd_idx, pte_ofs; unsigned long pfn; pgd_t *pgd; + pud_t *pud; pmd_t *pmd; pte_t *pte; unsigned pages_2m, pages_4k; @@ -281,8 +282,13 @@ repeat: pfn = start_pfn; pgd_idx = pgd_index((pfn<> PAGE_SHIFT); +#endif if (pfn >= end_pfn) continue; @@ -294,14 +300,13 @@ repeat: #endif for (; pmd_idx < PTRS_PER_PMD && pfn < end_pfn; pmd++, pmd_idx++) { - unsigned int addr = pfn * PAGE_SIZE + PAGE_OFFSET; + unsigned long address = pfn * PAGE_SIZE + PAGE_OFFSET; /* * Map with big pages if possible, otherwise * create normal page tables: */ if (use_pse) { - unsigned int addr2; pgprot_t prot = PAGE_KERNEL_LARGE; /* * first pass will use the same initial @@ -311,11 +316,7 @@ repeat: __pgprot(PTE_IDENT_ATTR | _PAGE_PSE); - addr2 = (pfn + PTRS_PER_PTE-1) * PAGE_SIZE + - PAGE_OFFSET + PAGE_SIZE-1; - - if (is_kernel_text(addr) || - is_kernel_text(addr2)) + if (is_kernel_text(address, address + PMD_SIZE)) prot = PAGE_KERNEL_LARGE_EXEC; pages_2m++; @@ -332,7 +333,7 @@ repeat: pte_ofs = pte_index((pfn<> 10, - (unsigned long)&_etext, (unsigned long)&_edata, - ((unsigned long)&_edata - (unsigned long)&_etext) >> 10, + (unsigned long)&_sdata, (unsigned long)&_edata, + ((unsigned long)&_edata - (unsigned long)&_sdata) >> 10, - (unsigned long)&_text, (unsigned long)&_etext, + ktla_ktva((unsigned long)&_text), ktla_ktva((unsigned long)&_etext), ((unsigned long)&_etext - (unsigned long)&_text) >> 10); /* @@ -883,6 +888,7 @@ void set_kernel_text_rw(void) if (!kernel_set_to_readonly) return; + start = ktla_ktva(start); pr_debug("Set kernel text: %lx - %lx for read write\n", start, start+size); @@ -897,6 +903,7 @@ void set_kernel_text_ro(void) if (!kernel_set_to_readonly) return; + start = ktla_ktva(start); pr_debug("Set kernel text: %lx - %lx for read only\n", start, start+size); @@ -925,6 +932,7 @@ void mark_rodata_ro(void) unsigned long start = PFN_ALIGN(_text); unsigned long size = PFN_ALIGN(_etext) - start; + start = ktla_ktva(start); set_pages_ro(virt_to_page(start), size >> PAGE_SHIFT); printk(KERN_INFO "Write protecting the kernel text: %luk\n", size >> 10); diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index 436a030..4f97ffc 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -75,7 +75,7 @@ early_param("gbpages", parse_direct_gbpages_on); * around without checking the pgd every time. */ -pteval_t __supported_pte_mask __read_mostly = ~_PAGE_IOMAP; +pteval_t __supported_pte_mask __read_only = ~(_PAGE_NX | _PAGE_IOMAP); EXPORT_SYMBOL_GPL(__supported_pte_mask); int force_personality32; @@ -108,12 +108,22 @@ void sync_global_pgds(unsigned long start, unsigned long end) for (address = start; address <= end; address += PGDIR_SIZE) { const pgd_t *pgd_ref = pgd_offset_k(address); + +#ifdef CONFIG_PAX_PER_CPU_PGD + unsigned long cpu; +#else struct page *page; +#endif if (pgd_none(*pgd_ref)) continue; spin_lock(&pgd_lock); + +#ifdef CONFIG_PAX_PER_CPU_PGD + for (cpu = 0; cpu < nr_cpu_ids; ++cpu) { + pgd_t *pgd = pgd_offset_cpu(cpu, address); +#else list_for_each_entry(page, &pgd_list, lru) { pgd_t *pgd; spinlock_t *pgt_lock; @@ -122,6 +132,7 @@ void sync_global_pgds(unsigned long start, unsigned long end) /* the pgt_lock only for Xen */ pgt_lock = &pgd_page_get_mm(page)->page_table_lock; spin_lock(pgt_lock); +#endif if (pgd_none(*pgd)) set_pgd(pgd, *pgd_ref); @@ -129,7 +140,10 @@ void sync_global_pgds(unsigned long start, unsigned long end) BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref)); +#ifndef CONFIG_PAX_PER_CPU_PGD spin_unlock(pgt_lock); +#endif + } spin_unlock(&pgd_lock); } @@ -162,7 +176,7 @@ static pud_t *fill_pud(pgd_t *pgd, unsigned long vaddr) { if (pgd_none(*pgd)) { pud_t *pud = (pud_t *)spp_getpage(); - pgd_populate(&init_mm, pgd, pud); + pgd_populate_kernel(&init_mm, pgd, pud); if (pud != pud_offset(pgd, 0)) printk(KERN_ERR "PAGETABLE BUG #00! %p <-> %p\n", pud, pud_offset(pgd, 0)); @@ -174,7 +188,7 @@ static pmd_t *fill_pmd(pud_t *pud, unsigned long vaddr) { if (pud_none(*pud)) { pmd_t *pmd = (pmd_t *) spp_getpage(); - pud_populate(&init_mm, pud, pmd); + pud_populate_kernel(&init_mm, pud, pmd); if (pmd != pmd_offset(pud, 0)) printk(KERN_ERR "PAGETABLE BUG #01! %p <-> %p\n", pmd, pmd_offset(pud, 0)); @@ -203,7 +217,9 @@ void set_pte_vaddr_pud(pud_t *pud_page, unsigned long vaddr, pte_t new_pte) pmd = fill_pmd(pud, vaddr); pte = fill_pte(pmd, vaddr); + pax_open_kernel(); set_pte(pte, new_pte); + pax_close_kernel(); /* * It's enough to flush this one mapping. @@ -262,14 +278,12 @@ static void __init __init_extra_mapping(unsigned long phys, unsigned long size, pgd = pgd_offset_k((unsigned long)__va(phys)); if (pgd_none(*pgd)) { pud = (pud_t *) spp_getpage(); - set_pgd(pgd, __pgd(__pa(pud) | _KERNPG_TABLE | - _PAGE_USER)); + set_pgd(pgd, __pgd(__pa(pud) | _PAGE_TABLE)); } pud = pud_offset(pgd, (unsigned long)__va(phys)); if (pud_none(*pud)) { pmd = (pmd_t *) spp_getpage(); - set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | - _PAGE_USER)); + set_pud(pud, __pud(__pa(pmd) | _PAGE_TABLE)); } pmd = pmd_offset(pud, phys); BUG_ON(!pmd_none(*pmd)); @@ -330,7 +344,7 @@ static __ref void *alloc_low_page(unsigned long *phys) if (pfn >= pgt_buf_top) panic("alloc_low_page: ran out of memory"); - adr = early_memremap(pfn * PAGE_SIZE, PAGE_SIZE); + adr = (void __force_kernel *)early_memremap(pfn * PAGE_SIZE, PAGE_SIZE); clear_page(adr); *phys = pfn * PAGE_SIZE; return adr; @@ -346,7 +360,7 @@ static __ref void *map_low_page(void *virt) phys = __pa(virt); left = phys & (PAGE_SIZE - 1); - adr = early_memremap(phys & PAGE_MASK, PAGE_SIZE); + adr = (void __force_kernel *)early_memremap(phys & PAGE_MASK, PAGE_SIZE); adr = (void *)(((unsigned long)adr) | left); return adr; @@ -546,7 +560,7 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end, unmap_low_page(pmd); spin_lock(&init_mm.page_table_lock); - pud_populate(&init_mm, pud, __va(pmd_phys)); + pud_populate_kernel(&init_mm, pud, __va(pmd_phys)); spin_unlock(&init_mm.page_table_lock); } __flush_tlb_all(); @@ -592,7 +606,7 @@ kernel_physical_mapping_init(unsigned long start, unmap_low_page(pud); spin_lock(&init_mm.page_table_lock); - pgd_populate(&init_mm, pgd, __va(pud_phys)); + pgd_populate_kernel(&init_mm, pgd, __va(pud_phys)); spin_unlock(&init_mm.page_table_lock); pgd_changed = true; } @@ -684,6 +698,12 @@ void __init mem_init(void) pci_iommu_alloc(); +#ifdef CONFIG_PAX_PER_CPU_PGD + clone_pgd_range(get_cpu_pgd(0) + KERNEL_PGD_BOUNDARY, + swapper_pg_dir + KERNEL_PGD_BOUNDARY, + KERNEL_PGD_PTRS); +#endif + /* clear_bss() already clear the empty_zero_page */ reservedpages = 0; @@ -844,8 +864,8 @@ int kern_addr_valid(unsigned long addr) static struct vm_area_struct gate_vma = { .vm_start = VSYSCALL_START, .vm_end = VSYSCALL_START + (VSYSCALL_MAPPED_PAGES * PAGE_SIZE), - .vm_page_prot = PAGE_READONLY_EXEC, - .vm_flags = VM_READ | VM_EXEC + .vm_page_prot = PAGE_READONLY, + .vm_flags = VM_READ }; struct vm_area_struct *get_gate_vma(struct mm_struct *mm) @@ -879,7 +899,7 @@ int in_gate_area_no_mm(unsigned long addr) const char *arch_vma_name(struct vm_area_struct *vma) { - if (vma->vm_mm && vma->vm_start == (long)vma->vm_mm->context.vdso) + if (vma->vm_mm && vma->vm_start == vma->vm_mm->context.vdso) return "[vdso]"; if (vma == &gate_vma) return "[vsyscall]"; diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index 7b179b4..6bd1777 100644 --- a/arch/x86/mm/iomap_32.c +++ b/arch/x86/mm/iomap_32.c @@ -64,7 +64,11 @@ void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) type = kmap_atomic_idx_push(); idx = type + KM_TYPE_NR * smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + + pax_open_kernel(); set_pte(kmap_pte - idx, pfn_pte(pfn, prot)); + pax_close_kernel(); + arch_flush_lazy_mmu_mode(); return (void *)vaddr; diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index be1ef57..55f0160 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -97,7 +97,7 @@ static void __iomem *__ioremap_caller(resource_size_t phys_addr, for (pfn = phys_addr >> PAGE_SHIFT; pfn <= last_pfn; pfn++) { int is_ram = page_is_ram(pfn); - if (is_ram && pfn_valid(pfn) && !PageReserved(pfn_to_page(pfn))) + if (is_ram && pfn_valid(pfn) && (pfn >= 0x100 || !PageReserved(pfn_to_page(pfn)))) return NULL; WARN_ON_ONCE(is_ram); } @@ -315,6 +315,9 @@ void *xlate_dev_mem_ptr(unsigned long phys) /* If page is RAM, we can use __va. Otherwise ioremap and unmap. */ if (page_is_ram(start >> PAGE_SHIFT)) +#ifdef CONFIG_HIGHMEM + if ((start >> PAGE_SHIFT) < max_low_pfn) +#endif return __va(phys); addr = (void __force *)ioremap_cache(start, PAGE_SIZE); @@ -344,7 +347,7 @@ static int __init early_ioremap_debug_setup(char *str) early_param("early_ioremap_debug", early_ioremap_debug_setup); static __initdata int after_paging_init; -static pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] __page_aligned_bss; +static pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] __read_only __aligned(PAGE_SIZE); static inline pmd_t * __init early_ioremap_pmd(unsigned long addr) { @@ -381,8 +384,7 @@ void __init early_ioremap_init(void) slot_virt[i] = __fix_to_virt(FIX_BTMAP_BEGIN - NR_FIX_BTMAPS*i); pmd = early_ioremap_pmd(fix_to_virt(FIX_BTMAP_BEGIN)); - memset(bm_pte, 0, sizeof(bm_pte)); - pmd_populate_kernel(&init_mm, pmd, bm_pte); + pmd_populate_user(&init_mm, pmd, bm_pte); /* * The boot-ioremap range spans multiple pmds, for which diff --git a/arch/x86/mm/kmemcheck/kmemcheck.c b/arch/x86/mm/kmemcheck/kmemcheck.c index d87dd6d..bf3fa66 100644 --- a/arch/x86/mm/kmemcheck/kmemcheck.c +++ b/arch/x86/mm/kmemcheck/kmemcheck.c @@ -622,9 +622,9 @@ bool kmemcheck_fault(struct pt_regs *regs, unsigned long address, * memory (e.g. tracked pages)? For now, we need this to avoid * invoking kmemcheck for PnP BIOS calls. */ - if (regs->flags & X86_VM_MASK) + if (v8086_mode(regs)) return false; - if (regs->cs != __KERNEL_CS) + if (regs->cs != __KERNEL_CS && regs->cs != __KERNEXEC_KERNEL_CS) return false; pte = kmemcheck_pte_lookup(address); diff --git a/arch/x86/mm/mmap.c b/arch/x86/mm/mmap.c index 845df68..1d8d29f 100644 --- a/arch/x86/mm/mmap.c +++ b/arch/x86/mm/mmap.c @@ -52,7 +52,7 @@ static unsigned int stack_maxrandom_size(void) * Leave an at least ~128 MB hole with possible stack randomization. */ #define MIN_GAP (128*1024*1024UL + stack_maxrandom_size()) -#define MAX_GAP (TASK_SIZE/6*5) +#define MAX_GAP (pax_task_size/6*5) static int mmap_is_legacy(void) { @@ -82,27 +82,40 @@ static unsigned long mmap_rnd(void) return rnd << PAGE_SHIFT; } -static unsigned long mmap_base(void) +static unsigned long mmap_base(struct mm_struct *mm) { unsigned long gap = rlimit(RLIMIT_STACK); + unsigned long pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif if (gap < MIN_GAP) gap = MIN_GAP; else if (gap > MAX_GAP) gap = MAX_GAP; - return PAGE_ALIGN(TASK_SIZE - gap - mmap_rnd()); + return PAGE_ALIGN(pax_task_size - gap - mmap_rnd()); } /* * Bottom-up (legacy) layout on X86_32 did not support randomization, X86_64 * does, but not when emulating X86_32 */ -static unsigned long mmap_legacy_base(void) +static unsigned long mmap_legacy_base(struct mm_struct *mm) { - if (mmap_is_ia32()) + if (mmap_is_ia32()) { + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + return SEGMEXEC_TASK_UNMAPPED_BASE; + else +#endif + return TASK_UNMAPPED_BASE; - else + } else return TASK_UNMAPPED_BASE + mmap_rnd(); } @@ -113,11 +126,23 @@ static unsigned long mmap_legacy_base(void) void arch_pick_mmap_layout(struct mm_struct *mm) { if (mmap_is_legacy()) { - mm->mmap_base = mmap_legacy_base(); + mm->mmap_base = mmap_legacy_base(mm); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { - mm->mmap_base = mmap_base(); + mm->mmap_base = mmap_base(mm); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff --git a/arch/x86/mm/mmio-mod.c b/arch/x86/mm/mmio-mod.c index dc0b727..dc9d71a 100644 --- a/arch/x86/mm/mmio-mod.c +++ b/arch/x86/mm/mmio-mod.c @@ -194,7 +194,7 @@ static void pre(struct kmmio_probe *p, struct pt_regs *regs, break; default: { - unsigned char *ip = (unsigned char *)instptr; + unsigned char *ip = (unsigned char *)ktla_ktva(instptr); my_trace->opcode = MMIO_UNKNOWN_OP; my_trace->width = 0; my_trace->value = (*ip) << 16 | *(ip + 1) << 8 | @@ -234,7 +234,7 @@ static void post(struct kmmio_probe *p, unsigned long condition, static void ioremap_trace_core(resource_size_t offset, unsigned long size, void __iomem *addr) { - static atomic_t next_id; + static atomic_unchecked_t next_id; struct remap_trace *trace = kmalloc(sizeof(*trace), GFP_KERNEL); /* These are page-unaligned. */ struct mmiotrace_map map = { @@ -258,7 +258,7 @@ static void ioremap_trace_core(resource_size_t offset, unsigned long size, .private = trace }, .phys = offset, - .id = atomic_inc_return(&next_id) + .id = atomic_inc_return_unchecked(&next_id) }; map.map_id = trace->id; diff --git a/arch/x86/mm/pageattr-test.c b/arch/x86/mm/pageattr-test.c index b008656..773eac2 100644 --- a/arch/x86/mm/pageattr-test.c +++ b/arch/x86/mm/pageattr-test.c @@ -36,7 +36,7 @@ enum { static int pte_testbit(pte_t pte) { - return pte_flags(pte) & _PAGE_UNUSED1; + return pte_flags(pte) & _PAGE_CPA_TEST; } struct split_state { diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index e1ebde3..b1e1db38 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -261,7 +261,7 @@ static inline pgprot_t static_protections(pgprot_t prot, unsigned long address, */ #ifdef CONFIG_PCI_BIOS if (pcibios_enabled && within(pfn, BIOS_BEGIN >> PAGE_SHIFT, BIOS_END >> PAGE_SHIFT)) - pgprot_val(forbidden) |= _PAGE_NX; + pgprot_val(forbidden) |= _PAGE_NX & __supported_pte_mask; #endif /* @@ -269,9 +269,10 @@ static inline pgprot_t static_protections(pgprot_t prot, unsigned long address, * Does not cover __inittext since that is gone later on. On * 64bit we do not enforce !NX on the low mapping */ - if (within(address, (unsigned long)_text, (unsigned long)_etext)) - pgprot_val(forbidden) |= _PAGE_NX; + if (within(address, ktla_ktva((unsigned long)_text), ktla_ktva((unsigned long)_etext))) + pgprot_val(forbidden) |= _PAGE_NX & __supported_pte_mask; +#ifdef CONFIG_DEBUG_RODATA /* * The .rodata section needs to be read-only. Using the pfn * catches all aliases. @@ -279,6 +280,7 @@ static inline pgprot_t static_protections(pgprot_t prot, unsigned long address, if (within(pfn, __pa((unsigned long)__start_rodata) >> PAGE_SHIFT, __pa((unsigned long)__end_rodata) >> PAGE_SHIFT)) pgprot_val(forbidden) |= _PAGE_RW; +#endif #if defined(CONFIG_X86_64) && defined(CONFIG_DEBUG_RODATA) /* @@ -317,6 +319,13 @@ static inline pgprot_t static_protections(pgprot_t prot, unsigned long address, } #endif +#ifdef CONFIG_PAX_KERNEXEC + if (within(pfn, __pa((unsigned long)&_text), __pa((unsigned long)&_sdata))) { + pgprot_val(forbidden) |= _PAGE_RW; + pgprot_val(forbidden) |= _PAGE_NX & __supported_pte_mask; + } +#endif + prot = __pgprot(pgprot_val(prot) & ~pgprot_val(forbidden)); return prot; @@ -369,23 +378,37 @@ EXPORT_SYMBOL_GPL(lookup_address); static void __set_pmd_pte(pte_t *kpte, unsigned long address, pte_t pte) { /* change init_mm */ + pax_open_kernel(); set_pte_atomic(kpte, pte); + #ifdef CONFIG_X86_32 if (!SHARED_KERNEL_PMD) { + +#ifdef CONFIG_PAX_PER_CPU_PGD + unsigned long cpu; +#else struct page *page; +#endif +#ifdef CONFIG_PAX_PER_CPU_PGD + for (cpu = 0; cpu < nr_cpu_ids; ++cpu) { + pgd_t *pgd = get_cpu_pgd(cpu); +#else list_for_each_entry(page, &pgd_list, lru) { - pgd_t *pgd; + pgd_t *pgd = (pgd_t *)page_address(page); +#endif + pud_t *pud; pmd_t *pmd; - pgd = (pgd_t *)page_address(page) + pgd_index(address); + pgd += pgd_index(address); pud = pud_offset(pgd, address); pmd = pmd_offset(pud, address); set_pte_atomic((pte_t *)pmd, pte); } } #endif + pax_close_kernel(); } static int diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c index f6ff57b..481690f 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -361,7 +361,7 @@ int free_memtype(u64 start, u64 end) if (!entry) { printk(KERN_INFO "%s:%d freeing invalid memtype %Lx-%Lx\n", - current->comm, current->pid, start, end); + current->comm, task_pid_nr(current), start, end); return -EINVAL; } @@ -492,8 +492,8 @@ static inline int range_is_allowed(unsigned long pfn, unsigned long size) while (cursor < to) { if (!devmem_is_allowed(pfn)) { printk(KERN_INFO - "Program %s tried to access /dev/mem between %Lx->%Lx.\n", - current->comm, from, to); + "Program %s tried to access /dev/mem between %Lx->%Lx (%Lx).\n", + current->comm, from, to, cursor); return 0; } cursor += PAGE_SIZE; @@ -557,7 +557,7 @@ int kernel_map_sync_memtype(u64 base, unsigned long size, unsigned long flags) printk(KERN_INFO "%s:%d ioremap_change_attr failed %s " "for %Lx-%Lx\n", - current->comm, current->pid, + current->comm, task_pid_nr(current), cattr_name(flags), base, (unsigned long long)(base + size)); return -EINVAL; @@ -593,7 +593,7 @@ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot, if (want_flags != flags) { printk(KERN_WARNING "%s:%d map pfn RAM range req %s for %Lx-%Lx, got %s\n", - current->comm, current->pid, + current->comm, task_pid_nr(current), cattr_name(want_flags), (unsigned long long)paddr, (unsigned long long)(paddr + size), @@ -615,7 +615,7 @@ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot, free_memtype(paddr, paddr + size); printk(KERN_ERR "%s:%d map pfn expected mapping type %s" " for %Lx-%Lx, got %s\n", - current->comm, current->pid, + current->comm, task_pid_nr(current), cattr_name(want_flags), (unsigned long long)paddr, (unsigned long long)(paddr + size), diff --git a/arch/x86/mm/pf_in.c b/arch/x86/mm/pf_in.c index 9f0614d..92ae64a 100644 --- a/arch/x86/mm/pf_in.c +++ b/arch/x86/mm/pf_in.c @@ -148,7 +148,7 @@ enum reason_type get_ins_type(unsigned long ins_addr) int i; enum reason_type rv = OTHERS; - p = (unsigned char *)ins_addr; + p = (unsigned char *)ktla_ktva(ins_addr); p += skip_prefix(p, &prf); p += get_opcode(p, &opcode); @@ -168,7 +168,7 @@ static unsigned int get_ins_reg_width(unsigned long ins_addr) struct prefix_bits prf; int i; - p = (unsigned char *)ins_addr; + p = (unsigned char *)ktla_ktva(ins_addr); p += skip_prefix(p, &prf); p += get_opcode(p, &opcode); @@ -191,7 +191,7 @@ unsigned int get_ins_mem_width(unsigned long ins_addr) struct prefix_bits prf; int i; - p = (unsigned char *)ins_addr; + p = (unsigned char *)ktla_ktva(ins_addr); p += skip_prefix(p, &prf); p += get_opcode(p, &opcode); @@ -415,7 +415,7 @@ unsigned long get_ins_reg_val(unsigned long ins_addr, struct pt_regs *regs) struct prefix_bits prf; int i; - p = (unsigned char *)ins_addr; + p = (unsigned char *)ktla_ktva(ins_addr); p += skip_prefix(p, &prf); p += get_opcode(p, &opcode); for (i = 0; i < ARRAY_SIZE(reg_rop); i++) @@ -470,7 +470,7 @@ unsigned long get_ins_imm_val(unsigned long ins_addr) struct prefix_bits prf; int i; - p = (unsigned char *)ins_addr; + p = (unsigned char *)ktla_ktva(ins_addr); p += skip_prefix(p, &prf); p += get_opcode(p, &opcode); for (i = 0; i < ARRAY_SIZE(imm_wop); i++) diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c index 8573b83..7d9628f 100644 --- a/arch/x86/mm/pgtable.c +++ b/arch/x86/mm/pgtable.c @@ -84,10 +84,60 @@ static inline void pgd_list_del(pgd_t *pgd) list_del(&page->lru); } -#define UNSHARED_PTRS_PER_PGD \ - (SHARED_KERNEL_PMD ? KERNEL_PGD_BOUNDARY : PTRS_PER_PGD) +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +pgdval_t clone_pgd_mask __read_only = ~_PAGE_PRESENT; +void __shadow_user_pgds(pgd_t *dst, const pgd_t *src, int count) +{ + while (count--) + *dst++ = __pgd((pgd_val(*src++) | (_PAGE_NX & __supported_pte_mask)) & ~_PAGE_USER); +} +#endif +#ifdef CONFIG_PAX_PER_CPU_PGD +void __clone_user_pgds(pgd_t *dst, const pgd_t *src, int count) +{ + while (count--) { + pgd_t pgd; + +#ifdef CONFIG_X86_64 + pgd = __pgd(pgd_val(*src++) | _PAGE_USER); +#else + pgd = *src++; +#endif + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + pgd = __pgd(pgd_val(pgd) & clone_pgd_mask); +#endif + + *dst++ = pgd; + } + +} +#endif + +#ifdef CONFIG_X86_64 +#define pxd_t pud_t +#define pyd_t pgd_t +#define paravirt_release_pxd(pfn) paravirt_release_pud(pfn) +#define pxd_free(mm, pud) pud_free((mm), (pud)) +#define pyd_populate(mm, pgd, pud) pgd_populate((mm), (pgd), (pud)) +#define pyd_offset(mm, address) pgd_offset((mm), (address)) +#define PYD_SIZE PGDIR_SIZE +#else +#define pxd_t pmd_t +#define pyd_t pud_t +#define paravirt_release_pxd(pfn) paravirt_release_pmd(pfn) +#define pxd_free(mm, pud) pmd_free((mm), (pud)) +#define pyd_populate(mm, pgd, pud) pud_populate((mm), (pgd), (pud)) +#define pyd_offset(mm, address) pud_offset((mm), (address)) +#define PYD_SIZE PUD_SIZE +#endif + +#ifdef CONFIG_PAX_PER_CPU_PGD +static inline void pgd_ctor(struct mm_struct *mm, pgd_t *pgd) {} +static inline void pgd_dtor(pgd_t *pgd) {} +#else static void pgd_set_mm(pgd_t *pgd, struct mm_struct *mm) { BUILD_BUG_ON(sizeof(virt_to_page(pgd)->index) < sizeof(mm)); @@ -128,6 +178,7 @@ static void pgd_dtor(pgd_t *pgd) pgd_list_del(pgd); spin_unlock(&pgd_lock); } +#endif /* * List of all pgd's needed for non-PAE so it can invalidate entries @@ -140,7 +191,7 @@ static void pgd_dtor(pgd_t *pgd) * -- wli */ -#ifdef CONFIG_X86_PAE +#if defined(CONFIG_X86_32) && defined(CONFIG_X86_PAE) /* * In PAE mode, we need to do a cr3 reload (=tlb flush) when * updating the top-level pagetable entries to guarantee the @@ -152,7 +203,7 @@ static void pgd_dtor(pgd_t *pgd) * not shared between pagetables (!SHARED_KERNEL_PMDS), we allocate * and initialize the kernel pmds here. */ -#define PREALLOCATED_PMDS UNSHARED_PTRS_PER_PGD +#define PREALLOCATED_PXDS (SHARED_KERNEL_PMD ? KERNEL_PGD_BOUNDARY : PTRS_PER_PGD) void pud_populate(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd) { @@ -170,36 +221,38 @@ void pud_populate(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd) */ flush_tlb_mm(mm); } +#elif defined(CONFIG_X86_64) && defined(CONFIG_PAX_PER_CPU_PGD) +#define PREALLOCATED_PXDS USER_PGD_PTRS #else /* !CONFIG_X86_PAE */ /* No need to prepopulate any pagetable entries in non-PAE modes. */ -#define PREALLOCATED_PMDS 0 +#define PREALLOCATED_PXDS 0 #endif /* CONFIG_X86_PAE */ -static void free_pmds(pmd_t *pmds[]) +static void free_pxds(pxd_t *pxds[]) { int i; - for(i = 0; i < PREALLOCATED_PMDS; i++) - if (pmds[i]) - free_page((unsigned long)pmds[i]); + for(i = 0; i < PREALLOCATED_PXDS; i++) + if (pxds[i]) + free_page((unsigned long)pxds[i]); } -static int preallocate_pmds(pmd_t *pmds[]) +static int preallocate_pxds(pxd_t *pxds[]) { int i; bool failed = false; - for(i = 0; i < PREALLOCATED_PMDS; i++) { - pmd_t *pmd = (pmd_t *)__get_free_page(PGALLOC_GFP); - if (pmd == NULL) + for(i = 0; i < PREALLOCATED_PXDS; i++) { + pxd_t *pxd = (pxd_t *)__get_free_page(PGALLOC_GFP); + if (pxd == NULL) failed = true; - pmds[i] = pmd; + pxds[i] = pxd; } if (failed) { - free_pmds(pmds); + free_pxds(pxds); return -ENOMEM; } @@ -212,51 +265,55 @@ static int preallocate_pmds(pmd_t *pmds[]) * preallocate which never got a corresponding vma will need to be * freed manually. */ -static void pgd_mop_up_pmds(struct mm_struct *mm, pgd_t *pgdp) +static void pgd_mop_up_pxds(struct mm_struct *mm, pgd_t *pgdp) { int i; - for(i = 0; i < PREALLOCATED_PMDS; i++) { + for(i = 0; i < PREALLOCATED_PXDS; i++) { pgd_t pgd = pgdp[i]; if (pgd_val(pgd) != 0) { - pmd_t *pmd = (pmd_t *)pgd_page_vaddr(pgd); + pxd_t *pxd = (pxd_t *)pgd_page_vaddr(pgd); - pgdp[i] = native_make_pgd(0); + set_pgd(pgdp + i, native_make_pgd(0)); - paravirt_release_pmd(pgd_val(pgd) >> PAGE_SHIFT); - pmd_free(mm, pmd); + paravirt_release_pxd(pgd_val(pgd) >> PAGE_SHIFT); + pxd_free(mm, pxd); } } } -static void pgd_prepopulate_pmd(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmds[]) +static void pgd_prepopulate_pxd(struct mm_struct *mm, pgd_t *pgd, pxd_t *pxds[]) { - pud_t *pud; + pyd_t *pyd; unsigned long addr; int i; - if (PREALLOCATED_PMDS == 0) /* Work around gcc-3.4.x bug */ + if (PREALLOCATED_PXDS == 0) /* Work around gcc-3.4.x bug */ return; - pud = pud_offset(pgd, 0); +#ifdef CONFIG_X86_64 + pyd = pyd_offset(mm, 0L); +#else + pyd = pyd_offset(pgd, 0L); +#endif - for (addr = i = 0; i < PREALLOCATED_PMDS; - i++, pud++, addr += PUD_SIZE) { - pmd_t *pmd = pmds[i]; + for (addr = i = 0; i < PREALLOCATED_PXDS; + i++, pyd++, addr += PYD_SIZE) { + pxd_t *pxd = pxds[i]; if (i >= KERNEL_PGD_BOUNDARY) - memcpy(pmd, (pmd_t *)pgd_page_vaddr(swapper_pg_dir[i]), - sizeof(pmd_t) * PTRS_PER_PMD); + memcpy(pxd, (pxd_t *)pgd_page_vaddr(swapper_pg_dir[i]), + sizeof(pxd_t) * PTRS_PER_PMD); - pud_populate(mm, pud, pmd); + pyd_populate(mm, pyd, pxd); } } pgd_t *pgd_alloc(struct mm_struct *mm) { pgd_t *pgd; - pmd_t *pmds[PREALLOCATED_PMDS]; + pxd_t *pxds[PREALLOCATED_PXDS]; pgd = (pgd_t *)__get_free_page(PGALLOC_GFP); @@ -265,11 +322,11 @@ pgd_t *pgd_alloc(struct mm_struct *mm) mm->pgd = pgd; - if (preallocate_pmds(pmds) != 0) + if (preallocate_pxds(pxds) != 0) goto out_free_pgd; if (paravirt_pgd_alloc(mm) != 0) - goto out_free_pmds; + goto out_free_pxds; /* * Make sure that pre-populating the pmds is atomic with @@ -279,14 +336,14 @@ pgd_t *pgd_alloc(struct mm_struct *mm) spin_lock(&pgd_lock); pgd_ctor(mm, pgd); - pgd_prepopulate_pmd(mm, pgd, pmds); + pgd_prepopulate_pxd(mm, pgd, pxds); spin_unlock(&pgd_lock); return pgd; -out_free_pmds: - free_pmds(pmds); +out_free_pxds: + free_pxds(pxds); out_free_pgd: free_page((unsigned long)pgd); out: @@ -295,7 +352,7 @@ out: void pgd_free(struct mm_struct *mm, pgd_t *pgd) { - pgd_mop_up_pmds(mm, pgd); + pgd_mop_up_pxds(mm, pgd); pgd_dtor(pgd); paravirt_pgd_free(mm, pgd); free_page((unsigned long)pgd); diff --git a/arch/x86/mm/pgtable_32.c b/arch/x86/mm/pgtable_32.c index cac7184..09a39fa 100644 --- a/arch/x86/mm/pgtable_32.c +++ b/arch/x86/mm/pgtable_32.c @@ -48,10 +48,13 @@ void set_pte_vaddr(unsigned long vaddr, pte_t pteval) return; } pte = pte_offset_kernel(pmd, vaddr); + + pax_open_kernel(); if (pte_val(pteval)) set_pte_at(&init_mm, vaddr, pte, pteval); else pte_clear(&init_mm, vaddr, pte); + pax_close_kernel(); /* * It's enough to flush this one mapping. diff --git a/arch/x86/mm/setup_nx.c b/arch/x86/mm/setup_nx.c index 410531d..0f16030 100644 --- a/arch/x86/mm/setup_nx.c +++ b/arch/x86/mm/setup_nx.c @@ -5,8 +5,10 @@ #include #include +#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) static int disable_nx __cpuinitdata; +#ifndef CONFIG_PAX_PAGEEXEC /* * noexec = on|off * @@ -28,12 +30,17 @@ static int __init noexec_setup(char *str) return 0; } early_param("noexec", noexec_setup); +#endif + +#endif void __cpuinit x86_configure_nx(void) { +#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) if (cpu_has_nx && !disable_nx) __supported_pte_mask |= _PAGE_NX; else +#endif __supported_pte_mask &= ~_PAGE_NX; } diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c index d6c0418..06a0ad5 100644 --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -65,7 +65,11 @@ void leave_mm(int cpu) BUG(); cpumask_clear_cpu(cpu, mm_cpumask(percpu_read(cpu_tlbstate.active_mm))); + +#ifndef CONFIG_PAX_PER_CPU_PGD load_cr3(swapper_pg_dir); +#endif + } EXPORT_SYMBOL_GPL(leave_mm); diff --git a/arch/x86/net/bpf_jit.S b/arch/x86/net/bpf_jit.S index 6687022..ceabcfa 100644 --- a/arch/x86/net/bpf_jit.S +++ b/arch/x86/net/bpf_jit.S @@ -9,6 +9,7 @@ */ #include #include +#include /* * Calling convention : @@ -35,6 +36,7 @@ sk_load_word: jle bpf_slow_path_word mov (SKBDATA,%rsi),%eax bswap %eax /* ntohl() */ + pax_force_retaddr ret @@ -53,6 +55,7 @@ sk_load_half: jle bpf_slow_path_half movzwl (SKBDATA,%rsi),%eax rol $8,%ax # ntohs() + pax_force_retaddr ret sk_load_byte_ind: @@ -66,6 +69,7 @@ sk_load_byte: cmp %esi,%r9d /* if (offset >= hlen) goto bpf_slow_path_byte */ jle bpf_slow_path_byte movzbl (SKBDATA,%rsi),%eax + pax_force_retaddr ret /** @@ -82,6 +86,7 @@ ENTRY(sk_load_byte_msh) movzbl (SKBDATA,%rsi),%ebx and $15,%bl shl $2,%bl + pax_force_retaddr ret CFI_ENDPROC ENDPROC(sk_load_byte_msh) @@ -91,6 +96,7 @@ bpf_error: xor %eax,%eax mov -8(%rbp),%rbx leaveq + pax_force_retaddr ret /* rsi contains offset and can be scratched */ @@ -113,6 +119,7 @@ bpf_slow_path_word: js bpf_error mov -12(%rbp),%eax bswap %eax + pax_force_retaddr ret bpf_slow_path_half: @@ -121,12 +128,14 @@ bpf_slow_path_half: mov -12(%rbp),%ax rol $8,%ax movzwl %ax,%eax + pax_force_retaddr ret bpf_slow_path_byte: bpf_slow_path_common(1) js bpf_error movzbl -12(%rbp),%eax + pax_force_retaddr ret bpf_slow_path_byte_msh: @@ -137,4 +146,5 @@ bpf_slow_path_byte_msh: and $15,%al shl $2,%al xchg %eax,%ebx + pax_force_retaddr ret diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index 5a5b6e4..201d42e 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -117,6 +117,10 @@ static inline void bpf_flush_icache(void *start, void *end) set_fs(old_fs); } +struct bpf_jit_work { + struct work_struct work; + void *image; +}; void bpf_jit_compile(struct sk_filter *fp) { @@ -141,6 +145,10 @@ void bpf_jit_compile(struct sk_filter *fp) if (addrs == NULL) return; + fp->work = kmalloc(sizeof(*fp->work), GFP_KERNEL); + if (!fp->work) + goto out; + /* Before first pass, make a rough estimation of addrs[] * each bpf instruction is translated to less than 64 bytes */ @@ -477,7 +485,7 @@ void bpf_jit_compile(struct sk_filter *fp) common_load: seen |= SEEN_DATAREF; if ((int)K < 0) { /* Abort the JIT because __load_pointer() is needed. */ - goto out; + goto error; } t_offset = func - (image + addrs[i]); EMIT1_off32(0xbe, K); /* mov imm32,%esi */ @@ -492,7 +500,7 @@ common_load: seen |= SEEN_DATAREF; case BPF_S_LDX_B_MSH: if ((int)K < 0) { /* Abort the JIT because __load_pointer() is needed. */ - goto out; + goto error; } seen |= SEEN_DATAREF | SEEN_XREG; t_offset = sk_load_byte_msh - (image + addrs[i]); @@ -582,17 +590,18 @@ cond_branch: f_offset = addrs[i + filter[i].jf] - addrs[i]; break; default: /* hmm, too complex filter, give up with jit compiler */ - goto out; + goto error; } ilen = prog - temp; if (image) { if (unlikely(proglen + ilen > oldproglen)) { pr_err("bpb_jit_compile fatal error\n"); - kfree(addrs); - module_free(NULL, image); - return; + module_free_exec(NULL, image); + goto error; } + pax_open_kernel(); memcpy(image + proglen, temp, ilen); + pax_close_kernel(); } proglen += ilen; addrs[i] = proglen; @@ -613,11 +622,9 @@ cond_branch: f_offset = addrs[i + filter[i].jf] - addrs[i]; break; } if (proglen == oldproglen) { - image = module_alloc(max_t(unsigned int, - proglen, - sizeof(struct work_struct))); + image = module_alloc_exec(proglen); if (!image) - goto out; + goto error; } oldproglen = proglen; } @@ -633,7 +640,10 @@ cond_branch: f_offset = addrs[i + filter[i].jf] - addrs[i]; bpf_flush_icache(image, image + proglen); fp->bpf_func = (void *)image; - } + } else +error: + kfree(fp->work); + out: kfree(addrs); return; @@ -641,18 +651,20 @@ out: static void jit_free_defer(struct work_struct *arg) { - module_free(NULL, arg); + module_free_exec(NULL, ((struct bpf_jit_work *)arg)->image); + kfree(arg); } /* run from softirq, we must use a work_struct to call - * module_free() from process context + * module_free_exec() from process context */ void bpf_jit_free(struct sk_filter *fp) { if (fp->bpf_func != sk_run_filter) { - struct work_struct *work = (struct work_struct *)fp->bpf_func; + struct work_struct *work = &fp->work->work; INIT_WORK(work, jit_free_defer); + fp->work->image = fp->bpf_func; schedule_work(work); } } diff --git a/arch/x86/oprofile/backtrace.c b/arch/x86/oprofile/backtrace.c index bff89df..377758a 100644 --- a/arch/x86/oprofile/backtrace.c +++ b/arch/x86/oprofile/backtrace.c @@ -46,11 +46,11 @@ dump_user_backtrace_32(struct stack_frame_ia32 *head) struct stack_frame_ia32 *fp; unsigned long bytes; - bytes = copy_from_user_nmi(bufhead, head, sizeof(bufhead)); + bytes = copy_from_user_nmi(bufhead, (const char __force_user *)head, sizeof(bufhead)); if (bytes != sizeof(bufhead)) return NULL; - fp = (struct stack_frame_ia32 *) compat_ptr(bufhead[0].next_frame); + fp = (struct stack_frame_ia32 __force_kernel *) compat_ptr(bufhead[0].next_frame); oprofile_add_trace(bufhead[0].return_address); @@ -92,7 +92,7 @@ static struct stack_frame *dump_user_backtrace(struct stack_frame *head) struct stack_frame bufhead[2]; unsigned long bytes; - bytes = copy_from_user_nmi(bufhead, head, sizeof(bufhead)); + bytes = copy_from_user_nmi(bufhead, (const char __force_user *)head, sizeof(bufhead)); if (bytes != sizeof(bufhead)) return NULL; @@ -111,7 +111,7 @@ x86_backtrace(struct pt_regs * const regs, unsigned int depth) { struct stack_frame *head = (struct stack_frame *)frame_pointer(regs); - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { unsigned long stack = kernel_stack_pointer(regs); if (depth) dump_trace(NULL, regs, (unsigned long *)stack, 0, diff --git a/arch/x86/pci/mrst.c b/arch/x86/pci/mrst.c index cb29191..036766d 100644 --- a/arch/x86/pci/mrst.c +++ b/arch/x86/pci/mrst.c @@ -234,7 +234,9 @@ int __init pci_mrst_init(void) printk(KERN_INFO "Moorestown platform detected, using MRST PCI ops\n"); pci_mmcfg_late_init(); pcibios_enable_irq = mrst_pci_irq_enable; - pci_root_ops = pci_mrst_ops; + pax_open_kernel(); + memcpy((void *)&pci_root_ops, &pci_mrst_ops, sizeof(pci_mrst_ops)); + pax_close_kernel(); /* Continue with standard init */ return 1; } diff --git a/arch/x86/pci/pcbios.c b/arch/x86/pci/pcbios.c index da8fe05..7ee6704 100644 --- a/arch/x86/pci/pcbios.c +++ b/arch/x86/pci/pcbios.c @@ -79,50 +79,93 @@ union bios32 { static struct { unsigned long address; unsigned short segment; -} bios32_indirect = { 0, __KERNEL_CS }; +} bios32_indirect __read_only = { 0, __PCIBIOS_CS }; /* * Returns the entry point for the given service, NULL on error */ -static unsigned long bios32_service(unsigned long service) +static unsigned long __devinit bios32_service(unsigned long service) { unsigned char return_code; /* %al */ unsigned long address; /* %ebx */ unsigned long length; /* %ecx */ unsigned long entry; /* %edx */ unsigned long flags; + struct desc_struct d, *gdt; local_irq_save(flags); - __asm__("lcall *(%%edi); cld" + + gdt = get_cpu_gdt_table(smp_processor_id()); + + pack_descriptor(&d, 0UL, 0xFFFFFUL, 0x9B, 0xC); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_CS, &d, DESCTYPE_S); + pack_descriptor(&d, 0UL, 0xFFFFFUL, 0x93, 0xC); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_DS, &d, DESCTYPE_S); + + __asm__("movw %w7, %%ds; lcall *(%%edi); push %%ss; pop %%ds; cld" : "=a" (return_code), "=b" (address), "=c" (length), "=d" (entry) : "0" (service), "1" (0), - "D" (&bios32_indirect)); + "D" (&bios32_indirect), + "r"(__PCIBIOS_DS) + : "memory"); + + pax_open_kernel(); + gdt[GDT_ENTRY_PCIBIOS_CS].a = 0; + gdt[GDT_ENTRY_PCIBIOS_CS].b = 0; + gdt[GDT_ENTRY_PCIBIOS_DS].a = 0; + gdt[GDT_ENTRY_PCIBIOS_DS].b = 0; + pax_close_kernel(); + local_irq_restore(flags); switch (return_code) { - case 0: - return address + entry; - case 0x80: /* Not present */ - printk(KERN_WARNING "bios32_service(0x%lx): not present\n", service); - return 0; - default: /* Shouldn't happen */ - printk(KERN_WARNING "bios32_service(0x%lx): returned 0x%x -- BIOS bug!\n", - service, return_code); + case 0: { + int cpu; + unsigned char flags; + + printk(KERN_INFO "bios32_service: base:%08lx length:%08lx entry:%08lx\n", address, length, entry); + if (address >= 0xFFFF0 || length > 0x100000 - address || length <= entry) { + printk(KERN_WARNING "bios32_service: not valid\n"); return 0; + } + address = address + PAGE_OFFSET; + length += 16UL; /* some BIOSs underreport this... */ + flags = 4; + if (length >= 64*1024*1024) { + length >>= PAGE_SHIFT; + flags |= 8; + } + + for (cpu = 0; cpu < nr_cpu_ids; cpu++) { + gdt = get_cpu_gdt_table(cpu); + pack_descriptor(&d, address, length, 0x9b, flags); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_CS, &d, DESCTYPE_S); + pack_descriptor(&d, address, length, 0x93, flags); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_DS, &d, DESCTYPE_S); + } + return entry; + } + case 0x80: /* Not present */ + printk(KERN_WARNING "bios32_service(0x%lx): not present\n", service); + return 0; + default: /* Shouldn't happen */ + printk(KERN_WARNING "bios32_service(0x%lx): returned 0x%x -- BIOS bug!\n", + service, return_code); + return 0; } } static struct { unsigned long address; unsigned short segment; -} pci_indirect = { 0, __KERNEL_CS }; +} pci_indirect __read_only = { 0, __PCIBIOS_CS }; -static int pci_bios_present; +static int pci_bios_present __read_only; static int __devinit check_pcibios(void) { @@ -131,11 +174,13 @@ static int __devinit check_pcibios(void) unsigned long flags, pcibios_entry; if ((pcibios_entry = bios32_service(PCI_SERVICE))) { - pci_indirect.address = pcibios_entry + PAGE_OFFSET; + pci_indirect.address = pcibios_entry; local_irq_save(flags); - __asm__( - "lcall *(%%edi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%edi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -144,7 +189,8 @@ static int __devinit check_pcibios(void) "=b" (ebx), "=c" (ecx) : "1" (PCIBIOS_PCI_BIOS_PRESENT), - "D" (&pci_indirect) + "D" (&pci_indirect), + "r" (__PCIBIOS_DS) : "memory"); local_irq_restore(flags); @@ -189,7 +235,10 @@ static int pci_bios_read(unsigned int seg, unsigned int bus, switch (len) { case 1: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -198,7 +247,8 @@ static int pci_bios_read(unsigned int seg, unsigned int bus, : "1" (PCIBIOS_READ_CONFIG_BYTE), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); /* * Zero-extend the result beyond 8 bits, do not trust the * BIOS having done it: @@ -206,7 +256,10 @@ static int pci_bios_read(unsigned int seg, unsigned int bus, *value &= 0xff; break; case 2: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -215,7 +268,8 @@ static int pci_bios_read(unsigned int seg, unsigned int bus, : "1" (PCIBIOS_READ_CONFIG_WORD), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); /* * Zero-extend the result beyond 16 bits, do not trust the * BIOS having done it: @@ -223,7 +277,10 @@ static int pci_bios_read(unsigned int seg, unsigned int bus, *value &= 0xffff; break; case 4: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -232,7 +289,8 @@ static int pci_bios_read(unsigned int seg, unsigned int bus, : "1" (PCIBIOS_READ_CONFIG_DWORD), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; } @@ -256,7 +314,10 @@ static int pci_bios_write(unsigned int seg, unsigned int bus, switch (len) { case 1: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -265,10 +326,14 @@ static int pci_bios_write(unsigned int seg, unsigned int bus, "c" (value), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; case 2: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -277,10 +342,14 @@ static int pci_bios_write(unsigned int seg, unsigned int bus, "c" (value), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; case 4: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -289,7 +358,8 @@ static int pci_bios_write(unsigned int seg, unsigned int bus, "c" (value), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; } @@ -394,10 +464,13 @@ struct irq_routing_table * pcibios_get_irq_routing_table(void) DBG("PCI: Fetching IRQ routing table... "); __asm__("push %%es\n\t" + "movw %w8, %%ds\n\t" "push %%ds\n\t" "pop %%es\n\t" - "lcall *(%%esi); cld\n\t" + "lcall *%%ss:(%%esi); cld\n\t" "pop %%es\n\t" + "push %%ss\n\t" + "pop %%ds\n" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -408,7 +481,8 @@ struct irq_routing_table * pcibios_get_irq_routing_table(void) "1" (0), "D" ((long) &opt), "S" (&pci_indirect), - "m" (opt) + "m" (opt), + "r" (__PCIBIOS_DS) : "memory"); DBG("OK ret=%d, size=%d, map=%x\n", ret, opt.size, map); if (ret & 0xff00) @@ -432,7 +506,10 @@ int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq) { int ret; - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w5, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -440,7 +517,8 @@ int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq) : "0" (PCIBIOS_SET_PCI_HW_INT), "b" ((dev->bus->number << 8) | dev->devfn), "c" ((irq << 8) | (pin + 10)), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); return !(ret & 0xff00); } EXPORT_SYMBOL(pcibios_set_irq_routing); diff --git a/arch/x86/platform/efi/efi_32.c b/arch/x86/platform/efi/efi_32.c index 40e4469..1ab536e 100644 --- a/arch/x86/platform/efi/efi_32.c +++ b/arch/x86/platform/efi/efi_32.c @@ -44,11 +44,22 @@ void efi_call_phys_prelog(void) { struct desc_ptr gdt_descr; +#ifdef CONFIG_PAX_KERNEXEC + struct desc_struct d; +#endif + local_irq_save(efi_rt_eflags); load_cr3(initial_page_table); __flush_tlb_all(); +#ifdef CONFIG_PAX_KERNEXEC + pack_descriptor(&d, 0, 0xFFFFF, 0x9B, 0xC); + write_gdt_entry(get_cpu_gdt_table(0), GDT_ENTRY_KERNEXEC_EFI_CS, &d, DESCTYPE_S); + pack_descriptor(&d, 0, 0xFFFFF, 0x93, 0xC); + write_gdt_entry(get_cpu_gdt_table(0), GDT_ENTRY_KERNEXEC_EFI_DS, &d, DESCTYPE_S); +#endif + gdt_descr.address = __pa(get_cpu_gdt_table(0)); gdt_descr.size = GDT_SIZE - 1; load_gdt(&gdt_descr); @@ -58,6 +69,14 @@ void efi_call_phys_epilog(void) { struct desc_ptr gdt_descr; +#ifdef CONFIG_PAX_KERNEXEC + struct desc_struct d; + + memset(&d, 0, sizeof d); + write_gdt_entry(get_cpu_gdt_table(0), GDT_ENTRY_KERNEXEC_EFI_CS, &d, DESCTYPE_S); + write_gdt_entry(get_cpu_gdt_table(0), GDT_ENTRY_KERNEXEC_EFI_DS, &d, DESCTYPE_S); +#endif + gdt_descr.address = (unsigned long)get_cpu_gdt_table(0); gdt_descr.size = GDT_SIZE - 1; load_gdt(&gdt_descr); diff --git a/arch/x86/platform/efi/efi_stub_32.S b/arch/x86/platform/efi/efi_stub_32.S index fbe66e6..c5c0dd2 100644 --- a/arch/x86/platform/efi/efi_stub_32.S +++ b/arch/x86/platform/efi/efi_stub_32.S @@ -6,7 +6,9 @@ */ #include +#include #include +#include /* * efi_call_phys(void *, ...) is a function with variable parameters. @@ -20,7 +22,7 @@ * service functions will comply with gcc calling convention, too. */ -.text +__INIT ENTRY(efi_call_phys) /* * 0. The function can only be called in Linux kernel. So CS has been @@ -36,9 +38,11 @@ ENTRY(efi_call_phys) * The mapping of lower virtual memory has been created in prelog and * epilog. */ - movl $1f, %edx - subl $__PAGE_OFFSET, %edx - jmp *%edx + movl $(__KERNEXEC_EFI_DS), %edx + mov %edx, %ds + mov %edx, %es + mov %edx, %ss + ljmp $(__KERNEXEC_EFI_CS),$1f-__PAGE_OFFSET 1: /* @@ -47,14 +51,8 @@ ENTRY(efi_call_phys) * parameter 2, ..., param n. To make things easy, we save the return * address of efi_call_phys in a global variable. */ - popl %edx - movl %edx, saved_return_addr - /* get the function pointer into ECX*/ - popl %ecx - movl %ecx, efi_rt_function_ptr - movl $2f, %edx - subl $__PAGE_OFFSET, %edx - pushl %edx + popl (saved_return_addr) + popl (efi_rt_function_ptr) /* * 3. Clear PG bit in %CR0. @@ -73,9 +71,8 @@ ENTRY(efi_call_phys) /* * 5. Call the physical function. */ - jmp *%ecx + call *(efi_rt_function_ptr-__PAGE_OFFSET) -2: /* * 6. After EFI runtime service returns, control will return to * following instruction. We'd better readjust stack pointer first. @@ -88,35 +85,32 @@ ENTRY(efi_call_phys) movl %cr0, %edx orl $0x80000000, %edx movl %edx, %cr0 - jmp 1f -1: + /* * 8. Now restore the virtual mode from flat mode by * adding EIP with PAGE_OFFSET. */ - movl $1f, %edx - jmp *%edx + ljmp $(__KERNEL_CS),$1f+__PAGE_OFFSET 1: + movl $(__KERNEL_DS), %edx + mov %edx, %ds + mov %edx, %es + mov %edx, %ss /* * 9. Balance the stack. And because EAX contain the return value, * we'd better not clobber it. */ - leal efi_rt_function_ptr, %edx - movl (%edx), %ecx - pushl %ecx + pushl (efi_rt_function_ptr) /* - * 10. Push the saved return address onto the stack and return. + * 10. Return to the saved return address. */ - leal saved_return_addr, %edx - movl (%edx), %ecx - pushl %ecx - ret + jmpl *(saved_return_addr) ENDPROC(efi_call_phys) .previous -.data +__INITDATA saved_return_addr: .long 0 efi_rt_function_ptr: diff --git a/arch/x86/platform/efi/efi_stub_64.S b/arch/x86/platform/efi/efi_stub_64.S index 4c07cca..2c8427d 100644 --- a/arch/x86/platform/efi/efi_stub_64.S +++ b/arch/x86/platform/efi/efi_stub_64.S @@ -7,6 +7,7 @@ */ #include +#include #define SAVE_XMM \ mov %rsp, %rax; \ @@ -40,6 +41,7 @@ ENTRY(efi_call0) call *%rdi addq $32, %rsp RESTORE_XMM + pax_force_retaddr 0, 1 ret ENDPROC(efi_call0) @@ -50,6 +52,7 @@ ENTRY(efi_call1) call *%rdi addq $32, %rsp RESTORE_XMM + pax_force_retaddr 0, 1 ret ENDPROC(efi_call1) @@ -60,6 +63,7 @@ ENTRY(efi_call2) call *%rdi addq $32, %rsp RESTORE_XMM + pax_force_retaddr 0, 1 ret ENDPROC(efi_call2) @@ -71,6 +75,7 @@ ENTRY(efi_call3) call *%rdi addq $32, %rsp RESTORE_XMM + pax_force_retaddr 0, 1 ret ENDPROC(efi_call3) @@ -83,6 +88,7 @@ ENTRY(efi_call4) call *%rdi addq $32, %rsp RESTORE_XMM + pax_force_retaddr 0, 1 ret ENDPROC(efi_call4) @@ -96,6 +102,7 @@ ENTRY(efi_call5) call *%rdi addq $48, %rsp RESTORE_XMM + pax_force_retaddr 0, 1 ret ENDPROC(efi_call5) @@ -112,5 +119,6 @@ ENTRY(efi_call6) call *%rdi addq $48, %rsp RESTORE_XMM + pax_force_retaddr 0, 1 ret ENDPROC(efi_call6) diff --git a/arch/x86/platform/mrst/mrst.c b/arch/x86/platform/mrst/mrst.c index 475e2cd..1b8e708 100644 --- a/arch/x86/platform/mrst/mrst.c +++ b/arch/x86/platform/mrst/mrst.c @@ -76,18 +76,20 @@ struct sfi_rtc_table_entry sfi_mrtc_array[SFI_MRTC_MAX]; EXPORT_SYMBOL_GPL(sfi_mrtc_array); int sfi_mrtc_num; -static void mrst_power_off(void) +static __noreturn void mrst_power_off(void) { if (__mrst_cpu_chip == MRST_CPU_CHIP_LINCROFT) intel_scu_ipc_simple_command(IPCMSG_COLD_RESET, 1); + BUG(); } -static void mrst_reboot(void) +static __noreturn void mrst_reboot(void) { if (__mrst_cpu_chip == MRST_CPU_CHIP_LINCROFT) intel_scu_ipc_simple_command(IPCMSG_COLD_RESET, 0); else intel_scu_ipc_simple_command(IPCMSG_COLD_BOOT, 0); + BUG(); } /* parse all the mtimer info to a static mtimer array */ diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index 3ae0e61..4202d86 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -1424,6 +1424,8 @@ static ssize_t tunables_read(struct file *file, char __user *userbuf, * 0: display meaning of the statistics */ static ssize_t ptc_proc_write(struct file *file, const char __user *user, + size_t count, loff_t *data) __size_overflow(3); +static ssize_t ptc_proc_write(struct file *file, const char __user *user, size_t count, loff_t *data) { int cpu; @@ -1539,6 +1541,8 @@ static int parse_tunables_write(struct bau_control *bcp, char *instr, * Handle a write to debugfs. (/sys/kernel/debug/sgi_uv/bau_tunables) */ static ssize_t tunables_write(struct file *file, const char __user *user, + size_t count, loff_t *data) __size_overflow(3); +static ssize_t tunables_write(struct file *file, const char __user *user, size_t count, loff_t *data) { int cpu; diff --git a/arch/x86/power/cpu.c b/arch/x86/power/cpu.c index f10c0af..3ec1f95 100644 --- a/arch/x86/power/cpu.c +++ b/arch/x86/power/cpu.c @@ -131,7 +131,7 @@ static void do_fpu_end(void) static void fix_processor_context(void) { int cpu = smp_processor_id(); - struct tss_struct *t = &per_cpu(init_tss, cpu); + struct tss_struct *t = init_tss + cpu; set_tss_desc(cpu, t); /* * This just modifies memory; should not be @@ -141,7 +141,9 @@ static void fix_processor_context(void) */ #ifdef CONFIG_X86_64 + pax_open_kernel(); get_cpu_gdt_table(cpu)[GDT_ENTRY_TSS].type = 9; + pax_close_kernel(); syscall_init(); /* This sets MSR_*STAR and related */ #endif diff --git a/arch/x86/vdso/Makefile b/arch/x86/vdso/Makefile index 5d17950..2253fc9 100644 --- a/arch/x86/vdso/Makefile +++ b/arch/x86/vdso/Makefile @@ -137,7 +137,7 @@ quiet_cmd_vdso = VDSO $@ -Wl,-T,$(filter %.lds,$^) $(filter %.o,$^) && \ sh $(srctree)/$(src)/checkundef.sh '$(NM)' '$@' -VDSO_LDFLAGS = -fPIC -shared $(call cc-ldoption, -Wl$(comma)--hash-style=sysv) +VDSO_LDFLAGS = -fPIC -shared -Wl,--no-undefined $(call cc-ldoption, -Wl$(comma)--hash-style=sysv) GCOV_PROFILE := n # diff --git a/arch/x86/vdso/vdso32-setup.c b/arch/x86/vdso/vdso32-setup.c index 468d591..8e80a0a 100644 --- a/arch/x86/vdso/vdso32-setup.c +++ b/arch/x86/vdso/vdso32-setup.c @@ -25,6 +25,7 @@ #include #include #include +#include enum { VDSO_DISABLED = 0, @@ -226,7 +227,7 @@ static inline void map_compat_vdso(int map) void enable_sep_cpu(void) { int cpu = get_cpu(); - struct tss_struct *tss = &per_cpu(init_tss, cpu); + struct tss_struct *tss = init_tss + cpu; if (!boot_cpu_has(X86_FEATURE_SEP)) { put_cpu(); @@ -249,7 +250,7 @@ static int __init gate_vma_init(void) gate_vma.vm_start = FIXADDR_USER_START; gate_vma.vm_end = FIXADDR_USER_END; gate_vma.vm_flags = VM_READ | VM_MAYREAD | VM_EXEC | VM_MAYEXEC; - gate_vma.vm_page_prot = __P101; + gate_vma.vm_page_prot = vm_get_page_prot(gate_vma.vm_flags); /* * Make sure the vDSO gets into every core dump. * Dumping its contents makes post-mortem fully interpretable later @@ -331,14 +332,14 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) if (compat) addr = VDSO_HIGH_BASE; else { - addr = get_unmapped_area(NULL, 0, PAGE_SIZE, 0, 0); + addr = get_unmapped_area(NULL, 0, PAGE_SIZE, 0, MAP_EXECUTABLE); if (IS_ERR_VALUE(addr)) { ret = addr; goto up_fail; } } - current->mm->context.vdso = (void *)addr; + current->mm->context.vdso = addr; if (compat_uses_vma || !compat) { /* @@ -361,11 +362,11 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) } current_thread_info()->sysenter_return = - VDSO32_SYMBOL(addr, SYSENTER_RETURN); + (__force void __user *)VDSO32_SYMBOL(addr, SYSENTER_RETURN); up_fail: if (ret) - current->mm->context.vdso = NULL; + current->mm->context.vdso = 0; up_write(&mm->mmap_sem); @@ -412,8 +413,14 @@ __initcall(ia32_binfmt_init); const char *arch_vma_name(struct vm_area_struct *vma) { - if (vma->vm_mm && vma->vm_start == (long)vma->vm_mm->context.vdso) + if (vma->vm_mm && vma->vm_start == vma->vm_mm->context.vdso) return "[vdso]"; + +#ifdef CONFIG_PAX_SEGMEXEC + if (vma->vm_mm && vma->vm_mirror && vma->vm_mirror->vm_start == vma->vm_mm->context.vdso) + return "[vdso]"; +#endif + return NULL; } @@ -423,7 +430,7 @@ struct vm_area_struct *get_gate_vma(struct mm_struct *mm) * Check to see if the corresponding task was created in compat vdso * mode. */ - if (mm && mm->context.vdso == (void *)VDSO_HIGH_BASE) + if (mm && mm->context.vdso == VDSO_HIGH_BASE) return &gate_vma; return NULL; } diff --git a/arch/x86/vdso/vma.c b/arch/x86/vdso/vma.c index 153407c..611cba9 100644 --- a/arch/x86/vdso/vma.c +++ b/arch/x86/vdso/vma.c @@ -16,8 +16,6 @@ #include #include -unsigned int __read_mostly vdso_enabled = 1; - extern char vdso_start[], vdso_end[]; extern unsigned short vdso_sync_cpuid; @@ -96,7 +94,6 @@ static unsigned long vdso_addr(unsigned long start, unsigned len) * unaligned here as a result of stack start randomization. */ addr = PAGE_ALIGN(addr); - addr = align_addr(addr, NULL, ALIGN_VDSO); return addr; } @@ -106,40 +103,35 @@ static unsigned long vdso_addr(unsigned long start, unsigned len) int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) { struct mm_struct *mm = current->mm; - unsigned long addr; + unsigned long addr = 0; int ret; - if (!vdso_enabled) - return 0; - down_write(&mm->mmap_sem); + +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + addr = vdso_addr(mm->start_stack, vdso_size); + addr = align_addr(addr, NULL, ALIGN_VDSO); addr = get_unmapped_area(NULL, addr, vdso_size, 0, 0); if (IS_ERR_VALUE(addr)) { ret = addr; goto up_fail; } - current->mm->context.vdso = (void *)addr; + mm->context.vdso = addr; ret = install_special_mapping(mm, addr, vdso_size, VM_READ|VM_EXEC| VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC| VM_ALWAYSDUMP, vdso_pages); - if (ret) { - current->mm->context.vdso = NULL; - goto up_fail; - } + + if (ret) + mm->context.vdso = 0; up_fail: up_write(&mm->mmap_sem); return ret; } - -static __init int vdso_setup(char *s) -{ - vdso_enabled = simple_strtoul(s, NULL, 0); - return 0; -} -__setup("vdso=", vdso_setup); diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index 4e517d4..68a48f5 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -86,8 +86,6 @@ EXPORT_SYMBOL_GPL(xen_start_info); struct shared_info xen_dummy_shared_info; -void *xen_initial_gdt; - RESERVE_BRK(shared_info_page_brk, PAGE_SIZE); __read_mostly int xen_have_vector_callback; EXPORT_SYMBOL_GPL(xen_have_vector_callback); @@ -1030,30 +1028,30 @@ static const struct pv_apic_ops xen_apic_ops __initconst = { #endif }; -static void xen_reboot(int reason) +static __noreturn void xen_reboot(int reason) { struct sched_shutdown r = { .reason = reason }; - if (HYPERVISOR_sched_op(SCHEDOP_shutdown, &r)) - BUG(); + HYPERVISOR_sched_op(SCHEDOP_shutdown, &r); + BUG(); } -static void xen_restart(char *msg) +static __noreturn void xen_restart(char *msg) { xen_reboot(SHUTDOWN_reboot); } -static void xen_emergency_restart(void) +static __noreturn void xen_emergency_restart(void) { xen_reboot(SHUTDOWN_reboot); } -static void xen_machine_halt(void) +static __noreturn void xen_machine_halt(void) { xen_reboot(SHUTDOWN_poweroff); } -static void xen_machine_power_off(void) +static __noreturn void xen_machine_power_off(void) { if (pm_power_off) pm_power_off(); @@ -1156,7 +1154,17 @@ asmlinkage void __init xen_start_kernel(void) __userpte_alloc_gfp &= ~__GFP_HIGHMEM; /* Work out if we support NX */ - x86_configure_nx(); +#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) + if ((cpuid_eax(0x80000000) & 0xffff0000) == 0x80000000 && + (cpuid_edx(0x80000001) & (1U << (X86_FEATURE_NX & 31)))) { + unsigned l, h; + + __supported_pte_mask |= _PAGE_NX; + rdmsr(MSR_EFER, l, h); + l |= EFER_NX; + wrmsr(MSR_EFER, l, h); + } +#endif xen_setup_features(); @@ -1187,13 +1195,6 @@ asmlinkage void __init xen_start_kernel(void) machine_ops = xen_machine_ops; - /* - * The only reliable way to retain the initial address of the - * percpu gdt_page is to remember it here, so we can go and - * mark it RW later, when the initial percpu area is freed. - */ - xen_initial_gdt = &per_cpu(gdt_page, 0); - xen_smp_init(); #ifdef CONFIG_ACPI_NUMA diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index dc19347..1b07a2c 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1738,6 +1738,9 @@ pgd_t * __init xen_setup_kernel_pagetable(pgd_t *pgd, convert_pfn_mfn(init_level4_pgt); convert_pfn_mfn(level3_ident_pgt); convert_pfn_mfn(level3_kernel_pgt); + convert_pfn_mfn(level3_vmalloc_start_pgt); + convert_pfn_mfn(level3_vmalloc_end_pgt); + convert_pfn_mfn(level3_vmemmap_pgt); l3 = m2v(pgd[pgd_index(__START_KERNEL_map)].pgd); l2 = m2v(l3[pud_index(__START_KERNEL_map)].pud); @@ -1756,7 +1759,11 @@ pgd_t * __init xen_setup_kernel_pagetable(pgd_t *pgd, set_page_prot(init_level4_pgt, PAGE_KERNEL_RO); set_page_prot(level3_ident_pgt, PAGE_KERNEL_RO); set_page_prot(level3_kernel_pgt, PAGE_KERNEL_RO); + set_page_prot(level3_vmalloc_start_pgt, PAGE_KERNEL_RO); + set_page_prot(level3_vmalloc_end_pgt, PAGE_KERNEL_RO); + set_page_prot(level3_vmemmap_pgt, PAGE_KERNEL_RO); set_page_prot(level3_user_vsyscall, PAGE_KERNEL_RO); + set_page_prot(level2_vmemmap_pgt, PAGE_KERNEL_RO); set_page_prot(level2_kernel_pgt, PAGE_KERNEL_RO); set_page_prot(level2_fixmap_pgt, PAGE_KERNEL_RO); @@ -1963,6 +1970,7 @@ static void __init xen_post_allocator_init(void) pv_mmu_ops.set_pud = xen_set_pud; #if PAGETABLE_LEVELS == 4 pv_mmu_ops.set_pgd = xen_set_pgd; + pv_mmu_ops.set_pgd_batched = xen_set_pgd; #endif /* This will work as long as patching hasn't happened yet @@ -2044,6 +2052,7 @@ static const struct pv_mmu_ops xen_mmu_ops __initconst = { .pud_val = PV_CALLEE_SAVE(xen_pud_val), .make_pud = PV_CALLEE_SAVE(xen_make_pud), .set_pgd = xen_set_pgd_hyper, + .set_pgd_batched = xen_set_pgd_hyper, .alloc_pud = xen_alloc_pmd_init, .release_pud = xen_release_pmd_init, diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c index f2ce60a..14e08dc 100644 --- a/arch/x86/xen/smp.c +++ b/arch/x86/xen/smp.c @@ -209,11 +209,6 @@ static void __init xen_smp_prepare_boot_cpu(void) { BUG_ON(smp_processor_id() != 0); native_smp_prepare_boot_cpu(); - - /* We've switched to the "real" per-cpu gdt, so make sure the - old memory can be recycled */ - make_lowmem_page_readwrite(xen_initial_gdt); - xen_filter_cpu_maps(); xen_setup_vcpu_info_placement(); } @@ -290,12 +285,12 @@ cpu_initialize_context(unsigned int cpu, struct task_struct *idle) gdt = get_cpu_gdt_table(cpu); ctxt->flags = VGCF_IN_KERNEL; - ctxt->user_regs.ds = __USER_DS; - ctxt->user_regs.es = __USER_DS; + ctxt->user_regs.ds = __KERNEL_DS; + ctxt->user_regs.es = __KERNEL_DS; ctxt->user_regs.ss = __KERNEL_DS; #ifdef CONFIG_X86_32 ctxt->user_regs.fs = __KERNEL_PERCPU; - ctxt->user_regs.gs = __KERNEL_STACK_CANARY; + savesegment(gs, ctxt->user_regs.gs); #else ctxt->gs_base_kernel = per_cpu_offset(cpu); #endif @@ -346,13 +341,12 @@ static int __cpuinit xen_cpu_up(unsigned int cpu) int rc; per_cpu(current_task, cpu) = idle; + per_cpu(current_tinfo, cpu) = &idle->tinfo; #ifdef CONFIG_X86_32 irq_ctx_init(cpu); #else clear_tsk_thread_flag(idle, TIF_FORK); - per_cpu(kernel_stack, cpu) = - (unsigned long)task_stack_page(idle) - - KERNEL_STACK_OFFSET + THREAD_SIZE; + per_cpu(kernel_stack, cpu) = (unsigned long)task_stack_page(idle) - 16 + THREAD_SIZE; #endif xen_setup_runstate_info(cpu); xen_setup_timer(cpu); diff --git a/arch/x86/xen/xen-asm_32.S b/arch/x86/xen/xen-asm_32.S index b040b0e..8cc4fe0 100644 --- a/arch/x86/xen/xen-asm_32.S +++ b/arch/x86/xen/xen-asm_32.S @@ -83,14 +83,14 @@ ENTRY(xen_iret) ESP_OFFSET=4 # bytes pushed onto stack /* - * Store vcpu_info pointer for easy access. Do it this way to - * avoid having to reload %fs + * Store vcpu_info pointer for easy access. */ #ifdef CONFIG_SMP - GET_THREAD_INFO(%eax) - movl TI_cpu(%eax), %eax - movl __per_cpu_offset(,%eax,4), %eax - mov xen_vcpu(%eax), %eax + push %fs + mov $(__KERNEL_PERCPU), %eax + mov %eax, %fs + mov PER_CPU_VAR(xen_vcpu), %eax + pop %fs #else movl xen_vcpu, %eax #endif diff --git a/arch/x86/xen/xen-head.S b/arch/x86/xen/xen-head.S index aaa7291..3f77960 100644 --- a/arch/x86/xen/xen-head.S +++ b/arch/x86/xen/xen-head.S @@ -19,6 +19,17 @@ ENTRY(startup_xen) #ifdef CONFIG_X86_32 mov %esi,xen_start_info mov $init_thread_union+THREAD_SIZE,%esp +#ifdef CONFIG_SMP + movl $cpu_gdt_table,%edi + movl $__per_cpu_load,%eax + movw %ax,__KERNEL_PERCPU + 2(%edi) + rorl $16,%eax + movb %al,__KERNEL_PERCPU + 4(%edi) + movb %ah,__KERNEL_PERCPU + 7(%edi) + movl $__per_cpu_end - 1,%eax + subl $__per_cpu_start,%eax + movw %ax,__KERNEL_PERCPU + 0(%edi) +#endif #else mov %rsi,xen_start_info mov $init_thread_union+THREAD_SIZE,%rsp diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h index b095739..8c17bcd 100644 --- a/arch/x86/xen/xen-ops.h +++ b/arch/x86/xen/xen-ops.h @@ -10,8 +10,6 @@ extern const char xen_hypervisor_callback[]; extern const char xen_failsafe_callback[]; -extern void *xen_initial_gdt; - struct trap_info; void xen_copy_trap_info(struct trap_info *traps); diff --git a/arch/xtensa/variants/dc232b/include/variant/core.h b/arch/xtensa/variants/dc232b/include/variant/core.h index 525bd3d..ef888b1 100644 --- a/arch/xtensa/variants/dc232b/include/variant/core.h +++ b/arch/xtensa/variants/dc232b/include/variant/core.h @@ -119,9 +119,9 @@ ----------------------------------------------------------------------*/ #define XCHAL_ICACHE_LINESIZE 32 /* I-cache line size in bytes */ -#define XCHAL_DCACHE_LINESIZE 32 /* D-cache line size in bytes */ #define XCHAL_ICACHE_LINEWIDTH 5 /* log2(I line size in bytes) */ #define XCHAL_DCACHE_LINEWIDTH 5 /* log2(D line size in bytes) */ +#define XCHAL_DCACHE_LINESIZE (_AC(1,UL) << XCHAL_DCACHE_LINEWIDTH) /* D-cache line size in bytes */ #define XCHAL_ICACHE_SIZE 16384 /* I-cache size in bytes or 0 */ #define XCHAL_DCACHE_SIZE 16384 /* D-cache size in bytes or 0 */ diff --git a/arch/xtensa/variants/fsf/include/variant/core.h b/arch/xtensa/variants/fsf/include/variant/core.h index 2f33760..835e50a 100644 --- a/arch/xtensa/variants/fsf/include/variant/core.h +++ b/arch/xtensa/variants/fsf/include/variant/core.h @@ -11,6 +11,7 @@ #ifndef _XTENSA_CORE_H #define _XTENSA_CORE_H +#include /**************************************************************************** Parameters Useful for Any Code, USER or PRIVILEGED @@ -112,9 +113,9 @@ ----------------------------------------------------------------------*/ #define XCHAL_ICACHE_LINESIZE 16 /* I-cache line size in bytes */ -#define XCHAL_DCACHE_LINESIZE 16 /* D-cache line size in bytes */ #define XCHAL_ICACHE_LINEWIDTH 4 /* log2(I line size in bytes) */ #define XCHAL_DCACHE_LINEWIDTH 4 /* log2(D line size in bytes) */ +#define XCHAL_DCACHE_LINESIZE (_AC(1,UL) << XCHAL_DCACHE_LINEWIDTH) /* D-cache line size in bytes */ #define XCHAL_ICACHE_SIZE 8192 /* I-cache size in bytes or 0 */ #define XCHAL_DCACHE_SIZE 8192 /* D-cache size in bytes or 0 */ diff --git a/arch/xtensa/variants/s6000/include/variant/core.h b/arch/xtensa/variants/s6000/include/variant/core.h index af00795..2bb8105 100644 --- a/arch/xtensa/variants/s6000/include/variant/core.h +++ b/arch/xtensa/variants/s6000/include/variant/core.h @@ -11,6 +11,7 @@ #ifndef _XTENSA_CORE_CONFIGURATION_H #define _XTENSA_CORE_CONFIGURATION_H +#include /**************************************************************************** Parameters Useful for Any Code, USER or PRIVILEGED @@ -118,9 +119,9 @@ ----------------------------------------------------------------------*/ #define XCHAL_ICACHE_LINESIZE 16 /* I-cache line size in bytes */ -#define XCHAL_DCACHE_LINESIZE 16 /* D-cache line size in bytes */ #define XCHAL_ICACHE_LINEWIDTH 4 /* log2(I line size in bytes) */ #define XCHAL_DCACHE_LINEWIDTH 4 /* log2(D line size in bytes) */ +#define XCHAL_DCACHE_LINESIZE (_AC(1,UL) << XCHAL_DCACHE_LINEWIDTH) /* D-cache line size in bytes */ #define XCHAL_ICACHE_SIZE 32768 /* I-cache size in bytes or 0 */ #define XCHAL_DCACHE_SIZE 32768 /* D-cache size in bytes or 0 */ diff --git a/block/blk-iopoll.c b/block/blk-iopoll.c index 58916af..9cb880b 100644 --- a/block/blk-iopoll.c +++ b/block/blk-iopoll.c @@ -77,7 +77,7 @@ void blk_iopoll_complete(struct blk_iopoll *iopoll) } EXPORT_SYMBOL(blk_iopoll_complete); -static void blk_iopoll_softirq(struct softirq_action *h) +static void blk_iopoll_softirq(void) { struct list_head *list = &__get_cpu_var(blk_cpu_iopoll); int rearm = 0, budget = blk_iopoll_budget; diff --git a/block/blk-map.c b/block/blk-map.c index 623e1cd..ca1e109 100644 --- a/block/blk-map.c +++ b/block/blk-map.c @@ -302,7 +302,7 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf, if (!len || !kbuf) return -EINVAL; - do_copy = !blk_rq_aligned(q, addr, len) || object_is_on_stack(kbuf); + do_copy = !blk_rq_aligned(q, addr, len) || object_starts_on_stack(kbuf); if (do_copy) bio = bio_copy_kern(q, kbuf, len, gfp_mask, reading); else diff --git a/block/blk-softirq.c b/block/blk-softirq.c index 1366a89..e17f54b 100644 --- a/block/blk-softirq.c +++ b/block/blk-softirq.c @@ -17,7 +17,7 @@ static DEFINE_PER_CPU(struct list_head, blk_cpu_done); * Softirq action handler - move entries to local list and loop over them * while passing them to the queue registered handler. */ -static void blk_done_softirq(struct softirq_action *h) +static void blk_done_softirq(void) { struct list_head *cpu_list, local_list; diff --git a/block/bsg.c b/block/bsg.c index ff64ae3..593560c 100644 --- a/block/bsg.c +++ b/block/bsg.c @@ -176,16 +176,24 @@ static int blk_fill_sgv4_hdr_rq(struct request_queue *q, struct request *rq, struct sg_io_v4 *hdr, struct bsg_device *bd, fmode_t has_write_perm) { + unsigned char tmpcmd[sizeof(rq->__cmd)]; + unsigned char *cmdptr; + if (hdr->request_len > BLK_MAX_CDB) { rq->cmd = kzalloc(hdr->request_len, GFP_KERNEL); if (!rq->cmd) return -ENOMEM; - } + cmdptr = rq->cmd; + } else + cmdptr = tmpcmd; - if (copy_from_user(rq->cmd, (void __user *)(unsigned long)hdr->request, + if (copy_from_user(cmdptr, (void __user *)(unsigned long)hdr->request, hdr->request_len)) return -EFAULT; + if (cmdptr != rq->cmd) + memcpy(rq->cmd, cmdptr, hdr->request_len); + if (hdr->subprotocol == BSG_SUB_PROTOCOL_SCSI_CMD) { if (blk_verify_command(rq->cmd, has_write_perm)) return -EPERM; diff --git a/block/compat_ioctl.c b/block/compat_ioctl.c index 7c668c8..db3521c 100644 --- a/block/compat_ioctl.c +++ b/block/compat_ioctl.c @@ -340,7 +340,7 @@ static int compat_fd_ioctl(struct block_device *bdev, fmode_t mode, err |= __get_user(f->spec1, &uf->spec1); err |= __get_user(f->fmt_gap, &uf->fmt_gap); err |= __get_user(name, &uf->name); - f->name = compat_ptr(name); + f->name = (void __force_kernel *)compat_ptr(name); if (err) { err = -EFAULT; goto out; diff --git a/block/partitions/efi.c b/block/partitions/efi.c index 6296b40..417c00f 100644 --- a/block/partitions/efi.c +++ b/block/partitions/efi.c @@ -234,14 +234,14 @@ static gpt_entry *alloc_read_gpt_entries(struct parsed_partitions *state, if (!gpt) return NULL; + if (!le32_to_cpu(gpt->num_partition_entries)) + return NULL; + pte = kcalloc(le32_to_cpu(gpt->num_partition_entries), le32_to_cpu(gpt->sizeof_partition_entry), GFP_KERNEL); + if (!pte) + return NULL; + count = le32_to_cpu(gpt->num_partition_entries) * le32_to_cpu(gpt->sizeof_partition_entry); - if (!count) - return NULL; - pte = kzalloc(count, GFP_KERNEL); - if (!pte) - return NULL; - if (read_lba(state, le64_to_cpu(gpt->partition_entry_lba), (u8 *) pte, count) < count) { diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c index 260fa80..e8f3caf 100644 --- a/block/scsi_ioctl.c +++ b/block/scsi_ioctl.c @@ -223,8 +223,20 @@ EXPORT_SYMBOL(blk_verify_command); static int blk_fill_sghdr_rq(struct request_queue *q, struct request *rq, struct sg_io_hdr *hdr, fmode_t mode) { - if (copy_from_user(rq->cmd, hdr->cmdp, hdr->cmd_len)) + unsigned char tmpcmd[sizeof(rq->__cmd)]; + unsigned char *cmdptr; + + if (rq->cmd != rq->__cmd) + cmdptr = rq->cmd; + else + cmdptr = tmpcmd; + + if (copy_from_user(cmdptr, hdr->cmdp, hdr->cmd_len)) return -EFAULT; + + if (cmdptr != rq->cmd) + memcpy(rq->cmd, cmdptr, hdr->cmd_len); + if (blk_verify_command(rq->cmd, mode & FMODE_WRITE)) return -EPERM; @@ -433,6 +445,8 @@ int sg_scsi_ioctl(struct request_queue *q, struct gendisk *disk, fmode_t mode, int err; unsigned int in_len, out_len, bytes, opcode, cmdlen; char *buffer = NULL, sense[SCSI_SENSE_BUFFERSIZE]; + unsigned char tmpcmd[sizeof(rq->__cmd)]; + unsigned char *cmdptr; if (!sic) return -EINVAL; @@ -466,9 +480,18 @@ int sg_scsi_ioctl(struct request_queue *q, struct gendisk *disk, fmode_t mode, */ err = -EFAULT; rq->cmd_len = cmdlen; - if (copy_from_user(rq->cmd, sic->data, cmdlen)) + + if (rq->cmd != rq->__cmd) + cmdptr = rq->cmd; + else + cmdptr = tmpcmd; + + if (copy_from_user(cmdptr, sic->data, cmdlen)) goto error; + if (rq->cmd != cmdptr) + memcpy(rq->cmd, cmdptr, cmdlen); + if (in_len && copy_from_user(buffer, sic->data + cmdlen, in_len)) goto error; diff --git a/crypto/ablkcipher.c b/crypto/ablkcipher.c index a0f768c..1da9c73 100644 --- a/crypto/ablkcipher.c +++ b/crypto/ablkcipher.c @@ -307,6 +307,8 @@ int ablkcipher_walk_phys(struct ablkcipher_request *req, EXPORT_SYMBOL_GPL(ablkcipher_walk_phys); static int setkey_unaligned(struct crypto_ablkcipher *tfm, const u8 *key, + unsigned int keylen) __size_overflow(3); +static int setkey_unaligned(struct crypto_ablkcipher *tfm, const u8 *key, unsigned int keylen) { struct ablkcipher_alg *cipher = crypto_ablkcipher_alg(tfm); @@ -329,6 +331,8 @@ static int setkey_unaligned(struct crypto_ablkcipher *tfm, const u8 *key, } static int setkey(struct crypto_ablkcipher *tfm, const u8 *key, + unsigned int keylen) __size_overflow(3); +static int setkey(struct crypto_ablkcipher *tfm, const u8 *key, unsigned int keylen) { struct ablkcipher_alg *cipher = crypto_ablkcipher_alg(tfm); diff --git a/crypto/aead.c b/crypto/aead.c index 04add3dc..983032f 100644 --- a/crypto/aead.c +++ b/crypto/aead.c @@ -27,6 +27,8 @@ #include "internal.h" static int setkey_unaligned(struct crypto_aead *tfm, const u8 *key, + unsigned int keylen) __size_overflow(3); +static int setkey_unaligned(struct crypto_aead *tfm, const u8 *key, unsigned int keylen) { struct aead_alg *aead = crypto_aead_alg(tfm); @@ -48,6 +50,7 @@ static int setkey_unaligned(struct crypto_aead *tfm, const u8 *key, return ret; } +static int setkey(struct crypto_aead *tfm, const u8 *key, unsigned int keylen) __size_overflow(3); static int setkey(struct crypto_aead *tfm, const u8 *key, unsigned int keylen) { struct aead_alg *aead = crypto_aead_alg(tfm); diff --git a/crypto/blkcipher.c b/crypto/blkcipher.c index 1e61d1a..cf06b86 100644 --- a/crypto/blkcipher.c +++ b/crypto/blkcipher.c @@ -359,6 +359,8 @@ int blkcipher_walk_virt_block(struct blkcipher_desc *desc, EXPORT_SYMBOL_GPL(blkcipher_walk_virt_block); static int setkey_unaligned(struct crypto_tfm *tfm, const u8 *key, + unsigned int keylen) __size_overflow(3); +static int setkey_unaligned(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen) { struct blkcipher_alg *cipher = &tfm->__crt_alg->cra_blkcipher; @@ -380,6 +382,7 @@ static int setkey_unaligned(struct crypto_tfm *tfm, const u8 *key, return ret; } +static int setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen) __size_overflow(3); static int setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen) { struct blkcipher_alg *cipher = &tfm->__crt_alg->cra_blkcipher; diff --git a/crypto/cipher.c b/crypto/cipher.c index 39541e0..802d956 100644 --- a/crypto/cipher.c +++ b/crypto/cipher.c @@ -21,6 +21,8 @@ #include "internal.h" static int setkey_unaligned(struct crypto_tfm *tfm, const u8 *key, + unsigned int keylen) __size_overflow(3); +static int setkey_unaligned(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen) { struct cipher_alg *cia = &tfm->__crt_alg->cra_cipher; @@ -43,6 +45,7 @@ static int setkey_unaligned(struct crypto_tfm *tfm, const u8 *key, } +static int setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen) __size_overflow(3); static int setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen) { struct cipher_alg *cia = &tfm->__crt_alg->cra_cipher; diff --git a/crypto/cryptd.c b/crypto/cryptd.c index 671d4d6..5f24030 100644 --- a/crypto/cryptd.c +++ b/crypto/cryptd.c @@ -63,7 +63,7 @@ struct cryptd_blkcipher_ctx { struct cryptd_blkcipher_request_ctx { crypto_completion_t complete; -}; +} __no_const; struct cryptd_hash_ctx { struct crypto_shash *child; @@ -80,7 +80,7 @@ struct cryptd_aead_ctx { struct cryptd_aead_request_ctx { crypto_completion_t complete; -}; +} __no_const; static void cryptd_queue_worker(struct work_struct *work); diff --git a/drivers/acpi/apei/cper.c b/drivers/acpi/apei/cper.c index 5d41894..22021e4 100644 --- a/drivers/acpi/apei/cper.c +++ b/drivers/acpi/apei/cper.c @@ -38,12 +38,12 @@ */ u64 cper_next_record_id(void) { - static atomic64_t seq; + static atomic64_unchecked_t seq; - if (!atomic64_read(&seq)) - atomic64_set(&seq, ((u64)get_seconds()) << 32); + if (!atomic64_read_unchecked(&seq)) + atomic64_set_unchecked(&seq, ((u64)get_seconds()) << 32); - return atomic64_inc_return(&seq); + return atomic64_inc_return_unchecked(&seq); } EXPORT_SYMBOL_GPL(cper_next_record_id); diff --git a/drivers/acpi/battery.c b/drivers/acpi/battery.c index 86933ca..5cb1a69 100644 --- a/drivers/acpi/battery.c +++ b/drivers/acpi/battery.c @@ -787,6 +787,9 @@ static int acpi_battery_print_alarm(struct seq_file *seq, int result) static ssize_t acpi_battery_write_alarm(struct file *file, const char __user * buffer, + size_t count, loff_t * ppos) __size_overflow(3); +static ssize_t acpi_battery_write_alarm(struct file *file, + const char __user * buffer, size_t count, loff_t * ppos) { int result = 0; diff --git a/drivers/acpi/ec_sys.c b/drivers/acpi/ec_sys.c index b258cab..3fb7da7 100644 --- a/drivers/acpi/ec_sys.c +++ b/drivers/acpi/ec_sys.c @@ -12,6 +12,7 @@ #include #include #include +#include #include "internal.h" MODULE_AUTHOR("Thomas Renninger "); @@ -40,7 +41,7 @@ static ssize_t acpi_ec_read_io(struct file *f, char __user *buf, * struct acpi_ec *ec = ((struct seq_file *)f->private_data)->private; */ unsigned int size = EC_SPACE_SIZE; - u8 *data = (u8 *) buf; + u8 data; loff_t init_off = *off; int err = 0; @@ -53,9 +54,11 @@ static ssize_t acpi_ec_read_io(struct file *f, char __user *buf, size = count; while (size) { - err = ec_read(*off, &data[*off - init_off]); + err = ec_read(*off, &data); if (err) return err; + if (put_user(data, &buf[*off - init_off])) + return -EFAULT; *off += 1; size--; } @@ -71,7 +74,6 @@ static ssize_t acpi_ec_write_io(struct file *f, const char __user *buf, unsigned int size = count; loff_t init_off = *off; - u8 *data = (u8 *) buf; int err = 0; if (*off >= EC_SPACE_SIZE) @@ -82,7 +84,9 @@ static ssize_t acpi_ec_write_io(struct file *f, const char __user *buf, } while (size) { - u8 byte_write = data[*off - init_off]; + u8 byte_write; + if (get_user(byte_write, &buf[*off - init_off])) + return -EFAULT; err = ec_write(*off, byte_write); if (err) return err; diff --git a/drivers/acpi/proc.c b/drivers/acpi/proc.c index 251c7b62..000462d 100644 --- a/drivers/acpi/proc.c +++ b/drivers/acpi/proc.c @@ -343,19 +343,13 @@ acpi_system_write_wakeup_device(struct file *file, size_t count, loff_t * ppos) { struct list_head *node, *next; - char strbuf[5]; - char str[5] = ""; - unsigned int len = count; + char strbuf[5] = {0}; - if (len > 4) - len = 4; - if (len < 0) + if (count > 4) + count = 4; + if (copy_from_user(strbuf, buffer, count)) return -EFAULT; - - if (copy_from_user(strbuf, buffer, len)) - return -EFAULT; - strbuf[len] = '\0'; - sscanf(strbuf, "%s", str); + strbuf[count] = '\0'; mutex_lock(&acpi_device_lock); list_for_each_safe(node, next, &acpi_wakeup_device_list) { @@ -364,7 +358,7 @@ acpi_system_write_wakeup_device(struct file *file, if (!dev->wakeup.flags.valid) continue; - if (!strncmp(dev->pnp.bus_id, str, 4)) { + if (!strncmp(dev->pnp.bus_id, strbuf, 4)) { if (device_can_wakeup(&dev->dev)) { bool enable = !device_may_wakeup(&dev->dev); device_set_wakeup_enable(&dev->dev, enable); diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index 8ae05ce..7dbbed9 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -555,7 +555,7 @@ static int __cpuinit acpi_processor_add(struct acpi_device *device) return 0; #endif - BUG_ON((pr->id >= nr_cpu_ids) || (pr->id < 0)); + BUG_ON(pr->id >= nr_cpu_ids); /* * Buggy BIOS check diff --git a/drivers/acpi/sbs.c b/drivers/acpi/sbs.c index 6e36d0c..f319944 100644 --- a/drivers/acpi/sbs.c +++ b/drivers/acpi/sbs.c @@ -655,6 +655,9 @@ static int acpi_battery_read_alarm(struct seq_file *seq, void *offset) static ssize_t acpi_battery_write_alarm(struct file *file, const char __user * buffer, + size_t count, loff_t * ppos) __size_overflow(3); +static ssize_t +acpi_battery_write_alarm(struct file *file, const char __user * buffer, size_t count, loff_t * ppos) { struct seq_file *seq = file->private_data; diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index c06e0ec..a2c06ba 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -4736,7 +4736,7 @@ void ata_qc_free(struct ata_queued_cmd *qc) struct ata_port *ap; unsigned int tag; - WARN_ON_ONCE(qc == NULL); /* ata_qc_from_tag _might_ return NULL */ + BUG_ON(qc == NULL); /* ata_qc_from_tag _might_ return NULL */ ap = qc->ap; qc->flags = 0; @@ -4752,7 +4752,7 @@ void __ata_qc_complete(struct ata_queued_cmd *qc) struct ata_port *ap; struct ata_link *link; - WARN_ON_ONCE(qc == NULL); /* ata_qc_from_tag _might_ return NULL */ + BUG_ON(qc == NULL); /* ata_qc_from_tag _might_ return NULL */ WARN_ON_ONCE(!(qc->flags & ATA_QCFLAG_ACTIVE)); ap = qc->ap; link = qc->dev->link; @@ -5816,6 +5816,7 @@ static void ata_finalize_port_ops(struct ata_port_operations *ops) return; spin_lock(&lock); + pax_open_kernel(); for (cur = ops->inherits; cur; cur = cur->inherits) { void **inherit = (void **)cur; @@ -5829,8 +5830,9 @@ static void ata_finalize_port_ops(struct ata_port_operations *ops) if (IS_ERR(*pp)) *pp = NULL; - ops->inherits = NULL; + *(struct ata_port_operations **)&ops->inherits = NULL; + pax_close_kernel(); spin_unlock(&lock); } diff --git a/drivers/ata/pata_arasan_cf.c b/drivers/ata/pata_arasan_cf.c index 048589f..4002b98 100644 --- a/drivers/ata/pata_arasan_cf.c +++ b/drivers/ata/pata_arasan_cf.c @@ -862,7 +862,9 @@ static int __devinit arasan_cf_probe(struct platform_device *pdev) /* Handle platform specific quirks */ if (pdata->quirk) { if (pdata->quirk & CF_BROKEN_PIO) { - ap->ops->set_piomode = NULL; + pax_open_kernel(); + *(void **)&ap->ops->set_piomode = NULL; + pax_close_kernel(); ap->pio_mask = 0; } if (pdata->quirk & CF_BROKEN_MWDMA) diff --git a/drivers/atm/adummy.c b/drivers/atm/adummy.c index f9b983a..887b9d8 100644 --- a/drivers/atm/adummy.c +++ b/drivers/atm/adummy.c @@ -114,7 +114,7 @@ adummy_send(struct atm_vcc *vcc, struct sk_buff *skb) vcc->pop(vcc, skb); else dev_kfree_skb_any(skb); - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); return 0; } diff --git a/drivers/atm/ambassador.c b/drivers/atm/ambassador.c index f8f41e0..1f987dd 100644 --- a/drivers/atm/ambassador.c +++ b/drivers/atm/ambassador.c @@ -454,7 +454,7 @@ static void tx_complete (amb_dev * dev, tx_out * tx) { PRINTD (DBG_FLOW|DBG_TX, "tx_complete %p %p", dev, tx); // VC layer stats - atomic_inc(&ATM_SKB(skb)->vcc->stats->tx); + atomic_inc_unchecked(&ATM_SKB(skb)->vcc->stats->tx); // free the descriptor kfree (tx_descr); @@ -495,7 +495,7 @@ static void rx_complete (amb_dev * dev, rx_out * rx) { dump_skb ("<<<", vc, skb); // VC layer stats - atomic_inc(&atm_vcc->stats->rx); + atomic_inc_unchecked(&atm_vcc->stats->rx); __net_timestamp(skb); // end of our responsibility atm_vcc->push (atm_vcc, skb); @@ -510,7 +510,7 @@ static void rx_complete (amb_dev * dev, rx_out * rx) { } else { PRINTK (KERN_INFO, "dropped over-size frame"); // should we count this? - atomic_inc(&atm_vcc->stats->rx_drop); + atomic_inc_unchecked(&atm_vcc->stats->rx_drop); } } else { @@ -1338,7 +1338,7 @@ static int amb_send (struct atm_vcc * atm_vcc, struct sk_buff * skb) { } if (check_area (skb->data, skb->len)) { - atomic_inc(&atm_vcc->stats->tx_err); + atomic_inc_unchecked(&atm_vcc->stats->tx_err); return -ENOMEM; // ? } diff --git a/drivers/atm/atmtcp.c b/drivers/atm/atmtcp.c index b22d71c..d6e1049 100644 --- a/drivers/atm/atmtcp.c +++ b/drivers/atm/atmtcp.c @@ -207,7 +207,7 @@ static int atmtcp_v_send(struct atm_vcc *vcc,struct sk_buff *skb) if (vcc->pop) vcc->pop(vcc,skb); else dev_kfree_skb(skb); if (dev_data) return 0; - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); return -ENOLINK; } size = skb->len+sizeof(struct atmtcp_hdr); @@ -215,7 +215,7 @@ static int atmtcp_v_send(struct atm_vcc *vcc,struct sk_buff *skb) if (!new_skb) { if (vcc->pop) vcc->pop(vcc,skb); else dev_kfree_skb(skb); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); return -ENOBUFS; } hdr = (void *) skb_put(new_skb,sizeof(struct atmtcp_hdr)); @@ -226,8 +226,8 @@ static int atmtcp_v_send(struct atm_vcc *vcc,struct sk_buff *skb) if (vcc->pop) vcc->pop(vcc,skb); else dev_kfree_skb(skb); out_vcc->push(out_vcc,new_skb); - atomic_inc(&vcc->stats->tx); - atomic_inc(&out_vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->tx); + atomic_inc_unchecked(&out_vcc->stats->rx); return 0; } @@ -301,7 +301,7 @@ static int atmtcp_c_send(struct atm_vcc *vcc,struct sk_buff *skb) out_vcc = find_vcc(dev, ntohs(hdr->vpi), ntohs(hdr->vci)); read_unlock(&vcc_sklist_lock); if (!out_vcc) { - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); goto done; } skb_pull(skb,sizeof(struct atmtcp_hdr)); @@ -313,8 +313,8 @@ static int atmtcp_c_send(struct atm_vcc *vcc,struct sk_buff *skb) __net_timestamp(new_skb); skb_copy_from_linear_data(skb, skb_put(new_skb, skb->len), skb->len); out_vcc->push(out_vcc,new_skb); - atomic_inc(&vcc->stats->tx); - atomic_inc(&out_vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->tx); + atomic_inc_unchecked(&out_vcc->stats->rx); done: if (vcc->pop) vcc->pop(vcc,skb); else dev_kfree_skb(skb); diff --git a/drivers/atm/eni.c b/drivers/atm/eni.c index 956e9ac..133516d 100644 --- a/drivers/atm/eni.c +++ b/drivers/atm/eni.c @@ -526,7 +526,7 @@ static int rx_aal0(struct atm_vcc *vcc) DPRINTK(DEV_LABEL "(itf %d): trashing empty cell\n", vcc->dev->number); length = 0; - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); } else { length = ATM_CELL_SIZE-1; /* no HEC */ @@ -581,7 +581,7 @@ static int rx_aal5(struct atm_vcc *vcc) size); } eff = length = 0; - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); } else { size = (descr & MID_RED_COUNT)*(ATM_CELL_PAYLOAD >> 2); @@ -598,7 +598,7 @@ static int rx_aal5(struct atm_vcc *vcc) "(VCI=%d,length=%ld,size=%ld (descr 0x%lx))\n", vcc->dev->number,vcc->vci,length,size << 2,descr); length = eff = 0; - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); } } skb = eff ? atm_alloc_charge(vcc,eff << 2,GFP_ATOMIC) : NULL; @@ -771,7 +771,7 @@ rx_dequeued++; vcc->push(vcc,skb); pushed++; } - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); } wake_up(&eni_dev->rx_wait); } @@ -1229,7 +1229,7 @@ static void dequeue_tx(struct atm_dev *dev) PCI_DMA_TODEVICE); if (vcc->pop) vcc->pop(vcc,skb); else dev_kfree_skb_irq(skb); - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); wake_up(&eni_dev->tx_wait); dma_complete++; } @@ -1569,7 +1569,7 @@ tx_complete++; /*--------------------------------- entries ---------------------------------*/ -static const char *media_name[] __devinitdata = { +static const char *media_name[] __devinitconst = { "MMF", "SMF", "MMF", "03?", /* 0- 3 */ "UTP", "05?", "06?", "07?", /* 4- 7 */ "TAXI","09?", "10?", "11?", /* 8-11 */ diff --git a/drivers/atm/firestream.c b/drivers/atm/firestream.c index 5072f8a..fa52520d 100644 --- a/drivers/atm/firestream.c +++ b/drivers/atm/firestream.c @@ -750,7 +750,7 @@ static void process_txdone_queue (struct fs_dev *dev, struct queue *q) } } - atomic_inc(&ATM_SKB(skb)->vcc->stats->tx); + atomic_inc_unchecked(&ATM_SKB(skb)->vcc->stats->tx); fs_dprintk (FS_DEBUG_TXMEM, "i"); fs_dprintk (FS_DEBUG_ALLOC, "Free t-skb: %p\n", skb); @@ -817,7 +817,7 @@ static void process_incoming (struct fs_dev *dev, struct queue *q) #endif skb_put (skb, qe->p1 & 0xffff); ATM_SKB(skb)->vcc = atm_vcc; - atomic_inc(&atm_vcc->stats->rx); + atomic_inc_unchecked(&atm_vcc->stats->rx); __net_timestamp(skb); fs_dprintk (FS_DEBUG_ALLOC, "Free rec-skb: %p (pushed)\n", skb); atm_vcc->push (atm_vcc, skb); @@ -838,12 +838,12 @@ static void process_incoming (struct fs_dev *dev, struct queue *q) kfree (pe); } if (atm_vcc) - atomic_inc(&atm_vcc->stats->rx_drop); + atomic_inc_unchecked(&atm_vcc->stats->rx_drop); break; case 0x1f: /* Reassembly abort: no buffers. */ /* Silently increment error counter. */ if (atm_vcc) - atomic_inc(&atm_vcc->stats->rx_drop); + atomic_inc_unchecked(&atm_vcc->stats->rx_drop); break; default: /* Hmm. Haven't written the code to handle the others yet... -- REW */ printk (KERN_WARNING "Don't know what to do with RX status %x: %s.\n", diff --git a/drivers/atm/fore200e.c b/drivers/atm/fore200e.c index 361f5ae..7fc552d 100644 --- a/drivers/atm/fore200e.c +++ b/drivers/atm/fore200e.c @@ -933,9 +933,9 @@ fore200e_tx_irq(struct fore200e* fore200e) #endif /* check error condition */ if (*entry->status & STATUS_ERROR) - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); else - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); } } @@ -1084,7 +1084,7 @@ fore200e_push_rpd(struct fore200e* fore200e, struct atm_vcc* vcc, struct rpd* rp if (skb == NULL) { DPRINTK(2, "unable to alloc new skb, rx PDU length = %d\n", pdu_len); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); return -ENOMEM; } @@ -1127,14 +1127,14 @@ fore200e_push_rpd(struct fore200e* fore200e, struct atm_vcc* vcc, struct rpd* rp dev_kfree_skb_any(skb); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); return -ENOMEM; } ASSERT(atomic_read(&sk_atm(vcc)->sk_wmem_alloc) >= 0); vcc->push(vcc, skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); ASSERT(atomic_read(&sk_atm(vcc)->sk_wmem_alloc) >= 0); @@ -1212,7 +1212,7 @@ fore200e_rx_irq(struct fore200e* fore200e) DPRINTK(2, "damaged PDU on %d.%d.%d\n", fore200e->atm_dev->number, entry->rpd->atm_header.vpi, entry->rpd->atm_header.vci); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); } } @@ -1657,7 +1657,7 @@ fore200e_send(struct atm_vcc *vcc, struct sk_buff *skb) goto retry_here; } - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); fore200e->tx_sat++; DPRINTK(2, "tx queue of device %s is saturated, PDU dropped - heartbeat is %08x\n", diff --git a/drivers/atm/he.c b/drivers/atm/he.c index b182c2f..1c6fa8a 100644 --- a/drivers/atm/he.c +++ b/drivers/atm/he.c @@ -1709,7 +1709,7 @@ he_service_rbrq(struct he_dev *he_dev, int group) if (RBRQ_HBUF_ERR(he_dev->rbrq_head)) { hprintk("HBUF_ERR! (cid 0x%x)\n", cid); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); goto return_host_buffers; } @@ -1736,7 +1736,7 @@ he_service_rbrq(struct he_dev *he_dev, int group) RBRQ_LEN_ERR(he_dev->rbrq_head) ? "LEN_ERR" : "", vcc->vpi, vcc->vci); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); goto return_host_buffers; } @@ -1788,7 +1788,7 @@ he_service_rbrq(struct he_dev *he_dev, int group) vcc->push(vcc, skb); spin_lock(&he_dev->global_lock); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); return_host_buffers: ++pdus_assembled; @@ -2114,7 +2114,7 @@ __enqueue_tpd(struct he_dev *he_dev, struct he_tpd *tpd, unsigned cid) tpd->vcc->pop(tpd->vcc, tpd->skb); else dev_kfree_skb_any(tpd->skb); - atomic_inc(&tpd->vcc->stats->tx_err); + atomic_inc_unchecked(&tpd->vcc->stats->tx_err); } pci_pool_free(he_dev->tpd_pool, tpd, TPD_ADDR(tpd->status)); return; @@ -2526,7 +2526,7 @@ he_send(struct atm_vcc *vcc, struct sk_buff *skb) vcc->pop(vcc, skb); else dev_kfree_skb_any(skb); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); return -EINVAL; } @@ -2537,7 +2537,7 @@ he_send(struct atm_vcc *vcc, struct sk_buff *skb) vcc->pop(vcc, skb); else dev_kfree_skb_any(skb); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); return -EINVAL; } #endif @@ -2549,7 +2549,7 @@ he_send(struct atm_vcc *vcc, struct sk_buff *skb) vcc->pop(vcc, skb); else dev_kfree_skb_any(skb); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); spin_unlock_irqrestore(&he_dev->global_lock, flags); return -ENOMEM; } @@ -2591,7 +2591,7 @@ he_send(struct atm_vcc *vcc, struct sk_buff *skb) vcc->pop(vcc, skb); else dev_kfree_skb_any(skb); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); spin_unlock_irqrestore(&he_dev->global_lock, flags); return -ENOMEM; } @@ -2622,7 +2622,7 @@ he_send(struct atm_vcc *vcc, struct sk_buff *skb) __enqueue_tpd(he_dev, tpd, cid); spin_unlock_irqrestore(&he_dev->global_lock, flags); - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); return 0; } diff --git a/drivers/atm/horizon.c b/drivers/atm/horizon.c index b812103..e391a49 100644 --- a/drivers/atm/horizon.c +++ b/drivers/atm/horizon.c @@ -1035,7 +1035,7 @@ static void rx_schedule (hrz_dev * dev, int irq) { { struct atm_vcc * vcc = ATM_SKB(skb)->vcc; // VC layer stats - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); __net_timestamp(skb); // end of our responsibility vcc->push (vcc, skb); @@ -1187,7 +1187,7 @@ static void tx_schedule (hrz_dev * const dev, int irq) { dev->tx_iovec = NULL; // VC layer stats - atomic_inc(&ATM_SKB(skb)->vcc->stats->tx); + atomic_inc_unchecked(&ATM_SKB(skb)->vcc->stats->tx); // free the skb hrz_kfree_skb (skb); diff --git a/drivers/atm/idt77252.c b/drivers/atm/idt77252.c index 1c05212..c28e200 100644 --- a/drivers/atm/idt77252.c +++ b/drivers/atm/idt77252.c @@ -812,7 +812,7 @@ drain_scq(struct idt77252_dev *card, struct vc_map *vc) else dev_kfree_skb(skb); - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); } atomic_dec(&scq->used); @@ -1075,13 +1075,13 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe) if ((sb = dev_alloc_skb(64)) == NULL) { printk("%s: Can't allocate buffers for aal0.\n", card->name); - atomic_add(i, &vcc->stats->rx_drop); + atomic_add_unchecked(i, &vcc->stats->rx_drop); break; } if (!atm_charge(vcc, sb->truesize)) { RXPRINTK("%s: atm_charge() dropped aal0 packets.\n", card->name); - atomic_add(i - 1, &vcc->stats->rx_drop); + atomic_add_unchecked(i - 1, &vcc->stats->rx_drop); dev_kfree_skb(sb); break; } @@ -1098,7 +1098,7 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe) ATM_SKB(sb)->vcc = vcc; __net_timestamp(sb); vcc->push(vcc, sb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); cell += ATM_CELL_PAYLOAD; } @@ -1135,13 +1135,13 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe) "(CDC: %08x)\n", card->name, len, rpp->len, readl(SAR_REG_CDC)); recycle_rx_pool_skb(card, rpp); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); return; } if (stat & SAR_RSQE_CRC) { RXPRINTK("%s: AAL5 CRC error.\n", card->name); recycle_rx_pool_skb(card, rpp); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); return; } if (skb_queue_len(&rpp->queue) > 1) { @@ -1152,7 +1152,7 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe) RXPRINTK("%s: Can't alloc RX skb.\n", card->name); recycle_rx_pool_skb(card, rpp); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); return; } if (!atm_charge(vcc, skb->truesize)) { @@ -1171,7 +1171,7 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe) __net_timestamp(skb); vcc->push(vcc, skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); return; } @@ -1193,7 +1193,7 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe) __net_timestamp(skb); vcc->push(vcc, skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); if (skb->truesize > SAR_FB_SIZE_3) add_rx_skb(card, 3, SAR_FB_SIZE_3, 1); @@ -1304,14 +1304,14 @@ idt77252_rx_raw(struct idt77252_dev *card) if (vcc->qos.aal != ATM_AAL0) { RPRINTK("%s: raw cell for non AAL0 vc %u.%u\n", card->name, vpi, vci); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); goto drop; } if ((sb = dev_alloc_skb(64)) == NULL) { printk("%s: Can't allocate buffers for AAL0.\n", card->name); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); goto drop; } @@ -1330,7 +1330,7 @@ idt77252_rx_raw(struct idt77252_dev *card) ATM_SKB(sb)->vcc = vcc; __net_timestamp(sb); vcc->push(vcc, sb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); drop: skb_pull(queue, 64); @@ -1955,13 +1955,13 @@ idt77252_send_skb(struct atm_vcc *vcc, struct sk_buff *skb, int oam) if (vc == NULL) { printk("%s: NULL connection in send().\n", card->name); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb(skb); return -EINVAL; } if (!test_bit(VCF_TX, &vc->flags)) { printk("%s: Trying to transmit on a non-tx VC.\n", card->name); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb(skb); return -EINVAL; } @@ -1973,14 +1973,14 @@ idt77252_send_skb(struct atm_vcc *vcc, struct sk_buff *skb, int oam) break; default: printk("%s: Unsupported AAL: %d\n", card->name, vcc->qos.aal); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb(skb); return -EINVAL; } if (skb_shinfo(skb)->nr_frags != 0) { printk("%s: No scatter-gather yet.\n", card->name); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb(skb); return -EINVAL; } @@ -1988,7 +1988,7 @@ idt77252_send_skb(struct atm_vcc *vcc, struct sk_buff *skb, int oam) err = queue_skb(card, vc, skb, oam); if (err) { - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb(skb); return err; } @@ -2011,7 +2011,7 @@ idt77252_send_oam(struct atm_vcc *vcc, void *cell, int flags) skb = dev_alloc_skb(64); if (!skb) { printk("%s: Out of memory in send_oam().\n", card->name); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); return -ENOMEM; } atomic_add(skb->truesize, &sk_atm(vcc)->sk_wmem_alloc); diff --git a/drivers/atm/iphase.c b/drivers/atm/iphase.c index 9e373ba..cf93727 100644 --- a/drivers/atm/iphase.c +++ b/drivers/atm/iphase.c @@ -1146,7 +1146,7 @@ static int rx_pkt(struct atm_dev *dev) status = (u_short) (buf_desc_ptr->desc_mode); if (status & (RX_CER | RX_PTE | RX_OFL)) { - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); IF_ERR(printk("IA: bad packet, dropping it");) if (status & RX_CER) { IF_ERR(printk(" cause: packet CRC error\n");) @@ -1169,7 +1169,7 @@ static int rx_pkt(struct atm_dev *dev) len = dma_addr - buf_addr; if (len > iadev->rx_buf_sz) { printk("Over %d bytes sdu received, dropped!!!\n", iadev->rx_buf_sz); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); goto out_free_desc; } @@ -1319,7 +1319,7 @@ static void rx_dle_intr(struct atm_dev *dev) ia_vcc = INPH_IA_VCC(vcc); if (ia_vcc == NULL) { - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); atm_return(vcc, skb->truesize); dev_kfree_skb_any(skb); goto INCR_DLE; @@ -1331,7 +1331,7 @@ static void rx_dle_intr(struct atm_dev *dev) if ((length > iadev->rx_buf_sz) || (length > (skb->len - sizeof(struct cpcs_trailer)))) { - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); IF_ERR(printk("rx_dle_intr: Bad AAL5 trailer %d (skb len %d)", length, skb->len);) atm_return(vcc, skb->truesize); @@ -1347,7 +1347,7 @@ static void rx_dle_intr(struct atm_dev *dev) IF_RX(printk("rx_dle_intr: skb push");) vcc->push(vcc,skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); iadev->rx_pkt_cnt++; } INCR_DLE: @@ -2827,15 +2827,15 @@ static int ia_ioctl(struct atm_dev *dev, unsigned int cmd, void __user *arg) { struct k_sonet_stats *stats; stats = &PRIV(_ia_dev[board])->sonet_stats; - printk("section_bip: %d\n", atomic_read(&stats->section_bip)); - printk("line_bip : %d\n", atomic_read(&stats->line_bip)); - printk("path_bip : %d\n", atomic_read(&stats->path_bip)); - printk("line_febe : %d\n", atomic_read(&stats->line_febe)); - printk("path_febe : %d\n", atomic_read(&stats->path_febe)); - printk("corr_hcs : %d\n", atomic_read(&stats->corr_hcs)); - printk("uncorr_hcs : %d\n", atomic_read(&stats->uncorr_hcs)); - printk("tx_cells : %d\n", atomic_read(&stats->tx_cells)); - printk("rx_cells : %d\n", atomic_read(&stats->rx_cells)); + printk("section_bip: %d\n", atomic_read_unchecked(&stats->section_bip)); + printk("line_bip : %d\n", atomic_read_unchecked(&stats->line_bip)); + printk("path_bip : %d\n", atomic_read_unchecked(&stats->path_bip)); + printk("line_febe : %d\n", atomic_read_unchecked(&stats->line_febe)); + printk("path_febe : %d\n", atomic_read_unchecked(&stats->path_febe)); + printk("corr_hcs : %d\n", atomic_read_unchecked(&stats->corr_hcs)); + printk("uncorr_hcs : %d\n", atomic_read_unchecked(&stats->uncorr_hcs)); + printk("tx_cells : %d\n", atomic_read_unchecked(&stats->tx_cells)); + printk("rx_cells : %d\n", atomic_read_unchecked(&stats->rx_cells)); } ia_cmds.status = 0; break; @@ -2940,7 +2940,7 @@ static int ia_pkt_tx (struct atm_vcc *vcc, struct sk_buff *skb) { if ((desc == 0) || (desc > iadev->num_tx_desc)) { IF_ERR(printk(DEV_LABEL "invalid desc for send: %d\n", desc);) - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); if (vcc->pop) vcc->pop(vcc, skb); else @@ -3045,14 +3045,14 @@ static int ia_pkt_tx (struct atm_vcc *vcc, struct sk_buff *skb) { ATM_DESC(skb) = vcc->vci; skb_queue_tail(&iadev->tx_dma_q, skb); - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); iadev->tx_pkt_cnt++; /* Increment transaction counter */ writel(2, iadev->dma+IPHASE5575_TX_COUNTER); #if 0 /* add flow control logic */ - if (atomic_read(&vcc->stats->tx) % 20 == 0) { + if (atomic_read_unchecked(&vcc->stats->tx) % 20 == 0) { if (iavcc->vc_desc_cnt > 10) { vcc->tx_quota = vcc->tx_quota * 3 / 4; printk("Tx1: vcc->tx_quota = %d \n", (u32)vcc->tx_quota ); diff --git a/drivers/atm/lanai.c b/drivers/atm/lanai.c index f556969..0da15eb 100644 --- a/drivers/atm/lanai.c +++ b/drivers/atm/lanai.c @@ -1303,7 +1303,7 @@ static void lanai_send_one_aal5(struct lanai_dev *lanai, vcc_tx_add_aal5_trailer(lvcc, skb->len, 0, 0); lanai_endtx(lanai, lvcc); lanai_free_skb(lvcc->tx.atmvcc, skb); - atomic_inc(&lvcc->tx.atmvcc->stats->tx); + atomic_inc_unchecked(&lvcc->tx.atmvcc->stats->tx); } /* Try to fill the buffer - don't call unless there is backlog */ @@ -1426,7 +1426,7 @@ static void vcc_rx_aal5(struct lanai_vcc *lvcc, int endptr) ATM_SKB(skb)->vcc = lvcc->rx.atmvcc; __net_timestamp(skb); lvcc->rx.atmvcc->push(lvcc->rx.atmvcc, skb); - atomic_inc(&lvcc->rx.atmvcc->stats->rx); + atomic_inc_unchecked(&lvcc->rx.atmvcc->stats->rx); out: lvcc->rx.buf.ptr = end; cardvcc_write(lvcc, endptr, vcc_rxreadptr); @@ -1667,7 +1667,7 @@ static int handle_service(struct lanai_dev *lanai, u32 s) DPRINTK("(itf %d) got RX service entry 0x%X for non-AAL5 " "vcc %d\n", lanai->number, (unsigned int) s, vci); lanai->stats.service_rxnotaal5++; - atomic_inc(&lvcc->rx.atmvcc->stats->rx_err); + atomic_inc_unchecked(&lvcc->rx.atmvcc->stats->rx_err); return 0; } if (likely(!(s & (SERVICE_TRASH | SERVICE_STREAM | SERVICE_CRCERR)))) { @@ -1679,7 +1679,7 @@ static int handle_service(struct lanai_dev *lanai, u32 s) int bytes; read_unlock(&vcc_sklist_lock); DPRINTK("got trashed rx pdu on vci %d\n", vci); - atomic_inc(&lvcc->rx.atmvcc->stats->rx_err); + atomic_inc_unchecked(&lvcc->rx.atmvcc->stats->rx_err); lvcc->stats.x.aal5.service_trash++; bytes = (SERVICE_GET_END(s) * 16) - (((unsigned long) lvcc->rx.buf.ptr) - @@ -1691,7 +1691,7 @@ static int handle_service(struct lanai_dev *lanai, u32 s) } if (s & SERVICE_STREAM) { read_unlock(&vcc_sklist_lock); - atomic_inc(&lvcc->rx.atmvcc->stats->rx_err); + atomic_inc_unchecked(&lvcc->rx.atmvcc->stats->rx_err); lvcc->stats.x.aal5.service_stream++; printk(KERN_ERR DEV_LABEL "(itf %d): Got AAL5 stream " "PDU on VCI %d!\n", lanai->number, vci); @@ -1699,7 +1699,7 @@ static int handle_service(struct lanai_dev *lanai, u32 s) return 0; } DPRINTK("got rx crc error on vci %d\n", vci); - atomic_inc(&lvcc->rx.atmvcc->stats->rx_err); + atomic_inc_unchecked(&lvcc->rx.atmvcc->stats->rx_err); lvcc->stats.x.aal5.service_rxcrc++; lvcc->rx.buf.ptr = &lvcc->rx.buf.start[SERVICE_GET_END(s) * 4]; cardvcc_write(lvcc, SERVICE_GET_END(s), vcc_rxreadptr); diff --git a/drivers/atm/nicstar.c b/drivers/atm/nicstar.c index 1c70c45..300718d 100644 --- a/drivers/atm/nicstar.c +++ b/drivers/atm/nicstar.c @@ -1654,7 +1654,7 @@ static int ns_send(struct atm_vcc *vcc, struct sk_buff *skb) if ((vc = (vc_map *) vcc->dev_data) == NULL) { printk("nicstar%d: vcc->dev_data == NULL on ns_send().\n", card->index); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb_any(skb); return -EINVAL; } @@ -1662,7 +1662,7 @@ static int ns_send(struct atm_vcc *vcc, struct sk_buff *skb) if (!vc->tx) { printk("nicstar%d: Trying to transmit on a non-tx VC.\n", card->index); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb_any(skb); return -EINVAL; } @@ -1670,14 +1670,14 @@ static int ns_send(struct atm_vcc *vcc, struct sk_buff *skb) if (vcc->qos.aal != ATM_AAL5 && vcc->qos.aal != ATM_AAL0) { printk("nicstar%d: Only AAL0 and AAL5 are supported.\n", card->index); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb_any(skb); return -EINVAL; } if (skb_shinfo(skb)->nr_frags != 0) { printk("nicstar%d: No scatter-gather yet.\n", card->index); - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb_any(skb); return -EINVAL; } @@ -1725,11 +1725,11 @@ static int ns_send(struct atm_vcc *vcc, struct sk_buff *skb) } if (push_scqe(card, vc, scq, &scqe, skb) != 0) { - atomic_inc(&vcc->stats->tx_err); + atomic_inc_unchecked(&vcc->stats->tx_err); dev_kfree_skb_any(skb); return -EIO; } - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); return 0; } @@ -2046,14 +2046,14 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) printk ("nicstar%d: Can't allocate buffers for aal0.\n", card->index); - atomic_add(i, &vcc->stats->rx_drop); + atomic_add_unchecked(i, &vcc->stats->rx_drop); break; } if (!atm_charge(vcc, sb->truesize)) { RXPRINTK ("nicstar%d: atm_charge() dropped aal0 packets.\n", card->index); - atomic_add(i - 1, &vcc->stats->rx_drop); /* already increased by 1 */ + atomic_add_unchecked(i - 1, &vcc->stats->rx_drop); /* already increased by 1 */ dev_kfree_skb_any(sb); break; } @@ -2068,7 +2068,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) ATM_SKB(sb)->vcc = vcc; __net_timestamp(sb); vcc->push(vcc, sb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); cell += ATM_CELL_PAYLOAD; } @@ -2085,7 +2085,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) if (iovb == NULL) { printk("nicstar%d: Out of iovec buffers.\n", card->index); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); recycle_rx_buf(card, skb); return; } @@ -2109,7 +2109,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) small or large buffer itself. */ } else if (NS_PRV_IOVCNT(iovb) >= NS_MAX_IOVECS) { printk("nicstar%d: received too big AAL5 SDU.\n", card->index); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); recycle_iovec_rx_bufs(card, (struct iovec *)iovb->data, NS_MAX_IOVECS); NS_PRV_IOVCNT(iovb) = 0; @@ -2129,7 +2129,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) ("nicstar%d: Expected a small buffer, and this is not one.\n", card->index); which_list(card, skb); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); recycle_rx_buf(card, skb); vc->rx_iov = NULL; recycle_iov_buf(card, iovb); @@ -2142,7 +2142,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) ("nicstar%d: Expected a large buffer, and this is not one.\n", card->index); which_list(card, skb); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); recycle_iovec_rx_bufs(card, (struct iovec *)iovb->data, NS_PRV_IOVCNT(iovb)); vc->rx_iov = NULL; @@ -2165,7 +2165,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) printk(" - PDU size mismatch.\n"); else printk(".\n"); - atomic_inc(&vcc->stats->rx_err); + atomic_inc_unchecked(&vcc->stats->rx_err); recycle_iovec_rx_bufs(card, (struct iovec *)iovb->data, NS_PRV_IOVCNT(iovb)); vc->rx_iov = NULL; @@ -2179,7 +2179,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) /* skb points to a small buffer */ if (!atm_charge(vcc, skb->truesize)) { push_rxbufs(card, skb); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); } else { skb_put(skb, len); dequeue_sm_buf(card, skb); @@ -2189,7 +2189,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) ATM_SKB(skb)->vcc = vcc; __net_timestamp(skb); vcc->push(vcc, skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); } } else if (NS_PRV_IOVCNT(iovb) == 2) { /* One small plus one large buffer */ struct sk_buff *sb; @@ -2200,7 +2200,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) if (len <= NS_SMBUFSIZE) { if (!atm_charge(vcc, sb->truesize)) { push_rxbufs(card, sb); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); } else { skb_put(sb, len); dequeue_sm_buf(card, sb); @@ -2210,7 +2210,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) ATM_SKB(sb)->vcc = vcc; __net_timestamp(sb); vcc->push(vcc, sb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); } push_rxbufs(card, skb); @@ -2219,7 +2219,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) if (!atm_charge(vcc, skb->truesize)) { push_rxbufs(card, skb); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); } else { dequeue_lg_buf(card, skb); #ifdef NS_USE_DESTRUCTORS @@ -2232,7 +2232,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) ATM_SKB(skb)->vcc = vcc; __net_timestamp(skb); vcc->push(vcc, skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); } push_rxbufs(card, sb); @@ -2253,7 +2253,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) printk ("nicstar%d: Out of huge buffers.\n", card->index); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); recycle_iovec_rx_bufs(card, (struct iovec *) iovb->data, @@ -2304,7 +2304,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) card->hbpool.count++; } else dev_kfree_skb_any(hb); - atomic_inc(&vcc->stats->rx_drop); + atomic_inc_unchecked(&vcc->stats->rx_drop); } else { /* Copy the small buffer to the huge buffer */ sb = (struct sk_buff *)iov->iov_base; @@ -2341,7 +2341,7 @@ static void dequeue_rx(ns_dev * card, ns_rsqe * rsqe) #endif /* NS_USE_DESTRUCTORS */ __net_timestamp(hb); vcc->push(vcc, hb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); } } diff --git a/drivers/atm/solos-pci.c b/drivers/atm/solos-pci.c index e8cd652..bbbd1fc 100644 --- a/drivers/atm/solos-pci.c +++ b/drivers/atm/solos-pci.c @@ -714,7 +714,7 @@ void solos_bh(unsigned long card_arg) } atm_charge(vcc, skb->truesize); vcc->push(vcc, skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); break; case PKT_STATUS: @@ -1008,7 +1008,7 @@ static uint32_t fpga_tx(struct solos_card *card) vcc = SKB_CB(oldskb)->vcc; if (vcc) { - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); solos_pop(vcc, oldskb); } else dev_kfree_skb_irq(oldskb); diff --git a/drivers/atm/suni.c b/drivers/atm/suni.c index 90f1ccc..04c4a1e 100644 --- a/drivers/atm/suni.c +++ b/drivers/atm/suni.c @@ -50,8 +50,8 @@ static DEFINE_SPINLOCK(sunis_lock); #define ADD_LIMITED(s,v) \ - atomic_add((v),&stats->s); \ - if (atomic_read(&stats->s) < 0) atomic_set(&stats->s,INT_MAX); + atomic_add_unchecked((v),&stats->s); \ + if (atomic_read_unchecked(&stats->s) < 0) atomic_set_unchecked(&stats->s,INT_MAX); static void suni_hz(unsigned long from_timer) diff --git a/drivers/atm/uPD98402.c b/drivers/atm/uPD98402.c index 5120a96..e2572bd 100644 --- a/drivers/atm/uPD98402.c +++ b/drivers/atm/uPD98402.c @@ -42,7 +42,7 @@ static int fetch_stats(struct atm_dev *dev,struct sonet_stats __user *arg,int ze struct sonet_stats tmp; int error = 0; - atomic_add(GET(HECCT),&PRIV(dev)->sonet_stats.uncorr_hcs); + atomic_add_unchecked(GET(HECCT),&PRIV(dev)->sonet_stats.uncorr_hcs); sonet_copy_stats(&PRIV(dev)->sonet_stats,&tmp); if (arg) error = copy_to_user(arg,&tmp,sizeof(tmp)); if (zero && !error) { @@ -161,9 +161,9 @@ static int uPD98402_ioctl(struct atm_dev *dev,unsigned int cmd,void __user *arg) #define ADD_LIMITED(s,v) \ - { atomic_add(GET(v),&PRIV(dev)->sonet_stats.s); \ - if (atomic_read(&PRIV(dev)->sonet_stats.s) < 0) \ - atomic_set(&PRIV(dev)->sonet_stats.s,INT_MAX); } + { atomic_add_unchecked(GET(v),&PRIV(dev)->sonet_stats.s); \ + if (atomic_read_unchecked(&PRIV(dev)->sonet_stats.s) < 0) \ + atomic_set_unchecked(&PRIV(dev)->sonet_stats.s,INT_MAX); } static void stat_event(struct atm_dev *dev) @@ -194,7 +194,7 @@ static void uPD98402_int(struct atm_dev *dev) if (reason & uPD98402_INT_PFM) stat_event(dev); if (reason & uPD98402_INT_PCO) { (void) GET(PCOCR); /* clear interrupt cause */ - atomic_add(GET(HECCT), + atomic_add_unchecked(GET(HECCT), &PRIV(dev)->sonet_stats.uncorr_hcs); } if ((reason & uPD98402_INT_RFO) && @@ -222,9 +222,9 @@ static int uPD98402_start(struct atm_dev *dev) PUT(~(uPD98402_INT_PFM | uPD98402_INT_ALM | uPD98402_INT_RFO | uPD98402_INT_LOS),PIMR); /* enable them */ (void) fetch_stats(dev,NULL,1); /* clear kernel counters */ - atomic_set(&PRIV(dev)->sonet_stats.corr_hcs,-1); - atomic_set(&PRIV(dev)->sonet_stats.tx_cells,-1); - atomic_set(&PRIV(dev)->sonet_stats.rx_cells,-1); + atomic_set_unchecked(&PRIV(dev)->sonet_stats.corr_hcs,-1); + atomic_set_unchecked(&PRIV(dev)->sonet_stats.tx_cells,-1); + atomic_set_unchecked(&PRIV(dev)->sonet_stats.rx_cells,-1); return 0; } diff --git a/drivers/atm/zatm.c b/drivers/atm/zatm.c index d889f56..17eb71e 100644 --- a/drivers/atm/zatm.c +++ b/drivers/atm/zatm.c @@ -460,7 +460,7 @@ printk("dummy: 0x%08lx, 0x%08lx\n",dummy[0],dummy[1]); } if (!size) { dev_kfree_skb_irq(skb); - if (vcc) atomic_inc(&vcc->stats->rx_err); + if (vcc) atomic_inc_unchecked(&vcc->stats->rx_err); continue; } if (!atm_charge(vcc,skb->truesize)) { @@ -470,7 +470,7 @@ printk("dummy: 0x%08lx, 0x%08lx\n",dummy[0],dummy[1]); skb->len = size; ATM_SKB(skb)->vcc = vcc; vcc->push(vcc,skb); - atomic_inc(&vcc->stats->rx); + atomic_inc_unchecked(&vcc->stats->rx); } zout(pos & 0xffff,MTA(mbx)); #if 0 /* probably a stupid idea */ @@ -734,7 +734,7 @@ if (*ZATM_PRV_DSC(skb) != (uPD98401_TXPD_V | uPD98401_TXPD_DP | skb_queue_head(&zatm_vcc->backlog,skb); break; } - atomic_inc(&vcc->stats->tx); + atomic_inc_unchecked(&vcc->stats->tx); wake_up(&zatm_vcc->tx_wait); } diff --git a/drivers/base/devtmpfs.c b/drivers/base/devtmpfs.c index 8493536..31adee0 100644 --- a/drivers/base/devtmpfs.c +++ b/drivers/base/devtmpfs.c @@ -368,7 +368,7 @@ int devtmpfs_mount(const char *mntdir) if (!thread) return 0; - err = sys_mount("devtmpfs", (char *)mntdir, "devtmpfs", MS_SILENT, NULL); + err = sys_mount((char __force_user *)"devtmpfs", (char __force_user *)mntdir, (char __force_user *)"devtmpfs", MS_SILENT, NULL); if (err) printk(KERN_INFO "devtmpfs: error mounting %i\n", err); else diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c index caf995f..6f76697 100644 --- a/drivers/base/power/wakeup.c +++ b/drivers/base/power/wakeup.c @@ -30,14 +30,14 @@ bool events_check_enabled; * They need to be modified together atomically, so it's better to use one * atomic variable to hold them both. */ -static atomic_t combined_event_count = ATOMIC_INIT(0); +static atomic_unchecked_t combined_event_count = ATOMIC_INIT(0); #define IN_PROGRESS_BITS (sizeof(int) * 4) #define MAX_IN_PROGRESS ((1 << IN_PROGRESS_BITS) - 1) static void split_counters(unsigned int *cnt, unsigned int *inpr) { - unsigned int comb = atomic_read(&combined_event_count); + unsigned int comb = atomic_read_unchecked(&combined_event_count); *cnt = (comb >> IN_PROGRESS_BITS); *inpr = comb & MAX_IN_PROGRESS; @@ -353,7 +353,7 @@ static void wakeup_source_activate(struct wakeup_source *ws) ws->last_time = ktime_get(); /* Increment the counter of events in progress. */ - atomic_inc(&combined_event_count); + atomic_inc_unchecked(&combined_event_count); } /** @@ -443,7 +443,7 @@ static void wakeup_source_deactivate(struct wakeup_source *ws) * Increment the counter of registered wakeup events and decrement the * couter of wakeup events in progress simultaneously. */ - atomic_add(MAX_IN_PROGRESS, &combined_event_count); + atomic_add_unchecked(MAX_IN_PROGRESS, &combined_event_count); } /** diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index b0f553b..77b928b 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -1198,6 +1198,8 @@ static int cciss_ioctl32_passthru(struct block_device *bdev, fmode_t mode, int err; u32 cp; + memset(&arg64, 0, sizeof(arg64)); + err = 0; err |= copy_from_user(&arg64.LUN_info, &arg32->LUN_info, @@ -3007,7 +3009,7 @@ static void start_io(ctlr_info_t *h) while (!list_empty(&h->reqQ)) { c = list_entry(h->reqQ.next, CommandList_struct, list); /* can't do anything if fifo is full */ - if ((h->access.fifo_full(h))) { + if ((h->access->fifo_full(h))) { dev_warn(&h->pdev->dev, "fifo full\n"); break; } @@ -3017,7 +3019,7 @@ static void start_io(ctlr_info_t *h) h->Qdepth--; /* Tell the controller execute command */ - h->access.submit_command(h, c); + h->access->submit_command(h, c); /* Put job onto the completed Q */ addQ(&h->cmpQ, c); @@ -3443,17 +3445,17 @@ startio: static inline unsigned long get_next_completion(ctlr_info_t *h) { - return h->access.command_completed(h); + return h->access->command_completed(h); } static inline int interrupt_pending(ctlr_info_t *h) { - return h->access.intr_pending(h); + return h->access->intr_pending(h); } static inline long interrupt_not_for_us(ctlr_info_t *h) { - return ((h->access.intr_pending(h) == 0) || + return ((h->access->intr_pending(h) == 0) || (h->interrupts_enabled == 0)); } @@ -3486,7 +3488,7 @@ static inline u32 next_command(ctlr_info_t *h) u32 a; if (unlikely(!(h->transMethod & CFGTBL_Trans_Performant))) - return h->access.command_completed(h); + return h->access->command_completed(h); if ((*(h->reply_pool_head) & 1) == (h->reply_pool_wraparound)) { a = *(h->reply_pool_head); /* Next cmd in ring buffer */ @@ -4044,7 +4046,7 @@ static void __devinit cciss_put_controller_into_performant_mode(ctlr_info_t *h) trans_support & CFGTBL_Trans_use_short_tags); /* Change the access methods to the performant access methods */ - h->access = SA5_performant_access; + h->access = &SA5_performant_access; h->transMethod = CFGTBL_Trans_Performant; return; @@ -4316,7 +4318,7 @@ static int __devinit cciss_pci_init(ctlr_info_t *h) if (prod_index < 0) return -ENODEV; h->product_name = products[prod_index].product_name; - h->access = *(products[prod_index].access); + h->access = products[prod_index].access; if (cciss_board_disabled(h)) { dev_warn(&h->pdev->dev, "controller appears to be disabled\n"); @@ -5041,7 +5043,7 @@ reinit_after_soft_reset: } /* make sure the board interrupts are off */ - h->access.set_intr_mask(h, CCISS_INTR_OFF); + h->access->set_intr_mask(h, CCISS_INTR_OFF); rc = cciss_request_irq(h, do_cciss_msix_intr, do_cciss_intx); if (rc) goto clean2; @@ -5093,7 +5095,7 @@ reinit_after_soft_reset: * fake ones to scoop up any residual completions. */ spin_lock_irqsave(&h->lock, flags); - h->access.set_intr_mask(h, CCISS_INTR_OFF); + h->access->set_intr_mask(h, CCISS_INTR_OFF); spin_unlock_irqrestore(&h->lock, flags); free_irq(h->intr[h->intr_mode], h); rc = cciss_request_irq(h, cciss_msix_discard_completions, @@ -5113,9 +5115,9 @@ reinit_after_soft_reset: dev_info(&h->pdev->dev, "Board READY.\n"); dev_info(&h->pdev->dev, "Waiting for stale completions to drain.\n"); - h->access.set_intr_mask(h, CCISS_INTR_ON); + h->access->set_intr_mask(h, CCISS_INTR_ON); msleep(10000); - h->access.set_intr_mask(h, CCISS_INTR_OFF); + h->access->set_intr_mask(h, CCISS_INTR_OFF); rc = controller_reset_failed(h->cfgtable); if (rc) @@ -5138,7 +5140,7 @@ reinit_after_soft_reset: cciss_scsi_setup(h); /* Turn the interrupts on so we can service requests */ - h->access.set_intr_mask(h, CCISS_INTR_ON); + h->access->set_intr_mask(h, CCISS_INTR_ON); /* Get the firmware version */ inq_buff = kzalloc(sizeof(InquiryData_struct), GFP_KERNEL); @@ -5211,7 +5213,7 @@ static void cciss_shutdown(struct pci_dev *pdev) kfree(flush_buf); if (return_code != IO_OK) dev_warn(&h->pdev->dev, "Error flushing cache\n"); - h->access.set_intr_mask(h, CCISS_INTR_OFF); + h->access->set_intr_mask(h, CCISS_INTR_OFF); free_irq(h->intr[h->intr_mode], h); } diff --git a/drivers/block/cciss.h b/drivers/block/cciss.h index 7fda30e..eb5dfe0 100644 --- a/drivers/block/cciss.h +++ b/drivers/block/cciss.h @@ -101,7 +101,7 @@ struct ctlr_info /* information about each logical volume */ drive_info_struct *drv[CISS_MAX_LUN]; - struct access_method access; + struct access_method *access; /* queue and queue Info */ struct list_head reqQ; diff --git a/drivers/block/cpqarray.c b/drivers/block/cpqarray.c index 9125bbe..eede5c8 100644 --- a/drivers/block/cpqarray.c +++ b/drivers/block/cpqarray.c @@ -404,7 +404,7 @@ static int __devinit cpqarray_register_ctlr( int i, struct pci_dev *pdev) if (register_blkdev(COMPAQ_SMART2_MAJOR+i, hba[i]->devname)) { goto Enomem4; } - hba[i]->access.set_intr_mask(hba[i], 0); + hba[i]->access->set_intr_mask(hba[i], 0); if (request_irq(hba[i]->intr, do_ida_intr, IRQF_DISABLED|IRQF_SHARED, hba[i]->devname, hba[i])) { @@ -459,7 +459,7 @@ static int __devinit cpqarray_register_ctlr( int i, struct pci_dev *pdev) add_timer(&hba[i]->timer); /* Enable IRQ now that spinlock and rate limit timer are set up */ - hba[i]->access.set_intr_mask(hba[i], FIFO_NOT_EMPTY); + hba[i]->access->set_intr_mask(hba[i], FIFO_NOT_EMPTY); for(j=0; jproduct_name = products[i].product_name; - c->access = *(products[i].access); + c->access = products[i].access; break; } } @@ -792,7 +792,7 @@ static int __devinit cpqarray_eisa_detect(void) hba[ctlr]->intr = intr; sprintf(hba[ctlr]->devname, "ida%d", nr_ctlr); hba[ctlr]->product_name = products[j].product_name; - hba[ctlr]->access = *(products[j].access); + hba[ctlr]->access = products[j].access; hba[ctlr]->ctlr = ctlr; hba[ctlr]->board_id = board_id; hba[ctlr]->pci_dev = NULL; /* not PCI */ @@ -980,7 +980,7 @@ static void start_io(ctlr_info_t *h) while((c = h->reqQ) != NULL) { /* Can't do anything if we're busy */ - if (h->access.fifo_full(h) == 0) + if (h->access->fifo_full(h) == 0) return; /* Get the first entry from the request Q */ @@ -988,7 +988,7 @@ static void start_io(ctlr_info_t *h) h->Qdepth--; /* Tell the controller to do our bidding */ - h->access.submit_command(h, c); + h->access->submit_command(h, c); /* Get onto the completion Q */ addQ(&h->cmpQ, c); @@ -1050,7 +1050,7 @@ static irqreturn_t do_ida_intr(int irq, void *dev_id) unsigned long flags; __u32 a,a1; - istat = h->access.intr_pending(h); + istat = h->access->intr_pending(h); /* Is this interrupt for us? */ if (istat == 0) return IRQ_NONE; @@ -1061,7 +1061,7 @@ static irqreturn_t do_ida_intr(int irq, void *dev_id) */ spin_lock_irqsave(IDA_LOCK(h->ctlr), flags); if (istat & FIFO_NOT_EMPTY) { - while((a = h->access.command_completed(h))) { + while((a = h->access->command_completed(h))) { a1 = a; a &= ~3; if ((c = h->cmpQ) == NULL) { @@ -1449,11 +1449,11 @@ static int sendcmd( /* * Disable interrupt */ - info_p->access.set_intr_mask(info_p, 0); + info_p->access->set_intr_mask(info_p, 0); /* Make sure there is room in the command FIFO */ /* Actually it should be completely empty at this time. */ for (i = 200000; i > 0; i--) { - temp = info_p->access.fifo_full(info_p); + temp = info_p->access->fifo_full(info_p); if (temp != 0) { break; } @@ -1466,7 +1466,7 @@ DBG( /* * Send the cmd */ - info_p->access.submit_command(info_p, c); + info_p->access->submit_command(info_p, c); complete = pollcomplete(ctlr); pci_unmap_single(info_p->pci_dev, (dma_addr_t) c->req.sg[0].addr, @@ -1549,9 +1549,9 @@ static int revalidate_allvol(ctlr_info_t *host) * we check the new geometry. Then turn interrupts back on when * we're done. */ - host->access.set_intr_mask(host, 0); + host->access->set_intr_mask(host, 0); getgeometry(ctlr); - host->access.set_intr_mask(host, FIFO_NOT_EMPTY); + host->access->set_intr_mask(host, FIFO_NOT_EMPTY); for(i=0; i 0; i--) { - done = hba[ctlr]->access.command_completed(hba[ctlr]); + done = hba[ctlr]->access->command_completed(hba[ctlr]); if (done == 0) { udelay(10); /* a short fixed delay */ } else diff --git a/drivers/block/cpqarray.h b/drivers/block/cpqarray.h index be73e9d..7fbf140 100644 --- a/drivers/block/cpqarray.h +++ b/drivers/block/cpqarray.h @@ -99,7 +99,7 @@ struct ctlr_info { drv_info_t drv[NWD]; struct proc_dir_entry *proc; - struct access_method access; + struct access_method *access; cmdlist_t *reqQ; cmdlist_t *cmpQ; diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 8d68056..e67050f 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -736,7 +736,7 @@ struct drbd_request; struct drbd_epoch { struct list_head list; unsigned int barrier_nr; - atomic_t epoch_size; /* increased on every request added. */ + atomic_unchecked_t epoch_size; /* increased on every request added. */ atomic_t active; /* increased on every req. added, and dec on every finished. */ unsigned long flags; }; @@ -1108,7 +1108,7 @@ struct drbd_conf { void *int_dig_in; void *int_dig_vv; wait_queue_head_t seq_wait; - atomic_t packet_seq; + atomic_unchecked_t packet_seq; unsigned int peer_seq; spinlock_t peer_seq_lock; unsigned int minor; @@ -1617,30 +1617,30 @@ static inline int drbd_setsockopt(struct socket *sock, int level, int optname, static inline void drbd_tcp_cork(struct socket *sock) { - int __user val = 1; + int val = 1; (void) drbd_setsockopt(sock, SOL_TCP, TCP_CORK, - (char __user *)&val, sizeof(val)); + (char __force_user *)&val, sizeof(val)); } static inline void drbd_tcp_uncork(struct socket *sock) { - int __user val = 0; + int val = 0; (void) drbd_setsockopt(sock, SOL_TCP, TCP_CORK, - (char __user *)&val, sizeof(val)); + (char __force_user *)&val, sizeof(val)); } static inline void drbd_tcp_nodelay(struct socket *sock) { - int __user val = 1; + int val = 1; (void) drbd_setsockopt(sock, SOL_TCP, TCP_NODELAY, - (char __user *)&val, sizeof(val)); + (char __force_user *)&val, sizeof(val)); } static inline void drbd_tcp_quickack(struct socket *sock) { - int __user val = 2; + int val = 2; (void) drbd_setsockopt(sock, SOL_TCP, TCP_QUICKACK, - (char __user *)&val, sizeof(val)); + (char __force_user *)&val, sizeof(val)); } void drbd_bump_write_ordering(struct drbd_conf *mdev, enum write_ordering_e wo); diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c index 211fc44..c5116f1 100644 --- a/drivers/block/drbd/drbd_main.c +++ b/drivers/block/drbd/drbd_main.c @@ -2397,7 +2397,7 @@ static int _drbd_send_ack(struct drbd_conf *mdev, enum drbd_packets cmd, p.sector = sector; p.block_id = block_id; p.blksize = blksize; - p.seq_num = cpu_to_be32(atomic_add_return(1, &mdev->packet_seq)); + p.seq_num = cpu_to_be32(atomic_add_return_unchecked(1, &mdev->packet_seq)); if (!mdev->meta.socket || mdev->state.conn < C_CONNECTED) return false; @@ -2696,7 +2696,7 @@ int drbd_send_dblock(struct drbd_conf *mdev, struct drbd_request *req) p.sector = cpu_to_be64(req->sector); p.block_id = (unsigned long)req; p.seq_num = cpu_to_be32(req->seq_num = - atomic_add_return(1, &mdev->packet_seq)); + atomic_add_return_unchecked(1, &mdev->packet_seq)); dp_flags = bio_flags_to_wire(mdev, req->master_bio->bi_rw); @@ -2981,7 +2981,7 @@ void drbd_init_set_defaults(struct drbd_conf *mdev) atomic_set(&mdev->unacked_cnt, 0); atomic_set(&mdev->local_cnt, 0); atomic_set(&mdev->net_cnt, 0); - atomic_set(&mdev->packet_seq, 0); + atomic_set_unchecked(&mdev->packet_seq, 0); atomic_set(&mdev->pp_in_use, 0); atomic_set(&mdev->pp_in_use_by_net, 0); atomic_set(&mdev->rs_sect_in, 0); @@ -3063,8 +3063,8 @@ void drbd_mdev_cleanup(struct drbd_conf *mdev) mdev->receiver.t_state); /* no need to lock it, I'm the only thread alive */ - if (atomic_read(&mdev->current_epoch->epoch_size) != 0) - dev_err(DEV, "epoch_size:%d\n", atomic_read(&mdev->current_epoch->epoch_size)); + if (atomic_read_unchecked(&mdev->current_epoch->epoch_size) != 0) + dev_err(DEV, "epoch_size:%d\n", atomic_read_unchecked(&mdev->current_epoch->epoch_size)); mdev->al_writ_cnt = mdev->bm_writ_cnt = mdev->read_cnt = diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c index af2a250..219c74b 100644 --- a/drivers/block/drbd/drbd_nl.c +++ b/drivers/block/drbd/drbd_nl.c @@ -2359,7 +2359,7 @@ static void drbd_connector_callback(struct cn_msg *req, struct netlink_skb_parms module_put(THIS_MODULE); } -static atomic_t drbd_nl_seq = ATOMIC_INIT(2); /* two. */ +static atomic_unchecked_t drbd_nl_seq = ATOMIC_INIT(2); /* two. */ static unsigned short * __tl_add_blob(unsigned short *tl, enum drbd_tags tag, const void *data, @@ -2430,7 +2430,7 @@ void drbd_bcast_state(struct drbd_conf *mdev, union drbd_state state) cn_reply->id.idx = CN_IDX_DRBD; cn_reply->id.val = CN_VAL_DRBD; - cn_reply->seq = atomic_add_return(1, &drbd_nl_seq); + cn_reply->seq = atomic_add_return_unchecked(1, &drbd_nl_seq); cn_reply->ack = 0; /* not used here. */ cn_reply->len = sizeof(struct drbd_nl_cfg_reply) + (int)((char *)tl - (char *)reply->tag_list); @@ -2462,7 +2462,7 @@ void drbd_bcast_ev_helper(struct drbd_conf *mdev, char *helper_name) cn_reply->id.idx = CN_IDX_DRBD; cn_reply->id.val = CN_VAL_DRBD; - cn_reply->seq = atomic_add_return(1, &drbd_nl_seq); + cn_reply->seq = atomic_add_return_unchecked(1, &drbd_nl_seq); cn_reply->ack = 0; /* not used here. */ cn_reply->len = sizeof(struct drbd_nl_cfg_reply) + (int)((char *)tl - (char *)reply->tag_list); @@ -2540,7 +2540,7 @@ void drbd_bcast_ee(struct drbd_conf *mdev, cn_reply->id.idx = CN_IDX_DRBD; cn_reply->id.val = CN_VAL_DRBD; - cn_reply->seq = atomic_add_return(1,&drbd_nl_seq); + cn_reply->seq = atomic_add_return_unchecked(1,&drbd_nl_seq); cn_reply->ack = 0; // not used here. cn_reply->len = sizeof(struct drbd_nl_cfg_reply) + (int)((char*)tl - (char*)reply->tag_list); @@ -2579,7 +2579,7 @@ void drbd_bcast_sync_progress(struct drbd_conf *mdev) cn_reply->id.idx = CN_IDX_DRBD; cn_reply->id.val = CN_VAL_DRBD; - cn_reply->seq = atomic_add_return(1, &drbd_nl_seq); + cn_reply->seq = atomic_add_return_unchecked(1, &drbd_nl_seq); cn_reply->ack = 0; /* not used here. */ cn_reply->len = sizeof(struct drbd_nl_cfg_reply) + (int)((char *)tl - (char *)reply->tag_list); diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index 43beaca..4a5b1dd 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -894,7 +894,7 @@ retry: sock->sk->sk_sndtimeo = mdev->net_conf->timeout*HZ/10; sock->sk->sk_rcvtimeo = MAX_SCHEDULE_TIMEOUT; - atomic_set(&mdev->packet_seq, 0); + atomic_set_unchecked(&mdev->packet_seq, 0); mdev->peer_seq = 0; drbd_thread_start(&mdev->asender); @@ -985,7 +985,7 @@ static enum finish_epoch drbd_may_finish_epoch(struct drbd_conf *mdev, do { next_epoch = NULL; - epoch_size = atomic_read(&epoch->epoch_size); + epoch_size = atomic_read_unchecked(&epoch->epoch_size); switch (ev & ~EV_CLEANUP) { case EV_PUT: @@ -1020,7 +1020,7 @@ static enum finish_epoch drbd_may_finish_epoch(struct drbd_conf *mdev, rv = FE_DESTROYED; } else { epoch->flags = 0; - atomic_set(&epoch->epoch_size, 0); + atomic_set_unchecked(&epoch->epoch_size, 0); /* atomic_set(&epoch->active, 0); is already zero */ if (rv == FE_STILL_LIVE) rv = FE_RECYCLED; @@ -1191,14 +1191,14 @@ static int receive_Barrier(struct drbd_conf *mdev, enum drbd_packets cmd, unsign drbd_wait_ee_list_empty(mdev, &mdev->active_ee); drbd_flush(mdev); - if (atomic_read(&mdev->current_epoch->epoch_size)) { + if (atomic_read_unchecked(&mdev->current_epoch->epoch_size)) { epoch = kmalloc(sizeof(struct drbd_epoch), GFP_NOIO); if (epoch) break; } epoch = mdev->current_epoch; - wait_event(mdev->ee_wait, atomic_read(&epoch->epoch_size) == 0); + wait_event(mdev->ee_wait, atomic_read_unchecked(&epoch->epoch_size) == 0); D_ASSERT(atomic_read(&epoch->active) == 0); D_ASSERT(epoch->flags == 0); @@ -1210,11 +1210,11 @@ static int receive_Barrier(struct drbd_conf *mdev, enum drbd_packets cmd, unsign } epoch->flags = 0; - atomic_set(&epoch->epoch_size, 0); + atomic_set_unchecked(&epoch->epoch_size, 0); atomic_set(&epoch->active, 0); spin_lock(&mdev->epoch_lock); - if (atomic_read(&mdev->current_epoch->epoch_size)) { + if (atomic_read_unchecked(&mdev->current_epoch->epoch_size)) { list_add(&epoch->list, &mdev->current_epoch->list); mdev->current_epoch = epoch; mdev->epochs++; @@ -1663,7 +1663,7 @@ static int receive_Data(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned spin_unlock(&mdev->peer_seq_lock); drbd_send_ack_dp(mdev, P_NEG_ACK, p, data_size); - atomic_inc(&mdev->current_epoch->epoch_size); + atomic_inc_unchecked(&mdev->current_epoch->epoch_size); return drbd_drain_block(mdev, data_size); } @@ -1689,7 +1689,7 @@ static int receive_Data(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned spin_lock(&mdev->epoch_lock); e->epoch = mdev->current_epoch; - atomic_inc(&e->epoch->epoch_size); + atomic_inc_unchecked(&e->epoch->epoch_size); atomic_inc(&e->epoch->active); spin_unlock(&mdev->epoch_lock); @@ -3885,7 +3885,7 @@ static void drbd_disconnect(struct drbd_conf *mdev) D_ASSERT(list_empty(&mdev->done_ee)); /* ok, no more ee's on the fly, it is safe to reset the epoch_size */ - atomic_set(&mdev->current_epoch->epoch_size, 0); + atomic_set_unchecked(&mdev->current_epoch->epoch_size, 0); D_ASSERT(list_empty(&mdev->current_epoch->list)); } diff --git a/drivers/block/loop.c b/drivers/block/loop.c index cd50435..ba1ffb5 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -226,7 +226,7 @@ static int __do_lo_send_write(struct file *file, mm_segment_t old_fs = get_fs(); set_fs(get_ds()); - bw = file->f_op->write(file, buf, len, &pos); + bw = file->f_op->write(file, (const char __force_user *)buf, len, &pos); set_fs(old_fs); if (likely(bw == len)) return 0; diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 4364303..9adf4ee 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -8,7 +8,8 @@ source "drivers/tty/Kconfig" config DEVKMEM bool "/dev/kmem virtual device support" - default y + default n + depends on !GRKERNSEC_KMEM help Say Y here if you want to support the /dev/kmem device. The /dev/kmem device is rarely used, but can be used for certain @@ -596,6 +597,7 @@ config DEVPORT bool depends on !M68K depends on ISA || PCI + depends on !GRKERNSEC_KMEM default y source "drivers/s390/char/Kconfig" diff --git a/drivers/char/agp/frontend.c b/drivers/char/agp/frontend.c index 2e04433..22afc64 100644 --- a/drivers/char/agp/frontend.c +++ b/drivers/char/agp/frontend.c @@ -817,7 +817,7 @@ static int agpioc_reserve_wrap(struct agp_file_private *priv, void __user *arg) if (copy_from_user(&reserve, arg, sizeof(struct agp_region))) return -EFAULT; - if ((unsigned) reserve.seg_count >= ~0U/sizeof(struct agp_segment)) + if ((unsigned) reserve.seg_count >= ~0U/sizeof(struct agp_segment_priv)) return -EFAULT; client = agp_find_client_by_pid(reserve.pid); diff --git a/drivers/char/briq_panel.c b/drivers/char/briq_panel.c index 095ab90..afad0a4 100644 --- a/drivers/char/briq_panel.c +++ b/drivers/char/briq_panel.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include #include @@ -34,6 +35,7 @@ static int vfd_is_open; static unsigned char vfd[40]; static int vfd_cursor; static unsigned char ledpb, led; +static DEFINE_MUTEX(vfd_mutex); static void update_vfd(void) { @@ -140,12 +142,15 @@ static ssize_t briq_panel_write(struct file *file, const char __user *buf, size_ if (!vfd_is_open) return -EBUSY; + mutex_lock(&vfd_mutex); for (;;) { char c; if (!indx) break; - if (get_user(c, buf)) + if (get_user(c, buf)) { + mutex_unlock(&vfd_mutex); return -EFAULT; + } if (esc) { set_led(c); esc = 0; @@ -175,6 +180,7 @@ static ssize_t briq_panel_write(struct file *file, const char __user *buf, size_ buf++; } update_vfd(); + mutex_unlock(&vfd_mutex); return len; } diff --git a/drivers/char/genrtc.c b/drivers/char/genrtc.c index f773a9d..65cd683 100644 --- a/drivers/char/genrtc.c +++ b/drivers/char/genrtc.c @@ -273,6 +273,7 @@ static int gen_rtc_ioctl(struct file *file, switch (cmd) { case RTC_PLL_GET: + memset(&pll, 0, sizeof(pll)); if (get_rtc_pll(&pll)) return -EINVAL; else diff --git a/drivers/char/hpet.c b/drivers/char/hpet.c index 0833896..cccce52 100644 --- a/drivers/char/hpet.c +++ b/drivers/char/hpet.c @@ -572,7 +572,7 @@ static inline unsigned long hpet_time_div(struct hpets *hpets, } static int -hpet_ioctl_common(struct hpet_dev *devp, int cmd, unsigned long arg, +hpet_ioctl_common(struct hpet_dev *devp, unsigned int cmd, unsigned long arg, struct hpet_info *info) { struct hpet_timer __iomem *timer; diff --git a/drivers/char/ipmi/ipmi_msghandler.c b/drivers/char/ipmi/ipmi_msghandler.c index 58c0e63..46c16bf 100644 --- a/drivers/char/ipmi/ipmi_msghandler.c +++ b/drivers/char/ipmi/ipmi_msghandler.c @@ -415,7 +415,7 @@ struct ipmi_smi { struct proc_dir_entry *proc_dir; char proc_dir_name[10]; - atomic_t stats[IPMI_NUM_STATS]; + atomic_unchecked_t stats[IPMI_NUM_STATS]; /* * run_to_completion duplicate of smb_info, smi_info @@ -448,9 +448,9 @@ static DEFINE_MUTEX(smi_watchers_mutex); #define ipmi_inc_stat(intf, stat) \ - atomic_inc(&(intf)->stats[IPMI_STAT_ ## stat]) + atomic_inc_unchecked(&(intf)->stats[IPMI_STAT_ ## stat]) #define ipmi_get_stat(intf, stat) \ - ((unsigned int) atomic_read(&(intf)->stats[IPMI_STAT_ ## stat])) + ((unsigned int) atomic_read_unchecked(&(intf)->stats[IPMI_STAT_ ## stat])) static int is_lan_addr(struct ipmi_addr *addr) { @@ -2868,7 +2868,7 @@ int ipmi_register_smi(struct ipmi_smi_handlers *handlers, INIT_LIST_HEAD(&intf->cmd_rcvrs); init_waitqueue_head(&intf->waitq); for (i = 0; i < IPMI_NUM_STATS; i++) - atomic_set(&intf->stats[i], 0); + atomic_set_unchecked(&intf->stats[i], 0); intf->proc_dir = NULL; diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 50fcf9c..91b5528 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -277,7 +277,7 @@ struct smi_info { unsigned char slave_addr; /* Counters and things for the proc filesystem. */ - atomic_t stats[SI_NUM_STATS]; + atomic_unchecked_t stats[SI_NUM_STATS]; struct task_struct *thread; @@ -286,9 +286,9 @@ struct smi_info { }; #define smi_inc_stat(smi, stat) \ - atomic_inc(&(smi)->stats[SI_STAT_ ## stat]) + atomic_inc_unchecked(&(smi)->stats[SI_STAT_ ## stat]) #define smi_get_stat(smi, stat) \ - ((unsigned int) atomic_read(&(smi)->stats[SI_STAT_ ## stat])) + ((unsigned int) atomic_read_unchecked(&(smi)->stats[SI_STAT_ ## stat])) #define SI_MAX_PARMS 4 @@ -3230,7 +3230,7 @@ static int try_smi_init(struct smi_info *new_smi) atomic_set(&new_smi->req_events, 0); new_smi->run_to_completion = 0; for (i = 0; i < SI_NUM_STATS; i++) - atomic_set(&new_smi->stats[i], 0); + atomic_set_unchecked(&new_smi->stats[i], 0); new_smi->interrupt_disabled = 1; atomic_set(&new_smi->stop_operation, 0); diff --git a/drivers/char/mbcs.c b/drivers/char/mbcs.c index 1aeaaba..e018570 100644 --- a/drivers/char/mbcs.c +++ b/drivers/char/mbcs.c @@ -800,7 +800,7 @@ static int mbcs_remove(struct cx_dev *dev) return 0; } -static const struct cx_device_id __devinitdata mbcs_id_table[] = { +static const struct cx_device_id __devinitconst mbcs_id_table[] = { { .part_num = MBCS_PART_NUM, .mfg_num = MBCS_MFG_NUM, diff --git a/drivers/char/mem.c b/drivers/char/mem.c index d6e9d08..4493e89 100644 --- a/drivers/char/mem.c +++ b/drivers/char/mem.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include #include @@ -35,6 +36,10 @@ # include #endif +#if defined(CONFIG_GRKERNSEC) && !defined(CONFIG_GRKERNSEC_NO_RBAC) +extern const struct file_operations grsec_fops; +#endif + static inline unsigned long size_inside_page(unsigned long start, unsigned long size) { @@ -66,9 +71,13 @@ static inline int range_is_allowed(unsigned long pfn, unsigned long size) while (cursor < to) { if (!devmem_is_allowed(pfn)) { +#ifdef CONFIG_GRKERNSEC_KMEM + gr_handle_mem_readwrite(from, to); +#else printk(KERN_INFO "Program %s tried to access /dev/mem between %Lx->%Lx.\n", current->comm, from, to); +#endif return 0; } cursor += PAGE_SIZE; @@ -76,6 +85,11 @@ static inline int range_is_allowed(unsigned long pfn, unsigned long size) } return 1; } +#elif defined(CONFIG_GRKERNSEC_KMEM) +static inline int range_is_allowed(unsigned long pfn, unsigned long size) +{ + return 0; +} #else static inline int range_is_allowed(unsigned long pfn, unsigned long size) { @@ -118,6 +132,7 @@ static ssize_t read_mem(struct file *file, char __user *buf, while (count > 0) { unsigned long remaining; + char *temp; sz = size_inside_page(p, count); @@ -133,7 +148,23 @@ static ssize_t read_mem(struct file *file, char __user *buf, if (!ptr) return -EFAULT; - remaining = copy_to_user(buf, ptr, sz); +#ifdef CONFIG_PAX_USERCOPY + temp = kmalloc(sz, GFP_KERNEL); + if (!temp) { + unxlate_dev_mem_ptr(p, ptr); + return -ENOMEM; + } + memcpy(temp, ptr, sz); +#else + temp = ptr; +#endif + + remaining = copy_to_user(buf, temp, sz); + +#ifdef CONFIG_PAX_USERCOPY + kfree(temp); +#endif + unxlate_dev_mem_ptr(p, ptr); if (remaining) return -EFAULT; @@ -396,9 +427,8 @@ static ssize_t read_kmem(struct file *file, char __user *buf, size_t count, loff_t *ppos) { unsigned long p = *ppos; - ssize_t low_count, read, sz; + ssize_t low_count, read, sz, err = 0; char * kbuf; /* k-addr because vread() takes vmlist_lock rwlock */ - int err = 0; read = 0; if (p < (unsigned long) high_memory) { @@ -420,6 +450,8 @@ static ssize_t read_kmem(struct file *file, char __user *buf, } #endif while (low_count > 0) { + char *temp; + sz = size_inside_page(p, low_count); /* @@ -429,7 +461,22 @@ static ssize_t read_kmem(struct file *file, char __user *buf, */ kbuf = xlate_dev_kmem_ptr((char *)p); - if (copy_to_user(buf, kbuf, sz)) +#ifdef CONFIG_PAX_USERCOPY + temp = kmalloc(sz, GFP_KERNEL); + if (!temp) + return -ENOMEM; + memcpy(temp, kbuf, sz); +#else + temp = kbuf; +#endif + + err = copy_to_user(buf, temp, sz); + +#ifdef CONFIG_PAX_USERCOPY + kfree(temp); +#endif + + if (err) return -EFAULT; buf += sz; p += sz; @@ -867,6 +914,9 @@ static const struct memdev { #ifdef CONFIG_CRASH_DUMP [12] = { "oldmem", 0, &oldmem_fops, NULL }, #endif +#if defined(CONFIG_GRKERNSEC) && !defined(CONFIG_GRKERNSEC_NO_RBAC) + [13] = { "grsec",S_IRUSR | S_IWUGO, &grsec_fops, NULL }, +#endif }; static int memory_open(struct inode *inode, struct file *filp) diff --git a/drivers/char/nvram.c b/drivers/char/nvram.c index da3cfee..a5a6606 100644 --- a/drivers/char/nvram.c +++ b/drivers/char/nvram.c @@ -248,7 +248,7 @@ static ssize_t nvram_read(struct file *file, char __user *buf, spin_unlock_irq(&rtc_lock); - if (copy_to_user(buf, contents, tmp - contents)) + if (tmp - contents > sizeof(contents) || copy_to_user(buf, contents, tmp - contents)) return -EFAULT; *ppos = i; diff --git a/drivers/char/random.c b/drivers/char/random.c index 54ca8b2..4a092ed 100644 --- a/drivers/char/random.c +++ b/drivers/char/random.c @@ -261,8 +261,13 @@ /* * Configuration information */ +#ifdef CONFIG_GRKERNSEC_RANDNET +#define INPUT_POOL_WORDS 512 +#define OUTPUT_POOL_WORDS 128 +#else #define INPUT_POOL_WORDS 128 #define OUTPUT_POOL_WORDS 32 +#endif #define SEC_XFER_SIZE 512 #define EXTRACT_SIZE 10 @@ -300,10 +305,17 @@ static struct poolinfo { int poolwords; int tap1, tap2, tap3, tap4, tap5; } poolinfo_table[] = { +#ifdef CONFIG_GRKERNSEC_RANDNET + /* x^512 + x^411 + x^308 + x^208 +x^104 + x + 1 -- 225 */ + { 512, 411, 308, 208, 104, 1 }, + /* x^128 + x^103 + x^76 + x^51 + x^25 + x + 1 -- 105 */ + { 128, 103, 76, 51, 25, 1 }, +#else /* x^128 + x^103 + x^76 + x^51 +x^25 + x + 1 -- 105 */ { 128, 103, 76, 51, 25, 1 }, /* x^32 + x^26 + x^20 + x^14 + x^7 + x + 1 -- 15 */ { 32, 26, 20, 14, 7, 1 }, +#endif #if 0 /* x^2048 + x^1638 + x^1231 + x^819 + x^411 + x + 1 -- 115 */ { 2048, 1638, 1231, 819, 411, 1 }, @@ -913,7 +925,7 @@ static ssize_t extract_entropy_user(struct entropy_store *r, void __user *buf, extract_buf(r, tmp); i = min_t(int, nbytes, EXTRACT_SIZE); - if (copy_to_user(buf, tmp, i)) { + if (i > sizeof(tmp) || copy_to_user(buf, tmp, i)) { ret = -EFAULT; break; } @@ -1238,7 +1250,7 @@ EXPORT_SYMBOL(generate_random_uuid); #include static int min_read_thresh = 8, min_write_thresh; -static int max_read_thresh = INPUT_POOL_WORDS * 32; +static int max_read_thresh = OUTPUT_POOL_WORDS * 32; static int max_write_thresh = INPUT_POOL_WORDS * 32; static char sysctl_bootid[16]; @@ -1260,10 +1272,15 @@ static int proc_do_uuid(ctl_table *table, int write, uuid = table->data; if (!uuid) { uuid = tmp_uuid; - uuid[8] = 0; - } - if (uuid[8] == 0) generate_random_uuid(uuid); + } else { + static DEFINE_SPINLOCK(bootid_spinlock); + + spin_lock(&bootid_spinlock); + if (!uuid[8]) + generate_random_uuid(uuid); + spin_unlock(&bootid_spinlock); + } sprintf(buf, "%pU", uuid); diff --git a/drivers/char/sonypi.c b/drivers/char/sonypi.c index 1ee8ce7..b778bef 100644 --- a/drivers/char/sonypi.c +++ b/drivers/char/sonypi.c @@ -55,6 +55,7 @@ #include #include #include +#include #include @@ -491,7 +492,7 @@ static struct sonypi_device { spinlock_t fifo_lock; wait_queue_head_t fifo_proc_list; struct fasync_struct *fifo_async; - int open_count; + local_t open_count; int model; struct input_dev *input_jog_dev; struct input_dev *input_key_dev; @@ -898,7 +899,7 @@ static int sonypi_misc_fasync(int fd, struct file *filp, int on) static int sonypi_misc_release(struct inode *inode, struct file *file) { mutex_lock(&sonypi_device.lock); - sonypi_device.open_count--; + local_dec(&sonypi_device.open_count); mutex_unlock(&sonypi_device.lock); return 0; } @@ -907,9 +908,9 @@ static int sonypi_misc_open(struct inode *inode, struct file *file) { mutex_lock(&sonypi_device.lock); /* Flush input queue on first open */ - if (!sonypi_device.open_count) + if (!local_read(&sonypi_device.open_count)) kfifo_reset(&sonypi_device.fifo); - sonypi_device.open_count++; + local_inc(&sonypi_device.open_count); mutex_unlock(&sonypi_device.lock); return 0; diff --git a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c index ad7c732..5aa8054 100644 --- a/drivers/char/tpm/tpm.c +++ b/drivers/char/tpm/tpm.c @@ -415,7 +415,7 @@ static ssize_t tpm_transmit(struct tpm_chip *chip, const char *buf, chip->vendor.req_complete_val) goto out_recv; - if ((status == chip->vendor.req_canceled)) { + if (status == chip->vendor.req_canceled) { dev_err(chip->dev, "Operation Canceled\n"); rc = -ECANCELED; goto out; diff --git a/drivers/char/tpm/tpm_bios.c b/drivers/char/tpm/tpm_bios.c index 0636520..169c1d0 100644 --- a/drivers/char/tpm/tpm_bios.c +++ b/drivers/char/tpm/tpm_bios.c @@ -173,7 +173,7 @@ static void *tpm_bios_measurements_start(struct seq_file *m, loff_t *pos) event = addr; if ((event->event_type == 0 && event->event_size == 0) || - ((addr + sizeof(struct tcpa_event) + event->event_size) >= limit)) + (event->event_size >= limit - addr - sizeof(struct tcpa_event))) return NULL; return addr; @@ -198,7 +198,7 @@ static void *tpm_bios_measurements_next(struct seq_file *m, void *v, return NULL; if ((event->event_type == 0 && event->event_size == 0) || - ((v + sizeof(struct tcpa_event) + event->event_size) >= limit)) + (event->event_size >= limit - v - sizeof(struct tcpa_event))) return NULL; (*pos)++; @@ -291,7 +291,8 @@ static int tpm_binary_bios_measurements_show(struct seq_file *m, void *v) int i; for (i = 0; i < sizeof(struct tcpa_event) + event->event_size; i++) - seq_putc(m, data[i]); + if (!seq_putc(m, data[i])) + return -EFAULT; return 0; } @@ -410,8 +411,13 @@ static int read_log(struct tpm_bios_log *log) log->bios_event_log_end = log->bios_event_log + len; virt = acpi_os_map_memory(start, len); + if (!virt) { + kfree(log->bios_event_log); + log->bios_event_log = NULL; + return -EFAULT; + } - memcpy(log->bios_event_log, virt, len); + memcpy(log->bios_event_log, (const char __force_kernel *)virt, len); acpi_os_unmap_memory(virt, len); return 0; diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c index b58b561..c9088c8 100644 --- a/drivers/char/virtio_console.c +++ b/drivers/char/virtio_console.c @@ -563,7 +563,7 @@ static ssize_t fill_readbuf(struct port *port, char *out_buf, size_t out_count, if (to_user) { ssize_t ret; - ret = copy_to_user(out_buf, buf->buf + buf->offset, out_count); + ret = copy_to_user((char __force_user *)out_buf, buf->buf + buf->offset, out_count); if (ret) return -EFAULT; } else { @@ -662,7 +662,7 @@ static ssize_t port_fops_read(struct file *filp, char __user *ubuf, if (!port_has_data(port) && !port->host_connected) return 0; - return fill_readbuf(port, ubuf, count, true); + return fill_readbuf(port, (char __force_kernel *)ubuf, count, true); } static ssize_t port_fops_write(struct file *filp, const char __user *ubuf, diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c index c9eee6d..f9d5280 100644 --- a/drivers/edac/amd64_edac.c +++ b/drivers/edac/amd64_edac.c @@ -2685,7 +2685,7 @@ static void __devexit amd64_remove_one_instance(struct pci_dev *pdev) * PCI core identifies what devices are on a system during boot, and then * inquiry this table to see if this driver is for a given device found. */ -static const struct pci_device_id amd64_pci_table[] __devinitdata = { +static const struct pci_device_id amd64_pci_table[] __devinitconst = { { .vendor = PCI_VENDOR_ID_AMD, .device = PCI_DEVICE_ID_AMD_K8_NB_MEMCTL, diff --git a/drivers/edac/amd76x_edac.c b/drivers/edac/amd76x_edac.c index e47e73b..348e0bd 100644 --- a/drivers/edac/amd76x_edac.c +++ b/drivers/edac/amd76x_edac.c @@ -321,7 +321,7 @@ static void __devexit amd76x_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id amd76x_pci_tbl[] __devinitdata = { +static const struct pci_device_id amd76x_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(AMD, FE_GATE_700C), PCI_ANY_ID, PCI_ANY_ID, 0, 0, AMD762}, diff --git a/drivers/edac/e752x_edac.c b/drivers/edac/e752x_edac.c index 1af531a..3a8ff27 100644 --- a/drivers/edac/e752x_edac.c +++ b/drivers/edac/e752x_edac.c @@ -1380,7 +1380,7 @@ static void __devexit e752x_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id e752x_pci_tbl[] __devinitdata = { +static const struct pci_device_id e752x_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, 7520_0), PCI_ANY_ID, PCI_ANY_ID, 0, 0, E7520}, diff --git a/drivers/edac/e7xxx_edac.c b/drivers/edac/e7xxx_edac.c index 6ffb6d2..383d8d7 100644 --- a/drivers/edac/e7xxx_edac.c +++ b/drivers/edac/e7xxx_edac.c @@ -525,7 +525,7 @@ static void __devexit e7xxx_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id e7xxx_pci_tbl[] __devinitdata = { +static const struct pci_device_id e7xxx_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, 7205_0), PCI_ANY_ID, PCI_ANY_ID, 0, 0, E7205}, diff --git a/drivers/edac/edac_pci_sysfs.c b/drivers/edac/edac_pci_sysfs.c index 97f5064..202b6e6 100644 --- a/drivers/edac/edac_pci_sysfs.c +++ b/drivers/edac/edac_pci_sysfs.c @@ -26,8 +26,8 @@ static int edac_pci_log_pe = 1; /* log PCI parity errors */ static int edac_pci_log_npe = 1; /* log PCI non-parity error errors */ static int edac_pci_poll_msec = 1000; /* one second workq period */ -static atomic_t pci_parity_count = ATOMIC_INIT(0); -static atomic_t pci_nonparity_count = ATOMIC_INIT(0); +static atomic_unchecked_t pci_parity_count = ATOMIC_INIT(0); +static atomic_unchecked_t pci_nonparity_count = ATOMIC_INIT(0); static struct kobject *edac_pci_top_main_kobj; static atomic_t edac_pci_sysfs_refcount = ATOMIC_INIT(0); @@ -582,7 +582,7 @@ static void edac_pci_dev_parity_test(struct pci_dev *dev) edac_printk(KERN_CRIT, EDAC_PCI, "Signaled System Error on %s\n", pci_name(dev)); - atomic_inc(&pci_nonparity_count); + atomic_inc_unchecked(&pci_nonparity_count); } if (status & (PCI_STATUS_PARITY)) { @@ -590,7 +590,7 @@ static void edac_pci_dev_parity_test(struct pci_dev *dev) "Master Data Parity Error on %s\n", pci_name(dev)); - atomic_inc(&pci_parity_count); + atomic_inc_unchecked(&pci_parity_count); } if (status & (PCI_STATUS_DETECTED_PARITY)) { @@ -598,7 +598,7 @@ static void edac_pci_dev_parity_test(struct pci_dev *dev) "Detected Parity Error on %s\n", pci_name(dev)); - atomic_inc(&pci_parity_count); + atomic_inc_unchecked(&pci_parity_count); } } @@ -619,7 +619,7 @@ static void edac_pci_dev_parity_test(struct pci_dev *dev) edac_printk(KERN_CRIT, EDAC_PCI, "Bridge " "Signaled System Error on %s\n", pci_name(dev)); - atomic_inc(&pci_nonparity_count); + atomic_inc_unchecked(&pci_nonparity_count); } if (status & (PCI_STATUS_PARITY)) { @@ -627,7 +627,7 @@ static void edac_pci_dev_parity_test(struct pci_dev *dev) "Master Data Parity Error on " "%s\n", pci_name(dev)); - atomic_inc(&pci_parity_count); + atomic_inc_unchecked(&pci_parity_count); } if (status & (PCI_STATUS_DETECTED_PARITY)) { @@ -635,7 +635,7 @@ static void edac_pci_dev_parity_test(struct pci_dev *dev) "Detected Parity Error on %s\n", pci_name(dev)); - atomic_inc(&pci_parity_count); + atomic_inc_unchecked(&pci_parity_count); } } } @@ -677,7 +677,7 @@ void edac_pci_do_parity_check(void) if (!check_pci_errors) return; - before_count = atomic_read(&pci_parity_count); + before_count = atomic_read_unchecked(&pci_parity_count); /* scan all PCI devices looking for a Parity Error on devices and * bridges. @@ -689,7 +689,7 @@ void edac_pci_do_parity_check(void) /* Only if operator has selected panic on PCI Error */ if (edac_pci_get_panic_on_pe()) { /* If the count is different 'after' from 'before' */ - if (before_count != atomic_read(&pci_parity_count)) + if (before_count != atomic_read_unchecked(&pci_parity_count)) panic("EDAC: PCI Parity Error"); } } diff --git a/drivers/edac/i3000_edac.c b/drivers/edac/i3000_edac.c index c0510b3..6e2a954 100644 --- a/drivers/edac/i3000_edac.c +++ b/drivers/edac/i3000_edac.c @@ -470,7 +470,7 @@ static void __devexit i3000_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id i3000_pci_tbl[] __devinitdata = { +static const struct pci_device_id i3000_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, 3000_HB), PCI_ANY_ID, PCI_ANY_ID, 0, 0, I3000}, diff --git a/drivers/edac/i3200_edac.c b/drivers/edac/i3200_edac.c index 73f55e200..5faaf59 100644 --- a/drivers/edac/i3200_edac.c +++ b/drivers/edac/i3200_edac.c @@ -445,7 +445,7 @@ static void __devexit i3200_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id i3200_pci_tbl[] __devinitdata = { +static const struct pci_device_id i3200_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, 3200_HB), PCI_ANY_ID, PCI_ANY_ID, 0, 0, I3200}, diff --git a/drivers/edac/i5000_edac.c b/drivers/edac/i5000_edac.c index 4dc3ac2..67d05a6 100644 --- a/drivers/edac/i5000_edac.c +++ b/drivers/edac/i5000_edac.c @@ -1516,7 +1516,7 @@ static void __devexit i5000_remove_one(struct pci_dev *pdev) * * The "E500P" device is the first device supported. */ -static const struct pci_device_id i5000_pci_tbl[] __devinitdata = { +static const struct pci_device_id i5000_pci_tbl[] __devinitconst = { {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_I5000_DEV16), .driver_data = I5000P}, diff --git a/drivers/edac/i5100_edac.c b/drivers/edac/i5100_edac.c index bcbdeec..9886d16 100644 --- a/drivers/edac/i5100_edac.c +++ b/drivers/edac/i5100_edac.c @@ -1051,7 +1051,7 @@ static void __devexit i5100_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id i5100_pci_tbl[] __devinitdata = { +static const struct pci_device_id i5100_pci_tbl[] __devinitconst = { /* Device 16, Function 0, Channel 0 Memory Map, Error Flag/Mask, ... */ { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_5100_16) }, { 0, } diff --git a/drivers/edac/i5400_edac.c b/drivers/edac/i5400_edac.c index 74d6ec34..baff517 100644 --- a/drivers/edac/i5400_edac.c +++ b/drivers/edac/i5400_edac.c @@ -1383,7 +1383,7 @@ static void __devexit i5400_remove_one(struct pci_dev *pdev) * * The "E500P" device is the first device supported. */ -static const struct pci_device_id i5400_pci_tbl[] __devinitdata = { +static const struct pci_device_id i5400_pci_tbl[] __devinitconst = { {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_5400_ERR)}, {0,} /* 0 terminated list. */ }; diff --git a/drivers/edac/i7300_edac.c b/drivers/edac/i7300_edac.c index 6104dba..e7ea8e1 100644 --- a/drivers/edac/i7300_edac.c +++ b/drivers/edac/i7300_edac.c @@ -1192,7 +1192,7 @@ static void __devexit i7300_remove_one(struct pci_dev *pdev) * * Has only 8086:360c PCI ID */ -static const struct pci_device_id i7300_pci_tbl[] __devinitdata = { +static const struct pci_device_id i7300_pci_tbl[] __devinitconst = { {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_I7300_MCH_ERR)}, {0,} /* 0 terminated list. */ }; diff --git a/drivers/edac/i7core_edac.c b/drivers/edac/i7core_edac.c index 8568d9b..42b2fa8 100644 --- a/drivers/edac/i7core_edac.c +++ b/drivers/edac/i7core_edac.c @@ -391,7 +391,7 @@ static const struct pci_id_table pci_dev_table[] = { /* * pci_device_id table for which devices we are looking for */ -static const struct pci_device_id i7core_pci_tbl[] __devinitdata = { +static const struct pci_device_id i7core_pci_tbl[] __devinitconst = { {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_X58_HUB_MGMT)}, {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_LYNNFIELD_QPI_LINK0)}, {0,} /* 0 terminated list. */ diff --git a/drivers/edac/i82443bxgx_edac.c b/drivers/edac/i82443bxgx_edac.c index 4329d39..f3022ef 100644 --- a/drivers/edac/i82443bxgx_edac.c +++ b/drivers/edac/i82443bxgx_edac.c @@ -380,7 +380,7 @@ static void __devexit i82443bxgx_edacmc_remove_one(struct pci_dev *pdev) EXPORT_SYMBOL_GPL(i82443bxgx_edacmc_remove_one); -static const struct pci_device_id i82443bxgx_pci_tbl[] __devinitdata = { +static const struct pci_device_id i82443bxgx_pci_tbl[] __devinitconst = { {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82443BX_0)}, {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82443BX_2)}, {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82443GX_0)}, diff --git a/drivers/edac/i82860_edac.c b/drivers/edac/i82860_edac.c index 931a057..fd28340 100644 --- a/drivers/edac/i82860_edac.c +++ b/drivers/edac/i82860_edac.c @@ -270,7 +270,7 @@ static void __devexit i82860_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id i82860_pci_tbl[] __devinitdata = { +static const struct pci_device_id i82860_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, 82860_0), PCI_ANY_ID, PCI_ANY_ID, 0, 0, I82860}, diff --git a/drivers/edac/i82875p_edac.c b/drivers/edac/i82875p_edac.c index 33864c6..01edc61 100644 --- a/drivers/edac/i82875p_edac.c +++ b/drivers/edac/i82875p_edac.c @@ -511,7 +511,7 @@ static void __devexit i82875p_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id i82875p_pci_tbl[] __devinitdata = { +static const struct pci_device_id i82875p_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, 82875_0), PCI_ANY_ID, PCI_ANY_ID, 0, 0, I82875P}, diff --git a/drivers/edac/i82975x_edac.c b/drivers/edac/i82975x_edac.c index 4184e01..dcb2cd3 100644 --- a/drivers/edac/i82975x_edac.c +++ b/drivers/edac/i82975x_edac.c @@ -612,7 +612,7 @@ static void __devexit i82975x_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id i82975x_pci_tbl[] __devinitdata = { +static const struct pci_device_id i82975x_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, 82975_0), PCI_ANY_ID, PCI_ANY_ID, 0, 0, I82975X diff --git a/drivers/edac/mce_amd.h b/drivers/edac/mce_amd.h index 0106747..0b40417 100644 --- a/drivers/edac/mce_amd.h +++ b/drivers/edac/mce_amd.h @@ -83,7 +83,7 @@ struct amd_decoder_ops { bool (*dc_mce)(u16, u8); bool (*ic_mce)(u16, u8); bool (*nb_mce)(u16, u8); -}; +} __no_const; void amd_report_gart_errors(bool); void amd_register_ecc_decoder(void (*f)(int, struct mce *)); diff --git a/drivers/edac/r82600_edac.c b/drivers/edac/r82600_edac.c index e294e1b..a41b05b 100644 --- a/drivers/edac/r82600_edac.c +++ b/drivers/edac/r82600_edac.c @@ -373,7 +373,7 @@ static void __devexit r82600_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id r82600_pci_tbl[] __devinitdata = { +static const struct pci_device_id r82600_pci_tbl[] __devinitconst = { { PCI_DEVICE(PCI_VENDOR_ID_RADISYS, R82600_BRIDGE_ID) }, diff --git a/drivers/edac/sb_edac.c b/drivers/edac/sb_edac.c index 1dc118d..8c68af9 100644 --- a/drivers/edac/sb_edac.c +++ b/drivers/edac/sb_edac.c @@ -367,7 +367,7 @@ static const struct pci_id_table pci_dev_descr_sbridge_table[] = { /* * pci_device_id table for which devices we are looking for */ -static const struct pci_device_id sbridge_pci_tbl[] __devinitdata = { +static const struct pci_device_id sbridge_pci_tbl[] __devinitconst = { {PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SBRIDGE_IMC_TA)}, {0,} /* 0 terminated list. */ }; diff --git a/drivers/edac/x38_edac.c b/drivers/edac/x38_edac.c index b6f47de..c5acf3a 100644 --- a/drivers/edac/x38_edac.c +++ b/drivers/edac/x38_edac.c @@ -440,7 +440,7 @@ static void __devexit x38_remove_one(struct pci_dev *pdev) edac_mc_free(mci); } -static const struct pci_device_id x38_pci_tbl[] __devinitdata = { +static const struct pci_device_id x38_pci_tbl[] __devinitconst = { { PCI_VEND_DEV(INTEL, X38_HB), PCI_ANY_ID, PCI_ANY_ID, 0, 0, X38}, diff --git a/drivers/firewire/core-card.c b/drivers/firewire/core-card.c index 85661b0..c784559a 100644 --- a/drivers/firewire/core-card.c +++ b/drivers/firewire/core-card.c @@ -657,7 +657,7 @@ void fw_card_release(struct kref *kref) void fw_core_remove_card(struct fw_card *card) { - struct fw_card_driver dummy_driver = dummy_driver_template; + fw_card_driver_no_const dummy_driver = dummy_driver_template; card->driver->update_phy_reg(card, 4, PHY_LINK_ACTIVE | PHY_CONTENDER, 0); diff --git a/drivers/firewire/core-cdev.c b/drivers/firewire/core-cdev.c index 4799393..37bd3ab 100644 --- a/drivers/firewire/core-cdev.c +++ b/drivers/firewire/core-cdev.c @@ -1331,8 +1331,7 @@ static int init_iso_resource(struct client *client, int ret; if ((request->channels == 0 && request->bandwidth == 0) || - request->bandwidth > BANDWIDTH_AVAILABLE_INITIAL || - request->bandwidth < 0) + request->bandwidth > BANDWIDTH_AVAILABLE_INITIAL) return -EINVAL; r = kmalloc(sizeof(*r), GFP_KERNEL); diff --git a/drivers/firewire/core-transaction.c b/drivers/firewire/core-transaction.c index 855ab3f..11f4bbd 100644 --- a/drivers/firewire/core-transaction.c +++ b/drivers/firewire/core-transaction.c @@ -37,6 +37,7 @@ #include #include #include +#include #include diff --git a/drivers/firewire/core.h b/drivers/firewire/core.h index b45be57..5fad18b 100644 --- a/drivers/firewire/core.h +++ b/drivers/firewire/core.h @@ -101,6 +101,7 @@ struct fw_card_driver { int (*stop_iso)(struct fw_iso_context *ctx); }; +typedef struct fw_card_driver __no_const fw_card_driver_no_const; void fw_card_initialize(struct fw_card *card, const struct fw_card_driver *driver, struct device *device); diff --git a/drivers/firmware/dmi_scan.c b/drivers/firmware/dmi_scan.c index 153980b..4b4d046 100644 --- a/drivers/firmware/dmi_scan.c +++ b/drivers/firmware/dmi_scan.c @@ -449,11 +449,6 @@ void __init dmi_scan_machine(void) } } else { - /* - * no iounmap() for that ioremap(); it would be a no-op, but - * it's so early in setup that sucker gets confused into doing - * what it shouldn't if we actually call it. - */ p = dmi_ioremap(0xF0000, 0x10000); if (p == NULL) goto error; @@ -723,7 +718,7 @@ int dmi_walk(void (*decode)(const struct dmi_header *, void *), if (buf == NULL) return -1; - dmi_table(buf, dmi_len, dmi_num, decode, private_data); + dmi_table((char __force_kernel *)buf, dmi_len, dmi_num, decode, private_data); iounmap(buf); return 0; diff --git a/drivers/gpio/gpio-vr41xx.c b/drivers/gpio/gpio-vr41xx.c index 82d5c20..44a7177 100644 --- a/drivers/gpio/gpio-vr41xx.c +++ b/drivers/gpio/gpio-vr41xx.c @@ -204,7 +204,7 @@ static int giu_get_irq(unsigned int irq) printk(KERN_ERR "spurious GIU interrupt: %04x(%04x),%04x(%04x)\n", maskl, pendl, maskh, pendh); - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); return -EINVAL; } diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index 84a4a80..ce0306e 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -280,7 +280,7 @@ static bool drm_encoder_crtc_ok(struct drm_encoder *encoder, struct drm_crtc *tmp; int crtc_mask = 1; - WARN(!crtc, "checking null crtc?\n"); + BUG_ON(!crtc); dev = crtc->dev; diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c index ebf7d3f..d64c436 100644 --- a/drivers/gpu/drm/drm_drv.c +++ b/drivers/gpu/drm/drm_drv.c @@ -312,7 +312,7 @@ module_exit(drm_core_exit); /** * Copy and IOCTL return string to user space */ -static int drm_copy_field(char *buf, size_t *buf_len, const char *value) +static int drm_copy_field(char __user *buf, size_t *buf_len, const char *value) { int len; @@ -391,7 +391,7 @@ long drm_ioctl(struct file *filp, dev = file_priv->minor->dev; atomic_inc(&dev->ioctl_count); - atomic_inc(&dev->counts[_DRM_STAT_IOCTLS]); + atomic_inc_unchecked(&dev->counts[_DRM_STAT_IOCTLS]); ++file_priv->ioctl_count; DRM_DEBUG("pid=%d, cmd=0x%02x, nr=0x%02x, dev 0x%lx, auth=%d\n", diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c index 6263b01..7987f55 100644 --- a/drivers/gpu/drm/drm_fops.c +++ b/drivers/gpu/drm/drm_fops.c @@ -71,7 +71,7 @@ static int drm_setup(struct drm_device * dev) } for (i = 0; i < ARRAY_SIZE(dev->counts); i++) - atomic_set(&dev->counts[i], 0); + atomic_set_unchecked(&dev->counts[i], 0); dev->sigdata.lock = NULL; @@ -135,8 +135,8 @@ int drm_open(struct inode *inode, struct file *filp) retcode = drm_open_helper(inode, filp, dev); if (!retcode) { - atomic_inc(&dev->counts[_DRM_STAT_OPENS]); - if (!dev->open_count++) + atomic_inc_unchecked(&dev->counts[_DRM_STAT_OPENS]); + if (local_inc_return(&dev->open_count) == 1) retcode = drm_setup(dev); } if (!retcode) { @@ -473,7 +473,7 @@ int drm_release(struct inode *inode, struct file *filp) mutex_lock(&drm_global_mutex); - DRM_DEBUG("open_count = %d\n", dev->open_count); + DRM_DEBUG("open_count = %ld\n", local_read(&dev->open_count)); if (dev->driver->preclose) dev->driver->preclose(dev, file_priv); @@ -482,10 +482,10 @@ int drm_release(struct inode *inode, struct file *filp) * Begin inline drm_release */ - DRM_DEBUG("pid = %d, device = 0x%lx, open_count = %d\n", + DRM_DEBUG("pid = %d, device = 0x%lx, open_count = %ld\n", task_pid_nr(current), (long)old_encode_dev(file_priv->minor->device), - dev->open_count); + local_read(&dev->open_count)); /* Release any auth tokens that might point to this file_priv, (do that under the drm_global_mutex) */ @@ -571,8 +571,8 @@ int drm_release(struct inode *inode, struct file *filp) * End inline drm_release */ - atomic_inc(&dev->counts[_DRM_STAT_CLOSES]); - if (!--dev->open_count) { + atomic_inc_unchecked(&dev->counts[_DRM_STAT_CLOSES]); + if (local_dec_and_test(&dev->open_count)) { if (atomic_read(&dev->ioctl_count)) { DRM_ERROR("Device busy: %d\n", atomic_read(&dev->ioctl_count)); diff --git a/drivers/gpu/drm/drm_global.c b/drivers/gpu/drm/drm_global.c index c87dc96..326055d 100644 --- a/drivers/gpu/drm/drm_global.c +++ b/drivers/gpu/drm/drm_global.c @@ -36,7 +36,7 @@ struct drm_global_item { struct mutex mutex; void *object; - int refcount; + atomic_t refcount; }; static struct drm_global_item glob[DRM_GLOBAL_NUM]; @@ -49,7 +49,7 @@ void drm_global_init(void) struct drm_global_item *item = &glob[i]; mutex_init(&item->mutex); item->object = NULL; - item->refcount = 0; + atomic_set(&item->refcount, 0); } } @@ -59,7 +59,7 @@ void drm_global_release(void) for (i = 0; i < DRM_GLOBAL_NUM; ++i) { struct drm_global_item *item = &glob[i]; BUG_ON(item->object != NULL); - BUG_ON(item->refcount != 0); + BUG_ON(atomic_read(&item->refcount) != 0); } } @@ -70,7 +70,7 @@ int drm_global_item_ref(struct drm_global_reference *ref) void *object; mutex_lock(&item->mutex); - if (item->refcount == 0) { + if (atomic_read(&item->refcount) == 0) { item->object = kzalloc(ref->size, GFP_KERNEL); if (unlikely(item->object == NULL)) { ret = -ENOMEM; @@ -83,7 +83,7 @@ int drm_global_item_ref(struct drm_global_reference *ref) goto out_err; } - ++item->refcount; + atomic_inc(&item->refcount); ref->object = item->object; object = item->object; mutex_unlock(&item->mutex); @@ -100,9 +100,9 @@ void drm_global_item_unref(struct drm_global_reference *ref) struct drm_global_item *item = &glob[ref->global_type]; mutex_lock(&item->mutex); - BUG_ON(item->refcount == 0); + BUG_ON(atomic_read(&item->refcount) == 0); BUG_ON(ref->object != item->object); - if (--item->refcount == 0) { + if (atomic_dec_and_test(&item->refcount)) { ref->release(ref); item->object = NULL; } diff --git a/drivers/gpu/drm/drm_info.c b/drivers/gpu/drm/drm_info.c index ab1162d..42587b2 100644 --- a/drivers/gpu/drm/drm_info.c +++ b/drivers/gpu/drm/drm_info.c @@ -75,10 +75,14 @@ int drm_vm_info(struct seq_file *m, void *data) struct drm_local_map *map; struct drm_map_list *r_list; - /* Hardcoded from _DRM_FRAME_BUFFER, - _DRM_REGISTERS, _DRM_SHM, _DRM_AGP, and - _DRM_SCATTER_GATHER and _DRM_CONSISTENT */ - const char *types[] = { "FB", "REG", "SHM", "AGP", "SG", "PCI" }; + static const char * const types[] = { + [_DRM_FRAME_BUFFER] = "FB", + [_DRM_REGISTERS] = "REG", + [_DRM_SHM] = "SHM", + [_DRM_AGP] = "AGP", + [_DRM_SCATTER_GATHER] = "SG", + [_DRM_CONSISTENT] = "PCI", + [_DRM_GEM] = "GEM" }; const char *type; int i; @@ -89,7 +93,7 @@ int drm_vm_info(struct seq_file *m, void *data) map = r_list->map; if (!map) continue; - if (map->type < 0 || map->type > 5) + if (map->type >= ARRAY_SIZE(types)) type = "??"; else type = types[map->type]; @@ -290,7 +294,11 @@ int drm_vma_info(struct seq_file *m, void *data) vma->vm_flags & VM_MAYSHARE ? 's' : 'p', vma->vm_flags & VM_LOCKED ? 'l' : '-', vma->vm_flags & VM_IO ? 'i' : '-', +#ifdef CONFIG_GRKERNSEC_HIDESYM + 0); +#else vma->vm_pgoff); +#endif #if defined(__i386__) pgprot = pgprot_val(vma->vm_page_prot); diff --git a/drivers/gpu/drm/drm_ioc32.c b/drivers/gpu/drm/drm_ioc32.c index 637fcc3..e890b33 100644 --- a/drivers/gpu/drm/drm_ioc32.c +++ b/drivers/gpu/drm/drm_ioc32.c @@ -457,7 +457,7 @@ static int compat_drm_infobufs(struct file *file, unsigned int cmd, request = compat_alloc_user_space(nbytes); if (!access_ok(VERIFY_WRITE, request, nbytes)) return -EFAULT; - list = (struct drm_buf_desc *) (request + 1); + list = (struct drm_buf_desc __user *) (request + 1); if (__put_user(count, &request->count) || __put_user(list, &request->list)) @@ -518,7 +518,7 @@ static int compat_drm_mapbufs(struct file *file, unsigned int cmd, request = compat_alloc_user_space(nbytes); if (!access_ok(VERIFY_WRITE, request, nbytes)) return -EFAULT; - list = (struct drm_buf_pub *) (request + 1); + list = (struct drm_buf_pub __user *) (request + 1); if (__put_user(count, &request->count) || __put_user(list, &request->list)) diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c index 956fd38..e52167a 100644 --- a/drivers/gpu/drm/drm_ioctl.c +++ b/drivers/gpu/drm/drm_ioctl.c @@ -251,7 +251,7 @@ int drm_getstats(struct drm_device *dev, void *data, stats->data[i].value = (file_priv->master->lock.hw_lock ? file_priv->master->lock.hw_lock->lock : 0); else - stats->data[i].value = atomic_read(&dev->counts[i]); + stats->data[i].value = atomic_read_unchecked(&dev->counts[i]); stats->data[i].type = dev->types[i]; } diff --git a/drivers/gpu/drm/drm_lock.c b/drivers/gpu/drm/drm_lock.c index c79c713..2048588 100644 --- a/drivers/gpu/drm/drm_lock.c +++ b/drivers/gpu/drm/drm_lock.c @@ -90,7 +90,7 @@ int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv) if (drm_lock_take(&master->lock, lock->context)) { master->lock.file_priv = file_priv; master->lock.lock_time = jiffies; - atomic_inc(&dev->counts[_DRM_STAT_LOCKS]); + atomic_inc_unchecked(&dev->counts[_DRM_STAT_LOCKS]); break; /* Got lock */ } @@ -161,7 +161,7 @@ int drm_unlock(struct drm_device *dev, void *data, struct drm_file *file_priv) return -EINVAL; } - atomic_inc(&dev->counts[_DRM_STAT_UNLOCKS]); + atomic_inc_unchecked(&dev->counts[_DRM_STAT_UNLOCKS]); if (drm_lock_free(&master->lock, lock->context)) { /* FIXME: Should really bail out here. */ diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c index 7f4b4e1..bf4def2 100644 --- a/drivers/gpu/drm/i810/i810_dma.c +++ b/drivers/gpu/drm/i810/i810_dma.c @@ -948,8 +948,8 @@ static int i810_dma_vertex(struct drm_device *dev, void *data, dma->buflist[vertex->idx], vertex->discard, vertex->used); - atomic_add(vertex->used, &dev->counts[_DRM_STAT_SECONDARY]); - atomic_inc(&dev->counts[_DRM_STAT_DMA]); + atomic_add_unchecked(vertex->used, &dev->counts[_DRM_STAT_SECONDARY]); + atomic_inc_unchecked(&dev->counts[_DRM_STAT_DMA]); sarea_priv->last_enqueue = dev_priv->counter - 1; sarea_priv->last_dispatch = (int)hw_status[5]; @@ -1109,8 +1109,8 @@ static int i810_dma_mc(struct drm_device *dev, void *data, i810_dma_dispatch_mc(dev, dma->buflist[mc->idx], mc->used, mc->last_render); - atomic_add(mc->used, &dev->counts[_DRM_STAT_SECONDARY]); - atomic_inc(&dev->counts[_DRM_STAT_DMA]); + atomic_add_unchecked(mc->used, &dev->counts[_DRM_STAT_SECONDARY]); + atomic_inc_unchecked(&dev->counts[_DRM_STAT_DMA]); sarea_priv->last_enqueue = dev_priv->counter - 1; sarea_priv->last_dispatch = (int)hw_status[5]; diff --git a/drivers/gpu/drm/i810/i810_drv.h b/drivers/gpu/drm/i810/i810_drv.h index c9339f4..f5e1b9d 100644 --- a/drivers/gpu/drm/i810/i810_drv.h +++ b/drivers/gpu/drm/i810/i810_drv.h @@ -108,8 +108,8 @@ typedef struct drm_i810_private { int page_flipping; wait_queue_head_t irq_queue; - atomic_t irq_received; - atomic_t irq_emitted; + atomic_unchecked_t irq_received; + atomic_unchecked_t irq_emitted; int front_offset; } drm_i810_private_t; diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index deaa657..e0fd296 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -499,7 +499,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data) I915_READ(GTIMR)); } seq_printf(m, "Interrupts received: %d\n", - atomic_read(&dev_priv->irq_received)); + atomic_read_unchecked(&dev_priv->irq_received)); for (i = 0; i < I915_NUM_RINGS; i++) { if (IS_GEN6(dev) || IS_GEN7(dev)) { seq_printf(m, "Graphics Interrupt mask (%s): %08x\n", @@ -1321,7 +1321,7 @@ static int i915_opregion(struct seq_file *m, void *unused) return ret; if (opregion->header) - seq_write(m, opregion->header, OPREGION_SIZE); + seq_write(m, (const void __force_kernel *)opregion->header, OPREGION_SIZE); mutex_unlock(&dev->struct_mutex); diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index ddfe3d9..f6e6b21 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -1175,7 +1175,7 @@ static bool i915_switcheroo_can_switch(struct pci_dev *pdev) bool can_switch; spin_lock(&dev->count_lock); - can_switch = (dev->open_count == 0); + can_switch = (local_read(&dev->open_count) == 0); spin_unlock(&dev->count_lock); return can_switch; } diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 9689ca3..294f9c1 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -231,7 +231,7 @@ struct drm_i915_display_funcs { /* render clock increase/decrease */ /* display clock increase/decrease */ /* pll clock increase/decrease */ -}; +} __no_const; struct intel_device_info { u8 gen; @@ -320,7 +320,7 @@ typedef struct drm_i915_private { int current_page; int page_flipping; - atomic_t irq_received; + atomic_unchecked_t irq_received; /* protects the irq masks */ spinlock_t irq_lock; @@ -896,7 +896,7 @@ struct drm_i915_gem_object { * will be page flipped away on the next vblank. When it * reaches 0, dev_priv->pending_flip_queue will be woken up. */ - atomic_t pending_flip; + atomic_unchecked_t pending_flip; }; #define to_intel_bo(x) container_of(x, struct drm_i915_gem_object, base) @@ -1276,7 +1276,7 @@ extern int intel_setup_gmbus(struct drm_device *dev); extern void intel_teardown_gmbus(struct drm_device *dev); extern void intel_gmbus_set_speed(struct i2c_adapter *adapter, int speed); extern void intel_gmbus_force_bit(struct i2c_adapter *adapter, bool force_bit); -extern inline bool intel_gmbus_is_forced_bit(struct i2c_adapter *adapter) +static inline bool intel_gmbus_is_forced_bit(struct i2c_adapter *adapter) { return container_of(adapter, struct intel_gmbus, adapter)->force_bit; } diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index e159e33..cdcc663 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -189,7 +189,7 @@ i915_gem_object_set_to_gpu_domain(struct drm_i915_gem_object *obj, i915_gem_clflush_object(obj); if (obj->base.pending_write_domain) - cd->flips |= atomic_read(&obj->pending_flip); + cd->flips |= atomic_read_unchecked(&obj->pending_flip); /* The actual obj->write_domain will be updated with * pending_write_domain after we emit the accumulated flush for all @@ -882,9 +882,9 @@ i915_gem_check_execbuffer(struct drm_i915_gem_execbuffer2 *exec) static int validate_exec_list(struct drm_i915_gem_exec_object2 *exec, - int count) + unsigned int count) { - int i; + unsigned int i; for (i = 0; i < count; i++) { char __user *ptr = (char __user *)(uintptr_t)exec[i].relocs_ptr; diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 5bd4361..0241a42 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -475,7 +475,7 @@ static irqreturn_t ivybridge_irq_handler(DRM_IRQ_ARGS) u32 de_iir, gt_iir, de_ier, pch_iir, pm_iir; struct drm_i915_master_private *master_priv; - atomic_inc(&dev_priv->irq_received); + atomic_inc_unchecked(&dev_priv->irq_received); /* disable master interrupt before clearing iir */ de_ier = I915_READ(DEIER); @@ -566,7 +566,7 @@ static irqreturn_t ironlake_irq_handler(DRM_IRQ_ARGS) struct drm_i915_master_private *master_priv; u32 bsd_usr_interrupt = GT_BSD_USER_INTERRUPT; - atomic_inc(&dev_priv->irq_received); + atomic_inc_unchecked(&dev_priv->irq_received); if (IS_GEN6(dev)) bsd_usr_interrupt = GT_GEN6_BSD_USER_INTERRUPT; @@ -1231,7 +1231,7 @@ static irqreturn_t i915_driver_irq_handler(DRM_IRQ_ARGS) int ret = IRQ_NONE, pipe; bool blc_event = false; - atomic_inc(&dev_priv->irq_received); + atomic_inc_unchecked(&dev_priv->irq_received); iir = I915_READ(IIR); @@ -1743,7 +1743,7 @@ static void ironlake_irq_preinstall(struct drm_device *dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; - atomic_set(&dev_priv->irq_received, 0); + atomic_set_unchecked(&dev_priv->irq_received, 0); INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func); INIT_WORK(&dev_priv->error_work, i915_error_work_func); @@ -1932,7 +1932,7 @@ static void i915_driver_irq_preinstall(struct drm_device * dev) drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; int pipe; - atomic_set(&dev_priv->irq_received, 0); + atomic_set_unchecked(&dev_priv->irq_received, 0); INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func); INIT_WORK(&dev_priv->error_work, i915_error_work_func); diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 2163818..cede019 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2238,7 +2238,7 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y, wait_event(dev_priv->pending_flip_queue, atomic_read(&dev_priv->mm.wedged) || - atomic_read(&obj->pending_flip) == 0); + atomic_read_unchecked(&obj->pending_flip) == 0); /* Big Hammer, we also need to ensure that any pending * MI_WAIT_FOR_EVENT inside a user batch buffer on the @@ -2859,7 +2859,7 @@ static void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc) obj = to_intel_framebuffer(crtc->fb)->obj; dev_priv = crtc->dev->dev_private; wait_event(dev_priv->pending_flip_queue, - atomic_read(&obj->pending_flip) == 0); + atomic_read_unchecked(&obj->pending_flip) == 0); } static bool intel_crtc_driving_pch(struct drm_crtc *crtc) @@ -7171,7 +7171,7 @@ static void do_intel_finish_page_flip(struct drm_device *dev, atomic_clear_mask(1 << intel_crtc->plane, &obj->pending_flip.counter); - if (atomic_read(&obj->pending_flip) == 0) + if (atomic_read_unchecked(&obj->pending_flip) == 0) wake_up(&dev_priv->pending_flip_queue); schedule_work(&work->work); @@ -7354,7 +7354,13 @@ static int intel_gen6_queue_flip(struct drm_device *dev, OUT_RING(fb->pitches[0] | obj->tiling_mode); OUT_RING(obj->gtt_offset); - pf = I915_READ(PF_CTL(intel_crtc->pipe)) & PF_ENABLE; + /* Contrary to the suggestions in the documentation, + * "Enable Panel Fitter" does not seem to be required when page + * flipping with a non-native mode, and worse causes a normal + * modeset to fail. + * pf = I915_READ(PF_CTL(intel_crtc->pipe)) & PF_ENABLE; + */ + pf = 0; pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff; OUT_RING(pf | pipesrc); ADVANCE_LP_RING(); @@ -7461,7 +7467,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, /* Block clients from rendering to the new back buffer until * the flip occurs and the object is no longer visible. */ - atomic_add(1 << intel_crtc->plane, &work->old_fb_obj->pending_flip); + atomic_add_unchecked(1 << intel_crtc->plane, &work->old_fb_obj->pending_flip); ret = dev_priv->display.queue_flip(dev, crtc, fb, obj); if (ret) @@ -7475,7 +7481,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, return 0; cleanup_pending: - atomic_sub(1 << intel_crtc->plane, &work->old_fb_obj->pending_flip); + atomic_sub_unchecked(1 << intel_crtc->plane, &work->old_fb_obj->pending_flip); drm_gem_object_unreference(&work->old_fb_obj->base); drm_gem_object_unreference(&obj->base); mutex_unlock(&dev->struct_mutex); diff --git a/drivers/gpu/drm/mga/mga_drv.h b/drivers/gpu/drm/mga/mga_drv.h index 54558a0..2d97005 100644 --- a/drivers/gpu/drm/mga/mga_drv.h +++ b/drivers/gpu/drm/mga/mga_drv.h @@ -120,9 +120,9 @@ typedef struct drm_mga_private { u32 clear_cmd; u32 maccess; - atomic_t vbl_received; /**< Number of vblanks received. */ + atomic_unchecked_t vbl_received; /**< Number of vblanks received. */ wait_queue_head_t fence_queue; - atomic_t last_fence_retired; + atomic_unchecked_t last_fence_retired; u32 next_fence_to_post; unsigned int fb_cpp; diff --git a/drivers/gpu/drm/mga/mga_irq.c b/drivers/gpu/drm/mga/mga_irq.c index 2581202..f230a8d9 100644 --- a/drivers/gpu/drm/mga/mga_irq.c +++ b/drivers/gpu/drm/mga/mga_irq.c @@ -44,7 +44,7 @@ u32 mga_get_vblank_counter(struct drm_device *dev, int crtc) if (crtc != 0) return 0; - return atomic_read(&dev_priv->vbl_received); + return atomic_read_unchecked(&dev_priv->vbl_received); } @@ -60,7 +60,7 @@ irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS) /* VBLANK interrupt */ if (status & MGA_VLINEPEN) { MGA_WRITE(MGA_ICLEAR, MGA_VLINEICLR); - atomic_inc(&dev_priv->vbl_received); + atomic_inc_unchecked(&dev_priv->vbl_received); drm_handle_vblank(dev, 0); handled = 1; } @@ -79,7 +79,7 @@ irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS) if ((prim_start & ~0x03) != (prim_end & ~0x03)) MGA_WRITE(MGA_PRIMEND, prim_end); - atomic_inc(&dev_priv->last_fence_retired); + atomic_inc_unchecked(&dev_priv->last_fence_retired); DRM_WAKEUP(&dev_priv->fence_queue); handled = 1; } @@ -130,7 +130,7 @@ int mga_driver_fence_wait(struct drm_device *dev, unsigned int *sequence) * using fences. */ DRM_WAIT_ON(ret, dev_priv->fence_queue, 3 * DRM_HZ, - (((cur_fence = atomic_read(&dev_priv->last_fence_retired)) + (((cur_fence = atomic_read_unchecked(&dev_priv->last_fence_retired)) - *sequence) <= (1 << 23))); *sequence = cur_fence; diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c index e5cbead..6c354a3 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bios.c +++ b/drivers/gpu/drm/nouveau/nouveau_bios.c @@ -199,7 +199,7 @@ struct methods { const char desc[8]; void (*loadbios)(struct drm_device *, uint8_t *); const bool rw; -}; +} __do_const; static struct methods shadow_methods[] = { { "PRAMIN", load_vbios_pramin, true }, @@ -5290,7 +5290,7 @@ parse_bit_U_tbl_entry(struct drm_device *dev, struct nvbios *bios, struct bit_table { const char id; int (* const parse_fn)(struct drm_device *, struct nvbios *, struct bit_entry *); -}; +} __no_const; #define BIT_TABLE(id, funcid) ((struct bit_table){ id, parse_bit_##funcid##_tbl_entry }) diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index b827098..c31a797 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -242,7 +242,7 @@ struct nouveau_channel { struct list_head pending; uint32_t sequence; uint32_t sequence_ack; - atomic_t last_sequence_irq; + atomic_unchecked_t last_sequence_irq; struct nouveau_vma vma; } fence; @@ -323,7 +323,7 @@ struct nouveau_exec_engine { u32 handle, u16 class); void (*set_tile_region)(struct drm_device *dev, int i); void (*tlb_flush)(struct drm_device *, int engine); -}; +} __no_const; struct nouveau_instmem_engine { void *priv; @@ -345,13 +345,13 @@ struct nouveau_instmem_engine { struct nouveau_mc_engine { int (*init)(struct drm_device *dev); void (*takedown)(struct drm_device *dev); -}; +} __no_const; struct nouveau_timer_engine { int (*init)(struct drm_device *dev); void (*takedown)(struct drm_device *dev); uint64_t (*read)(struct drm_device *dev); -}; +} __no_const; struct nouveau_fb_engine { int num_tiles; @@ -566,7 +566,7 @@ struct nouveau_vram_engine { void (*put)(struct drm_device *, struct nouveau_mem **); bool (*flags_valid)(struct drm_device *, u32 tile_flags); -}; +} __no_const; struct nouveau_engine { struct nouveau_instmem_engine instmem; @@ -714,7 +714,7 @@ struct drm_nouveau_private { struct drm_global_reference mem_global_ref; struct ttm_bo_global_ref bo_global_ref; struct ttm_bo_device bdev; - atomic_t validate_sequence; + atomic_unchecked_t validate_sequence; } ttm; struct { diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c index 2f6daae..c9d7b9e 100644 --- a/drivers/gpu/drm/nouveau/nouveau_fence.c +++ b/drivers/gpu/drm/nouveau/nouveau_fence.c @@ -85,7 +85,7 @@ nouveau_fence_update(struct nouveau_channel *chan) if (USE_REFCNT(dev)) sequence = nvchan_rd32(chan, 0x48); else - sequence = atomic_read(&chan->fence.last_sequence_irq); + sequence = atomic_read_unchecked(&chan->fence.last_sequence_irq); if (chan->fence.sequence_ack == sequence) goto out; @@ -539,7 +539,7 @@ nouveau_fence_channel_init(struct nouveau_channel *chan) return ret; } - atomic_set(&chan->fence.last_sequence_irq, 0); + atomic_set_unchecked(&chan->fence.last_sequence_irq, 0); return 0; } diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c index 7ce3fde..cb3ea04 100644 --- a/drivers/gpu/drm/nouveau/nouveau_gem.c +++ b/drivers/gpu/drm/nouveau/nouveau_gem.c @@ -314,7 +314,7 @@ validate_init(struct nouveau_channel *chan, struct drm_file *file_priv, int trycnt = 0; int ret, i; - sequence = atomic_add_return(1, &dev_priv->ttm.validate_sequence); + sequence = atomic_add_return_unchecked(1, &dev_priv->ttm.validate_sequence); retry: if (++trycnt > 100000) { NV_ERROR(dev, "%s failed and gave up.\n", __func__); diff --git a/drivers/gpu/drm/nouveau/nouveau_state.c b/drivers/gpu/drm/nouveau/nouveau_state.c index f80c5e0..936baa7 100644 --- a/drivers/gpu/drm/nouveau/nouveau_state.c +++ b/drivers/gpu/drm/nouveau/nouveau_state.c @@ -543,7 +543,7 @@ static bool nouveau_switcheroo_can_switch(struct pci_dev *pdev) bool can_switch; spin_lock(&dev->count_lock); - can_switch = (dev->open_count == 0); + can_switch = (local_read(&dev->open_count) == 0); spin_unlock(&dev->count_lock); return can_switch; } diff --git a/drivers/gpu/drm/nouveau/nv04_graph.c b/drivers/gpu/drm/nouveau/nv04_graph.c index dbdea8e..cd6eeeb 100644 --- a/drivers/gpu/drm/nouveau/nv04_graph.c +++ b/drivers/gpu/drm/nouveau/nv04_graph.c @@ -554,7 +554,7 @@ static int nv04_graph_mthd_set_ref(struct nouveau_channel *chan, u32 class, u32 mthd, u32 data) { - atomic_set(&chan->fence.last_sequence_irq, data); + atomic_set_unchecked(&chan->fence.last_sequence_irq, data); return 0; } diff --git a/drivers/gpu/drm/r128/r128_cce.c b/drivers/gpu/drm/r128/r128_cce.c index bcac90b..53bfc76 100644 --- a/drivers/gpu/drm/r128/r128_cce.c +++ b/drivers/gpu/drm/r128/r128_cce.c @@ -378,7 +378,7 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init) /* GH: Simple idle check. */ - atomic_set(&dev_priv->idle_count, 0); + atomic_set_unchecked(&dev_priv->idle_count, 0); /* We don't support anything other than bus-mastering ring mode, * but the ring can be in either AGP or PCI space for the ring diff --git a/drivers/gpu/drm/r128/r128_drv.h b/drivers/gpu/drm/r128/r128_drv.h index 930c71b..499aded 100644 --- a/drivers/gpu/drm/r128/r128_drv.h +++ b/drivers/gpu/drm/r128/r128_drv.h @@ -90,14 +90,14 @@ typedef struct drm_r128_private { int is_pci; unsigned long cce_buffers_offset; - atomic_t idle_count; + atomic_unchecked_t idle_count; int page_flipping; int current_page; u32 crtc_offset; u32 crtc_offset_cntl; - atomic_t vbl_received; + atomic_unchecked_t vbl_received; u32 color_fmt; unsigned int front_offset; diff --git a/drivers/gpu/drm/r128/r128_irq.c b/drivers/gpu/drm/r128/r128_irq.c index 429d5a0..7e899ed 100644 --- a/drivers/gpu/drm/r128/r128_irq.c +++ b/drivers/gpu/drm/r128/r128_irq.c @@ -42,7 +42,7 @@ u32 r128_get_vblank_counter(struct drm_device *dev, int crtc) if (crtc != 0) return 0; - return atomic_read(&dev_priv->vbl_received); + return atomic_read_unchecked(&dev_priv->vbl_received); } irqreturn_t r128_driver_irq_handler(DRM_IRQ_ARGS) @@ -56,7 +56,7 @@ irqreturn_t r128_driver_irq_handler(DRM_IRQ_ARGS) /* VBLANK interrupt */ if (status & R128_CRTC_VBLANK_INT) { R128_WRITE(R128_GEN_INT_STATUS, R128_CRTC_VBLANK_INT_AK); - atomic_inc(&dev_priv->vbl_received); + atomic_inc_unchecked(&dev_priv->vbl_received); drm_handle_vblank(dev, 0); return IRQ_HANDLED; } diff --git a/drivers/gpu/drm/r128/r128_state.c b/drivers/gpu/drm/r128/r128_state.c index a9e33ce..09edd4b 100644 --- a/drivers/gpu/drm/r128/r128_state.c +++ b/drivers/gpu/drm/r128/r128_state.c @@ -321,10 +321,10 @@ static void r128_clear_box(drm_r128_private_t *dev_priv, static void r128_cce_performance_boxes(drm_r128_private_t *dev_priv) { - if (atomic_read(&dev_priv->idle_count) == 0) + if (atomic_read_unchecked(&dev_priv->idle_count) == 0) r128_clear_box(dev_priv, 64, 4, 8, 8, 0, 255, 0); else - atomic_set(&dev_priv->idle_count, 0); + atomic_set_unchecked(&dev_priv->idle_count, 0); } #endif diff --git a/drivers/gpu/drm/radeon/mkregtable.c b/drivers/gpu/drm/radeon/mkregtable.c index 5a82b6b..9e69c73 100644 --- a/drivers/gpu/drm/radeon/mkregtable.c +++ b/drivers/gpu/drm/radeon/mkregtable.c @@ -637,14 +637,14 @@ static int parser_auth(struct table *t, const char *filename) regex_t mask_rex; regmatch_t match[4]; char buf[1024]; - size_t end; + long end; int len; int done = 0; int r; unsigned o; struct offset *offset; char last_reg_s[10]; - int last_reg; + unsigned long last_reg; if (regcomp (&mask_rex, "(0x[0-9a-fA-F]*) *([_a-zA-Z0-9]*)", REG_EXTENDED)) { diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 1668ec1..30ebdab 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -250,7 +250,7 @@ struct radeon_fence_driver { uint32_t scratch_reg; uint64_t gpu_addr; volatile uint32_t *cpu_addr; - atomic_t seq; + atomic_unchecked_t seq; uint32_t last_seq; unsigned long last_jiffies; unsigned long last_timeout; @@ -752,7 +752,7 @@ struct r600_blit_cp_primitives { int x2, int y2); void (*draw_auto)(struct radeon_device *rdev); void (*set_default_state)(struct radeon_device *rdev); -}; +} __no_const; struct r600_blit { struct mutex mutex; @@ -1201,7 +1201,7 @@ struct radeon_asic { void (*pre_page_flip)(struct radeon_device *rdev, int crtc); u32 (*page_flip)(struct radeon_device *rdev, int crtc, u64 crtc_base); void (*post_page_flip)(struct radeon_device *rdev, int crtc); -}; +} __no_const; /* * Asic structures diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 49f7cb7..2fcb48f 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -687,7 +687,7 @@ static bool radeon_switcheroo_can_switch(struct pci_dev *pdev) bool can_switch; spin_lock(&dev->count_lock); - can_switch = (dev->open_count == 0); + can_switch = (local_read(&dev->open_count) == 0); spin_unlock(&dev->count_lock); return can_switch; } diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h index a1b59ca..86f2d44 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.h +++ b/drivers/gpu/drm/radeon/radeon_drv.h @@ -255,7 +255,7 @@ typedef struct drm_radeon_private { /* SW interrupt */ wait_queue_head_t swi_queue; - atomic_t swi_emitted; + atomic_unchecked_t swi_emitted; int vblank_crtc; uint32_t irq_enable_reg; uint32_t r500_disp_irq_reg; diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 4bd36a3..e66fe9c 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -70,7 +70,7 @@ int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence *fence) write_unlock_irqrestore(&rdev->fence_lock, irq_flags); return 0; } - fence->seq = atomic_add_return(1, &rdev->fence_drv[fence->ring].seq); + fence->seq = atomic_add_return_unchecked(1, &rdev->fence_drv[fence->ring].seq); if (!rdev->ring[fence->ring].ready) /* FIXME: cp is not running assume everythings is done right * away @@ -405,7 +405,7 @@ int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring) } rdev->fence_drv[ring].cpu_addr = &rdev->wb.wb[index/4]; rdev->fence_drv[ring].gpu_addr = rdev->wb.gpu_addr + index; - radeon_fence_write(rdev, atomic_read(&rdev->fence_drv[ring].seq), ring); + radeon_fence_write(rdev, atomic_read_unchecked(&rdev->fence_drv[ring].seq), ring); rdev->fence_drv[ring].initialized = true; DRM_INFO("fence driver on ring %d use gpu addr 0x%08Lx and cpu addr 0x%p\n", ring, rdev->fence_drv[ring].gpu_addr, rdev->fence_drv[ring].cpu_addr); @@ -418,7 +418,7 @@ static void radeon_fence_driver_init_ring(struct radeon_device *rdev, int ring) rdev->fence_drv[ring].scratch_reg = -1; rdev->fence_drv[ring].cpu_addr = NULL; rdev->fence_drv[ring].gpu_addr = 0; - atomic_set(&rdev->fence_drv[ring].seq, 0); + atomic_set_unchecked(&rdev->fence_drv[ring].seq, 0); INIT_LIST_HEAD(&rdev->fence_drv[ring].created); INIT_LIST_HEAD(&rdev->fence_drv[ring].emitted); INIT_LIST_HEAD(&rdev->fence_drv[ring].signaled); diff --git a/drivers/gpu/drm/radeon/radeon_ioc32.c b/drivers/gpu/drm/radeon/radeon_ioc32.c index 48b7cea..342236f 100644 --- a/drivers/gpu/drm/radeon/radeon_ioc32.c +++ b/drivers/gpu/drm/radeon/radeon_ioc32.c @@ -359,7 +359,7 @@ static int compat_radeon_cp_setparam(struct file *file, unsigned int cmd, request = compat_alloc_user_space(sizeof(*request)); if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) || __put_user(req32.param, &request->param) - || __put_user((void __user *)(unsigned long)req32.value, + || __put_user((unsigned long)req32.value, &request->value)) return -EFAULT; diff --git a/drivers/gpu/drm/radeon/radeon_irq.c b/drivers/gpu/drm/radeon/radeon_irq.c index 00da384..32f972d 100644 --- a/drivers/gpu/drm/radeon/radeon_irq.c +++ b/drivers/gpu/drm/radeon/radeon_irq.c @@ -225,8 +225,8 @@ static int radeon_emit_irq(struct drm_device * dev) unsigned int ret; RING_LOCALS; - atomic_inc(&dev_priv->swi_emitted); - ret = atomic_read(&dev_priv->swi_emitted); + atomic_inc_unchecked(&dev_priv->swi_emitted); + ret = atomic_read_unchecked(&dev_priv->swi_emitted); BEGIN_RING(4); OUT_RING_REG(RADEON_LAST_SWI_REG, ret); @@ -352,7 +352,7 @@ int radeon_driver_irq_postinstall(struct drm_device *dev) drm_radeon_private_t *dev_priv = (drm_radeon_private_t *) dev->dev_private; - atomic_set(&dev_priv->swi_emitted, 0); + atomic_set_unchecked(&dev_priv->swi_emitted, 0); DRM_INIT_WAITQUEUE(&dev_priv->swi_queue); dev->max_vblank_count = 0x001fffff; diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c index e8422ae..d22d4a8 100644 --- a/drivers/gpu/drm/radeon/radeon_state.c +++ b/drivers/gpu/drm/radeon/radeon_state.c @@ -2168,7 +2168,7 @@ static int radeon_cp_clear(struct drm_device *dev, void *data, struct drm_file * if (sarea_priv->nbox > RADEON_NR_SAREA_CLIPRECTS) sarea_priv->nbox = RADEON_NR_SAREA_CLIPRECTS; - if (DRM_COPY_FROM_USER(&depth_boxes, clear->depth_boxes, + if (sarea_priv->nbox > RADEON_NR_SAREA_CLIPRECTS || DRM_COPY_FROM_USER(&depth_boxes, clear->depth_boxes, sarea_priv->nbox * sizeof(depth_boxes[0]))) return -EFAULT; @@ -3031,7 +3031,7 @@ static int radeon_cp_getparam(struct drm_device *dev, void *data, struct drm_fil { drm_radeon_private_t *dev_priv = dev->dev_private; drm_radeon_getparam_t *param = data; - int value; + int value = 0; DRM_DEBUG("pid=%d\n", DRM_CURRENTPID); diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index c421e77..e6bf2e8 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -842,8 +842,10 @@ int radeon_mmap(struct file *filp, struct vm_area_struct *vma) } if (unlikely(ttm_vm_ops == NULL)) { ttm_vm_ops = vma->vm_ops; - radeon_ttm_vm_ops = *ttm_vm_ops; - radeon_ttm_vm_ops.fault = &radeon_ttm_fault; + pax_open_kernel(); + memcpy((void *)&radeon_ttm_vm_ops, ttm_vm_ops, sizeof(radeon_ttm_vm_ops)); + *(void **)&radeon_ttm_vm_ops.fault = &radeon_ttm_fault; + pax_close_kernel(); } vma->vm_ops = &radeon_ttm_vm_ops; return 0; diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c index f68dff2..8df955c 100644 --- a/drivers/gpu/drm/radeon/rs690.c +++ b/drivers/gpu/drm/radeon/rs690.c @@ -304,9 +304,11 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, if (rdev->pm.max_bandwidth.full > rdev->pm.sideport_bandwidth.full && rdev->pm.sideport_bandwidth.full) rdev->pm.max_bandwidth = rdev->pm.sideport_bandwidth; - read_delay_latency.full = dfixed_const(370 * 800 * 1000); + read_delay_latency.full = dfixed_const(800 * 1000); read_delay_latency.full = dfixed_div(read_delay_latency, rdev->pm.igp_sideport_mclk); + a.full = dfixed_const(370); + read_delay_latency.full = dfixed_mul(read_delay_latency, a); } else { if (rdev->pm.max_bandwidth.full > rdev->pm.k8_bandwidth.full && rdev->pm.k8_bandwidth.full) diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c index 499debd..66fce72 100644 --- a/drivers/gpu/drm/ttm/ttm_page_alloc.c +++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c @@ -398,9 +398,9 @@ static int ttm_pool_get_num_unused_pages(void) static int ttm_pool_mm_shrink(struct shrinker *shrink, struct shrink_control *sc) { - static atomic_t start_pool = ATOMIC_INIT(0); + static atomic_unchecked_t start_pool = ATOMIC_INIT(0); unsigned i; - unsigned pool_offset = atomic_add_return(1, &start_pool); + unsigned pool_offset = atomic_add_return_unchecked(1, &start_pool); struct ttm_page_pool *pool; int shrink_pages = sc->nr_to_scan; diff --git a/drivers/gpu/drm/via/via_drv.h b/drivers/gpu/drm/via/via_drv.h index 88edacc..1e5412b 100644 --- a/drivers/gpu/drm/via/via_drv.h +++ b/drivers/gpu/drm/via/via_drv.h @@ -51,7 +51,7 @@ typedef struct drm_via_ring_buffer { typedef uint32_t maskarray_t[5]; typedef struct drm_via_irq { - atomic_t irq_received; + atomic_unchecked_t irq_received; uint32_t pending_mask; uint32_t enable_mask; wait_queue_head_t irq_queue; @@ -75,7 +75,7 @@ typedef struct drm_via_private { struct timeval last_vblank; int last_vblank_valid; unsigned usec_per_vblank; - atomic_t vbl_received; + atomic_unchecked_t vbl_received; drm_via_state_t hc_state; char pci_buf[VIA_PCI_BUF_SIZE]; const uint32_t *fire_offsets[VIA_FIRE_BUF_SIZE]; diff --git a/drivers/gpu/drm/via/via_irq.c b/drivers/gpu/drm/via/via_irq.c index d391f48..10c8ca3 100644 --- a/drivers/gpu/drm/via/via_irq.c +++ b/drivers/gpu/drm/via/via_irq.c @@ -102,7 +102,7 @@ u32 via_get_vblank_counter(struct drm_device *dev, int crtc) if (crtc != 0) return 0; - return atomic_read(&dev_priv->vbl_received); + return atomic_read_unchecked(&dev_priv->vbl_received); } irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS) @@ -117,8 +117,8 @@ irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS) status = VIA_READ(VIA_REG_INTERRUPT); if (status & VIA_IRQ_VBLANK_PENDING) { - atomic_inc(&dev_priv->vbl_received); - if (!(atomic_read(&dev_priv->vbl_received) & 0x0F)) { + atomic_inc_unchecked(&dev_priv->vbl_received); + if (!(atomic_read_unchecked(&dev_priv->vbl_received) & 0x0F)) { do_gettimeofday(&cur_vblank); if (dev_priv->last_vblank_valid) { dev_priv->usec_per_vblank = @@ -128,7 +128,7 @@ irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS) dev_priv->last_vblank = cur_vblank; dev_priv->last_vblank_valid = 1; } - if (!(atomic_read(&dev_priv->vbl_received) & 0xFF)) { + if (!(atomic_read_unchecked(&dev_priv->vbl_received) & 0xFF)) { DRM_DEBUG("US per vblank is: %u\n", dev_priv->usec_per_vblank); } @@ -138,7 +138,7 @@ irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS) for (i = 0; i < dev_priv->num_irqs; ++i) { if (status & cur_irq->pending_mask) { - atomic_inc(&cur_irq->irq_received); + atomic_inc_unchecked(&cur_irq->irq_received); DRM_WAKEUP(&cur_irq->irq_queue); handled = 1; if (dev_priv->irq_map[drm_via_irq_dma0_td] == i) @@ -243,11 +243,11 @@ via_driver_irq_wait(struct drm_device *dev, unsigned int irq, int force_sequence DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * DRM_HZ, ((VIA_READ(masks[irq][2]) & masks[irq][3]) == masks[irq][4])); - cur_irq_sequence = atomic_read(&cur_irq->irq_received); + cur_irq_sequence = atomic_read_unchecked(&cur_irq->irq_received); } else { DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * DRM_HZ, (((cur_irq_sequence = - atomic_read(&cur_irq->irq_received)) - + atomic_read_unchecked(&cur_irq->irq_received)) - *sequence) <= (1 << 23))); } *sequence = cur_irq_sequence; @@ -285,7 +285,7 @@ void via_driver_irq_preinstall(struct drm_device *dev) } for (i = 0; i < dev_priv->num_irqs; ++i) { - atomic_set(&cur_irq->irq_received, 0); + atomic_set_unchecked(&cur_irq->irq_received, 0); cur_irq->enable_mask = dev_priv->irq_masks[i][0]; cur_irq->pending_mask = dev_priv->irq_masks[i][1]; DRM_INIT_WAITQUEUE(&cur_irq->irq_queue); @@ -367,7 +367,7 @@ int via_wait_irq(struct drm_device *dev, void *data, struct drm_file *file_priv) switch (irqwait->request.type & ~VIA_IRQ_FLAGS_MASK) { case VIA_IRQ_RELATIVE: irqwait->request.sequence += - atomic_read(&cur_irq->irq_received); + atomic_read_unchecked(&cur_irq->irq_received); irqwait->request.type &= ~_DRM_VBLANK_RELATIVE; case VIA_IRQ_ABSOLUTE: break; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index dc27970..f18b008 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -260,7 +260,7 @@ struct vmw_private { * Fencing and IRQs. */ - atomic_t marker_seq; + atomic_unchecked_t marker_seq; wait_queue_head_t fence_queue; wait_queue_head_t fifo_queue; int fence_queue_waiters; /* Protected by hw_mutex */ diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c index a0c2f12..68ae6cb 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c @@ -137,7 +137,7 @@ int vmw_fifo_init(struct vmw_private *dev_priv, struct vmw_fifo_state *fifo) (unsigned int) min, (unsigned int) fifo->capabilities); - atomic_set(&dev_priv->marker_seq, dev_priv->last_read_seqno); + atomic_set_unchecked(&dev_priv->marker_seq, dev_priv->last_read_seqno); iowrite32(dev_priv->last_read_seqno, fifo_mem + SVGA_FIFO_FENCE); vmw_marker_queue_init(&fifo->marker_queue); return vmw_fifo_send_fence(dev_priv, &dummy); @@ -355,7 +355,7 @@ void *vmw_fifo_reserve(struct vmw_private *dev_priv, uint32_t bytes) if (reserveable) iowrite32(bytes, fifo_mem + SVGA_FIFO_RESERVED); - return fifo_mem + (next_cmd >> 2); + return (__le32 __force_kernel *)fifo_mem + (next_cmd >> 2); } else { need_bounce = true; } @@ -475,7 +475,7 @@ int vmw_fifo_send_fence(struct vmw_private *dev_priv, uint32_t *seqno) fm = vmw_fifo_reserve(dev_priv, bytes); if (unlikely(fm == NULL)) { - *seqno = atomic_read(&dev_priv->marker_seq); + *seqno = atomic_read_unchecked(&dev_priv->marker_seq); ret = -ENOMEM; (void)vmw_fallback_wait(dev_priv, false, true, *seqno, false, 3*HZ); @@ -483,7 +483,7 @@ int vmw_fifo_send_fence(struct vmw_private *dev_priv, uint32_t *seqno) } do { - *seqno = atomic_add_return(1, &dev_priv->marker_seq); + *seqno = atomic_add_return_unchecked(1, &dev_priv->marker_seq); } while (*seqno == 0); if (!(fifo_state->capabilities & SVGA_FIFO_CAP_FENCE)) { diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c index cabc95f..14b3d77 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c @@ -107,7 +107,7 @@ bool vmw_seqno_passed(struct vmw_private *dev_priv, * emitted. Then the fence is stale and signaled. */ - ret = ((atomic_read(&dev_priv->marker_seq) - seqno) + ret = ((atomic_read_unchecked(&dev_priv->marker_seq) - seqno) > VMW_FENCE_WRAP); return ret; @@ -138,7 +138,7 @@ int vmw_fallback_wait(struct vmw_private *dev_priv, if (fifo_idle) down_read(&fifo_state->rwsem); - signal_seq = atomic_read(&dev_priv->marker_seq); + signal_seq = atomic_read_unchecked(&dev_priv->marker_seq); ret = 0; for (;;) { diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c b/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c index 8a8725c..afed796 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c @@ -151,7 +151,7 @@ int vmw_wait_lag(struct vmw_private *dev_priv, while (!vmw_lag_lt(queue, us)) { spin_lock(&queue->lock); if (list_empty(&queue->head)) - seqno = atomic_read(&dev_priv->marker_seq); + seqno = atomic_read_unchecked(&dev_priv->marker_seq); else { marker = list_first_entry(&queue->head, struct vmw_marker, head); diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index 75dbe34..f9204a8 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -2021,7 +2021,7 @@ static bool hid_ignore(struct hid_device *hdev) int hid_add_device(struct hid_device *hdev) { - static atomic_t id = ATOMIC_INIT(0); + static atomic_unchecked_t id = ATOMIC_INIT(0); int ret; if (WARN_ON(hdev->status & HID_STAT_ADDED)) @@ -2036,7 +2036,7 @@ int hid_add_device(struct hid_device *hdev) /* XXX hack, any other cleaner solution after the driver core * is converted to allow more than 20 bytes as the device name? */ dev_set_name(&hdev->dev, "%04X:%04X:%04X.%04X", hdev->bus, - hdev->vendor, hdev->product, atomic_inc_return(&id)); + hdev->vendor, hdev->product, atomic_inc_return_unchecked(&id)); hid_debug_register(hdev, dev_name(&hdev->dev)); ret = device_add(&hdev->dev); diff --git a/drivers/hid/usbhid/hiddev.c b/drivers/hid/usbhid/hiddev.c index b1ec0e2..c295a61 100644 --- a/drivers/hid/usbhid/hiddev.c +++ b/drivers/hid/usbhid/hiddev.c @@ -624,7 +624,7 @@ static long hiddev_ioctl(struct file *file, unsigned int cmd, unsigned long arg) break; case HIDIOCAPPLICATION: - if (arg < 0 || arg >= hid->maxapplication) + if (arg >= hid->maxapplication) break; for (i = 0; i < hid->maxcollection; i++) diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c index 4065374..10ed7dc 100644 --- a/drivers/hv/channel.c +++ b/drivers/hv/channel.c @@ -400,8 +400,8 @@ int vmbus_establish_gpadl(struct vmbus_channel *channel, void *kbuffer, int ret = 0; int t; - next_gpadl_handle = atomic_read(&vmbus_connection.next_gpadl_handle); - atomic_inc(&vmbus_connection.next_gpadl_handle); + next_gpadl_handle = atomic_read_unchecked(&vmbus_connection.next_gpadl_handle); + atomic_inc_unchecked(&vmbus_connection.next_gpadl_handle); ret = create_gpadl_header(kbuffer, size, &msginfo, &msgcount); if (ret) diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c index 12aa97f..c0679f7 100644 --- a/drivers/hv/hv.c +++ b/drivers/hv/hv.c @@ -132,7 +132,7 @@ static u64 do_hypercall(u64 control, void *input, void *output) u64 output_address = (output) ? virt_to_phys(output) : 0; u32 output_address_hi = output_address >> 32; u32 output_address_lo = output_address & 0xFFFFFFFF; - void *hypercall_page = hv_context.hypercall_page; + void *hypercall_page = ktva_ktla(hv_context.hypercall_page); __asm__ __volatile__ ("call *%8" : "=d"(hv_status_hi), "=a"(hv_status_lo) : "d" (control_hi), diff --git a/drivers/hv/hyperv_vmbus.h b/drivers/hv/hyperv_vmbus.h index 6d7d286..92b0873 100644 --- a/drivers/hv/hyperv_vmbus.h +++ b/drivers/hv/hyperv_vmbus.h @@ -556,7 +556,7 @@ enum vmbus_connect_state { struct vmbus_connection { enum vmbus_connect_state conn_state; - atomic_t next_gpadl_handle; + atomic_unchecked_t next_gpadl_handle; /* * Represents channel interrupts. Each bit position represents a diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c index a220e57..428f54d 100644 --- a/drivers/hv/vmbus_drv.c +++ b/drivers/hv/vmbus_drv.c @@ -663,10 +663,10 @@ int vmbus_device_register(struct hv_device *child_device_obj) { int ret = 0; - static atomic_t device_num = ATOMIC_INIT(0); + static atomic_unchecked_t device_num = ATOMIC_INIT(0); dev_set_name(&child_device_obj->device, "vmbus_0_%d", - atomic_inc_return(&device_num)); + atomic_inc_return_unchecked(&device_num)); child_device_obj->device.bus = &hv_bus; child_device_obj->device.parent = &hv_acpi_dev->dev; diff --git a/drivers/hwmon/acpi_power_meter.c b/drivers/hwmon/acpi_power_meter.c index 554f046..f8b4729 100644 --- a/drivers/hwmon/acpi_power_meter.c +++ b/drivers/hwmon/acpi_power_meter.c @@ -316,8 +316,6 @@ static ssize_t set_trip(struct device *dev, struct device_attribute *devattr, return res; temp /= 1000; - if (temp < 0) - return -EINVAL; mutex_lock(&resource->lock); resource->trip[attr->index - 7] = temp; diff --git a/drivers/hwmon/sht15.c b/drivers/hwmon/sht15.c index 91fdd1f..b66a686 100644 --- a/drivers/hwmon/sht15.c +++ b/drivers/hwmon/sht15.c @@ -166,7 +166,7 @@ struct sht15_data { int supply_uV; bool supply_uV_valid; struct work_struct update_supply_work; - atomic_t interrupt_handled; + atomic_unchecked_t interrupt_handled; }; /** @@ -509,13 +509,13 @@ static int sht15_measurement(struct sht15_data *data, return ret; gpio_direction_input(data->pdata->gpio_data); - atomic_set(&data->interrupt_handled, 0); + atomic_set_unchecked(&data->interrupt_handled, 0); enable_irq(gpio_to_irq(data->pdata->gpio_data)); if (gpio_get_value(data->pdata->gpio_data) == 0) { disable_irq_nosync(gpio_to_irq(data->pdata->gpio_data)); /* Only relevant if the interrupt hasn't occurred. */ - if (!atomic_read(&data->interrupt_handled)) + if (!atomic_read_unchecked(&data->interrupt_handled)) schedule_work(&data->read_work); } ret = wait_event_timeout(data->wait_queue, @@ -782,7 +782,7 @@ static irqreturn_t sht15_interrupt_fired(int irq, void *d) /* First disable the interrupt */ disable_irq_nosync(irq); - atomic_inc(&data->interrupt_handled); + atomic_inc_unchecked(&data->interrupt_handled); /* Then schedule a reading work struct */ if (data->state != SHT15_READING_NOTHING) schedule_work(&data->read_work); @@ -804,11 +804,11 @@ static void sht15_bh_read_data(struct work_struct *work_s) * If not, then start the interrupt again - care here as could * have gone low in meantime so verify it hasn't! */ - atomic_set(&data->interrupt_handled, 0); + atomic_set_unchecked(&data->interrupt_handled, 0); enable_irq(gpio_to_irq(data->pdata->gpio_data)); /* If still not occurred or another handler has been scheduled */ if (gpio_get_value(data->pdata->gpio_data) - || atomic_read(&data->interrupt_handled)) + || atomic_read_unchecked(&data->interrupt_handled)) return; } diff --git a/drivers/i2c/busses/i2c-amd756-s4882.c b/drivers/i2c/busses/i2c-amd756-s4882.c index 378fcb5..5e91fa8 100644 --- a/drivers/i2c/busses/i2c-amd756-s4882.c +++ b/drivers/i2c/busses/i2c-amd756-s4882.c @@ -43,7 +43,7 @@ extern struct i2c_adapter amd756_smbus; static struct i2c_adapter *s4882_adapter; -static struct i2c_algorithm *s4882_algo; +static i2c_algorithm_no_const *s4882_algo; /* Wrapper access functions for multiplexed SMBus */ static DEFINE_MUTEX(amd756_lock); diff --git a/drivers/i2c/busses/i2c-nforce2-s4985.c b/drivers/i2c/busses/i2c-nforce2-s4985.c index 29015eb..af2d8e9 100644 --- a/drivers/i2c/busses/i2c-nforce2-s4985.c +++ b/drivers/i2c/busses/i2c-nforce2-s4985.c @@ -41,7 +41,7 @@ extern struct i2c_adapter *nforce2_smbus; static struct i2c_adapter *s4985_adapter; -static struct i2c_algorithm *s4985_algo; +static i2c_algorithm_no_const *s4985_algo; /* Wrapper access functions for multiplexed SMBus */ static DEFINE_MUTEX(nforce2_lock); diff --git a/drivers/i2c/i2c-mux.c b/drivers/i2c/i2c-mux.c index d7a4833..7fae376 100644 --- a/drivers/i2c/i2c-mux.c +++ b/drivers/i2c/i2c-mux.c @@ -28,7 +28,7 @@ /* multiplexer per channel data */ struct i2c_mux_priv { struct i2c_adapter adap; - struct i2c_algorithm algo; + i2c_algorithm_no_const algo; struct i2c_adapter *parent; void *mux_dev; /* the mux chip/device */ diff --git a/drivers/ide/aec62xx.c b/drivers/ide/aec62xx.c index 57d00ca..0145194 100644 --- a/drivers/ide/aec62xx.c +++ b/drivers/ide/aec62xx.c @@ -181,7 +181,7 @@ static const struct ide_port_ops atp86x_port_ops = { .cable_detect = atp86x_cable_detect, }; -static const struct ide_port_info aec62xx_chipsets[] __devinitdata = { +static const struct ide_port_info aec62xx_chipsets[] __devinitconst = { { /* 0: AEC6210 */ .name = DRV_NAME, .init_chipset = init_chipset_aec62xx, diff --git a/drivers/ide/alim15x3.c b/drivers/ide/alim15x3.c index 2c8016a..911a27c 100644 --- a/drivers/ide/alim15x3.c +++ b/drivers/ide/alim15x3.c @@ -512,7 +512,7 @@ static const struct ide_dma_ops ali_dma_ops = { .dma_sff_read_status = ide_dma_sff_read_status, }; -static const struct ide_port_info ali15x3_chipset __devinitdata = { +static const struct ide_port_info ali15x3_chipset __devinitconst = { .name = DRV_NAME, .init_chipset = init_chipset_ali15x3, .init_hwif = init_hwif_ali15x3, diff --git a/drivers/ide/amd74xx.c b/drivers/ide/amd74xx.c index 3747b25..56fc995 100644 --- a/drivers/ide/amd74xx.c +++ b/drivers/ide/amd74xx.c @@ -223,7 +223,7 @@ static const struct ide_port_ops amd_port_ops = { .udma_mask = udma, \ } -static const struct ide_port_info amd74xx_chipsets[] __devinitdata = { +static const struct ide_port_info amd74xx_chipsets[] __devinitconst = { /* 0: AMD7401 */ DECLARE_AMD_DEV(0x00, ATA_UDMA2), /* 1: AMD7409 */ DECLARE_AMD_DEV(ATA_SWDMA2, ATA_UDMA4), /* 2: AMD7411/7441 */ DECLARE_AMD_DEV(ATA_SWDMA2, ATA_UDMA5), diff --git a/drivers/ide/atiixp.c b/drivers/ide/atiixp.c index 15f0ead..cb43480 100644 --- a/drivers/ide/atiixp.c +++ b/drivers/ide/atiixp.c @@ -139,7 +139,7 @@ static const struct ide_port_ops atiixp_port_ops = { .cable_detect = atiixp_cable_detect, }; -static const struct ide_port_info atiixp_pci_info[] __devinitdata = { +static const struct ide_port_info atiixp_pci_info[] __devinitconst = { { /* 0: IXP200/300/400/700 */ .name = DRV_NAME, .enablebits = {{0x48,0x01,0x00}, {0x48,0x08,0x00}}, diff --git a/drivers/ide/cmd64x.c b/drivers/ide/cmd64x.c index 5f80312..d1fc438 100644 --- a/drivers/ide/cmd64x.c +++ b/drivers/ide/cmd64x.c @@ -327,7 +327,7 @@ static const struct ide_dma_ops cmd646_rev1_dma_ops = { .dma_sff_read_status = ide_dma_sff_read_status, }; -static const struct ide_port_info cmd64x_chipsets[] __devinitdata = { +static const struct ide_port_info cmd64x_chipsets[] __devinitconst = { { /* 0: CMD643 */ .name = DRV_NAME, .init_chipset = init_chipset_cmd64x, diff --git a/drivers/ide/cs5520.c b/drivers/ide/cs5520.c index 2c1e5f7..1444762 100644 --- a/drivers/ide/cs5520.c +++ b/drivers/ide/cs5520.c @@ -94,7 +94,7 @@ static const struct ide_port_ops cs5520_port_ops = { .set_dma_mode = cs5520_set_dma_mode, }; -static const struct ide_port_info cyrix_chipset __devinitdata = { +static const struct ide_port_info cyrix_chipset __devinitconst = { .name = DRV_NAME, .enablebits = { { 0x60, 0x01, 0x01 }, { 0x60, 0x02, 0x02 } }, .port_ops = &cs5520_port_ops, diff --git a/drivers/ide/cs5530.c b/drivers/ide/cs5530.c index 4dc4eb9..49b40ad 100644 --- a/drivers/ide/cs5530.c +++ b/drivers/ide/cs5530.c @@ -245,7 +245,7 @@ static const struct ide_port_ops cs5530_port_ops = { .udma_filter = cs5530_udma_filter, }; -static const struct ide_port_info cs5530_chipset __devinitdata = { +static const struct ide_port_info cs5530_chipset __devinitconst = { .name = DRV_NAME, .init_chipset = init_chipset_cs5530, .init_hwif = init_hwif_cs5530, diff --git a/drivers/ide/cs5535.c b/drivers/ide/cs5535.c index 5059faf..18d4c85 100644 --- a/drivers/ide/cs5535.c +++ b/drivers/ide/cs5535.c @@ -170,7 +170,7 @@ static const struct ide_port_ops cs5535_port_ops = { .cable_detect = cs5535_cable_detect, }; -static const struct ide_port_info cs5535_chipset __devinitdata = { +static const struct ide_port_info cs5535_chipset __devinitconst = { .name = DRV_NAME, .port_ops = &cs5535_port_ops, .host_flags = IDE_HFLAG_SINGLE | IDE_HFLAG_POST_SET_MODE, diff --git a/drivers/ide/cy82c693.c b/drivers/ide/cy82c693.c index 847553f..3ffb49d 100644 --- a/drivers/ide/cy82c693.c +++ b/drivers/ide/cy82c693.c @@ -163,7 +163,7 @@ static const struct ide_port_ops cy82c693_port_ops = { .set_dma_mode = cy82c693_set_dma_mode, }; -static const struct ide_port_info cy82c693_chipset __devinitdata = { +static const struct ide_port_info cy82c693_chipset __devinitconst = { .name = DRV_NAME, .init_iops = init_iops_cy82c693, .port_ops = &cy82c693_port_ops, diff --git a/drivers/ide/hpt366.c b/drivers/ide/hpt366.c index 58c51cd..4aec3b8 100644 --- a/drivers/ide/hpt366.c +++ b/drivers/ide/hpt366.c @@ -443,7 +443,7 @@ static struct hpt_timings hpt37x_timings = { } }; -static const struct hpt_info hpt36x __devinitdata = { +static const struct hpt_info hpt36x __devinitconst = { .chip_name = "HPT36x", .chip_type = HPT36x, .udma_mask = HPT366_ALLOW_ATA66_3 ? (HPT366_ALLOW_ATA66_4 ? ATA_UDMA4 : ATA_UDMA3) : ATA_UDMA2, @@ -451,7 +451,7 @@ static const struct hpt_info hpt36x __devinitdata = { .timings = &hpt36x_timings }; -static const struct hpt_info hpt370 __devinitdata = { +static const struct hpt_info hpt370 __devinitconst = { .chip_name = "HPT370", .chip_type = HPT370, .udma_mask = HPT370_ALLOW_ATA100_5 ? ATA_UDMA5 : ATA_UDMA4, @@ -459,7 +459,7 @@ static const struct hpt_info hpt370 __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt370a __devinitdata = { +static const struct hpt_info hpt370a __devinitconst = { .chip_name = "HPT370A", .chip_type = HPT370A, .udma_mask = HPT370_ALLOW_ATA100_5 ? ATA_UDMA5 : ATA_UDMA4, @@ -467,7 +467,7 @@ static const struct hpt_info hpt370a __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt374 __devinitdata = { +static const struct hpt_info hpt374 __devinitconst = { .chip_name = "HPT374", .chip_type = HPT374, .udma_mask = ATA_UDMA5, @@ -475,7 +475,7 @@ static const struct hpt_info hpt374 __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt372 __devinitdata = { +static const struct hpt_info hpt372 __devinitconst = { .chip_name = "HPT372", .chip_type = HPT372, .udma_mask = HPT372_ALLOW_ATA133_6 ? ATA_UDMA6 : ATA_UDMA5, @@ -483,7 +483,7 @@ static const struct hpt_info hpt372 __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt372a __devinitdata = { +static const struct hpt_info hpt372a __devinitconst = { .chip_name = "HPT372A", .chip_type = HPT372A, .udma_mask = HPT372_ALLOW_ATA133_6 ? ATA_UDMA6 : ATA_UDMA5, @@ -491,7 +491,7 @@ static const struct hpt_info hpt372a __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt302 __devinitdata = { +static const struct hpt_info hpt302 __devinitconst = { .chip_name = "HPT302", .chip_type = HPT302, .udma_mask = HPT302_ALLOW_ATA133_6 ? ATA_UDMA6 : ATA_UDMA5, @@ -499,7 +499,7 @@ static const struct hpt_info hpt302 __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt371 __devinitdata = { +static const struct hpt_info hpt371 __devinitconst = { .chip_name = "HPT371", .chip_type = HPT371, .udma_mask = HPT371_ALLOW_ATA133_6 ? ATA_UDMA6 : ATA_UDMA5, @@ -507,7 +507,7 @@ static const struct hpt_info hpt371 __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt372n __devinitdata = { +static const struct hpt_info hpt372n __devinitconst = { .chip_name = "HPT372N", .chip_type = HPT372N, .udma_mask = HPT372_ALLOW_ATA133_6 ? ATA_UDMA6 : ATA_UDMA5, @@ -515,7 +515,7 @@ static const struct hpt_info hpt372n __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt302n __devinitdata = { +static const struct hpt_info hpt302n __devinitconst = { .chip_name = "HPT302N", .chip_type = HPT302N, .udma_mask = HPT302_ALLOW_ATA133_6 ? ATA_UDMA6 : ATA_UDMA5, @@ -523,7 +523,7 @@ static const struct hpt_info hpt302n __devinitdata = { .timings = &hpt37x_timings }; -static const struct hpt_info hpt371n __devinitdata = { +static const struct hpt_info hpt371n __devinitconst = { .chip_name = "HPT371N", .chip_type = HPT371N, .udma_mask = HPT371_ALLOW_ATA133_6 ? ATA_UDMA6 : ATA_UDMA5, @@ -1361,7 +1361,7 @@ static const struct ide_dma_ops hpt36x_dma_ops = { .dma_sff_read_status = ide_dma_sff_read_status, }; -static const struct ide_port_info hpt366_chipsets[] __devinitdata = { +static const struct ide_port_info hpt366_chipsets[] __devinitconst = { { /* 0: HPT36x */ .name = DRV_NAME, .init_chipset = init_chipset_hpt366, diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c index 8126824..55a2798 100644 --- a/drivers/ide/ide-cd.c +++ b/drivers/ide/ide-cd.c @@ -768,7 +768,7 @@ static void cdrom_do_block_pc(ide_drive_t *drive, struct request *rq) alignment = queue_dma_alignment(q) | q->dma_pad_mask; if ((unsigned long)buf & alignment || blk_rq_bytes(rq) & q->dma_pad_mask - || object_is_on_stack(buf)) + || object_starts_on_stack(buf)) drive->dma = 0; } } diff --git a/drivers/ide/ide-pci-generic.c b/drivers/ide/ide-pci-generic.c index 7f56b73..dab5b67 100644 --- a/drivers/ide/ide-pci-generic.c +++ b/drivers/ide/ide-pci-generic.c @@ -53,7 +53,7 @@ static const struct ide_port_ops netcell_port_ops = { .udma_mask = ATA_UDMA6, \ } -static const struct ide_port_info generic_chipsets[] __devinitdata = { +static const struct ide_port_info generic_chipsets[] __devinitconst = { /* 0: Unknown */ DECLARE_GENERIC_PCI_DEV(0), diff --git a/drivers/ide/it8172.c b/drivers/ide/it8172.c index 560e66d..d5dd180 100644 --- a/drivers/ide/it8172.c +++ b/drivers/ide/it8172.c @@ -115,7 +115,7 @@ static const struct ide_port_ops it8172_port_ops = { .set_dma_mode = it8172_set_dma_mode, }; -static const struct ide_port_info it8172_port_info __devinitdata = { +static const struct ide_port_info it8172_port_info __devinitconst = { .name = DRV_NAME, .port_ops = &it8172_port_ops, .enablebits = { {0x41, 0x80, 0x80}, {0x00, 0x00, 0x00} }, diff --git a/drivers/ide/it8213.c b/drivers/ide/it8213.c index 46816ba..1847aeb 100644 --- a/drivers/ide/it8213.c +++ b/drivers/ide/it8213.c @@ -156,7 +156,7 @@ static const struct ide_port_ops it8213_port_ops = { .cable_detect = it8213_cable_detect, }; -static const struct ide_port_info it8213_chipset __devinitdata = { +static const struct ide_port_info it8213_chipset __devinitconst = { .name = DRV_NAME, .enablebits = { {0x41, 0x80, 0x80} }, .port_ops = &it8213_port_ops, diff --git a/drivers/ide/it821x.c b/drivers/ide/it821x.c index 2e3169f..c5611db 100644 --- a/drivers/ide/it821x.c +++ b/drivers/ide/it821x.c @@ -630,7 +630,7 @@ static const struct ide_port_ops it821x_port_ops = { .cable_detect = it821x_cable_detect, }; -static const struct ide_port_info it821x_chipset __devinitdata = { +static const struct ide_port_info it821x_chipset __devinitconst = { .name = DRV_NAME, .init_chipset = init_chipset_it821x, .init_hwif = init_hwif_it821x, diff --git a/drivers/ide/jmicron.c b/drivers/ide/jmicron.c index 74c2c4a..efddd7d 100644 --- a/drivers/ide/jmicron.c +++ b/drivers/ide/jmicron.c @@ -102,7 +102,7 @@ static const struct ide_port_ops jmicron_port_ops = { .cable_detect = jmicron_cable_detect, }; -static const struct ide_port_info jmicron_chipset __devinitdata = { +static const struct ide_port_info jmicron_chipset __devinitconst = { .name = DRV_NAME, .enablebits = { { 0x40, 0x01, 0x01 }, { 0x40, 0x10, 0x10 } }, .port_ops = &jmicron_port_ops, diff --git a/drivers/ide/ns87415.c b/drivers/ide/ns87415.c index 95327a2..73f78d8 100644 --- a/drivers/ide/ns87415.c +++ b/drivers/ide/ns87415.c @@ -293,7 +293,7 @@ static const struct ide_dma_ops ns87415_dma_ops = { .dma_sff_read_status = superio_dma_sff_read_status, }; -static const struct ide_port_info ns87415_chipset __devinitdata = { +static const struct ide_port_info ns87415_chipset __devinitconst = { .name = DRV_NAME, .init_hwif = init_hwif_ns87415, .tp_ops = &ns87415_tp_ops, diff --git a/drivers/ide/opti621.c b/drivers/ide/opti621.c index 1a53a4c..39edc66 100644 --- a/drivers/ide/opti621.c +++ b/drivers/ide/opti621.c @@ -131,7 +131,7 @@ static const struct ide_port_ops opti621_port_ops = { .set_pio_mode = opti621_set_pio_mode, }; -static const struct ide_port_info opti621_chipset __devinitdata = { +static const struct ide_port_info opti621_chipset __devinitconst = { .name = DRV_NAME, .enablebits = { {0x45, 0x80, 0x00}, {0x40, 0x08, 0x00} }, .port_ops = &opti621_port_ops, diff --git a/drivers/ide/pdc202xx_new.c b/drivers/ide/pdc202xx_new.c index 9546fe2..2e5ceb6 100644 --- a/drivers/ide/pdc202xx_new.c +++ b/drivers/ide/pdc202xx_new.c @@ -465,7 +465,7 @@ static const struct ide_port_ops pdcnew_port_ops = { .udma_mask = udma, \ } -static const struct ide_port_info pdcnew_chipsets[] __devinitdata = { +static const struct ide_port_info pdcnew_chipsets[] __devinitconst = { /* 0: PDC202{68,70} */ DECLARE_PDCNEW_DEV(ATA_UDMA5), /* 1: PDC202{69,71,75,76,77} */ DECLARE_PDCNEW_DEV(ATA_UDMA6), }; diff --git a/drivers/ide/pdc202xx_old.c b/drivers/ide/pdc202xx_old.c index 3a35ec6..5634510 100644 --- a/drivers/ide/pdc202xx_old.c +++ b/drivers/ide/pdc202xx_old.c @@ -270,7 +270,7 @@ static const struct ide_dma_ops pdc2026x_dma_ops = { .max_sectors = sectors, \ } -static const struct ide_port_info pdc202xx_chipsets[] __devinitdata = { +static const struct ide_port_info pdc202xx_chipsets[] __devinitconst = { { /* 0: PDC20246 */ .name = DRV_NAME, .init_chipset = init_chipset_pdc202xx, diff --git a/drivers/ide/piix.c b/drivers/ide/piix.c index 1892e81..fe0fd60 100644 --- a/drivers/ide/piix.c +++ b/drivers/ide/piix.c @@ -344,7 +344,7 @@ static const struct ide_port_ops ich_port_ops = { .udma_mask = udma, \ } -static const struct ide_port_info piix_pci_info[] __devinitdata = { +static const struct ide_port_info piix_pci_info[] __devinitconst = { /* 0: MPIIX */ { /* * MPIIX actually has only a single IDE channel mapped to diff --git a/drivers/ide/rz1000.c b/drivers/ide/rz1000.c index a6414a8..c04173e 100644 --- a/drivers/ide/rz1000.c +++ b/drivers/ide/rz1000.c @@ -38,7 +38,7 @@ static int __devinit rz1000_disable_readahead(struct pci_dev *dev) } } -static const struct ide_port_info rz1000_chipset __devinitdata = { +static const struct ide_port_info rz1000_chipset __devinitconst = { .name = DRV_NAME, .host_flags = IDE_HFLAG_NO_DMA, }; diff --git a/drivers/ide/sc1200.c b/drivers/ide/sc1200.c index 356b9b5..d4758eb 100644 --- a/drivers/ide/sc1200.c +++ b/drivers/ide/sc1200.c @@ -291,7 +291,7 @@ static const struct ide_dma_ops sc1200_dma_ops = { .dma_sff_read_status = ide_dma_sff_read_status, }; -static const struct ide_port_info sc1200_chipset __devinitdata = { +static const struct ide_port_info sc1200_chipset __devinitconst = { .name = DRV_NAME, .port_ops = &sc1200_port_ops, .dma_ops = &sc1200_dma_ops, diff --git a/drivers/ide/scc_pata.c b/drivers/ide/scc_pata.c index b7f5b0c..9701038 100644 --- a/drivers/ide/scc_pata.c +++ b/drivers/ide/scc_pata.c @@ -811,7 +811,7 @@ static const struct ide_dma_ops scc_dma_ops = { .dma_sff_read_status = scc_dma_sff_read_status, }; -static const struct ide_port_info scc_chipset __devinitdata = { +static const struct ide_port_info scc_chipset __devinitconst = { .name = "sccIDE", .init_iops = init_iops_scc, .init_dma = scc_init_dma, diff --git a/drivers/ide/serverworks.c b/drivers/ide/serverworks.c index 35fb8da..24d72ef 100644 --- a/drivers/ide/serverworks.c +++ b/drivers/ide/serverworks.c @@ -337,7 +337,7 @@ static const struct ide_port_ops svwks_port_ops = { .cable_detect = svwks_cable_detect, }; -static const struct ide_port_info serverworks_chipsets[] __devinitdata = { +static const struct ide_port_info serverworks_chipsets[] __devinitconst = { { /* 0: OSB4 */ .name = DRV_NAME, .init_chipset = init_chipset_svwks, diff --git a/drivers/ide/siimage.c b/drivers/ide/siimage.c index ddeda44..46f7e30 100644 --- a/drivers/ide/siimage.c +++ b/drivers/ide/siimage.c @@ -719,7 +719,7 @@ static const struct ide_dma_ops sil_dma_ops = { .udma_mask = ATA_UDMA6, \ } -static const struct ide_port_info siimage_chipsets[] __devinitdata = { +static const struct ide_port_info siimage_chipsets[] __devinitconst = { /* 0: SiI680 */ DECLARE_SII_DEV(&sil_pata_port_ops), /* 1: SiI3112 */ DECLARE_SII_DEV(&sil_sata_port_ops) }; diff --git a/drivers/ide/sis5513.c b/drivers/ide/sis5513.c index 4a00225..09e61b4 100644 --- a/drivers/ide/sis5513.c +++ b/drivers/ide/sis5513.c @@ -563,7 +563,7 @@ static const struct ide_port_ops sis_ata133_port_ops = { .cable_detect = sis_cable_detect, }; -static const struct ide_port_info sis5513_chipset __devinitdata = { +static const struct ide_port_info sis5513_chipset __devinitconst = { .name = DRV_NAME, .init_chipset = init_chipset_sis5513, .enablebits = { {0x4a, 0x02, 0x02}, {0x4a, 0x04, 0x04} }, diff --git a/drivers/ide/sl82c105.c b/drivers/ide/sl82c105.c index f21dc2a..d051cd2 100644 --- a/drivers/ide/sl82c105.c +++ b/drivers/ide/sl82c105.c @@ -299,7 +299,7 @@ static const struct ide_dma_ops sl82c105_dma_ops = { .dma_sff_read_status = ide_dma_sff_read_status, }; -static const struct ide_port_info sl82c105_chipset __devinitdata = { +static const struct ide_port_info sl82c105_chipset __devinitconst = { .name = DRV_NAME, .init_chipset = init_chipset_sl82c105, .enablebits = {{0x40,0x01,0x01}, {0x40,0x10,0x10}}, diff --git a/drivers/ide/slc90e66.c b/drivers/ide/slc90e66.c index 864ffe0..863a5e9 100644 --- a/drivers/ide/slc90e66.c +++ b/drivers/ide/slc90e66.c @@ -132,7 +132,7 @@ static const struct ide_port_ops slc90e66_port_ops = { .cable_detect = slc90e66_cable_detect, }; -static const struct ide_port_info slc90e66_chipset __devinitdata = { +static const struct ide_port_info slc90e66_chipset __devinitconst = { .name = DRV_NAME, .enablebits = { {0x41, 0x80, 0x80}, {0x43, 0x80, 0x80} }, .port_ops = &slc90e66_port_ops, diff --git a/drivers/ide/tc86c001.c b/drivers/ide/tc86c001.c index 4799d5c..1794678 100644 --- a/drivers/ide/tc86c001.c +++ b/drivers/ide/tc86c001.c @@ -192,7 +192,7 @@ static const struct ide_dma_ops tc86c001_dma_ops = { .dma_sff_read_status = ide_dma_sff_read_status, }; -static const struct ide_port_info tc86c001_chipset __devinitdata = { +static const struct ide_port_info tc86c001_chipset __devinitconst = { .name = DRV_NAME, .init_hwif = init_hwif_tc86c001, .port_ops = &tc86c001_port_ops, diff --git a/drivers/ide/triflex.c b/drivers/ide/triflex.c index 281c914..55ce1b8 100644 --- a/drivers/ide/triflex.c +++ b/drivers/ide/triflex.c @@ -92,7 +92,7 @@ static const struct ide_port_ops triflex_port_ops = { .set_dma_mode = triflex_set_mode, }; -static const struct ide_port_info triflex_device __devinitdata = { +static const struct ide_port_info triflex_device __devinitconst = { .name = DRV_NAME, .enablebits = {{0x80, 0x01, 0x01}, {0x80, 0x02, 0x02}}, .port_ops = &triflex_port_ops, diff --git a/drivers/ide/trm290.c b/drivers/ide/trm290.c index 4b42ca0..e494a98 100644 --- a/drivers/ide/trm290.c +++ b/drivers/ide/trm290.c @@ -324,7 +324,7 @@ static struct ide_dma_ops trm290_dma_ops = { .dma_check = trm290_dma_check, }; -static const struct ide_port_info trm290_chipset __devinitdata = { +static const struct ide_port_info trm290_chipset __devinitconst = { .name = DRV_NAME, .init_hwif = init_hwif_trm290, .tp_ops = &trm290_tp_ops, diff --git a/drivers/ide/via82cxxx.c b/drivers/ide/via82cxxx.c index f46f49c..eb77678 100644 --- a/drivers/ide/via82cxxx.c +++ b/drivers/ide/via82cxxx.c @@ -403,7 +403,7 @@ static const struct ide_port_ops via_port_ops = { .cable_detect = via82cxxx_cable_detect, }; -static const struct ide_port_info via82cxxx_chipset __devinitdata = { +static const struct ide_port_info via82cxxx_chipset __devinitconst = { .name = DRV_NAME, .init_chipset = init_chipset_via82cxxx, .enablebits = { { 0x40, 0x02, 0x02 }, { 0x40, 0x01, 0x01 } }, diff --git a/drivers/ieee802154/fakehard.c b/drivers/ieee802154/fakehard.c index 73d4531..c90cd2d 100644 --- a/drivers/ieee802154/fakehard.c +++ b/drivers/ieee802154/fakehard.c @@ -386,7 +386,7 @@ static int __devinit ieee802154fake_probe(struct platform_device *pdev) phy->transmit_power = 0xbf; dev->netdev_ops = &fake_ops; - dev->ml_priv = &fake_mlme; + dev->ml_priv = (void *)&fake_mlme; priv = netdev_priv(dev); priv->phy = phy; diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index c889aae..6cf5aa7 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -114,7 +114,7 @@ static char const counter_group_names[CM_COUNTER_GROUPS] struct cm_counter_group { struct kobject obj; - atomic_long_t counter[CM_ATTR_COUNT]; + atomic_long_unchecked_t counter[CM_ATTR_COUNT]; }; struct cm_counter_attribute { @@ -1394,7 +1394,7 @@ static void cm_dup_req_handler(struct cm_work *work, struct ib_mad_send_buf *msg = NULL; int ret; - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_REQ_COUNTER]); /* Quick state check to discard duplicate REQs. */ @@ -1778,7 +1778,7 @@ static void cm_dup_rep_handler(struct cm_work *work) if (!cm_id_priv) return; - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_REP_COUNTER]); ret = cm_alloc_response_msg(work->port, work->mad_recv_wc, &msg); if (ret) @@ -1945,7 +1945,7 @@ static int cm_rtu_handler(struct cm_work *work) if (cm_id_priv->id.state != IB_CM_REP_SENT && cm_id_priv->id.state != IB_CM_MRA_REP_RCVD) { spin_unlock_irq(&cm_id_priv->lock); - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_RTU_COUNTER]); goto out; } @@ -2128,7 +2128,7 @@ static int cm_dreq_handler(struct cm_work *work) cm_id_priv = cm_acquire_id(dreq_msg->remote_comm_id, dreq_msg->local_comm_id); if (!cm_id_priv) { - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_DREQ_COUNTER]); cm_issue_drep(work->port, work->mad_recv_wc); return -EINVAL; @@ -2153,7 +2153,7 @@ static int cm_dreq_handler(struct cm_work *work) case IB_CM_MRA_REP_RCVD: break; case IB_CM_TIMEWAIT: - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_DREQ_COUNTER]); if (cm_alloc_response_msg(work->port, work->mad_recv_wc, &msg)) goto unlock; @@ -2167,7 +2167,7 @@ static int cm_dreq_handler(struct cm_work *work) cm_free_msg(msg); goto deref; case IB_CM_DREQ_RCVD: - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_DREQ_COUNTER]); goto unlock; default: @@ -2534,7 +2534,7 @@ static int cm_mra_handler(struct cm_work *work) ib_modify_mad(cm_id_priv->av.port->mad_agent, cm_id_priv->msg, timeout)) { if (cm_id_priv->id.lap_state == IB_CM_MRA_LAP_RCVD) - atomic_long_inc(&work->port-> + atomic_long_inc_unchecked(&work->port-> counter_group[CM_RECV_DUPLICATES]. counter[CM_MRA_COUNTER]); goto out; @@ -2543,7 +2543,7 @@ static int cm_mra_handler(struct cm_work *work) break; case IB_CM_MRA_REQ_RCVD: case IB_CM_MRA_REP_RCVD: - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_MRA_COUNTER]); /* fall through */ default: @@ -2705,7 +2705,7 @@ static int cm_lap_handler(struct cm_work *work) case IB_CM_LAP_IDLE: break; case IB_CM_MRA_LAP_SENT: - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_LAP_COUNTER]); if (cm_alloc_response_msg(work->port, work->mad_recv_wc, &msg)) goto unlock; @@ -2721,7 +2721,7 @@ static int cm_lap_handler(struct cm_work *work) cm_free_msg(msg); goto deref; case IB_CM_LAP_RCVD: - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_LAP_COUNTER]); goto unlock; default: @@ -3005,7 +3005,7 @@ static int cm_sidr_req_handler(struct cm_work *work) cur_cm_id_priv = cm_insert_remote_sidr(cm_id_priv); if (cur_cm_id_priv) { spin_unlock_irq(&cm.lock); - atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. + atomic_long_inc_unchecked(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_SIDR_REQ_COUNTER]); goto out; /* Duplicate message. */ } @@ -3217,10 +3217,10 @@ static void cm_send_handler(struct ib_mad_agent *mad_agent, if (!msg->context[0] && (attr_index != CM_REJ_COUNTER)) msg->retries = 1; - atomic_long_add(1 + msg->retries, + atomic_long_add_unchecked(1 + msg->retries, &port->counter_group[CM_XMIT].counter[attr_index]); if (msg->retries) - atomic_long_add(msg->retries, + atomic_long_add_unchecked(msg->retries, &port->counter_group[CM_XMIT_RETRIES]. counter[attr_index]); @@ -3430,7 +3430,7 @@ static void cm_recv_handler(struct ib_mad_agent *mad_agent, } attr_id = be16_to_cpu(mad_recv_wc->recv_buf.mad->mad_hdr.attr_id); - atomic_long_inc(&port->counter_group[CM_RECV]. + atomic_long_inc_unchecked(&port->counter_group[CM_RECV]. counter[attr_id - CM_ATTR_ID_OFFSET]); work = kmalloc(sizeof *work + sizeof(struct ib_sa_path_rec) * paths, @@ -3635,7 +3635,7 @@ static ssize_t cm_show_counter(struct kobject *obj, struct attribute *attr, cm_attr = container_of(attr, struct cm_counter_attribute, attr); return sprintf(buf, "%ld\n", - atomic_long_read(&group->counter[cm_attr->index])); + atomic_long_read_unchecked(&group->counter[cm_attr->index])); } static const struct sysfs_ops cm_counter_ops = { diff --git a/drivers/infiniband/core/fmr_pool.c b/drivers/infiniband/core/fmr_pool.c index 176c8f9..2627b62 100644 --- a/drivers/infiniband/core/fmr_pool.c +++ b/drivers/infiniband/core/fmr_pool.c @@ -98,8 +98,8 @@ struct ib_fmr_pool { struct task_struct *thread; - atomic_t req_ser; - atomic_t flush_ser; + atomic_unchecked_t req_ser; + atomic_unchecked_t flush_ser; wait_queue_head_t force_wait; }; @@ -180,10 +180,10 @@ static int ib_fmr_cleanup_thread(void *pool_ptr) struct ib_fmr_pool *pool = pool_ptr; do { - if (atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) < 0) { + if (atomic_read_unchecked(&pool->flush_ser) - atomic_read_unchecked(&pool->req_ser) < 0) { ib_fmr_batch_release(pool); - atomic_inc(&pool->flush_ser); + atomic_inc_unchecked(&pool->flush_ser); wake_up_interruptible(&pool->force_wait); if (pool->flush_function) @@ -191,7 +191,7 @@ static int ib_fmr_cleanup_thread(void *pool_ptr) } set_current_state(TASK_INTERRUPTIBLE); - if (atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) >= 0 && + if (atomic_read_unchecked(&pool->flush_ser) - atomic_read_unchecked(&pool->req_ser) >= 0 && !kthread_should_stop()) schedule(); __set_current_state(TASK_RUNNING); @@ -283,8 +283,8 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd *pd, pool->dirty_watermark = params->dirty_watermark; pool->dirty_len = 0; spin_lock_init(&pool->pool_lock); - atomic_set(&pool->req_ser, 0); - atomic_set(&pool->flush_ser, 0); + atomic_set_unchecked(&pool->req_ser, 0); + atomic_set_unchecked(&pool->flush_ser, 0); init_waitqueue_head(&pool->force_wait); pool->thread = kthread_run(ib_fmr_cleanup_thread, @@ -412,11 +412,11 @@ int ib_flush_fmr_pool(struct ib_fmr_pool *pool) } spin_unlock_irq(&pool->pool_lock); - serial = atomic_inc_return(&pool->req_ser); + serial = atomic_inc_return_unchecked(&pool->req_ser); wake_up_process(pool->thread); if (wait_event_interruptible(pool->force_wait, - atomic_read(&pool->flush_ser) - serial >= 0)) + atomic_read_unchecked(&pool->flush_ser) - serial >= 0)) return -EINTR; return 0; @@ -526,7 +526,7 @@ int ib_fmr_pool_unmap(struct ib_pool_fmr *fmr) } else { list_add_tail(&fmr->list, &pool->dirty_list); if (++pool->dirty_len >= pool->dirty_watermark) { - atomic_inc(&pool->req_ser); + atomic_inc_unchecked(&pool->req_ser); wake_up_process(pool->thread); } } diff --git a/drivers/infiniband/hw/cxgb4/mem.c b/drivers/infiniband/hw/cxgb4/mem.c index 40c8353..946b0e4 100644 --- a/drivers/infiniband/hw/cxgb4/mem.c +++ b/drivers/infiniband/hw/cxgb4/mem.c @@ -122,7 +122,7 @@ static int write_tpt_entry(struct c4iw_rdev *rdev, u32 reset_tpt_entry, int err; struct fw_ri_tpte tpt; u32 stag_idx; - static atomic_t key; + static atomic_unchecked_t key; if (c4iw_fatal_error(rdev)) return -EIO; @@ -135,7 +135,7 @@ static int write_tpt_entry(struct c4iw_rdev *rdev, u32 reset_tpt_entry, &rdev->resource.tpt_fifo_lock); if (!stag_idx) return -ENOMEM; - *stag = (stag_idx << 8) | (atomic_inc_return(&key) & 0xff); + *stag = (stag_idx << 8) | (atomic_inc_return_unchecked(&key) & 0xff); } PDBG("%s stag_state 0x%0x type 0x%0x pdid 0x%0x, stag_idx 0x%x\n", __func__, stag_state, type, pdid, stag_idx); diff --git a/drivers/infiniband/hw/ipath/ipath_fs.c b/drivers/infiniband/hw/ipath/ipath_fs.c index a4de9d5..5fa20c3 100644 --- a/drivers/infiniband/hw/ipath/ipath_fs.c +++ b/drivers/infiniband/hw/ipath/ipath_fs.c @@ -126,6 +126,8 @@ static const struct file_operations atomic_counters_ops = { }; static ssize_t flash_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) __size_overflow(3); +static ssize_t flash_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct ipath_devdata *dd; @@ -177,6 +179,8 @@ bail: } static ssize_t flash_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) __size_overflow(3); +static ssize_t flash_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { struct ipath_devdata *dd; diff --git a/drivers/infiniband/hw/ipath/ipath_rc.c b/drivers/infiniband/hw/ipath/ipath_rc.c index 79b3dbc..96e5fcc 100644 --- a/drivers/infiniband/hw/ipath/ipath_rc.c +++ b/drivers/infiniband/hw/ipath/ipath_rc.c @@ -1868,7 +1868,7 @@ void ipath_rc_rcv(struct ipath_ibdev *dev, struct ipath_ib_header *hdr, struct ib_atomic_eth *ateth; struct ipath_ack_entry *e; u64 vaddr; - atomic64_t *maddr; + atomic64_unchecked_t *maddr; u64 sdata; u32 rkey; u8 next; @@ -1903,11 +1903,11 @@ void ipath_rc_rcv(struct ipath_ibdev *dev, struct ipath_ib_header *hdr, IB_ACCESS_REMOTE_ATOMIC))) goto nack_acc_unlck; /* Perform atomic OP and save result. */ - maddr = (atomic64_t *) qp->r_sge.sge.vaddr; + maddr = (atomic64_unchecked_t *) qp->r_sge.sge.vaddr; sdata = be64_to_cpu(ateth->swap_data); e = &qp->s_ack_queue[qp->r_head_ack_queue]; e->atomic_data = (opcode == OP(FETCH_ADD)) ? - (u64) atomic64_add_return(sdata, maddr) - sdata : + (u64) atomic64_add_return_unchecked(sdata, maddr) - sdata : (u64) cmpxchg((u64 *) qp->r_sge.sge.vaddr, be64_to_cpu(ateth->compare_data), sdata); diff --git a/drivers/infiniband/hw/ipath/ipath_ruc.c b/drivers/infiniband/hw/ipath/ipath_ruc.c index 1f95bba..9530f87 100644 --- a/drivers/infiniband/hw/ipath/ipath_ruc.c +++ b/drivers/infiniband/hw/ipath/ipath_ruc.c @@ -266,7 +266,7 @@ static void ipath_ruc_loopback(struct ipath_qp *sqp) unsigned long flags; struct ib_wc wc; u64 sdata; - atomic64_t *maddr; + atomic64_unchecked_t *maddr; enum ib_wc_status send_status; /* @@ -382,11 +382,11 @@ again: IB_ACCESS_REMOTE_ATOMIC))) goto acc_err; /* Perform atomic OP and save result. */ - maddr = (atomic64_t *) qp->r_sge.sge.vaddr; + maddr = (atomic64_unchecked_t *) qp->r_sge.sge.vaddr; sdata = wqe->wr.wr.atomic.compare_add; *(u64 *) sqp->s_sge.sge.vaddr = (wqe->wr.opcode == IB_WR_ATOMIC_FETCH_AND_ADD) ? - (u64) atomic64_add_return(sdata, maddr) - sdata : + (u64) atomic64_add_return_unchecked(sdata, maddr) - sdata : (u64) cmpxchg((u64 *) qp->r_sge.sge.vaddr, sdata, wqe->wr.wr.atomic.swap); goto send_comp; diff --git a/drivers/infiniband/hw/nes/nes.c b/drivers/infiniband/hw/nes/nes.c index 7140199..da60063 100644 --- a/drivers/infiniband/hw/nes/nes.c +++ b/drivers/infiniband/hw/nes/nes.c @@ -103,7 +103,7 @@ MODULE_PARM_DESC(limit_maxrdreqsz, "Limit max read request size to 256 Bytes"); LIST_HEAD(nes_adapter_list); static LIST_HEAD(nes_dev_list); -atomic_t qps_destroyed; +atomic_unchecked_t qps_destroyed; static unsigned int ee_flsh_adapter; static unsigned int sysfs_nonidx_addr; @@ -272,7 +272,7 @@ static void nes_cqp_rem_ref_callback(struct nes_device *nesdev, struct nes_cqp_r struct nes_qp *nesqp = cqp_request->cqp_callback_pointer; struct nes_adapter *nesadapter = nesdev->nesadapter; - atomic_inc(&qps_destroyed); + atomic_inc_unchecked(&qps_destroyed); /* Free the control structures */ diff --git a/drivers/infiniband/hw/nes/nes.h b/drivers/infiniband/hw/nes/nes.h index c438e46..ca30356 100644 --- a/drivers/infiniband/hw/nes/nes.h +++ b/drivers/infiniband/hw/nes/nes.h @@ -178,17 +178,17 @@ extern unsigned int nes_debug_level; extern unsigned int wqm_quanta; extern struct list_head nes_adapter_list; -extern atomic_t cm_connects; -extern atomic_t cm_accepts; -extern atomic_t cm_disconnects; -extern atomic_t cm_closes; -extern atomic_t cm_connecteds; -extern atomic_t cm_connect_reqs; -extern atomic_t cm_rejects; -extern atomic_t mod_qp_timouts; -extern atomic_t qps_created; -extern atomic_t qps_destroyed; -extern atomic_t sw_qps_destroyed; +extern atomic_unchecked_t cm_connects; +extern atomic_unchecked_t cm_accepts; +extern atomic_unchecked_t cm_disconnects; +extern atomic_unchecked_t cm_closes; +extern atomic_unchecked_t cm_connecteds; +extern atomic_unchecked_t cm_connect_reqs; +extern atomic_unchecked_t cm_rejects; +extern atomic_unchecked_t mod_qp_timouts; +extern atomic_unchecked_t qps_created; +extern atomic_unchecked_t qps_destroyed; +extern atomic_unchecked_t sw_qps_destroyed; extern u32 mh_detected; extern u32 mh_pauses_sent; extern u32 cm_packets_sent; @@ -197,16 +197,16 @@ extern u32 cm_packets_created; extern u32 cm_packets_received; extern u32 cm_packets_dropped; extern u32 cm_packets_retrans; -extern atomic_t cm_listens_created; -extern atomic_t cm_listens_destroyed; +extern atomic_unchecked_t cm_listens_created; +extern atomic_unchecked_t cm_listens_destroyed; extern u32 cm_backlog_drops; -extern atomic_t cm_loopbacks; -extern atomic_t cm_nodes_created; -extern atomic_t cm_nodes_destroyed; -extern atomic_t cm_accel_dropped_pkts; -extern atomic_t cm_resets_recvd; -extern atomic_t pau_qps_created; -extern atomic_t pau_qps_destroyed; +extern atomic_unchecked_t cm_loopbacks; +extern atomic_unchecked_t cm_nodes_created; +extern atomic_unchecked_t cm_nodes_destroyed; +extern atomic_unchecked_t cm_accel_dropped_pkts; +extern atomic_unchecked_t cm_resets_recvd; +extern atomic_unchecked_t pau_qps_created; +extern atomic_unchecked_t pau_qps_destroyed; extern u32 int_mod_timer_init; extern u32 int_mod_cq_depth_256; diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c index a4972ab..1bcfc31 100644 --- a/drivers/infiniband/hw/nes/nes_cm.c +++ b/drivers/infiniband/hw/nes/nes_cm.c @@ -68,14 +68,14 @@ u32 cm_packets_dropped; u32 cm_packets_retrans; u32 cm_packets_created; u32 cm_packets_received; -atomic_t cm_listens_created; -atomic_t cm_listens_destroyed; +atomic_unchecked_t cm_listens_created; +atomic_unchecked_t cm_listens_destroyed; u32 cm_backlog_drops; -atomic_t cm_loopbacks; -atomic_t cm_nodes_created; -atomic_t cm_nodes_destroyed; -atomic_t cm_accel_dropped_pkts; -atomic_t cm_resets_recvd; +atomic_unchecked_t cm_loopbacks; +atomic_unchecked_t cm_nodes_created; +atomic_unchecked_t cm_nodes_destroyed; +atomic_unchecked_t cm_accel_dropped_pkts; +atomic_unchecked_t cm_resets_recvd; static inline int mini_cm_accelerated(struct nes_cm_core *, struct nes_cm_node *); static struct nes_cm_listener *mini_cm_listen(struct nes_cm_core *, struct nes_vnic *, struct nes_cm_info *); @@ -148,13 +148,13 @@ static struct nes_cm_ops nes_cm_api = { static struct nes_cm_core *g_cm_core; -atomic_t cm_connects; -atomic_t cm_accepts; -atomic_t cm_disconnects; -atomic_t cm_closes; -atomic_t cm_connecteds; -atomic_t cm_connect_reqs; -atomic_t cm_rejects; +atomic_unchecked_t cm_connects; +atomic_unchecked_t cm_accepts; +atomic_unchecked_t cm_disconnects; +atomic_unchecked_t cm_closes; +atomic_unchecked_t cm_connecteds; +atomic_unchecked_t cm_connect_reqs; +atomic_unchecked_t cm_rejects; int nes_add_ref_cm_node(struct nes_cm_node *cm_node) { @@ -1274,7 +1274,7 @@ static int mini_cm_dec_refcnt_listen(struct nes_cm_core *cm_core, kfree(listener); listener = NULL; ret = 0; - atomic_inc(&cm_listens_destroyed); + atomic_inc_unchecked(&cm_listens_destroyed); } else { spin_unlock_irqrestore(&cm_core->listen_list_lock, flags); } @@ -1473,7 +1473,7 @@ static struct nes_cm_node *make_cm_node(struct nes_cm_core *cm_core, cm_node->rem_mac); add_hte_node(cm_core, cm_node); - atomic_inc(&cm_nodes_created); + atomic_inc_unchecked(&cm_nodes_created); return cm_node; } @@ -1531,7 +1531,7 @@ static int rem_ref_cm_node(struct nes_cm_core *cm_core, } atomic_dec(&cm_core->node_cnt); - atomic_inc(&cm_nodes_destroyed); + atomic_inc_unchecked(&cm_nodes_destroyed); nesqp = cm_node->nesqp; if (nesqp) { nesqp->cm_node = NULL; @@ -1595,7 +1595,7 @@ static int process_options(struct nes_cm_node *cm_node, u8 *optionsloc, static void drop_packet(struct sk_buff *skb) { - atomic_inc(&cm_accel_dropped_pkts); + atomic_inc_unchecked(&cm_accel_dropped_pkts); dev_kfree_skb_any(skb); } @@ -1658,7 +1658,7 @@ static void handle_rst_pkt(struct nes_cm_node *cm_node, struct sk_buff *skb, { int reset = 0; /* whether to send reset in case of err.. */ - atomic_inc(&cm_resets_recvd); + atomic_inc_unchecked(&cm_resets_recvd); nes_debug(NES_DBG_CM, "Received Reset, cm_node = %p, state = %u." " refcnt=%d\n", cm_node, cm_node->state, atomic_read(&cm_node->ref_count)); @@ -2299,7 +2299,7 @@ static struct nes_cm_node *mini_cm_connect(struct nes_cm_core *cm_core, rem_ref_cm_node(cm_node->cm_core, cm_node); return NULL; } - atomic_inc(&cm_loopbacks); + atomic_inc_unchecked(&cm_loopbacks); loopbackremotenode->loopbackpartner = cm_node; loopbackremotenode->tcp_cntxt.rcv_wscale = NES_CM_DEFAULT_RCV_WND_SCALE; @@ -2574,7 +2574,7 @@ static int mini_cm_recv_pkt(struct nes_cm_core *cm_core, nes_queue_mgt_skbs(skb, nesvnic, cm_node->nesqp); else { rem_ref_cm_node(cm_core, cm_node); - atomic_inc(&cm_accel_dropped_pkts); + atomic_inc_unchecked(&cm_accel_dropped_pkts); dev_kfree_skb_any(skb); } break; @@ -2881,7 +2881,7 @@ static int nes_cm_disconn_true(struct nes_qp *nesqp) if ((cm_id) && (cm_id->event_handler)) { if (issue_disconn) { - atomic_inc(&cm_disconnects); + atomic_inc_unchecked(&cm_disconnects); cm_event.event = IW_CM_EVENT_DISCONNECT; cm_event.status = disconn_status; cm_event.local_addr = cm_id->local_addr; @@ -2903,7 +2903,7 @@ static int nes_cm_disconn_true(struct nes_qp *nesqp) } if (issue_close) { - atomic_inc(&cm_closes); + atomic_inc_unchecked(&cm_closes); nes_disconnect(nesqp, 1); cm_id->provider_data = nesqp; @@ -3039,7 +3039,7 @@ int nes_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) nes_debug(NES_DBG_CM, "QP%u, cm_node=%p, jiffies = %lu listener = %p\n", nesqp->hwqp.qp_id, cm_node, jiffies, cm_node->listener); - atomic_inc(&cm_accepts); + atomic_inc_unchecked(&cm_accepts); nes_debug(NES_DBG_CM, "netdev refcnt = %u.\n", netdev_refcnt_read(nesvnic->netdev)); @@ -3241,7 +3241,7 @@ int nes_reject(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) struct nes_cm_core *cm_core; u8 *start_buff; - atomic_inc(&cm_rejects); + atomic_inc_unchecked(&cm_rejects); cm_node = (struct nes_cm_node *)cm_id->provider_data; loopback = cm_node->loopbackpartner; cm_core = cm_node->cm_core; @@ -3301,7 +3301,7 @@ int nes_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ntohl(cm_id->local_addr.sin_addr.s_addr), ntohs(cm_id->local_addr.sin_port)); - atomic_inc(&cm_connects); + atomic_inc_unchecked(&cm_connects); nesqp->active_conn = 1; /* cache the cm_id in the qp */ @@ -3407,7 +3407,7 @@ int nes_create_listen(struct iw_cm_id *cm_id, int backlog) g_cm_core->api->stop_listener(g_cm_core, (void *)cm_node); return err; } - atomic_inc(&cm_listens_created); + atomic_inc_unchecked(&cm_listens_created); } cm_id->add_ref(cm_id); @@ -3508,7 +3508,7 @@ static void cm_event_connected(struct nes_cm_event *event) if (nesqp->destroyed) return; - atomic_inc(&cm_connecteds); + atomic_inc_unchecked(&cm_connecteds); nes_debug(NES_DBG_CM, "QP%u attempting to connect to 0x%08X:0x%04X on" " local port 0x%04X. jiffies = %lu.\n", nesqp->hwqp.qp_id, @@ -3695,7 +3695,7 @@ static void cm_event_reset(struct nes_cm_event *event) cm_id->add_ref(cm_id); ret = cm_id->event_handler(cm_id, &cm_event); - atomic_inc(&cm_closes); + atomic_inc_unchecked(&cm_closes); cm_event.event = IW_CM_EVENT_CLOSE; cm_event.status = 0; cm_event.provider_data = cm_id->provider_data; @@ -3731,7 +3731,7 @@ static void cm_event_mpa_req(struct nes_cm_event *event) return; cm_id = cm_node->cm_id; - atomic_inc(&cm_connect_reqs); + atomic_inc_unchecked(&cm_connect_reqs); nes_debug(NES_DBG_CM, "cm_node = %p - cm_id = %p, jiffies = %lu\n", cm_node, cm_id, jiffies); @@ -3771,7 +3771,7 @@ static void cm_event_mpa_reject(struct nes_cm_event *event) return; cm_id = cm_node->cm_id; - atomic_inc(&cm_connect_reqs); + atomic_inc_unchecked(&cm_connect_reqs); nes_debug(NES_DBG_CM, "cm_node = %p - cm_id = %p, jiffies = %lu\n", cm_node, cm_id, jiffies); diff --git a/drivers/infiniband/hw/nes/nes_mgt.c b/drivers/infiniband/hw/nes/nes_mgt.c index 3ba7be3..c81f6ff 100644 --- a/drivers/infiniband/hw/nes/nes_mgt.c +++ b/drivers/infiniband/hw/nes/nes_mgt.c @@ -40,8 +40,8 @@ #include "nes.h" #include "nes_mgt.h" -atomic_t pau_qps_created; -atomic_t pau_qps_destroyed; +atomic_unchecked_t pau_qps_created; +atomic_unchecked_t pau_qps_destroyed; static void nes_replenish_mgt_rq(struct nes_vnic_mgt *mgtvnic) { @@ -621,7 +621,7 @@ void nes_destroy_pau_qp(struct nes_device *nesdev, struct nes_qp *nesqp) { struct sk_buff *skb; unsigned long flags; - atomic_inc(&pau_qps_destroyed); + atomic_inc_unchecked(&pau_qps_destroyed); /* Free packets that have not yet been forwarded */ /* Lock is acquired by skb_dequeue when removing the skb */ @@ -812,7 +812,7 @@ static void nes_mgt_ce_handler(struct nes_device *nesdev, struct nes_hw_nic_cq * cq->cq_vbase[head].cqe_words[NES_NIC_CQE_HASH_RCVNXT]); skb_queue_head_init(&nesqp->pau_list); spin_lock_init(&nesqp->pau_lock); - atomic_inc(&pau_qps_created); + atomic_inc_unchecked(&pau_qps_created); nes_change_quad_hash(nesdev, mgtvnic->nesvnic, nesqp); } diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c index f3a3ecf..57d311d 100644 --- a/drivers/infiniband/hw/nes/nes_nic.c +++ b/drivers/infiniband/hw/nes/nes_nic.c @@ -1277,39 +1277,39 @@ static void nes_netdev_get_ethtool_stats(struct net_device *netdev, target_stat_values[++index] = mh_detected; target_stat_values[++index] = mh_pauses_sent; target_stat_values[++index] = nesvnic->endnode_ipv4_tcp_retransmits; - target_stat_values[++index] = atomic_read(&cm_connects); - target_stat_values[++index] = atomic_read(&cm_accepts); - target_stat_values[++index] = atomic_read(&cm_disconnects); - target_stat_values[++index] = atomic_read(&cm_connecteds); - target_stat_values[++index] = atomic_read(&cm_connect_reqs); - target_stat_values[++index] = atomic_read(&cm_rejects); - target_stat_values[++index] = atomic_read(&mod_qp_timouts); - target_stat_values[++index] = atomic_read(&qps_created); - target_stat_values[++index] = atomic_read(&sw_qps_destroyed); - target_stat_values[++index] = atomic_read(&qps_destroyed); - target_stat_values[++index] = atomic_read(&cm_closes); + target_stat_values[++index] = atomic_read_unchecked(&cm_connects); + target_stat_values[++index] = atomic_read_unchecked(&cm_accepts); + target_stat_values[++index] = atomic_read_unchecked(&cm_disconnects); + target_stat_values[++index] = atomic_read_unchecked(&cm_connecteds); + target_stat_values[++index] = atomic_read_unchecked(&cm_connect_reqs); + target_stat_values[++index] = atomic_read_unchecked(&cm_rejects); + target_stat_values[++index] = atomic_read_unchecked(&mod_qp_timouts); + target_stat_values[++index] = atomic_read_unchecked(&qps_created); + target_stat_values[++index] = atomic_read_unchecked(&sw_qps_destroyed); + target_stat_values[++index] = atomic_read_unchecked(&qps_destroyed); + target_stat_values[++index] = atomic_read_unchecked(&cm_closes); target_stat_values[++index] = cm_packets_sent; target_stat_values[++index] = cm_packets_bounced; target_stat_values[++index] = cm_packets_created; target_stat_values[++index] = cm_packets_received; target_stat_values[++index] = cm_packets_dropped; target_stat_values[++index] = cm_packets_retrans; - target_stat_values[++index] = atomic_read(&cm_listens_created); - target_stat_values[++index] = atomic_read(&cm_listens_destroyed); + target_stat_values[++index] = atomic_read_unchecked(&cm_listens_created); + target_stat_values[++index] = atomic_read_unchecked(&cm_listens_destroyed); target_stat_values[++index] = cm_backlog_drops; - target_stat_values[++index] = atomic_read(&cm_loopbacks); - target_stat_values[++index] = atomic_read(&cm_nodes_created); - target_stat_values[++index] = atomic_read(&cm_nodes_destroyed); - target_stat_values[++index] = atomic_read(&cm_accel_dropped_pkts); - target_stat_values[++index] = atomic_read(&cm_resets_recvd); + target_stat_values[++index] = atomic_read_unchecked(&cm_loopbacks); + target_stat_values[++index] = atomic_read_unchecked(&cm_nodes_created); + target_stat_values[++index] = atomic_read_unchecked(&cm_nodes_destroyed); + target_stat_values[++index] = atomic_read_unchecked(&cm_accel_dropped_pkts); + target_stat_values[++index] = atomic_read_unchecked(&cm_resets_recvd); target_stat_values[++index] = nesadapter->free_4kpbl; target_stat_values[++index] = nesadapter->free_256pbl; target_stat_values[++index] = int_mod_timer_init; target_stat_values[++index] = nesvnic->lro_mgr.stats.aggregated; target_stat_values[++index] = nesvnic->lro_mgr.stats.flushed; target_stat_values[++index] = nesvnic->lro_mgr.stats.no_desc; - target_stat_values[++index] = atomic_read(&pau_qps_created); - target_stat_values[++index] = atomic_read(&pau_qps_destroyed); + target_stat_values[++index] = atomic_read_unchecked(&pau_qps_created); + target_stat_values[++index] = atomic_read_unchecked(&pau_qps_destroyed); } /** diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c index 0927b5c..ed67986 100644 --- a/drivers/infiniband/hw/nes/nes_verbs.c +++ b/drivers/infiniband/hw/nes/nes_verbs.c @@ -46,9 +46,9 @@ #include -atomic_t mod_qp_timouts; -atomic_t qps_created; -atomic_t sw_qps_destroyed; +atomic_unchecked_t mod_qp_timouts; +atomic_unchecked_t qps_created; +atomic_unchecked_t sw_qps_destroyed; static void nes_unregister_ofa_device(struct nes_ib_device *nesibdev); @@ -1131,7 +1131,7 @@ static struct ib_qp *nes_create_qp(struct ib_pd *ibpd, if (init_attr->create_flags) return ERR_PTR(-EINVAL); - atomic_inc(&qps_created); + atomic_inc_unchecked(&qps_created); switch (init_attr->qp_type) { case IB_QPT_RC: if (nes_drv_opt & NES_DRV_OPT_NO_INLINE_DATA) { @@ -1460,7 +1460,7 @@ static int nes_destroy_qp(struct ib_qp *ibqp) struct iw_cm_event cm_event; int ret = 0; - atomic_inc(&sw_qps_destroyed); + atomic_inc_unchecked(&sw_qps_destroyed); nesqp->destroyed = 1; /* Blow away the connection if it exists. */ diff --git a/drivers/infiniband/hw/qib/qib.h b/drivers/infiniband/hw/qib/qib.h index b881bdc..c2e360c 100644 --- a/drivers/infiniband/hw/qib/qib.h +++ b/drivers/infiniband/hw/qib/qib.h @@ -51,6 +51,7 @@ #include #include #include +#include #include "qib_common.h" #include "qib_verbs.h" diff --git a/drivers/infiniband/hw/qib/qib_fs.c b/drivers/infiniband/hw/qib/qib_fs.c index 05e0f17..0275789 100644 --- a/drivers/infiniband/hw/qib/qib_fs.c +++ b/drivers/infiniband/hw/qib/qib_fs.c @@ -267,6 +267,8 @@ static const struct file_operations qsfp_ops[] = { }; static ssize_t flash_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) __size_overflow(3); +static ssize_t flash_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct qib_devdata *dd; @@ -318,6 +320,8 @@ bail: } static ssize_t flash_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) __size_overflow(3); +static ssize_t flash_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { struct qib_devdata *dd; diff --git a/drivers/input/gameport/gameport.c b/drivers/input/gameport/gameport.c index c351aa4..e6967c2 100644 --- a/drivers/input/gameport/gameport.c +++ b/drivers/input/gameport/gameport.c @@ -488,14 +488,14 @@ EXPORT_SYMBOL(gameport_set_phys); */ static void gameport_init_port(struct gameport *gameport) { - static atomic_t gameport_no = ATOMIC_INIT(0); + static atomic_unchecked_t gameport_no = ATOMIC_INIT(0); __module_get(THIS_MODULE); mutex_init(&gameport->drv_mutex); device_initialize(&gameport->dev); dev_set_name(&gameport->dev, "gameport%lu", - (unsigned long)atomic_inc_return(&gameport_no) - 1); + (unsigned long)atomic_inc_return_unchecked(&gameport_no) - 1); gameport->dev.bus = &gameport_bus; gameport->dev.release = gameport_release_port; if (gameport->parent) diff --git a/drivers/input/input.c b/drivers/input/input.c index 1f78c95..3cddc6c 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -1814,7 +1814,7 @@ static void input_cleanse_bitmasks(struct input_dev *dev) */ int input_register_device(struct input_dev *dev) { - static atomic_t input_no = ATOMIC_INIT(0); + static atomic_unchecked_t input_no = ATOMIC_INIT(0); struct input_handler *handler; const char *path; int error; @@ -1851,7 +1851,7 @@ int input_register_device(struct input_dev *dev) dev->setkeycode = input_default_setkeycode; dev_set_name(&dev->dev, "input%ld", - (unsigned long) atomic_inc_return(&input_no) - 1); + (unsigned long) atomic_inc_return_unchecked(&input_no) - 1); error = device_add(&dev->dev); if (error) diff --git a/drivers/input/joystick/sidewinder.c b/drivers/input/joystick/sidewinder.c index b8d8611..7a4a04b 100644 --- a/drivers/input/joystick/sidewinder.c +++ b/drivers/input/joystick/sidewinder.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/input/joystick/xpad.c b/drivers/input/joystick/xpad.c index fd7a0d5..a4af10c 100644 --- a/drivers/input/joystick/xpad.c +++ b/drivers/input/joystick/xpad.c @@ -710,7 +710,7 @@ static void xpad_led_set(struct led_classdev *led_cdev, static int xpad_led_probe(struct usb_xpad *xpad) { - static atomic_t led_seq = ATOMIC_INIT(0); + static atomic_unchecked_t led_seq = ATOMIC_INIT(0); long led_no; struct xpad_led *led; struct led_classdev *led_cdev; @@ -723,7 +723,7 @@ static int xpad_led_probe(struct usb_xpad *xpad) if (!led) return -ENOMEM; - led_no = (long)atomic_inc_return(&led_seq) - 1; + led_no = (long)atomic_inc_return_unchecked(&led_seq) - 1; snprintf(led->name, sizeof(led->name), "xpad%ld", led_no); led->xpad = xpad; diff --git a/drivers/input/mousedev.c b/drivers/input/mousedev.c index 0110b5a..d3ad144 100644 --- a/drivers/input/mousedev.c +++ b/drivers/input/mousedev.c @@ -763,7 +763,7 @@ static ssize_t mousedev_read(struct file *file, char __user *buffer, spin_unlock_irq(&client->packet_lock); - if (copy_to_user(buffer, data, count)) + if (count > sizeof(data) || copy_to_user(buffer, data, count)) return -EFAULT; return count; diff --git a/drivers/input/serio/serio.c b/drivers/input/serio/serio.c index ba70058..571d25d 100644 --- a/drivers/input/serio/serio.c +++ b/drivers/input/serio/serio.c @@ -497,7 +497,7 @@ static void serio_release_port(struct device *dev) */ static void serio_init_port(struct serio *serio) { - static atomic_t serio_no = ATOMIC_INIT(0); + static atomic_unchecked_t serio_no = ATOMIC_INIT(0); __module_get(THIS_MODULE); @@ -508,7 +508,7 @@ static void serio_init_port(struct serio *serio) mutex_init(&serio->drv_mutex); device_initialize(&serio->dev); dev_set_name(&serio->dev, "serio%ld", - (long)atomic_inc_return(&serio_no) - 1); + (long)atomic_inc_return_unchecked(&serio_no) - 1); serio->dev.bus = &serio_bus; serio->dev.release = serio_release_port; serio->dev.groups = serio_device_attr_groups; diff --git a/drivers/isdn/capi/capi.c b/drivers/isdn/capi/capi.c index e44933d..9ba484a 100644 --- a/drivers/isdn/capi/capi.c +++ b/drivers/isdn/capi/capi.c @@ -83,8 +83,8 @@ struct capiminor { struct capi20_appl *ap; u32 ncci; - atomic_t datahandle; - atomic_t msgid; + atomic_unchecked_t datahandle; + atomic_unchecked_t msgid; struct tty_port port; int ttyinstop; @@ -397,7 +397,7 @@ gen_data_b3_resp_for(struct capiminor *mp, struct sk_buff *skb) capimsg_setu16(s, 2, mp->ap->applid); capimsg_setu8 (s, 4, CAPI_DATA_B3); capimsg_setu8 (s, 5, CAPI_RESP); - capimsg_setu16(s, 6, atomic_inc_return(&mp->msgid)); + capimsg_setu16(s, 6, atomic_inc_return_unchecked(&mp->msgid)); capimsg_setu32(s, 8, mp->ncci); capimsg_setu16(s, 12, datahandle); } @@ -518,14 +518,14 @@ static void handle_minor_send(struct capiminor *mp) mp->outbytes -= len; spin_unlock_bh(&mp->outlock); - datahandle = atomic_inc_return(&mp->datahandle); + datahandle = atomic_inc_return_unchecked(&mp->datahandle); skb_push(skb, CAPI_DATA_B3_REQ_LEN); memset(skb->data, 0, CAPI_DATA_B3_REQ_LEN); capimsg_setu16(skb->data, 0, CAPI_DATA_B3_REQ_LEN); capimsg_setu16(skb->data, 2, mp->ap->applid); capimsg_setu8 (skb->data, 4, CAPI_DATA_B3); capimsg_setu8 (skb->data, 5, CAPI_REQ); - capimsg_setu16(skb->data, 6, atomic_inc_return(&mp->msgid)); + capimsg_setu16(skb->data, 6, atomic_inc_return_unchecked(&mp->msgid)); capimsg_setu32(skb->data, 8, mp->ncci); /* NCCI */ capimsg_setu32(skb->data, 12, (u32)(long)skb->data);/* Data32 */ capimsg_setu16(skb->data, 16, len); /* Data length */ diff --git a/drivers/isdn/gigaset/common.c b/drivers/isdn/gigaset/common.c index db621db..825ea1a 100644 --- a/drivers/isdn/gigaset/common.c +++ b/drivers/isdn/gigaset/common.c @@ -723,7 +723,7 @@ struct cardstate *gigaset_initcs(struct gigaset_driver *drv, int channels, cs->commands_pending = 0; cs->cur_at_seq = 0; cs->gotfwver = -1; - cs->open_count = 0; + local_set(&cs->open_count, 0); cs->dev = NULL; cs->tty = NULL; cs->tty_dev = NULL; diff --git a/drivers/isdn/gigaset/gigaset.h b/drivers/isdn/gigaset/gigaset.h index 212efaf..f187c6b 100644 --- a/drivers/isdn/gigaset/gigaset.h +++ b/drivers/isdn/gigaset/gigaset.h @@ -35,6 +35,7 @@ #include #include #include +#include #define GIG_VERSION {0, 5, 0, 0} #define GIG_COMPAT {0, 4, 0, 0} @@ -433,7 +434,7 @@ struct cardstate { spinlock_t cmdlock; unsigned curlen, cmdbytes; - unsigned open_count; + local_t open_count; struct tty_struct *tty; struct tasklet_struct if_wake_tasklet; unsigned control_state; diff --git a/drivers/isdn/gigaset/interface.c b/drivers/isdn/gigaset/interface.c index ee0a549..a7c9798 100644 --- a/drivers/isdn/gigaset/interface.c +++ b/drivers/isdn/gigaset/interface.c @@ -163,9 +163,7 @@ static int if_open(struct tty_struct *tty, struct file *filp) } tty->driver_data = cs; - ++cs->open_count; - - if (cs->open_count == 1) { + if (local_inc_return(&cs->open_count) == 1) { spin_lock_irqsave(&cs->lock, flags); cs->tty = tty; spin_unlock_irqrestore(&cs->lock, flags); @@ -193,10 +191,10 @@ static void if_close(struct tty_struct *tty, struct file *filp) if (!cs->connected) gig_dbg(DEBUG_IF, "not connected"); /* nothing to do */ - else if (!cs->open_count) + else if (!local_read(&cs->open_count)) dev_warn(cs->dev, "%s: device not opened\n", __func__); else { - if (!--cs->open_count) { + if (!local_dec_return(&cs->open_count)) { spin_lock_irqsave(&cs->lock, flags); cs->tty = NULL; spin_unlock_irqrestore(&cs->lock, flags); @@ -231,7 +229,7 @@ static int if_ioctl(struct tty_struct *tty, if (!cs->connected) { gig_dbg(DEBUG_IF, "not connected"); retval = -ENODEV; - } else if (!cs->open_count) + } else if (!local_read(&cs->open_count)) dev_warn(cs->dev, "%s: device not opened\n", __func__); else { retval = 0; @@ -361,7 +359,7 @@ static int if_write(struct tty_struct *tty, const unsigned char *buf, int count) retval = -ENODEV; goto done; } - if (!cs->open_count) { + if (!local_read(&cs->open_count)) { dev_warn(cs->dev, "%s: device not opened\n", __func__); retval = -ENODEV; goto done; @@ -414,7 +412,7 @@ static int if_write_room(struct tty_struct *tty) if (!cs->connected) { gig_dbg(DEBUG_IF, "not connected"); retval = -ENODEV; - } else if (!cs->open_count) + } else if (!local_read(&cs->open_count)) dev_warn(cs->dev, "%s: device not opened\n", __func__); else if (cs->mstate != MS_LOCKED) { dev_warn(cs->dev, "can't write to unlocked device\n"); @@ -444,7 +442,7 @@ static int if_chars_in_buffer(struct tty_struct *tty) if (!cs->connected) gig_dbg(DEBUG_IF, "not connected"); - else if (!cs->open_count) + else if (!local_read(&cs->open_count)) dev_warn(cs->dev, "%s: device not opened\n", __func__); else if (cs->mstate != MS_LOCKED) dev_warn(cs->dev, "can't write to unlocked device\n"); @@ -472,7 +470,7 @@ static void if_throttle(struct tty_struct *tty) if (!cs->connected) gig_dbg(DEBUG_IF, "not connected"); /* nothing to do */ - else if (!cs->open_count) + else if (!local_read(&cs->open_count)) dev_warn(cs->dev, "%s: device not opened\n", __func__); else gig_dbg(DEBUG_IF, "%s: not implemented\n", __func__); @@ -496,7 +494,7 @@ static void if_unthrottle(struct tty_struct *tty) if (!cs->connected) gig_dbg(DEBUG_IF, "not connected"); /* nothing to do */ - else if (!cs->open_count) + else if (!local_read(&cs->open_count)) dev_warn(cs->dev, "%s: device not opened\n", __func__); else gig_dbg(DEBUG_IF, "%s: not implemented\n", __func__); @@ -527,7 +525,7 @@ static void if_set_termios(struct tty_struct *tty, struct ktermios *old) goto out; } - if (!cs->open_count) { + if (!local_read(&cs->open_count)) { dev_warn(cs->dev, "%s: device not opened\n", __func__); goto out; } diff --git a/drivers/isdn/hardware/avm/b1.c b/drivers/isdn/hardware/avm/b1.c index 2a57da59..e7a12ed 100644 --- a/drivers/isdn/hardware/avm/b1.c +++ b/drivers/isdn/hardware/avm/b1.c @@ -176,7 +176,7 @@ int b1_load_t4file(avmcard *card, capiloaddatapart * t4file) } if (left) { if (t4file->user) { - if (copy_from_user(buf, dp, left)) + if (left > sizeof buf || copy_from_user(buf, dp, left)) return -EFAULT; } else { memcpy(buf, dp, left); @@ -224,7 +224,7 @@ int b1_load_config(avmcard *card, capiloaddatapart * config) } if (left) { if (config->user) { - if (copy_from_user(buf, dp, left)) + if (left > sizeof buf || copy_from_user(buf, dp, left)) return -EFAULT; } else { memcpy(buf, dp, left); diff --git a/drivers/isdn/hardware/eicon/divasync.h b/drivers/isdn/hardware/eicon/divasync.h index 85784a7..a19ca98 100644 --- a/drivers/isdn/hardware/eicon/divasync.h +++ b/drivers/isdn/hardware/eicon/divasync.h @@ -146,7 +146,7 @@ typedef struct _diva_didd_add_adapter { } diva_didd_add_adapter_t; typedef struct _diva_didd_remove_adapter { IDI_CALL p_request; -} diva_didd_remove_adapter_t; +} __no_const diva_didd_remove_adapter_t; typedef struct _diva_didd_read_adapter_array { void * buffer; dword length; diff --git a/drivers/isdn/hardware/eicon/xdi_adapter.h b/drivers/isdn/hardware/eicon/xdi_adapter.h index a3bd163..8956575 100644 --- a/drivers/isdn/hardware/eicon/xdi_adapter.h +++ b/drivers/isdn/hardware/eicon/xdi_adapter.h @@ -44,7 +44,7 @@ typedef struct _xdi_mbox_t { typedef struct _diva_os_idi_adapter_interface { diva_init_card_proc_t cleanup_adapter_proc; diva_cmd_card_proc_t cmd_proc; -} diva_os_idi_adapter_interface_t; +} __no_const diva_os_idi_adapter_interface_t; typedef struct _diva_os_xdi_adapter { struct list_head link; diff --git a/drivers/isdn/icn/icn.c b/drivers/isdn/icn/icn.c index 1f355bb..43f1fea 100644 --- a/drivers/isdn/icn/icn.c +++ b/drivers/isdn/icn/icn.c @@ -1045,7 +1045,7 @@ icn_writecmd(const u_char * buf, int len, int user, icn_card * card) if (count > len) count = len; if (user) { - if (copy_from_user(msg, buf, count)) + if (count > sizeof msg || copy_from_user(msg, buf, count)) return -EFAULT; } else memcpy(msg, buf, count); diff --git a/drivers/lguest/core.c b/drivers/lguest/core.c index b5fdcb7..5b6c59f 100644 --- a/drivers/lguest/core.c +++ b/drivers/lguest/core.c @@ -92,9 +92,17 @@ static __init int map_switcher(void) * it's worked so far. The end address needs +1 because __get_vm_area * allocates an extra guard page, so we need space for that. */ + +#if defined(CONFIG_MODULES) && defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + switcher_vma = __get_vm_area(TOTAL_SWITCHER_PAGES * PAGE_SIZE, + VM_ALLOC | VM_KERNEXEC, SWITCHER_ADDR, SWITCHER_ADDR + + (TOTAL_SWITCHER_PAGES+1) * PAGE_SIZE); +#else switcher_vma = __get_vm_area(TOTAL_SWITCHER_PAGES * PAGE_SIZE, VM_ALLOC, SWITCHER_ADDR, SWITCHER_ADDR + (TOTAL_SWITCHER_PAGES+1) * PAGE_SIZE); +#endif + if (!switcher_vma) { err = -ENOMEM; printk("lguest: could not map switcher pages high\n"); @@ -119,7 +127,7 @@ static __init int map_switcher(void) * Now the Switcher is mapped at the right address, we can't fail! * Copy in the compiled-in Switcher code (from x86/switcher_32.S). */ - memcpy(switcher_vma->addr, start_switcher_text, + memcpy(switcher_vma->addr, ktla_ktva(start_switcher_text), end_switcher_text - start_switcher_text); printk(KERN_INFO "lguest: mapped switcher at %p\n", diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c index ff4a0bc..f5fdd9c 100644 --- a/drivers/lguest/lguest_user.c +++ b/drivers/lguest/lguest_user.c @@ -198,6 +198,7 @@ static int user_send_irq(struct lg_cpu *cpu, const unsigned long __user *input) * Once our Guest is initialized, the Launcher makes it run by reading * from /dev/lguest. */ +static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o) __size_overflow(3); static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o) { struct lguest *lg = file->private_data; diff --git a/drivers/lguest/x86/core.c b/drivers/lguest/x86/core.c index 3980903..ce25c5e 100644 --- a/drivers/lguest/x86/core.c +++ b/drivers/lguest/x86/core.c @@ -59,7 +59,7 @@ static struct { /* Offset from where switcher.S was compiled to where we've copied it */ static unsigned long switcher_offset(void) { - return SWITCHER_ADDR - (unsigned long)start_switcher_text; + return SWITCHER_ADDR - (unsigned long)ktla_ktva(start_switcher_text); } /* This cpu's struct lguest_pages. */ @@ -100,7 +100,13 @@ static void copy_in_guest_info(struct lg_cpu *cpu, struct lguest_pages *pages) * These copies are pretty cheap, so we do them unconditionally: */ /* Save the current Host top-level page directory. */ + +#ifdef CONFIG_PAX_PER_CPU_PGD + pages->state.host_cr3 = read_cr3(); +#else pages->state.host_cr3 = __pa(current->mm->pgd); +#endif + /* * Set up the Guest's page tables to see this CPU's pages (and no * other CPU's pages). @@ -472,7 +478,7 @@ void __init lguest_arch_host_init(void) * compiled-in switcher code and the high-mapped copy we just made. */ for (i = 0; i < IDT_ENTRIES; i++) - default_idt_entries[i] += switcher_offset(); + default_idt_entries[i] = ktla_ktva(default_idt_entries[i]) + switcher_offset(); /* * Set up the Switcher's per-cpu areas. @@ -555,7 +561,7 @@ void __init lguest_arch_host_init(void) * it will be undisturbed when we switch. To change %cs and jump we * need this structure to feed to Intel's "lcall" instruction. */ - lguest_entry.offset = (long)switch_to_guest + switcher_offset(); + lguest_entry.offset = (long)ktla_ktva(switch_to_guest) + switcher_offset(); lguest_entry.segment = LGUEST_CS; /* diff --git a/drivers/lguest/x86/switcher_32.S b/drivers/lguest/x86/switcher_32.S index 40634b0..4f5855e 100644 --- a/drivers/lguest/x86/switcher_32.S +++ b/drivers/lguest/x86/switcher_32.S @@ -87,6 +87,7 @@ #include #include #include +#include // We mark the start of the code to copy // It's placed in .text tho it's never run here @@ -149,6 +150,13 @@ ENTRY(switch_to_guest) // Changes type when we load it: damn Intel! // For after we switch over our page tables // That entry will be read-only: we'd crash. + +#ifdef CONFIG_PAX_KERNEXEC + mov %cr0, %edx + xor $X86_CR0_WP, %edx + mov %edx, %cr0 +#endif + movl $(GDT_ENTRY_TSS*8), %edx ltr %dx @@ -157,9 +165,15 @@ ENTRY(switch_to_guest) // Let's clear it again for our return. // The GDT descriptor of the Host // Points to the table after two "size" bytes - movl (LGUEST_PAGES_host_gdt_desc+2)(%eax), %edx + movl (LGUEST_PAGES_host_gdt_desc+2)(%eax), %eax // Clear "used" from type field (byte 5, bit 2) - andb $0xFD, (GDT_ENTRY_TSS*8 + 5)(%edx) + andb $0xFD, (GDT_ENTRY_TSS*8 + 5)(%eax) + +#ifdef CONFIG_PAX_KERNEXEC + mov %cr0, %eax + xor $X86_CR0_WP, %eax + mov %eax, %cr0 +#endif // Once our page table's switched, the Guest is live! // The Host fades as we run this final step. @@ -295,13 +309,12 @@ deliver_to_host: // I consulted gcc, and it gave // These instructions, which I gladly credit: leal (%edx,%ebx,8), %eax - movzwl (%eax),%edx - movl 4(%eax), %eax - xorw %ax, %ax - orl %eax, %edx + movl 4(%eax), %edx + movw (%eax), %dx // Now the address of the handler's in %edx // We call it now: its "iret" drops us home. - jmp *%edx + ljmp $__KERNEL_CS, $1f +1: jmp *%edx // Every interrupt can come to us here // But we must truly tell each apart. diff --git a/drivers/macintosh/macio_asic.c b/drivers/macintosh/macio_asic.c index 4daf9e5..b8d1d0f 100644 --- a/drivers/macintosh/macio_asic.c +++ b/drivers/macintosh/macio_asic.c @@ -748,7 +748,7 @@ static void __devexit macio_pci_remove(struct pci_dev* pdev) * MacIO is matched against any Apple ID, it's probe() function * will then decide wether it applies or not */ -static const struct pci_device_id __devinitdata pci_ids [] = { { +static const struct pci_device_id __devinitconst pci_ids [] = { { .vendor = PCI_VENDOR_ID_APPLE, .device = PCI_ANY_ID, .subvendor = PCI_ANY_ID, diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 1ce84ed..0fdd40a 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -1589,7 +1589,7 @@ static int validate_params(uint cmd, struct dm_ioctl *param) cmd == DM_LIST_VERSIONS_CMD) return 0; - if ((cmd == DM_DEV_CREATE_CMD)) { + if (cmd == DM_DEV_CREATE_CMD) { if (!*param->name) { DMWARN("name not supplied when creating device"); return -EINVAL; diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c index 9bfd057..5373ff3 100644 --- a/drivers/md/dm-raid1.c +++ b/drivers/md/dm-raid1.c @@ -40,7 +40,7 @@ enum dm_raid1_error { struct mirror { struct mirror_set *ms; - atomic_t error_count; + atomic_unchecked_t error_count; unsigned long error_type; struct dm_dev *dev; sector_t offset; @@ -185,7 +185,7 @@ static struct mirror *get_valid_mirror(struct mirror_set *ms) struct mirror *m; for (m = ms->mirror; m < ms->mirror + ms->nr_mirrors; m++) - if (!atomic_read(&m->error_count)) + if (!atomic_read_unchecked(&m->error_count)) return m; return NULL; @@ -217,7 +217,7 @@ static void fail_mirror(struct mirror *m, enum dm_raid1_error error_type) * simple way to tell if a device has encountered * errors. */ - atomic_inc(&m->error_count); + atomic_inc_unchecked(&m->error_count); if (test_and_set_bit(error_type, &m->error_type)) return; @@ -408,7 +408,7 @@ static struct mirror *choose_mirror(struct mirror_set *ms, sector_t sector) struct mirror *m = get_default_mirror(ms); do { - if (likely(!atomic_read(&m->error_count))) + if (likely(!atomic_read_unchecked(&m->error_count))) return m; if (m-- == ms->mirror) @@ -422,7 +422,7 @@ static int default_ok(struct mirror *m) { struct mirror *default_mirror = get_default_mirror(m->ms); - return !atomic_read(&default_mirror->error_count); + return !atomic_read_unchecked(&default_mirror->error_count); } static int mirror_available(struct mirror_set *ms, struct bio *bio) @@ -559,7 +559,7 @@ static void do_reads(struct mirror_set *ms, struct bio_list *reads) */ if (likely(region_in_sync(ms, region, 1))) m = choose_mirror(ms, bio->bi_sector); - else if (m && atomic_read(&m->error_count)) + else if (m && atomic_read_unchecked(&m->error_count)) m = NULL; if (likely(m)) @@ -848,6 +848,10 @@ static void do_mirror(struct work_struct *work) static struct mirror_set *alloc_context(unsigned int nr_mirrors, uint32_t region_size, struct dm_target *ti, + struct dm_dirty_log *dl) __size_overflow(1); +static struct mirror_set *alloc_context(unsigned int nr_mirrors, + uint32_t region_size, + struct dm_target *ti, struct dm_dirty_log *dl) { size_t len; @@ -937,7 +941,7 @@ static int get_mirror(struct mirror_set *ms, struct dm_target *ti, } ms->mirror[mirror].ms = ms; - atomic_set(&(ms->mirror[mirror].error_count), 0); + atomic_set_unchecked(&(ms->mirror[mirror].error_count), 0); ms->mirror[mirror].error_type = 0; ms->mirror[mirror].offset = offset; @@ -1347,7 +1351,7 @@ static void mirror_resume(struct dm_target *ti) */ static char device_status_char(struct mirror *m) { - if (!atomic_read(&(m->error_count))) + if (!atomic_read_unchecked(&(m->error_count))) return 'A'; return (test_bit(DM_RAID1_FLUSH_ERROR, &(m->error_type))) ? 'F' : diff --git a/drivers/md/dm-stripe.c b/drivers/md/dm-stripe.c index 3d80cf0..7d98e1a 100644 --- a/drivers/md/dm-stripe.c +++ b/drivers/md/dm-stripe.c @@ -20,7 +20,7 @@ struct stripe { struct dm_dev *dev; sector_t physical_start; - atomic_t error_count; + atomic_unchecked_t error_count; }; struct stripe_c { @@ -55,6 +55,7 @@ static void trigger_event(struct work_struct *work) dm_table_event(sc->ti->table); } +static inline struct stripe_c *alloc_context(unsigned int stripes) __size_overflow(1); static inline struct stripe_c *alloc_context(unsigned int stripes) { size_t len; @@ -192,7 +193,7 @@ static int stripe_ctr(struct dm_target *ti, unsigned int argc, char **argv) kfree(sc); return r; } - atomic_set(&(sc->stripe[i].error_count), 0); + atomic_set_unchecked(&(sc->stripe[i].error_count), 0); } ti->private = sc; @@ -314,7 +315,7 @@ static int stripe_status(struct dm_target *ti, DMEMIT("%d ", sc->stripes); for (i = 0; i < sc->stripes; i++) { DMEMIT("%s ", sc->stripe[i].dev->name); - buffer[i] = atomic_read(&(sc->stripe[i].error_count)) ? + buffer[i] = atomic_read_unchecked(&(sc->stripe[i].error_count)) ? 'D' : 'A'; } buffer[i] = '\0'; @@ -361,8 +362,8 @@ static int stripe_end_io(struct dm_target *ti, struct bio *bio, */ for (i = 0; i < sc->stripes; i++) if (!strcmp(sc->stripe[i].dev->name, major_minor)) { - atomic_inc(&(sc->stripe[i].error_count)); - if (atomic_read(&(sc->stripe[i].error_count)) < + atomic_inc_unchecked(&(sc->stripe[i].error_count)); + if (atomic_read_unchecked(&(sc->stripe[i].error_count)) < DM_IO_ERROR_THRESHOLD) schedule_work(&sc->trigger_event); } diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c index 63cc542..8d45caf3 100644 --- a/drivers/md/dm-table.c +++ b/drivers/md/dm-table.c @@ -391,7 +391,7 @@ static int device_area_is_invalid(struct dm_target *ti, struct dm_dev *dev, if (!dev_size) return 0; - if ((start >= dev_size) || (start + len > dev_size)) { + if ((start >= dev_size) || (len > dev_size - start)) { DMWARN("%s: %s too small for target: " "start=%llu, len=%llu, dev_size=%llu", dm_device_name(ti->table->md), bdevname(bdev, b), diff --git a/drivers/md/dm-thin-metadata.c b/drivers/md/dm-thin-metadata.c index 237571a..fb6d19b 100644 --- a/drivers/md/dm-thin-metadata.c +++ b/drivers/md/dm-thin-metadata.c @@ -432,7 +432,7 @@ static int init_pmd(struct dm_pool_metadata *pmd, pmd->info.tm = tm; pmd->info.levels = 2; - pmd->info.value_type.context = pmd->data_sm; + pmd->info.value_type.context = (dm_space_map_no_const *)pmd->data_sm; pmd->info.value_type.size = sizeof(__le64); pmd->info.value_type.inc = data_block_inc; pmd->info.value_type.dec = data_block_dec; @@ -451,7 +451,7 @@ static int init_pmd(struct dm_pool_metadata *pmd, pmd->bl_info.tm = tm; pmd->bl_info.levels = 1; - pmd->bl_info.value_type.context = pmd->data_sm; + pmd->bl_info.value_type.context = (dm_space_map_no_const *)pmd->data_sm; pmd->bl_info.value_type.size = sizeof(__le64); pmd->bl_info.value_type.inc = data_block_inc; pmd->bl_info.value_type.dec = data_block_dec; diff --git a/drivers/md/dm.c b/drivers/md/dm.c index b89c548..2af3ce4 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -176,9 +176,9 @@ struct mapped_device { /* * Event handling. */ - atomic_t event_nr; + atomic_unchecked_t event_nr; wait_queue_head_t eventq; - atomic_t uevent_seq; + atomic_unchecked_t uevent_seq; struct list_head uevent_list; spinlock_t uevent_lock; /* Protect access to uevent_list */ @@ -1844,8 +1844,8 @@ static struct mapped_device *alloc_dev(int minor) rwlock_init(&md->map_lock); atomic_set(&md->holders, 1); atomic_set(&md->open_count, 0); - atomic_set(&md->event_nr, 0); - atomic_set(&md->uevent_seq, 0); + atomic_set_unchecked(&md->event_nr, 0); + atomic_set_unchecked(&md->uevent_seq, 0); INIT_LIST_HEAD(&md->uevent_list); spin_lock_init(&md->uevent_lock); @@ -1979,7 +1979,7 @@ static void event_callback(void *context) dm_send_uevents(&uevents, &disk_to_dev(md->disk)->kobj); - atomic_inc(&md->event_nr); + atomic_inc_unchecked(&md->event_nr); wake_up(&md->eventq); } @@ -2621,18 +2621,18 @@ int dm_kobject_uevent(struct mapped_device *md, enum kobject_action action, uint32_t dm_next_uevent_seq(struct mapped_device *md) { - return atomic_add_return(1, &md->uevent_seq); + return atomic_add_return_unchecked(1, &md->uevent_seq); } uint32_t dm_get_event_nr(struct mapped_device *md) { - return atomic_read(&md->event_nr); + return atomic_read_unchecked(&md->event_nr); } int dm_wait_event(struct mapped_device *md, int event_nr) { return wait_event_interruptible(md->eventq, - (event_nr != atomic_read(&md->event_nr))); + (event_nr != atomic_read_unchecked(&md->event_nr))); } void dm_uevent_add(struct mapped_device *md, struct list_head *elist) diff --git a/drivers/md/md.c b/drivers/md/md.c index 58027d8..d9cddcd 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -277,10 +277,10 @@ EXPORT_SYMBOL_GPL(md_trim_bio); * start build, activate spare */ static DECLARE_WAIT_QUEUE_HEAD(md_event_waiters); -static atomic_t md_event_count; +static atomic_unchecked_t md_event_count; void md_new_event(struct mddev *mddev) { - atomic_inc(&md_event_count); + atomic_inc_unchecked(&md_event_count); wake_up(&md_event_waiters); } EXPORT_SYMBOL_GPL(md_new_event); @@ -290,7 +290,7 @@ EXPORT_SYMBOL_GPL(md_new_event); */ static void md_new_event_inintr(struct mddev *mddev) { - atomic_inc(&md_event_count); + atomic_inc_unchecked(&md_event_count); wake_up(&md_event_waiters); } @@ -1524,7 +1524,7 @@ static int super_1_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor_ rdev->preferred_minor = 0xffff; rdev->data_offset = le64_to_cpu(sb->data_offset); - atomic_set(&rdev->corrected_errors, le32_to_cpu(sb->cnt_corrected_read)); + atomic_set_unchecked(&rdev->corrected_errors, le32_to_cpu(sb->cnt_corrected_read)); rdev->sb_size = le32_to_cpu(sb->max_dev) * 2 + 256; bmask = queue_logical_block_size(rdev->bdev->bd_disk->queue)-1; @@ -1743,7 +1743,7 @@ static void super_1_sync(struct mddev *mddev, struct md_rdev *rdev) else sb->resync_offset = cpu_to_le64(0); - sb->cnt_corrected_read = cpu_to_le32(atomic_read(&rdev->corrected_errors)); + sb->cnt_corrected_read = cpu_to_le32(atomic_read_unchecked(&rdev->corrected_errors)); sb->raid_disks = cpu_to_le32(mddev->raid_disks); sb->size = cpu_to_le64(mddev->dev_sectors); @@ -2689,7 +2689,7 @@ __ATTR(state, S_IRUGO|S_IWUSR, state_show, state_store); static ssize_t errors_show(struct md_rdev *rdev, char *page) { - return sprintf(page, "%d\n", atomic_read(&rdev->corrected_errors)); + return sprintf(page, "%d\n", atomic_read_unchecked(&rdev->corrected_errors)); } static ssize_t @@ -2698,7 +2698,7 @@ errors_store(struct md_rdev *rdev, const char *buf, size_t len) char *e; unsigned long n = simple_strtoul(buf, &e, 10); if (*buf && (*e == 0 || *e == '\n')) { - atomic_set(&rdev->corrected_errors, n); + atomic_set_unchecked(&rdev->corrected_errors, n); return len; } return -EINVAL; @@ -3084,8 +3084,8 @@ int md_rdev_init(struct md_rdev *rdev) rdev->sb_loaded = 0; rdev->bb_page = NULL; atomic_set(&rdev->nr_pending, 0); - atomic_set(&rdev->read_errors, 0); - atomic_set(&rdev->corrected_errors, 0); + atomic_set_unchecked(&rdev->read_errors, 0); + atomic_set_unchecked(&rdev->corrected_errors, 0); INIT_LIST_HEAD(&rdev->same_set); init_waitqueue_head(&rdev->blocked_wait); @@ -6736,7 +6736,7 @@ static int md_seq_show(struct seq_file *seq, void *v) spin_unlock(&pers_lock); seq_printf(seq, "\n"); - seq->poll_event = atomic_read(&md_event_count); + seq->poll_event = atomic_read_unchecked(&md_event_count); return 0; } if (v == (void*)2) { @@ -6828,7 +6828,7 @@ static int md_seq_show(struct seq_file *seq, void *v) chunk_kb ? "KB" : "B"); if (bitmap->file) { seq_printf(seq, ", file: "); - seq_path(seq, &bitmap->file->f_path, " \t\n"); + seq_path(seq, &bitmap->file->f_path, " \t\n\\"); } seq_printf(seq, "\n"); @@ -6859,7 +6859,7 @@ static int md_seq_open(struct inode *inode, struct file *file) return error; seq = file->private_data; - seq->poll_event = atomic_read(&md_event_count); + seq->poll_event = atomic_read_unchecked(&md_event_count); return error; } @@ -6873,7 +6873,7 @@ static unsigned int mdstat_poll(struct file *filp, poll_table *wait) /* always allow read */ mask = POLLIN | POLLRDNORM; - if (seq->poll_event != atomic_read(&md_event_count)) + if (seq->poll_event != atomic_read_unchecked(&md_event_count)) mask |= POLLERR | POLLPRI; return mask; } @@ -6917,7 +6917,7 @@ static int is_mddev_idle(struct mddev *mddev, int init) struct gendisk *disk = rdev->bdev->bd_contains->bd_disk; curr_events = (int)part_stat_read(&disk->part0, sectors[0]) + (int)part_stat_read(&disk->part0, sectors[1]) - - atomic_read(&disk->sync_io); + atomic_read_unchecked(&disk->sync_io); /* sync IO will cause sync_io to increase before the disk_stats * as sync_io is counted when a request starts, and * disk_stats is counted when it completes. diff --git a/drivers/md/md.h b/drivers/md/md.h index 44c63df..b795d1a 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -93,13 +93,13 @@ struct md_rdev { * only maintained for arrays that * support hot removal */ - atomic_t read_errors; /* number of consecutive read errors that + atomic_unchecked_t read_errors; /* number of consecutive read errors that * we have tried to ignore. */ struct timespec last_read_error; /* monotonic time since our * last read error */ - atomic_t corrected_errors; /* number of corrected read errors, + atomic_unchecked_t corrected_errors; /* number of corrected read errors, * for reporting to userspace and storing * in superblock. */ @@ -421,7 +421,7 @@ static inline void rdev_dec_pending(struct md_rdev *rdev, struct mddev *mddev) static inline void md_sync_acct(struct block_device *bdev, unsigned long nr_sectors) { - atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io); + atomic_add_unchecked(nr_sectors, &bdev->bd_contains->bd_disk->sync_io); } struct md_personality diff --git a/drivers/md/persistent-data/dm-space-map-checker.c b/drivers/md/persistent-data/dm-space-map-checker.c index 50ed53b..4f29d7d 100644 --- a/drivers/md/persistent-data/dm-space-map-checker.c +++ b/drivers/md/persistent-data/dm-space-map-checker.c @@ -159,7 +159,7 @@ static void ca_destroy(struct count_array *ca) /*----------------------------------------------------------------*/ struct sm_checker { - struct dm_space_map sm; + dm_space_map_no_const sm; struct count_array old_counts; struct count_array counts; diff --git a/drivers/md/persistent-data/dm-space-map-disk.c b/drivers/md/persistent-data/dm-space-map-disk.c index fc469ba..2d91555 100644 --- a/drivers/md/persistent-data/dm-space-map-disk.c +++ b/drivers/md/persistent-data/dm-space-map-disk.c @@ -23,7 +23,7 @@ * Space map interface. */ struct sm_disk { - struct dm_space_map sm; + dm_space_map_no_const sm; struct ll_disk ll; struct ll_disk old_ll; diff --git a/drivers/md/persistent-data/dm-space-map-metadata.c b/drivers/md/persistent-data/dm-space-map-metadata.c index e89ae5e..062e4c2 100644 --- a/drivers/md/persistent-data/dm-space-map-metadata.c +++ b/drivers/md/persistent-data/dm-space-map-metadata.c @@ -43,7 +43,7 @@ struct block_op { }; struct sm_metadata { - struct dm_space_map sm; + dm_space_map_no_const sm; struct ll_disk ll; struct ll_disk old_ll; diff --git a/drivers/md/persistent-data/dm-space-map.h b/drivers/md/persistent-data/dm-space-map.h index 1cbfc6b..56e1dbb 100644 --- a/drivers/md/persistent-data/dm-space-map.h +++ b/drivers/md/persistent-data/dm-space-map.h @@ -60,6 +60,7 @@ struct dm_space_map { int (*root_size)(struct dm_space_map *sm, size_t *result); int (*copy_root)(struct dm_space_map *sm, void *copy_to_here_le,