diff --git a/Documentation/dontdiff b/Documentation/dontdiff index 9de9813..1462492 100644 --- a/Documentation/dontdiff +++ b/Documentation/dontdiff @@ -3,9 +3,11 @@ *.bc *.bin *.bz2 +*.c.[012]*.* *.cis *.cpio *.csp +*.dbg *.dsp *.dvi *.elf @@ -15,6 +17,7 @@ *.gcov *.gen.S *.gif +*.gmo *.grep *.grp *.gz @@ -51,14 +54,17 @@ *.tab.h *.tex *.ver +*.vim *.xml *.xz *_MODULES +*_reg_safe.h *_vga16.c *~ \#*# *.9 -.* +.[^g]* +.gen* .*.d .mm 53c700_d.h @@ -72,9 +78,11 @@ Image Module.markers Module.symvers PENDING +PERF* SCCS System.map* TAGS +TRACEEVENT-CFLAGS aconf af_names.h aic7*reg.h* @@ -83,6 +91,7 @@ aic7*seq.h* aicasm aicdb.h* altivec*.c +ashldi3.S asm-offsets.h asm_offsets.h autoconf.h* @@ -95,32 +104,40 @@ 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 crc32table.h* cscope.* defkeymap.c +devicetable-offsets.h 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 @@ -128,12 +145,15 @@ 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 @@ -148,14 +168,14 @@ int32.c int4.c int8.c kallsyms -kconfig +kern_constants.h keywords.c ksym.c* ksym.h* kxgettext lex.c lex.*.c -linux +lib1funcs.S logo_*.c logo_*_clut224.c logo_*_mono.c @@ -165,14 +185,15 @@ mach-types.h machtypes.h map map_hugetlb -media mconf +mdp miboot* mk_elfconfig mkboot mkbugboot mkcpustr mkdep +mkpiggy mkprep mkregtable mktables @@ -188,6 +209,8 @@ oui.c* page-types parse.c parse.h +parse-events* +pasyms.h patches* pca200e.bin pca200e_ecd.bin2 @@ -197,6 +220,7 @@ perf-archive piggyback piggy.gzip piggy.S +pmu-* pnmtologo ppc_defs.h* pss_boot.h @@ -206,7 +230,12 @@ r200_reg_safe.h r300_reg_safe.h r420_reg_safe.h r600_reg_safe.h +randomize_layout_hash.h +randomize_layout_seed.h +realmode.lds +realmode.relocs recordmcount +regdb.c relocs rlim_names.h rn50_reg_safe.h @@ -216,8 +245,12 @@ series setup setup.bin setup.elf +signing_key* +size_overflow_hash.h sImage +slabinfo sm_tbl* +sortextable split-include syscalltab.h tables.c @@ -227,6 +260,7 @@ tftpboot.img timeconst.h times.h* trix_boot.h +user_constants.h utsrelease.h* vdso-syms.lds vdso.lds @@ -238,13 +272,17 @@ vdso32.lds vdso32.so.dbg vdso64.lds vdso64.so.dbg +vdsox32.lds +vdsox32-syms.lds version.h* vmImage vmlinux vmlinux-* vmlinux.aout vmlinux.bin.all +vmlinux.bin.bz2 vmlinux.lds +vmlinux.relocs vmlinuz voffset.h vsyscall.lds @@ -252,9 +290,12 @@ vsyscall_32.lds wanxlfw.inc uImage unifdef +utsrelease.h wakeup.bin wakeup.elf wakeup.lds +x509* zImage* zconf.hash.c +zconf.lex.c zoffset.h diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt index a1d0d7a..61d65cc 100644 --- a/Documentation/filesystems/vfs.txt +++ b/Documentation/filesystems/vfs.txt @@ -1053,7 +1053,8 @@ struct dentry_operations { If the 'rcu_walk' parameter is true, then the caller is doing a pathwalk in RCU-walk mode. Sleeping is not permitted in this mode, and the caller can be asked to leave it and call again by returning - -ECHILD. + -ECHILD. -EISDIR may also be returned to tell pathwalk to + ignore d_automount or any mounts. This function is only used if DCACHE_MANAGE_TRANSIT is set on the dentry being transited from. diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index f896f68..817e3ea 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1138,6 +1138,10 @@ bytes respectively. Such letter suffixes can also be entirely omitted. Format: such that (rxsize & ~0x1fffc0) == 0. Default: 1024 + grsec_proc_gid= [GRKERNSEC_PROC_USERGROUP] Chooses GID to + ignore grsecurity's /proc restrictions + + hashdist= [KNL,NUMA] Large hashes allocated during boot are distributed across NUMA nodes. Defaults on for 64-bit NUMA, off otherwise. @@ -2141,6 +2145,10 @@ bytes respectively. Such letter suffixes can also be entirely omitted. noexec=on: enable non-executable mappings (default) noexec=off: disable non-executable mappings + nopcid [X86-64] + Disable PCID (Process-Context IDentifier) even if it + is supported by the processor. + nosmap [X86] Disable SMAP (Supervisor Mode Access Prevention) even if it is supported by processor. @@ -2418,6 +2426,25 @@ 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_sanitize_slab= + 0/1 to disable/enable slab object sanitization (enabled by + default). + + pax_softmode= 0/1 to disable/enable PaX softmode on boot already. + + pax_extra_latent_entropy + Enable a very simple form of latent entropy extraction + from the first 4GB of memory as the bootmem allocator + passes the memory pages to the buddy allocator. + + pax_weakuderef [X86-64] enables the weaker but faster form of UDEREF + when the processor supports PCID. + pcbit= [HW,ISDN] pcd. [PARIDE] diff --git a/Documentation/networking/filter.txt b/Documentation/networking/filter.txt index ee78eba..a06b48d 100644 --- a/Documentation/networking/filter.txt +++ b/Documentation/networking/filter.txt @@ -277,11 +277,10 @@ Possible BPF extensions are shown in the following table: mark skb->mark queue skb->queue_mapping hatype skb->dev->type - rxhash skb->hash + rxhash skb->rxhash cpu raw_smp_processor_id() vlan_tci vlan_tx_tag_get(skb) vlan_pr vlan_tx_tag_present(skb) - rand prandom_u32() These extensions can also be prefixed with '#'. Examples for low-level BPF: @@ -309,18 +308,6 @@ Examples for low-level BPF: ret #-1 drop: ret #0 -** icmp random packet sampling, 1 in 4 - ldh [12] - jne #0x800, drop - ldb [23] - jneq #1, drop - # get a random uint32 number - ld rand - mod #4 - jneq #1, drop - ret #-1 - drop: ret #0 - ** SECCOMP filter example: ld [4] /* offsetof(struct seccomp_data, arch) */ @@ -559,456 +546,6 @@ ffffffffa0069c8f + : For BPF JIT developers, bpf_jit_disasm, bpf_asm and bpf_dbg provides a useful toolchain for developing and testing the kernel's JIT compiler. -BPF kernel internals --------------------- -Internally, for the kernel interpreter, a different instruction set -format with similar underlying principles from BPF described in previous -paragraphs is being used. However, the instruction set format is modelled -closer to the underlying architecture to mimic native instruction sets, so -that a better performance can be achieved (more details later). This new -ISA is called 'eBPF' or 'internal BPF' interchangeably. (Note: eBPF which -originates from [e]xtended BPF is not the same as BPF extensions! While -eBPF is an ISA, BPF extensions date back to classic BPF's 'overloading' -of BPF_LD | BPF_{B,H,W} | BPF_ABS instruction.) - -It is designed to be JITed with one to one mapping, which can also open up -the possibility for GCC/LLVM compilers to generate optimized eBPF code through -an eBPF backend that performs almost as fast as natively compiled code. - -The new instruction set was originally designed with the possible goal in -mind to write programs in "restricted C" and compile into eBPF with a optional -GCC/LLVM backend, so that it can just-in-time map to modern 64-bit CPUs with -minimal performance overhead over two steps, that is, C -> eBPF -> native code. - -Currently, the new format is being used for running user BPF programs, which -includes seccomp BPF, classic socket filters, cls_bpf traffic classifier, -team driver's classifier for its load-balancing mode, netfilter's xt_bpf -extension, PTP dissector/classifier, and much more. They are all internally -converted by the kernel into the new instruction set representation and run -in the eBPF interpreter. For in-kernel handlers, this all works transparently -by using sk_unattached_filter_create() for setting up the filter, resp. -sk_unattached_filter_destroy() for destroying it. The macro -SK_RUN_FILTER(filter, ctx) transparently invokes eBPF interpreter or JITed -code to run the filter. 'filter' is a pointer to struct sk_filter that we -got from sk_unattached_filter_create(), and 'ctx' the given context (e.g. -skb pointer). All constraints and restrictions from sk_chk_filter() apply -before a conversion to the new layout is being done behind the scenes! - -Currently, the classic BPF format is being used for JITing on most of the -architectures. Only x86-64 performs JIT compilation from eBPF instruction set, -however, future work will migrate other JIT compilers as well, so that they -will profit from the very same benefits. - -Some core changes of the new internal format: - -- Number of registers increase from 2 to 10: - - The old format had two registers A and X, and a hidden frame pointer. The - new layout extends this to be 10 internal registers and a read-only frame - pointer. Since 64-bit CPUs are passing arguments to functions via registers - the number of args from eBPF program to in-kernel function is restricted - to 5 and one register is used to accept return value from an in-kernel - function. Natively, x86_64 passes first 6 arguments in registers, aarch64/ - sparcv9/mips64 have 7 - 8 registers for arguments; x86_64 has 6 callee saved - registers, and aarch64/sparcv9/mips64 have 11 or more callee saved registers. - - Therefore, eBPF calling convention is defined as: - - * R0 - return value from in-kernel function, and exit value for eBPF program - * R1 - R5 - arguments from eBPF program to in-kernel function - * R6 - R9 - callee saved registers that in-kernel function will preserve - * R10 - read-only frame pointer to access stack - - Thus, all eBPF registers map one to one to HW registers on x86_64, aarch64, - etc, and eBPF calling convention maps directly to ABIs used by the kernel on - 64-bit architectures. - - On 32-bit architectures JIT may map programs that use only 32-bit arithmetic - and may let more complex programs to be interpreted. - - R0 - R5 are scratch registers and eBPF program needs spill/fill them if - necessary across calls. Note that there is only one eBPF program (== one - eBPF main routine) and it cannot call other eBPF functions, it can only - call predefined in-kernel functions, though. - -- Register width increases from 32-bit to 64-bit: - - Still, the semantics of the original 32-bit ALU operations are preserved - via 32-bit subregisters. All eBPF registers are 64-bit with 32-bit lower - subregisters that zero-extend into 64-bit if they are being written to. - That behavior maps directly to x86_64 and arm64 subregister definition, but - makes other JITs more difficult. - - 32-bit architectures run 64-bit internal BPF programs via interpreter. - Their JITs may convert BPF programs that only use 32-bit subregisters into - native instruction set and let the rest being interpreted. - - Operation is 64-bit, because on 64-bit architectures, pointers are also - 64-bit wide, and we want to pass 64-bit values in/out of kernel functions, - so 32-bit eBPF registers would otherwise require to define register-pair - ABI, thus, there won't be able to use a direct eBPF register to HW register - mapping and JIT would need to do combine/split/move operations for every - register in and out of the function, which is complex, bug prone and slow. - Another reason is the use of atomic 64-bit counters. - -- Conditional jt/jf targets replaced with jt/fall-through: - - While the original design has constructs such as "if (cond) jump_true; - else jump_false;", they are being replaced into alternative constructs like - "if (cond) jump_true; /* else fall-through */". - -- Introduces bpf_call insn and register passing convention for zero overhead - calls from/to other kernel functions: - - Before an in-kernel function call, the internal BPF program needs to - place function arguments into R1 to R5 registers to satisfy calling - convention, then the interpreter will take them from registers and pass - to in-kernel function. If R1 - R5 registers are mapped to CPU registers - that are used for argument passing on given architecture, the JIT compiler - doesn't need to emit extra moves. Function arguments will be in the correct - registers and BPF_CALL instruction will be JITed as single 'call' HW - instruction. This calling convention was picked to cover common call - situations without performance penalty. - - After an in-kernel function call, R1 - R5 are reset to unreadable and R0 has - a return value of the function. Since R6 - R9 are callee saved, their state - is preserved across the call. - - For example, consider three C functions: - - u64 f1() { return (*_f2)(1); } - u64 f2(u64 a) { return f3(a + 1, a); } - u64 f3(u64 a, u64 b) { return a - b; } - - GCC can compile f1, f3 into x86_64: - - f1: - movl $1, %edi - movq _f2(%rip), %rax - jmp *%rax - f3: - movq %rdi, %rax - subq %rsi, %rax - ret - - Function f2 in eBPF may look like: - - f2: - bpf_mov R2, R1 - bpf_add R1, 1 - bpf_call f3 - bpf_exit - - If f2 is JITed and the pointer stored to '_f2'. The calls f1 -> f2 -> f3 and - returns will be seamless. Without JIT, __sk_run_filter() interpreter needs to - be used to call into f2. - - For practical reasons all eBPF programs have only one argument 'ctx' which is - already placed into R1 (e.g. on __sk_run_filter() startup) and the programs - can call kernel functions with up to 5 arguments. Calls with 6 or more arguments - are currently not supported, but these restrictions can be lifted if necessary - in the future. - - On 64-bit architectures all register map to HW registers one to one. For - example, x86_64 JIT compiler can map them as ... - - R0 - rax - R1 - rdi - R2 - rsi - R3 - rdx - R4 - rcx - R5 - r8 - R6 - rbx - R7 - r13 - R8 - r14 - R9 - r15 - R10 - rbp - - ... since x86_64 ABI mandates rdi, rsi, rdx, rcx, r8, r9 for argument passing - and rbx, r12 - r15 are callee saved. - - Then the following internal BPF pseudo-program: - - bpf_mov R6, R1 /* save ctx */ - bpf_mov R2, 2 - bpf_mov R3, 3 - bpf_mov R4, 4 - bpf_mov R5, 5 - bpf_call foo - bpf_mov R7, R0 /* save foo() return value */ - bpf_mov R1, R6 /* restore ctx for next call */ - bpf_mov R2, 6 - bpf_mov R3, 7 - bpf_mov R4, 8 - bpf_mov R5, 9 - bpf_call bar - bpf_add R0, R7 - bpf_exit - - After JIT to x86_64 may look like: - - push %rbp - mov %rsp,%rbp - sub $0x228,%rsp - mov %rbx,-0x228(%rbp) - mov %r13,-0x220(%rbp) - mov %rdi,%rbx - mov $0x2,%esi - mov $0x3,%edx - mov $0x4,%ecx - mov $0x5,%r8d - callq foo - mov %rax,%r13 - mov %rbx,%rdi - mov $0x2,%esi - mov $0x3,%edx - mov $0x4,%ecx - mov $0x5,%r8d - callq bar - add %r13,%rax - mov -0x228(%rbp),%rbx - mov -0x220(%rbp),%r13 - leaveq - retq - - Which is in this example equivalent in C to: - - u64 bpf_filter(u64 ctx) - { - return foo(ctx, 2, 3, 4, 5) + bar(ctx, 6, 7, 8, 9); - } - - In-kernel functions foo() and bar() with prototype: u64 (*)(u64 arg1, u64 - arg2, u64 arg3, u64 arg4, u64 arg5); will receive arguments in proper - registers and place their return value into '%rax' which is R0 in eBPF. - Prologue and epilogue are emitted by JIT and are implicit in the - interpreter. R0-R5 are scratch registers, so eBPF program needs to preserve - them across the calls as defined by calling convention. - - For example the following program is invalid: - - bpf_mov R1, 1 - bpf_call foo - bpf_mov R0, R1 - bpf_exit - - After the call the registers R1-R5 contain junk values and cannot be read. - In the future an eBPF verifier can be used to validate internal BPF programs. - -Also in the new design, eBPF is limited to 4096 insns, which means that any -program will terminate quickly and will only call a fixed number of kernel -functions. Original BPF and the new format are two operand instructions, -which helps to do one-to-one mapping between eBPF insn and x86 insn during JIT. - -The input context pointer for invoking the interpreter function is generic, -its content is defined by a specific use case. For seccomp register R1 points -to seccomp_data, for converted BPF filters R1 points to a skb. - -A program, that is translated internally consists of the following elements: - - op:16, jt:8, jf:8, k:32 ==> op:8, dst_reg:4, src_reg:4, off:16, imm:32 - -So far 87 internal BPF instructions were implemented. 8-bit 'op' opcode field -has room for new instructions. Some of them may use 16/24/32 byte encoding. New -instructions must be multiple of 8 bytes to preserve backward compatibility. - -Internal BPF is a general purpose RISC instruction set. Not every register and -every instruction are used during translation from original BPF to new format. -For example, socket filters are not using 'exclusive add' instruction, but -tracing filters may do to maintain counters of events, for example. Register R9 -is not used by socket filters either, but more complex filters may be running -out of registers and would have to resort to spill/fill to stack. - -Internal BPF can used as generic assembler for last step performance -optimizations, socket filters and seccomp are using it as assembler. Tracing -filters may use it as assembler to generate code from kernel. In kernel usage -may not be bounded by security considerations, since generated internal BPF code -may be optimizing internal code path and not being exposed to the user space. -Safety of internal BPF can come from a verifier (TBD). In such use cases as -described, it may be used as safe instruction set. - -Just like the original BPF, the new format runs within a controlled environment, -is deterministic and the kernel can easily prove that. The safety of the program -can be determined in two steps: first step does depth-first-search to disallow -loops and other CFG validation; second step starts from the first insn and -descends all possible paths. It simulates execution of every insn and observes -the state change of registers and stack. - -eBPF opcode encoding --------------------- - -eBPF is reusing most of the opcode encoding from classic to simplify conversion -of classic BPF to eBPF. For arithmetic and jump instructions the 8-bit 'code' -field is divided into three parts: - - +----------------+--------+--------------------+ - | 4 bits | 1 bit | 3 bits | - | operation code | source | instruction class | - +----------------+--------+--------------------+ - (MSB) (LSB) - -Three LSB bits store instruction class which is one of: - - Classic BPF classes: eBPF classes: - - BPF_LD 0x00 BPF_LD 0x00 - BPF_LDX 0x01 BPF_LDX 0x01 - BPF_ST 0x02 BPF_ST 0x02 - BPF_STX 0x03 BPF_STX 0x03 - BPF_ALU 0x04 BPF_ALU 0x04 - BPF_JMP 0x05 BPF_JMP 0x05 - BPF_RET 0x06 [ class 6 unused, for future if needed ] - BPF_MISC 0x07 BPF_ALU64 0x07 - -When BPF_CLASS(code) == BPF_ALU or BPF_JMP, 4th bit encodes source operand ... - - BPF_K 0x00 - BPF_X 0x08 - - * in classic BPF, this means: - - BPF_SRC(code) == BPF_X - use register X as source operand - BPF_SRC(code) == BPF_K - use 32-bit immediate as source operand - - * in eBPF, this means: - - BPF_SRC(code) == BPF_X - use 'src_reg' register as source operand - BPF_SRC(code) == BPF_K - use 32-bit immediate as source operand - -... and four MSB bits store operation code. - -If BPF_CLASS(code) == BPF_ALU or BPF_ALU64 [ in eBPF ], BPF_OP(code) is one of: - - BPF_ADD 0x00 - BPF_SUB 0x10 - BPF_MUL 0x20 - BPF_DIV 0x30 - BPF_OR 0x40 - BPF_AND 0x50 - BPF_LSH 0x60 - BPF_RSH 0x70 - BPF_NEG 0x80 - BPF_MOD 0x90 - BPF_XOR 0xa0 - BPF_MOV 0xb0 /* eBPF only: mov reg to reg */ - BPF_ARSH 0xc0 /* eBPF only: sign extending shift right */ - BPF_END 0xd0 /* eBPF only: endianness conversion */ - -If BPF_CLASS(code) == BPF_JMP, BPF_OP(code) is one of: - - BPF_JA 0x00 - BPF_JEQ 0x10 - BPF_JGT 0x20 - BPF_JGE 0x30 - BPF_JSET 0x40 - BPF_JNE 0x50 /* eBPF only: jump != */ - BPF_JSGT 0x60 /* eBPF only: signed '>' */ - BPF_JSGE 0x70 /* eBPF only: signed '>=' */ - BPF_CALL 0x80 /* eBPF only: function call */ - BPF_EXIT 0x90 /* eBPF only: function return */ - -So BPF_ADD | BPF_X | BPF_ALU means 32-bit addition in both classic BPF -and eBPF. There are only two registers in classic BPF, so it means A += X. -In eBPF it means dst_reg = (u32) dst_reg + (u32) src_reg; similarly, -BPF_XOR | BPF_K | BPF_ALU means A ^= imm32 in classic BPF and analogous -src_reg = (u32) src_reg ^ (u32) imm32 in eBPF. - -Classic BPF is using BPF_MISC class to represent A = X and X = A moves. -eBPF is using BPF_MOV | BPF_X | BPF_ALU code instead. Since there are no -BPF_MISC operations in eBPF, the class 7 is used as BPF_ALU64 to mean -exactly the same operations as BPF_ALU, but with 64-bit wide operands -instead. So BPF_ADD | BPF_X | BPF_ALU64 means 64-bit addition, i.e.: -dst_reg = dst_reg + src_reg - -Classic BPF wastes the whole BPF_RET class to represent a single 'ret' -operation. Classic BPF_RET | BPF_K means copy imm32 into return register -and perform function exit. eBPF is modeled to match CPU, so BPF_JMP | BPF_EXIT -in eBPF means function exit only. The eBPF program needs to store return -value into register R0 before doing a BPF_EXIT. Class 6 in eBPF is currently -unused and reserved for future use. - -For load and store instructions the 8-bit 'code' field is divided as: - - +--------+--------+-------------------+ - | 3 bits | 2 bits | 3 bits | - | mode | size | instruction class | - +--------+--------+-------------------+ - (MSB) (LSB) - -Size modifier is one of ... - - BPF_W 0x00 /* word */ - BPF_H 0x08 /* half word */ - BPF_B 0x10 /* byte */ - BPF_DW 0x18 /* eBPF only, double word */ - -... which encodes size of load/store operation: - - B - 1 byte - H - 2 byte - W - 4 byte - DW - 8 byte (eBPF only) - -Mode modifier is one of: - - BPF_IMM 0x00 /* classic BPF only, reserved in eBPF */ - BPF_ABS 0x20 - BPF_IND 0x40 - BPF_MEM 0x60 - BPF_LEN 0x80 /* classic BPF only, reserved in eBPF */ - BPF_MSH 0xa0 /* classic BPF only, reserved in eBPF */ - BPF_XADD 0xc0 /* eBPF only, exclusive add */ - -eBPF has two non-generic instructions: (BPF_ABS | | BPF_LD) and -(BPF_IND | | BPF_LD) which are used to access packet data. - -They had to be carried over from classic to have strong performance of -socket filters running in eBPF interpreter. These instructions can only -be used when interpreter context is a pointer to 'struct sk_buff' and -have seven implicit operands. Register R6 is an implicit input that must -contain pointer to sk_buff. Register R0 is an implicit output which contains -the data fetched from the packet. Registers R1-R5 are scratch registers -and must not be used to store the data across BPF_ABS | BPF_LD or -BPF_IND | BPF_LD instructions. - -These instructions have implicit program exit condition as well. When -eBPF program is trying to access the data beyond the packet boundary, -the interpreter will abort the execution of the program. JIT compilers -therefore must preserve this property. src_reg and imm32 fields are -explicit inputs to these instructions. - -For example: - - BPF_IND | BPF_W | BPF_LD means: - - R0 = ntohl(*(u32 *) (((struct sk_buff *) R6)->data + src_reg + imm32)) - and R1 - R5 were scratched. - -Unlike classic BPF instruction set, eBPF has generic load/store operations: - -BPF_MEM | | BPF_STX: *(size *) (dst_reg + off) = src_reg -BPF_MEM | | BPF_ST: *(size *) (dst_reg + off) = imm32 -BPF_MEM | | BPF_LDX: dst_reg = *(size *) (src_reg + off) -BPF_XADD | BPF_W | BPF_STX: lock xadd *(u32 *)(dst_reg + off16) += src_reg -BPF_XADD | BPF_DW | BPF_STX: lock xadd *(u64 *)(dst_reg + off16) += src_reg - -Where size is one of: BPF_B or BPF_H or BPF_W or BPF_DW. Note that 1 and -2 byte atomic increments are not supported. - -Testing -------- - -Next to the BPF toolchain, the kernel also ships a test module that contains -various test cases for classic and internal BPF that can be executed against -the BPF interpreter and JIT compiler. It can be found in lib/test_bpf.c and -enabled via Kconfig: - - CONFIG_TEST_BPF=m - -After the module has been built and installed, the test suite can be executed -via insmod or modprobe against 'test_bpf' module. Results of the test cases -including timings in nsec can be found in the kernel log (dmesg). - Misc ---- @@ -1024,4 +561,3 @@ the underlying architecture. Jay Schulist Daniel Borkmann -Alexei Starovoitov diff --git a/Makefile b/Makefile index e75c75f..ebe05e8 100644 --- a/Makefile +++ b/Makefile @@ -303,8 +303,9 @@ 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 +HOSTCFLAGS += $(call cc-option, -Wno-empty-body) +HOSTCXXFLAGS = -O2 -Wall -W -Wno-array-bounds ifeq ($(shell $(HOSTCC) -v 2>&1 | grep -c "clang version"), 1) HOSTCFLAGS += -Wno-unused-value -Wno-unused-parameter \ @@ -449,8 +450,8 @@ export RCS_TAR_IGNORE := --exclude SCCS --exclude BitKeeper --exclude .svn \ # 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 @@ -621,6 +622,75 @@ else KBUILD_CFLAGS += -O2 endif +# Tell gcc to never replace conditional load with a non-conditional one +KBUILD_CFLAGS += $(call cc-option,--param=allow-store-data-races=0) + +ifndef DISABLE_PAX_PLUGINS +ifeq ($(call cc-ifversion, -ge, 0408, y), y) +PLUGINCC := $(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-plugin.sh "$(HOSTCXX)" "$(HOSTCXX)" "$(CC)") +else +PLUGINCC := $(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-plugin.sh "$(HOSTCC)" "$(HOSTCXX)" "$(CC)") +endif +ifneq ($(PLUGINCC),) +ifdef CONFIG_PAX_CONSTIFY_PLUGIN +CONSTIFY_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/constify_plugin.so -DCONSTIFY_PLUGIN +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_GRKERNSEC_RANDSTRUCT +RANDSTRUCT_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/randomize_layout_plugin.so -DRANDSTRUCT_PLUGIN +ifdef CONFIG_GRKERNSEC_RANDSTRUCT_PERFORMANCE +RANDSTRUCT_PLUGIN_CFLAGS += -fplugin-arg-randomize_layout_plugin-performance-mode +endif +endif +ifdef CONFIG_CHECKER_PLUGIN +ifeq ($(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_CFLAGS := -fplugin=$(objtree)/tools/gcc/size_overflow_plugin/size_overflow_plugin.so -DSIZE_OVERFLOW_PLUGIN +endif +ifdef CONFIG_PAX_LATENT_ENTROPY +LATENT_ENTROPY_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/latent_entropy_plugin.so -DLATENT_ENTROPY_PLUGIN +endif +ifdef CONFIG_PAX_MEMORY_STRUCTLEAK +STRUCTLEAK_PLUGIN_CFLAGS := -fplugin=$(objtree)/tools/gcc/structleak_plugin.so -DSTRUCTLEAK_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) +GCC_PLUGINS_CFLAGS += $(SIZE_OVERFLOW_PLUGIN_CFLAGS) $(LATENT_ENTROPY_PLUGIN_CFLAGS) $(STRUCTLEAK_PLUGIN_CFLAGS) +GCC_PLUGINS_CFLAGS += $(RANDSTRUCT_PLUGIN_CFLAGS) +GCC_PLUGINS_AFLAGS := $(KERNEXEC_PLUGIN_AFLAGS) +export PLUGINCC GCC_PLUGINS_CFLAGS GCC_PLUGINS_AFLAGS CONSTIFY_PLUGIN LATENT_ENTROPY_PLUGIN_CFLAGS +ifeq ($(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, constification, PAX_LATENT_ENTROPY and other features will be less secure. PAX_SIZE_OVERFLOW will not be active." +endif +endif + ifdef CONFIG_READABLE_ASM # Disable optimizations that make assembler listings hard to read. # reorder blocks reorders the control in the function @@ -692,7 +762,7 @@ KBUILD_CFLAGS += $(call cc-option, -fno-var-tracking-assignments) ifdef CONFIG_DEBUG_INFO KBUILD_CFLAGS += -g -KBUILD_AFLAGS += -Wa,-gdwarf-2 +KBUILD_AFLAGS += -Wa,--gdwarf-2 endif ifdef CONFIG_DEBUG_INFO_REDUCED @@ -839,7 +909,7 @@ export mod_sign_cmd ifeq ($(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) $(drivers-m) \ @@ -888,6 +958,8 @@ endif # The actual objects are generated when descending, # make sure no implicit rule kicks in +$(filter-out $(init-y),$(vmlinux-deps)): KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +$(filter-out $(init-y),$(vmlinux-deps)): KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) $(sort $(vmlinux-deps)): $(vmlinux-dirs) ; # Handle descending into subdirectories listed in $(vmlinux-dirs) @@ -897,7 +969,7 @@ $(sort $(vmlinux-deps)): $(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) $(build)=$@ define filechk_kernel.release @@ -940,10 +1012,13 @@ prepare1: prepare2 $(version_h) include/generated/utsrelease.h \ archprepare: archheaders archscripts prepare1 scripts_basic +prepare0: KBUILD_CFLAGS += $(GCC_PLUGINS_CFLAGS) +prepare0: KBUILD_AFLAGS += $(GCC_PLUGINS_AFLAGS) prepare0: archprepare FORCE $(Q)$(MAKE) $(build)=. # All the preparing.. +prepare: KBUILD_CFLAGS := $(filter-out $(GCC_PLUGINS_CFLAGS),$(KBUILD_CFLAGS)) prepare: prepare0 # Generate some files @@ -1051,6 +1126,8 @@ 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.'; @@ -1066,7 +1143,7 @@ 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 @@ -1132,7 +1209,10 @@ MRPROPER_FILES += .config .config.old .version .old_version $(version_h) \ Module.symvers tags TAGS cscope* GPATH GTAGS GRTAGS GSYMS \ signing_key.priv signing_key.x509 x509.genkey \ extra_certificates signing_key.x509.keyid \ - signing_key.x509.signer include/linux/version.h + signing_key.x509.signer include/linux/version.h \ + tools/gcc/size_overflow_plugin/size_overflow_hash_aux.h \ + tools/gcc/size_overflow_plugin/size_overflow_hash.h \ + tools/gcc/randomize_layout_seed.h # clean - Delete most, but leave enough to build external modules # @@ -1171,7 +1251,7 @@ distclean: mrproper @find $(srctree) $(RCS_FIND_IGNORE) \ \( -name '*.orig' -o -name '*.rej' -o -name '*~' \ -o -name '*.bak' -o -name '#*#' -o -name '.*.orig' \ - -o -name '.*.rej' -o -name '*%' -o -name 'core' \) \ + -o -name '.*.rej' -o -name '*.so' -o -name '*%' -o -name 'core' \) \ -type f -print | xargs rm -f @@ -1332,6 +1412,8 @@ 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 @@ -1471,17 +1553,21 @@ 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 $@) @@ -1491,11 +1577,15 @@ 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 ed60a1e..47f1a55 100644 --- a/arch/alpha/include/asm/atomic.h +++ b/arch/alpha/include/asm/atomic.h @@ -292,4 +292,14 @@ static inline long atomic64_dec_if_positive(atomic64_t *v) #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)) + #endif /* _ALPHA_ATOMIC_H */ 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 968d999..d36b2df 100644 --- a/arch/alpha/include/asm/elf.h +++ b/arch/alpha/include/asm/elf.h @@ -91,6 +91,13 @@ 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 aab14a0..b4fa3e7 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 d8f9b7e..f6222fa 100644 --- a/arch/alpha/include/asm/pgtable.h +++ b/arch/alpha/include/asm/pgtable.h @@ -102,6 +102,17 @@ 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 1402fcc..0b1abd2 100644 --- a/arch/alpha/kernel/osf_sys.c +++ b/arch/alpha/kernel/osf_sys.c @@ -1298,10 +1298,11 @@ SYSCALL_DEFINE1(old_adjtimex, struct timex32 __user *, txc_p) generic version except that we know how to honor ADDR_LIMIT_32BIT. */ static unsigned long -arch_get_unmapped_area_1(unsigned long addr, unsigned long len, - unsigned long limit) +arch_get_unmapped_area_1(struct file *filp, unsigned long addr, unsigned long len, + unsigned long limit, unsigned long flags) { struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(current->mm, filp, flags); info.flags = 0; info.length = len; @@ -1309,6 +1310,7 @@ arch_get_unmapped_area_1(unsigned long addr, unsigned long len, info.high_limit = limit; info.align_mask = 0; info.align_offset = 0; + info.threadstack_offset = offset; return vm_unmapped_area(&info); } @@ -1341,20 +1343,24 @@ 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); + addr = arch_get_unmapped_area_1 (filp, PAGE_ALIGN(addr), len, limit, flags); if (addr != (unsigned long) -ENOMEM) return 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 (filp, PAGE_ALIGN(current->mm->mmap_base), len, limit, flags); + if (addr != (unsigned long) -ENOMEM) return addr; /* Finally, try allocating in low memory. */ - addr = arch_get_unmapped_area_1 (PAGE_SIZE, len, limit); + addr = arch_get_unmapped_area_1 (filp, PAGE_SIZE, len, limit, flags); return addr; } diff --git a/arch/alpha/mm/fault.c b/arch/alpha/mm/fault.c index 98838a0..b304fb4 100644 --- a/arch/alpha/mm/fault.c +++ b/arch/alpha/mm/fault.c @@ -53,6 +53,124 @@ __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, @@ -133,8 +251,29 @@ retry: 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/Kconfig b/arch/arm/Kconfig index 290f02ee..a639059 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1787,7 +1787,7 @@ config ALIGNMENT_TRAP config UACCESS_WITH_MEMCPY bool "Use kernel mem{cpy,set}() for {copy_to,clear}_user()" - depends on MMU + depends on MMU && !PAX_MEMORY_UDEREF default y if CPU_FEROCEON help Implement faster copy_to_user and clear_user methods for CPU @@ -2051,6 +2051,7 @@ config XIP_PHYS_ADDR config KEXEC bool "Kexec system call (EXPERIMENTAL)" depends on (!SMP || PM_SLEEP_SMP) + depends on !GRKERNSEC_KMEM help kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot diff --git a/arch/arm/include/asm/atomic.h b/arch/arm/include/asm/atomic.h index 3040359..cf3bab0 100644 --- a/arch/arm/include/asm/atomic.h +++ b/arch/arm/include/asm/atomic.h @@ -18,17 +18,35 @@ #include #include +#ifdef CONFIG_GENERIC_ATOMIC64 +#include +#endif + #define ATOMIC_INIT(i) { (i) } #ifdef __KERNEL__ +#define _ASM_EXTABLE(from, to) \ +" .pushsection __ex_table,\"a\"\n"\ +" .align 3\n" \ +" .long " #from ", " #to"\n" \ +" .popsection" + /* * On ARM, ordinary assignment (str instruction) doesn't clear the local * strex/ldrex monitor on some implementations. The reason we can use it for * 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 @@ -44,6 +62,36 @@ static inline void atomic_add(int i, atomic_t *v) prefetchw(&v->counter); __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; + + prefetchw(&v->counter); + __asm__ __volatile__("@ atomic_add_unchecked\n" "1: ldrex %0, [%3]\n" " add %0, %0, %4\n" " strex %1, %0, [%3]\n" @@ -63,6 +111,43 @@ static inline int atomic_add_return(int i, atomic_t *v) prefetchw(&v->counter); __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(); + prefetchw(&v->counter); + + __asm__ __volatile__("@ atomic_add_return_unchecked\n" "1: ldrex %0, [%3]\n" " add %0, %0, %4\n" " strex %1, %0, [%3]\n" @@ -84,6 +169,36 @@ static inline void atomic_sub(int i, atomic_t *v) prefetchw(&v->counter); __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; + + prefetchw(&v->counter); + __asm__ __volatile__("@ atomic_sub_unchecked\n" "1: ldrex %0, [%3]\n" " sub %0, %0, %4\n" " strex %1, %0, [%3]\n" @@ -103,11 +218,25 @@ static inline int atomic_sub_return(int i, atomic_t *v) prefetchw(&v->counter); __asm__ __volatile__("@ atomic_sub_return\n" -"1: ldrex %0, [%3]\n" -" sub %0, %0, %4\n" +"1: ldrex %1, [%3]\n" +" subs %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"); @@ -152,12 +281,24 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) __asm__ __volatile__ ("@ atomic_add_unless\n" "1: ldrex %0, [%4]\n" " teq %0, %5\n" -" beq 2f\n" -" add %1, %0, %6\n" +" beq 4f\n" +" adds %1, %0, %6\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + " strex %2, %1, [%4]\n" " teq %2, #0\n" " bne 1b\n" -"2:" +"4:" + +#ifdef CONFIG_PAX_REFCOUNT + _ASM_EXTABLE(2b, 4b) +#endif + : "=&r" (oldval), "=&r" (newval), "=&r" (tmp), "+Qo" (v->counter) : "r" (&v->counter), "r" (u), "r" (a) : "cc"); @@ -168,6 +309,28 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) 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; +} + #else /* ARM_ARCH_6 */ #ifdef CONFIG_SMP @@ -186,7 +349,17 @@ static inline int atomic_add_return(int i, atomic_t *v) return val; } + +static inline int atomic_add_return_unchecked(int i, atomic_unchecked_t *v) +{ + return atomic_add_return(i, v); +} + #define atomic_add(i, v) (void) atomic_add_return(i, v) +static inline void atomic_add_unchecked(int i, atomic_unchecked_t *v) +{ + (void) atomic_add_return(i, v); +} static inline int atomic_sub_return(int i, atomic_t *v) { @@ -201,6 +374,10 @@ static inline int atomic_sub_return(int i, atomic_t *v) return val; } #define atomic_sub(i, v) (void) atomic_sub_return(i, v) +static inline void atomic_sub_unchecked(int i, atomic_unchecked_t *v) +{ + (void) atomic_sub_return(i, v); +} static inline int atomic_cmpxchg(atomic_t *v, int old, int new) { @@ -216,6 +393,11 @@ static inline int atomic_cmpxchg(atomic_t *v, int old, int new) return ret; } +static inline int atomic_cmpxchg_unchecked(atomic_unchecked_t *v, int old, int new) +{ + return atomic_cmpxchg(v, old, new); +} + static inline int __atomic_add_unless(atomic_t *v, int a, int u) { int c, old; @@ -229,13 +411,33 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) #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); +} #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) @@ -246,6 +448,14 @@ typedef struct { long long counter; } atomic64_t; +#ifdef CONFIG_PAX_REFCOUNT +typedef struct { + long long counter; +} atomic64_unchecked_t; +#else +typedef atomic64_t atomic64_unchecked_t; +#endif + #define ATOMIC64_INIT(i) { (i) } #ifdef CONFIG_ARM_LPAE @@ -262,6 +472,19 @@ static inline long long atomic64_read(const atomic64_t *v) return result; } +static inline long long atomic64_read_unchecked(const atomic64_unchecked_t *v) +{ + long long result; + + __asm__ __volatile__("@ atomic64_read_unchecked\n" +" ldrd %0, %H0, [%1]" + : "=&r" (result) + : "r" (&v->counter), "Qo" (v->counter) + ); + + return result; +} + static inline void atomic64_set(atomic64_t *v, long long i) { __asm__ __volatile__("@ atomic64_set\n" @@ -270,6 +493,15 @@ static inline void atomic64_set(atomic64_t *v, long long i) : "r" (&v->counter), "r" (i) ); } + +static inline void atomic64_set_unchecked(atomic64_unchecked_t *v, long long i) +{ + __asm__ __volatile__("@ atomic64_set_unchecked\n" +" strd %2, %H2, [%1]" + : "=Qo" (v->counter) + : "r" (&v->counter), "r" (i) + ); +} #else static inline long long atomic64_read(const atomic64_t *v) { @@ -284,6 +516,19 @@ static inline long long atomic64_read(const atomic64_t *v) return result; } +static inline long long atomic64_read_unchecked(const atomic64_unchecked_t *v) +{ + long long 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, long long i) { long long tmp; @@ -298,6 +543,21 @@ static inline void atomic64_set(atomic64_t *v, long long i) : "r" (&v->counter), "r" (i) : "cc"); } + +static inline void atomic64_set_unchecked(atomic64_unchecked_t *v, long long i) +{ + long long tmp; + + prefetchw(&v->counter); + __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"); +} #endif static inline void atomic64_add(long long i, atomic64_t *v) @@ -309,6 +569,37 @@ static inline void atomic64_add(long long i, atomic64_t *v) __asm__ __volatile__("@ atomic64_add\n" "1: ldrexd %0, %H0, [%3]\n" " adds %Q0, %Q0, %Q4\n" +" adcs %R0, %R0, %R4\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(long long i, atomic64_unchecked_t *v) +{ + long long result; + unsigned long tmp; + + prefetchw(&v->counter); + __asm__ __volatile__("@ atomic64_add_unchecked\n" +"1: ldrexd %0, %H0, [%3]\n" +" adds %Q0, %Q0, %Q4\n" " adc %R0, %R0, %R4\n" " strexd %1, %0, %H0, [%3]\n" " teq %1, #0\n" @@ -329,6 +620,44 @@ static inline long long atomic64_add_return(long long i, atomic64_t *v) __asm__ __volatile__("@ atomic64_add_return\n" "1: ldrexd %0, %H0, [%3]\n" " adds %Q0, %Q0, %Q4\n" +" adcs %R0, %R0, %R4\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 long long atomic64_add_return_unchecked(long long i, atomic64_unchecked_t *v) +{ + long long result; + unsigned long tmp; + + smp_mb(); + + __asm__ __volatile__("@ atomic64_add_return_unchecked\n" +"1: ldrexd %0, %H0, [%3]\n" +" adds %Q0, %Q0, %Q4\n" " adc %R0, %R0, %R4\n" " strexd %1, %0, %H0, [%3]\n" " teq %1, #0\n" @@ -351,6 +680,37 @@ static inline void atomic64_sub(long long i, atomic64_t *v) __asm__ __volatile__("@ atomic64_sub\n" "1: ldrexd %0, %H0, [%3]\n" " subs %Q0, %Q0, %Q4\n" +" sbcs %R0, %R0, %R4\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(long long i, atomic64_unchecked_t *v) +{ + long long result; + unsigned long tmp; + + prefetchw(&v->counter); + __asm__ __volatile__("@ atomic64_sub_unchecked\n" +"1: ldrexd %0, %H0, [%3]\n" +" subs %Q0, %Q0, %Q4\n" " sbc %R0, %R0, %R4\n" " strexd %1, %0, %H0, [%3]\n" " teq %1, #0\n" @@ -371,10 +731,25 @@ static inline long long atomic64_sub_return(long long i, atomic64_t *v) __asm__ __volatile__("@ atomic64_sub_return\n" "1: ldrexd %0, %H0, [%3]\n" " subs %Q0, %Q0, %Q4\n" -" sbc %R0, %R0, %R4\n" +" sbcs %R0, %R0, %R4\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"); @@ -410,6 +785,31 @@ static inline long long atomic64_cmpxchg(atomic64_t *ptr, long long old, return oldval; } +static inline long long atomic64_cmpxchg_unchecked(atomic64_unchecked_t *ptr, long long old, + long long new) +{ + long long 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 long long atomic64_xchg(atomic64_t *ptr, long long new) { long long result; @@ -435,21 +835,35 @@ static inline long long atomic64_xchg(atomic64_t *ptr, long long new) static inline long long atomic64_dec_if_positive(atomic64_t *v) { long long result; - unsigned long tmp; + u64 tmp; smp_mb(); prefetchw(&v->counter); __asm__ __volatile__("@ atomic64_dec_if_positive\n" -"1: ldrexd %0, %H0, [%3]\n" -" subs %Q0, %Q0, #1\n" -" sbc %R0, %R0, #0\n" +"1: ldrexd %1, %H1, [%3]\n" +" subs %Q0, %Q1, #1\n" +" sbcs %R0, %R1, #0\n" + +#ifdef CONFIG_PAX_REFCOUNT +" bvc 3f\n" +" mov %Q0, %Q1\n" +" mov %R0, %R1\n" +"2: bkpt 0xf103\n" +"3:\n" +#endif + " teq %R0, #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"); @@ -473,13 +887,25 @@ static inline int atomic64_add_unless(atomic64_t *v, long long a, long long u) " teq %0, %5\n" " teqeq %H0, %H5\n" " moveq %1, #0\n" -" beq 2f\n" +" beq 4f\n" " adds %Q0, %Q0, %Q6\n" -" adc %R0, %R0, %R6\n" +" adcs %R0, %R0, %R6\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"); @@ -492,10 +918,13 @@ static inline int atomic64_add_unless(atomic64_t *v, long long a, long long 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/barrier.h b/arch/arm/include/asm/barrier.h index c6a3e73..35cca85 100644 --- a/arch/arm/include/asm/barrier.h +++ b/arch/arm/include/asm/barrier.h @@ -63,7 +63,7 @@ do { \ compiletime_assert_atomic_type(*p); \ smp_mb(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ diff --git a/arch/arm/include/asm/cache.h b/arch/arm/include/asm/cache.h index 75fe66b..ba3dee4 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 @@ -24,5 +26,6 @@ #endif #define __read_mostly __attribute__((__section__(".data..read_mostly"))) +#define __read_only __attribute__ ((__section__(".data..read_only"))) #endif diff --git a/arch/arm/include/asm/cacheflush.h b/arch/arm/include/asm/cacheflush.h index 79ecb4f..6b0bbdd 100644 --- a/arch/arm/include/asm/cacheflush.h +++ b/arch/arm/include/asm/cacheflush.h @@ -116,7 +116,7 @@ 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/checksum.h b/arch/arm/include/asm/checksum.h index 5233151..87a71fa 100644 --- a/arch/arm/include/asm/checksum.h +++ b/arch/arm/include/asm/checksum.h @@ -37,7 +37,19 @@ __wsum csum_partial_copy_nocheck(const void *src, void *dst, int len, __wsum sum); __wsum -csum_partial_copy_from_user(const void __user *src, void *dst, int len, __wsum sum, int *err_ptr); +__csum_partial_copy_from_user(const void __user *src, void *dst, int len, __wsum sum, int *err_ptr); + +static inline __wsum +csum_partial_copy_from_user(const void __user *src, void *dst, int len, __wsum sum, int *err_ptr) +{ + __wsum ret; + pax_open_userland(); + ret = __csum_partial_copy_from_user(src, dst, len, sum, err_ptr); + pax_close_userland(); + return ret; +} + + /* * Fold a partial checksum without adding pseudo headers diff --git a/arch/arm/include/asm/cmpxchg.h b/arch/arm/include/asm/cmpxchg.h index abb2c37..96db950 100644 --- a/arch/arm/include/asm/cmpxchg.h +++ b/arch/arm/include/asm/cmpxchg.h @@ -104,6 +104,8 @@ static inline unsigned long __xchg(unsigned long x, volatile void *ptr, int size #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)))) #include diff --git a/arch/arm/include/asm/domain.h b/arch/arm/include/asm/domain.h index 6ddbe44..b5e38b1 100644 --- a/arch/arm/include/asm/domain.h +++ b/arch/arm/include/asm/domain.h @@ -48,18 +48,37 @@ * Domain types */ #define DOMAIN_NOACCESS 0 -#define DOMAIN_CLIENT 1 #ifdef CONFIG_CPU_USE_DOMAINS +#define DOMAIN_USERCLIENT 1 +#define DOMAIN_KERNELCLIENT 1 #define DOMAIN_MANAGER 3 +#define DOMAIN_VECTORS DOMAIN_USER #else + +#ifdef CONFIG_PAX_KERNEXEC #define DOMAIN_MANAGER 1 +#define DOMAIN_KERNEXEC 3 +#else +#define DOMAIN_MANAGER 1 +#endif + +#ifdef CONFIG_PAX_MEMORY_UDEREF +#define DOMAIN_USERCLIENT 0 +#define DOMAIN_UDEREF 1 +#define DOMAIN_VECTORS DOMAIN_KERNEL +#else +#define DOMAIN_USERCLIENT 1 +#define DOMAIN_VECTORS DOMAIN_USER +#endif +#define DOMAIN_KERNELCLIENT 1 + #endif #define domain_val(dom,type) ((type) << (2*(dom))) #ifndef __ASSEMBLY__ -#ifdef CONFIG_CPU_USE_DOMAINS +#if defined(CONFIG_CPU_USE_DOMAINS) || defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) static inline void set_domain(unsigned val) { asm volatile( @@ -68,15 +87,7 @@ static inline void set_domain(unsigned val) isb(); } -#define modify_domain(dom,type) \ - do { \ - struct thread_info *thread = current_thread_info(); \ - unsigned int domain = thread->cpu_domain; \ - domain &= ~domain_val(dom, DOMAIN_MANAGER); \ - thread->cpu_domain = domain | domain_val(dom, type); \ - set_domain(thread->cpu_domain); \ - } while (0) - +extern void modify_domain(unsigned int dom, unsigned int type); #else static inline void set_domain(unsigned val) { } static inline void modify_domain(unsigned dom, unsigned type) { } diff --git a/arch/arm/include/asm/elf.h b/arch/arm/include/asm/elf.h index f4b46d3..abc9b2b 100644 --- a/arch/arm/include/asm/elf.h +++ b/arch/arm/include/asm/elf.h @@ -114,7 +114,14 @@ 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 @@ -124,10 +131,6 @@ 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 - #ifdef CONFIG_MMU #define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1 struct linux_binprm; diff --git a/arch/arm/include/asm/fncpy.h b/arch/arm/include/asm/fncpy.h index de53547..52b9a28 100644 --- a/arch/arm/include/asm/fncpy.h +++ b/arch/arm/include/asm/fncpy.h @@ -81,7 +81,9 @@ BUG_ON((uintptr_t)(dest_buf) & (FNCPY_ALIGN - 1) || \ (__funcp_address & ~(uintptr_t)1 & (FNCPY_ALIGN - 1))); \ \ + pax_open_kernel(); \ memcpy(dest_buf, (void const *)(__funcp_address & ~1), size); \ + pax_close_kernel(); \ flush_icache_range((unsigned long)(dest_buf), \ (unsigned long)(dest_buf) + (size)); \ \ diff --git a/arch/arm/include/asm/futex.h b/arch/arm/include/asm/futex.h index 53e69da..3fdc896 100644 --- a/arch/arm/include/asm/futex.h +++ b/arch/arm/include/asm/futex.h @@ -46,6 +46,8 @@ futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr, if (!access_ok(VERIFY_WRITE, uaddr, sizeof(u32))) return -EFAULT; + pax_open_userland(); + smp_mb(); /* Prefetching cannot fault */ prefetchw(uaddr); @@ -63,6 +65,8 @@ futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr, : "cc", "memory"); smp_mb(); + pax_close_userland(); + *uval = val; return ret; } @@ -93,6 +97,8 @@ futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr, if (!access_ok(VERIFY_WRITE, uaddr, sizeof(u32))) return -EFAULT; + pax_open_userland(); + __asm__ __volatile__("@futex_atomic_cmpxchg_inatomic\n" "1: " TUSER(ldr) " %1, [%4]\n" " teq %1, %2\n" @@ -103,6 +109,8 @@ futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr, : "r" (oldval), "r" (newval), "r" (uaddr), "Ir" (-EFAULT) : "cc", "memory"); + pax_close_userland(); + *uval = val; return ret; } @@ -125,6 +133,7 @@ futex_atomic_op_inuser (int encoded_op, u32 __user *uaddr) return -EFAULT; pagefault_disable(); /* implies preempt_disable() */ + pax_open_userland(); switch (op) { case FUTEX_OP_SET: @@ -146,6 +155,7 @@ futex_atomic_op_inuser (int encoded_op, u32 __user *uaddr) ret = -ENOSYS; } + pax_close_userland(); pagefault_enable(); /* subsumes preempt_enable() */ if (!ret) { diff --git a/arch/arm/include/asm/kmap_types.h b/arch/arm/include/asm/kmap_types.h index 83eb2f7..ed77159 100644 --- a/arch/arm/include/asm/kmap_types.h +++ b/arch/arm/include/asm/kmap_types.h @@ -4,6 +4,6 @@ /* * This is the "bare minimum". AIO seems to require this. */ -#define KM_TYPE_NR 16 +#define KM_TYPE_NR 17 #endif diff --git a/arch/arm/include/asm/mach/dma.h b/arch/arm/include/asm/mach/dma.h index 9e614a1..3302cca 100644 --- a/arch/arm/include/asm/mach/dma.h +++ b/arch/arm/include/asm/mach/dma.h @@ -22,7 +22,7 @@ struct dma_ops { int (*residue)(unsigned int, dma_t *); /* optional */ int (*setspeed)(unsigned int, dma_t *, int); /* optional */ const char *type; -}; +} __do_const; struct dma_struct { void *addr; /* single DMA address */ diff --git a/arch/arm/include/asm/mach/map.h b/arch/arm/include/asm/mach/map.h index f98c7f3..e5c626d 100644 --- a/arch/arm/include/asm/mach/map.h +++ b/arch/arm/include/asm/mach/map.h @@ -23,17 +23,19 @@ struct map_desc { /* types 0-3 are defined in asm/io.h */ enum { - MT_UNCACHED = 4, - MT_CACHECLEAN, - MT_MINICLEAN, + MT_UNCACHED_RW = 4, + MT_CACHECLEAN_RO, + MT_MINICLEAN_RO, MT_LOW_VECTORS, MT_HIGH_VECTORS, - MT_MEMORY_RWX, + __MT_MEMORY_RWX, MT_MEMORY_RW, - MT_ROM, - MT_MEMORY_RWX_NONCACHED, + MT_MEMORY_RX, + MT_ROM_RX, + MT_MEMORY_RW_NONCACHED, + MT_MEMORY_RX_NONCACHED, MT_MEMORY_RW_DTCM, - MT_MEMORY_RWX_ITCM, + MT_MEMORY_RX_ITCM, MT_MEMORY_RW_SO, MT_MEMORY_DMA_READY, }; diff --git a/arch/arm/include/asm/outercache.h b/arch/arm/include/asm/outercache.h index 891a56b..48f337e 100644 --- a/arch/arm/include/asm/outercache.h +++ b/arch/arm/include/asm/outercache.h @@ -36,7 +36,7 @@ struct outer_cache_fns { /* This is an ARM L2C thing */ void (*write_sec)(unsigned long, unsigned); -}; +} __no_const; extern struct outer_cache_fns outer_cache; diff --git a/arch/arm/include/asm/page.h b/arch/arm/include/asm/page.h index 4355f0e..cd9168e 100644 --- a/arch/arm/include/asm/page.h +++ b/arch/arm/include/asm/page.h @@ -23,6 +23,7 @@ #else +#include #include /* @@ -114,7 +115,7 @@ 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 78a7793..e3dc06c 100644 --- a/arch/arm/include/asm/pgalloc.h +++ b/arch/arm/include/asm/pgalloc.h @@ -17,6 +17,7 @@ #include #include #include +#include #define check_pgt_cache() do { } while (0) @@ -43,6 +44,11 @@ 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 +57,7 @@ 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 */ @@ -128,6 +135,19 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte) __free_page(pte); } +static inline void __section_update(pmd_t *pmdp, unsigned long addr, pmdval_t prot) +{ +#ifdef CONFIG_ARM_LPAE + pmdp[0] = __pmd(pmd_val(pmdp[0]) | prot); +#else + if (addr & SECTION_SIZE) + pmdp[1] = __pmd(pmd_val(pmdp[1]) | prot); + else + pmdp[0] = __pmd(pmd_val(pmdp[0]) | prot); +#endif + flush_pmd_entry(pmdp); +} + static inline void __pmd_populate(pmd_t *pmdp, phys_addr_t pte, pmdval_t prot) { @@ -157,7 +177,7 @@ pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmdp, pte_t *ptep) static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmdp, pgtable_t ptep) { - __pmd_populate(pmdp, page_to_phys(ptep), _PAGE_USER_TABLE); + __pmd_populate(pmdp, page_to_phys(ptep), _PAGE_USER_TABLE | __supported_pmd_mask); } #define pmd_pgtable(pmd) pmd_page(pmd) diff --git a/arch/arm/include/asm/pgtable-2level-hwdef.h b/arch/arm/include/asm/pgtable-2level-hwdef.h index 5cfba15..f415e1a 100644 --- a/arch/arm/include/asm/pgtable-2level-hwdef.h +++ b/arch/arm/include/asm/pgtable-2level-hwdef.h @@ -20,12 +20,15 @@ #define PMD_TYPE_FAULT (_AT(pmdval_t, 0) << 0) #define PMD_TYPE_TABLE (_AT(pmdval_t, 1) << 0) #define PMD_TYPE_SECT (_AT(pmdval_t, 2) << 0) +#define PMD_PXNTABLE (_AT(pmdval_t, 1) << 2) /* v7 */ #define PMD_BIT4 (_AT(pmdval_t, 1) << 4) #define PMD_DOMAIN(x) (_AT(pmdval_t, (x)) << 5) #define PMD_PROTECTION (_AT(pmdval_t, 1) << 9) /* v5 */ + /* * - section */ +#define PMD_SECT_PXN (_AT(pmdval_t, 1) << 0) /* v7 */ #define PMD_SECT_BUFFERABLE (_AT(pmdval_t, 1) << 2) #define PMD_SECT_CACHEABLE (_AT(pmdval_t, 1) << 3) #define PMD_SECT_XN (_AT(pmdval_t, 1) << 4) /* v6 */ @@ -37,6 +40,7 @@ #define PMD_SECT_nG (_AT(pmdval_t, 1) << 17) /* v6 */ #define PMD_SECT_SUPER (_AT(pmdval_t, 1) << 18) /* v6 */ #define PMD_SECT_AF (_AT(pmdval_t, 0)) +#define PMD_SECT_RDONLY (_AT(pmdval_t, 0)) #define PMD_SECT_UNCACHED (_AT(pmdval_t, 0)) #define PMD_SECT_BUFFERED (PMD_SECT_BUFFERABLE) @@ -66,6 +70,7 @@ * - extended small page/tiny page */ #define PTE_EXT_XN (_AT(pteval_t, 1) << 0) /* v6 */ +#define PTE_EXT_PXN (_AT(pteval_t, 1) << 2) /* v7 */ #define PTE_EXT_AP_MASK (_AT(pteval_t, 3) << 4) #define PTE_EXT_AP0 (_AT(pteval_t, 1) << 4) #define PTE_EXT_AP1 (_AT(pteval_t, 2) << 4) diff --git a/arch/arm/include/asm/pgtable-2level.h b/arch/arm/include/asm/pgtable-2level.h index 219ac88..73ec32a 100644 --- a/arch/arm/include/asm/pgtable-2level.h +++ b/arch/arm/include/asm/pgtable-2level.h @@ -126,6 +126,9 @@ #define L_PTE_SHARED (_AT(pteval_t, 1) << 10) /* shared(v6), coherent(xsc3) */ #define L_PTE_NONE (_AT(pteval_t, 1) << 11) +/* Two-level page tables only have PXN in the PGD, not in the PTE. */ +#define L_PTE_PXN (_AT(pteval_t, 0)) + /* * These are the memory types, defined to be compatible with * pre-ARMv6 CPUs cacheable and bufferable bits: XXCB diff --git a/arch/arm/include/asm/pgtable-3level-hwdef.h b/arch/arm/include/asm/pgtable-3level-hwdef.h index 626989f..9d67a33 100644 --- a/arch/arm/include/asm/pgtable-3level-hwdef.h +++ b/arch/arm/include/asm/pgtable-3level-hwdef.h @@ -75,6 +75,7 @@ #define PTE_EXT_SHARED (_AT(pteval_t, 3) << 8) /* SH[1:0], inner shareable */ #define PTE_EXT_AF (_AT(pteval_t, 1) << 10) /* Access Flag */ #define PTE_EXT_NG (_AT(pteval_t, 1) << 11) /* nG */ +#define PTE_EXT_PXN (_AT(pteval_t, 1) << 53) /* PXN */ #define PTE_EXT_XN (_AT(pteval_t, 1) << 54) /* XN */ /* diff --git a/arch/arm/include/asm/pgtable-3level.h b/arch/arm/include/asm/pgtable-3level.h index 85c60ad..b0bbd7e 100644 --- a/arch/arm/include/asm/pgtable-3level.h +++ b/arch/arm/include/asm/pgtable-3level.h @@ -82,6 +82,7 @@ #define L_PTE_RDONLY (_AT(pteval_t, 1) << 7) /* AP[2] */ #define L_PTE_SHARED (_AT(pteval_t, 3) << 8) /* SH[1:0], inner shareable */ #define L_PTE_YOUNG (_AT(pteval_t, 1) << 10) /* AF */ +#define L_PTE_PXN (_AT(pteval_t, 1) << 53) /* PXN */ #define L_PTE_XN (_AT(pteval_t, 1) << 54) /* XN */ #define L_PTE_DIRTY (_AT(pteval_t, 1) << 55) /* unused */ #define L_PTE_SPECIAL (_AT(pteval_t, 1) << 56) /* unused */ @@ -95,6 +96,7 @@ /* * To be used in assembly code with the upper page attributes. */ +#define L_PTE_PXN_HIGH (1 << (53 - 32)) #define L_PTE_XN_HIGH (1 << (54 - 32)) #define L_PTE_DIRTY_HIGH (1 << (55 - 32)) diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h index 5478e5d..f5b5cb3 100644 --- a/arch/arm/include/asm/pgtable.h +++ b/arch/arm/include/asm/pgtable.h @@ -33,6 +33,9 @@ #include #endif +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) + /* * Just any arbitrary offset to the start of the vmalloc VM area: the * current 8MB value just means that there will be a 8MB "hole" after the @@ -48,6 +51,9 @@ #define LIBRARY_TEXT_START 0x0c000000 #ifndef __ASSEMBLY__ +extern pteval_t __supported_pte_mask; +extern pmdval_t __supported_pmd_mask; + extern void __pte_error(const char *file, int line, pte_t); extern void __pmd_error(const char *file, int line, pmd_t); extern void __pgd_error(const char *file, int line, pgd_t); @@ -56,6 +62,48 @@ extern void __pgd_error(const char *file, int line, pgd_t); #define pmd_ERROR(pmd) __pmd_error(__FILE__, __LINE__, pmd) #define pgd_ERROR(pgd) __pgd_error(__FILE__, __LINE__, pgd) +#define __HAVE_ARCH_PAX_OPEN_KERNEL +#define __HAVE_ARCH_PAX_CLOSE_KERNEL + +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) +#include +#include +#include + +static inline int test_domain(int domain, int domaintype) +{ + return ((current_thread_info()->cpu_domain) & domain_val(domain, 3)) == domain_val(domain, domaintype); +} +#endif + +#ifdef CONFIG_PAX_KERNEXEC +static inline unsigned long pax_open_kernel(void) { +#ifdef CONFIG_ARM_LPAE + /* TODO */ +#else + preempt_disable(); + BUG_ON(test_domain(DOMAIN_KERNEL, DOMAIN_KERNEXEC)); + modify_domain(DOMAIN_KERNEL, DOMAIN_KERNEXEC); +#endif + return 0; +} + +static inline unsigned long pax_close_kernel(void) { +#ifdef CONFIG_ARM_LPAE + /* TODO */ +#else + BUG_ON(test_domain(DOMAIN_KERNEL, DOMAIN_MANAGER)); + /* DOMAIN_MANAGER = "client" under KERNEXEC */ + modify_domain(DOMAIN_KERNEL, DOMAIN_MANAGER); + preempt_enable_no_resched(); +#endif + return 0; +} +#else +static inline unsigned long pax_open_kernel(void) { return 0; } +static inline unsigned long pax_close_kernel(void) { return 0; } +#endif + /* * This is the lowest virtual address we can permit any user space * mapping to be mapped at. This is particularly important for @@ -75,8 +123,8 @@ extern void __pgd_error(const char *file, int line, pgd_t); /* * The pgprot_* and protection_map entries will be fixed up in runtime * to include the cachable and bufferable bits based on memory policy, - * as well as any architecture dependent bits like global/ASID and SMP - * shared mapping bits. + * as well as any architecture dependent bits like global/ASID, PXN, + * and SMP shared mapping bits. */ #define _L_PTE_DEFAULT L_PTE_PRESENT | L_PTE_YOUNG @@ -265,7 +313,7 @@ static inline pte_t pte_mkspecial(pte_t pte) { return pte; } static inline pte_t pte_modify(pte_t pte, pgprot_t newprot) { const pteval_t mask = L_PTE_XN | L_PTE_RDONLY | L_PTE_USER | - L_PTE_NONE | L_PTE_VALID; + L_PTE_NONE | L_PTE_VALID | __supported_pte_mask; pte_val(pte) = (pte_val(pte) & ~mask) | (pgprot_val(newprot) & mask); return pte; } diff --git a/arch/arm/include/asm/psci.h b/arch/arm/include/asm/psci.h index c25ef3e..735f14b 100644 --- a/arch/arm/include/asm/psci.h +++ b/arch/arm/include/asm/psci.h @@ -32,7 +32,7 @@ struct psci_operations { int (*affinity_info)(unsigned long target_affinity, unsigned long lowest_affinity_level); int (*migrate_info_type)(void); -}; +} __no_const; extern struct psci_operations psci_ops; extern struct smp_operations psci_smp_ops; diff --git a/arch/arm/include/asm/smp.h b/arch/arm/include/asm/smp.h index 2ec765c..beb1fe16 100644 --- a/arch/arm/include/asm/smp.h +++ b/arch/arm/include/asm/smp.h @@ -113,7 +113,7 @@ struct smp_operations { int (*cpu_disable)(unsigned int cpu); #endif #endif -}; +} __no_const; struct of_cpu_method { const char *method; diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h index e4e4208..086684a 100644 --- a/arch/arm/include/asm/thread_info.h +++ b/arch/arm/include/asm/thread_info.h @@ -88,9 +88,9 @@ struct thread_info { .flags = 0, \ .preempt_count = INIT_PREEMPT_COUNT, \ .addr_limit = KERNEL_DS, \ - .cpu_domain = domain_val(DOMAIN_USER, DOMAIN_MANAGER) | \ - domain_val(DOMAIN_KERNEL, DOMAIN_MANAGER) | \ - domain_val(DOMAIN_IO, DOMAIN_CLIENT), \ + .cpu_domain = domain_val(DOMAIN_USER, DOMAIN_USERCLIENT) | \ + domain_val(DOMAIN_KERNEL, DOMAIN_KERNELCLIENT) | \ + domain_val(DOMAIN_IO, DOMAIN_KERNELCLIENT), \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ @@ -164,7 +164,11 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, #define TIF_SYSCALL_AUDIT 9 #define TIF_SYSCALL_TRACEPOINT 10 #define TIF_SECCOMP 11 /* seccomp syscall filtering active */ -#define TIF_NOHZ 12 /* in adaptive nohz mode */ +/* within 8 bits of TIF_SYSCALL_TRACE + * to meet flexible second operand requirements + */ +#define TIF_GRSEC_SETXID 12 +#define TIF_NOHZ 13 /* in adaptive nohz mode */ #define TIF_USING_IWMMXT 17 #define TIF_MEMDIE 18 /* is terminating due to OOM killer */ #define TIF_RESTORE_SIGMASK 20 @@ -178,10 +182,11 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, #define _TIF_SYSCALL_TRACEPOINT (1 << TIF_SYSCALL_TRACEPOINT) #define _TIF_SECCOMP (1 << TIF_SECCOMP) #define _TIF_USING_IWMMXT (1 << TIF_USING_IWMMXT) +#define _TIF_GRSEC_SETXID (1 << TIF_GRSEC_SETXID) /* Checks for any syscall work in entry-common.S */ #define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ - _TIF_SYSCALL_TRACEPOINT | _TIF_SECCOMP) + _TIF_SYSCALL_TRACEPOINT | _TIF_SECCOMP | _TIF_GRSEC_SETXID) /* * Change these and you break ASM code in entry-common.S diff --git a/arch/arm/include/asm/tls.h b/arch/arm/include/asm/tls.h index 5f833f7..76e6644 100644 --- a/arch/arm/include/asm/tls.h +++ b/arch/arm/include/asm/tls.h @@ -3,6 +3,7 @@ #include #include +#include #ifdef __ASSEMBLY__ #include @@ -89,7 +90,9 @@ static inline void set_tls(unsigned long val) * at 0xffff0fe0 must be used instead. (see * entry-armv.S for details) */ + pax_open_kernel(); *((unsigned int *)0xffff0ff0) = val; + pax_close_kernel(); #endif } diff --git a/arch/arm/include/asm/uaccess.h b/arch/arm/include/asm/uaccess.h index 75d9579..b5b40e4 100644 --- a/arch/arm/include/asm/uaccess.h +++ b/arch/arm/include/asm/uaccess.h @@ -18,6 +18,7 @@ #include #include #include +#include #ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS #include @@ -70,11 +71,38 @@ extern int __put_user_bad(void); static inline void set_fs(mm_segment_t fs) { current_thread_info()->addr_limit = fs; - modify_domain(DOMAIN_KERNEL, fs ? DOMAIN_CLIENT : DOMAIN_MANAGER); + modify_domain(DOMAIN_KERNEL, fs ? DOMAIN_KERNELCLIENT : DOMAIN_MANAGER); } #define segment_eq(a,b) ((a) == (b)) +#define __HAVE_ARCH_PAX_OPEN_USERLAND +#define __HAVE_ARCH_PAX_CLOSE_USERLAND + +static inline void pax_open_userland(void) +{ + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (segment_eq(get_fs(), USER_DS)) { + BUG_ON(test_domain(DOMAIN_USER, DOMAIN_UDEREF)); + modify_domain(DOMAIN_USER, DOMAIN_UDEREF); + } +#endif + +} + +static inline void pax_close_userland(void) +{ + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (segment_eq(get_fs(), USER_DS)) { + BUG_ON(test_domain(DOMAIN_USER, DOMAIN_NOACCESS)); + modify_domain(DOMAIN_USER, DOMAIN_NOACCESS); + } +#endif + +} + #define __addr_ok(addr) ({ \ unsigned long flag; \ __asm__("cmp %2, %0; movlo %0, #0" \ @@ -150,8 +178,12 @@ extern int __get_user_4(void *); #define get_user(x,p) \ ({ \ + int __e; \ might_fault(); \ - __get_user_check(x,p); \ + pax_open_userland(); \ + __e = __get_user_check(x,p); \ + pax_close_userland(); \ + __e; \ }) extern int __put_user_1(void *, unsigned int); @@ -196,8 +228,12 @@ extern int __put_user_8(void *, unsigned long long); #define put_user(x,p) \ ({ \ + int __e; \ might_fault(); \ - __put_user_check(x,p); \ + pax_open_userland(); \ + __e = __put_user_check(x,p); \ + pax_close_userland(); \ + __e; \ }) #else /* CONFIG_MMU */ @@ -221,6 +257,7 @@ static inline void set_fs(mm_segment_t fs) #endif /* CONFIG_MMU */ +#define access_ok_noprefault(type,addr,size) access_ok((type),(addr),(size)) #define access_ok(type,addr,size) (__range_ok(addr,size) == 0) #define user_addr_max() \ @@ -238,13 +275,17 @@ static inline void set_fs(mm_segment_t fs) #define __get_user(x,ptr) \ ({ \ long __gu_err = 0; \ + pax_open_userland(); \ __get_user_err((x),(ptr),__gu_err); \ + pax_close_userland(); \ __gu_err; \ }) #define __get_user_error(x,ptr,err) \ ({ \ + pax_open_userland(); \ __get_user_err((x),(ptr),err); \ + pax_close_userland(); \ (void) 0; \ }) @@ -320,13 +361,17 @@ do { \ #define __put_user(x,ptr) \ ({ \ long __pu_err = 0; \ + pax_open_userland(); \ __put_user_err((x),(ptr),__pu_err); \ + pax_close_userland(); \ __pu_err; \ }) #define __put_user_error(x,ptr,err) \ ({ \ + pax_open_userland(); \ __put_user_err((x),(ptr),err); \ + pax_close_userland(); \ (void) 0; \ }) @@ -426,11 +471,44 @@ 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) +{ + unsigned long ret; + + check_object_size(to, n, false); + pax_open_userland(); + ret = ___copy_from_user(to, from, n); + pax_close_userland(); + return ret; +} + +static inline unsigned long __must_check __copy_to_user(void __user *to, const void *from, unsigned long n) +{ + unsigned long ret; + + check_object_size(from, n, true); + pax_open_userland(); + ret = ___copy_to_user(to, from, n); + pax_close_userland(); + return ret; +} + 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(void __user *addr, unsigned long n); extern unsigned long __must_check __clear_user_std(void __user *addr, unsigned long n); + +static inline unsigned long __must_check __clear_user(void __user *addr, unsigned long n) +{ + unsigned long ret; + pax_open_userland(); + ret = ___clear_user(addr, n); + pax_close_userland(); + return ret; +} + #else #define __copy_from_user(to,from,n) (memcpy(to, (void __force *)from, n), 0) #define __copy_to_user(to,from,n) (memcpy((void __force *)to, from, n), 0) @@ -439,6 +517,9 @@ extern unsigned long __must_check __clear_user_std(void __user *addr, unsigned l 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 */ @@ -448,6 +529,9 @@ 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/include/uapi/asm/ptrace.h b/arch/arm/include/uapi/asm/ptrace.h index 5af0ed1..cea83883 100644 --- a/arch/arm/include/uapi/asm/ptrace.h +++ b/arch/arm/include/uapi/asm/ptrace.h @@ -92,7 +92,7 @@ * ARMv7 groups of PSR bits */ #define APSR_MASK 0xf80f0000 /* N, Z, C, V, Q and GE flags */ -#define PSR_ISET_MASK 0x01000010 /* ISA state (J, T) mask */ +#define PSR_ISET_MASK 0x01000020 /* ISA state (J, T) mask */ #define PSR_IT_MASK 0x0600fc00 /* If-Then execution state mask */ #define PSR_ENDIAN_MASK 0x00000200 /* Endianness state mask */ diff --git a/arch/arm/kernel/armksyms.c b/arch/arm/kernel/armksyms.c index f7b450f..f5364c5 100644 --- a/arch/arm/kernel/armksyms.c +++ b/arch/arm/kernel/armksyms.c @@ -55,7 +55,7 @@ EXPORT_SYMBOL(arm_delay_ops); /* networking */ EXPORT_SYMBOL(csum_partial); -EXPORT_SYMBOL(csum_partial_copy_from_user); +EXPORT_SYMBOL(__csum_partial_copy_from_user); EXPORT_SYMBOL(csum_partial_copy_nocheck); EXPORT_SYMBOL(__csum_ipv6_magic); @@ -91,9 +91,9 @@ EXPORT_SYMBOL(__memzero); #ifdef CONFIG_MMU EXPORT_SYMBOL(copy_page); -EXPORT_SYMBOL(__copy_from_user); -EXPORT_SYMBOL(__copy_to_user); -EXPORT_SYMBOL(__clear_user); +EXPORT_SYMBOL(___copy_from_user); +EXPORT_SYMBOL(___copy_to_user); +EXPORT_SYMBOL(___clear_user); EXPORT_SYMBOL(__get_user_1); EXPORT_SYMBOL(__get_user_2); diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 52a949a..d8bbcab 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -47,6 +47,87 @@ 9997: .endm + .macro pax_enter_kernel +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + @ make aligned space for saved DACR + sub sp, sp, #8 + @ save regs + stmdb sp!, {r1, r2} + @ read DACR from cpu_domain into r1 + mov r2, sp + @ assume 8K pages, since we have to split the immediate in two + bic r2, r2, #(0x1fc0) + bic r2, r2, #(0x3f) + ldr r1, [r2, #TI_CPU_DOMAIN] + @ store old DACR on stack + str r1, [sp, #8] +#ifdef CONFIG_PAX_KERNEXEC + @ set type of DOMAIN_KERNEL to DOMAIN_KERNELCLIENT + bic r1, r1, #(domain_val(DOMAIN_KERNEL, 3)) + orr r1, r1, #(domain_val(DOMAIN_KERNEL, DOMAIN_KERNELCLIENT)) +#endif +#ifdef CONFIG_PAX_MEMORY_UDEREF + @ set current DOMAIN_USER to DOMAIN_NOACCESS + bic r1, r1, #(domain_val(DOMAIN_USER, 3)) +#endif + @ write r1 to current_thread_info()->cpu_domain + str r1, [r2, #TI_CPU_DOMAIN] + @ write r1 to DACR + mcr p15, 0, r1, c3, c0, 0 + @ instruction sync + instr_sync + @ restore regs + ldmia sp!, {r1, r2} +#endif + .endm + + .macro pax_open_userland +#ifdef CONFIG_PAX_MEMORY_UDEREF + @ save regs + stmdb sp!, {r0, r1} + @ read DACR from cpu_domain into r1 + mov r0, sp + @ assume 8K pages, since we have to split the immediate in two + bic r0, r0, #(0x1fc0) + bic r0, r0, #(0x3f) + ldr r1, [r0, #TI_CPU_DOMAIN] + @ set current DOMAIN_USER to DOMAIN_CLIENT + bic r1, r1, #(domain_val(DOMAIN_USER, 3)) + orr r1, r1, #(domain_val(DOMAIN_USER, DOMAIN_UDEREF)) + @ write r1 to current_thread_info()->cpu_domain + str r1, [r0, #TI_CPU_DOMAIN] + @ write r1 to DACR + mcr p15, 0, r1, c3, c0, 0 + @ instruction sync + instr_sync + @ restore regs + ldmia sp!, {r0, r1} +#endif + .endm + + .macro pax_close_userland +#ifdef CONFIG_PAX_MEMORY_UDEREF + @ save regs + stmdb sp!, {r0, r1} + @ read DACR from cpu_domain into r1 + mov r0, sp + @ assume 8K pages, since we have to split the immediate in two + bic r0, r0, #(0x1fc0) + bic r0, r0, #(0x3f) + ldr r1, [r0, #TI_CPU_DOMAIN] + @ set current DOMAIN_USER to DOMAIN_NOACCESS + bic r1, r1, #(domain_val(DOMAIN_USER, 3)) + @ write r1 to current_thread_info()->cpu_domain + str r1, [r0, #TI_CPU_DOMAIN] + @ write r1 to DACR + mcr p15, 0, r1, c3, c0, 0 + @ instruction sync + instr_sync + @ restore regs + ldmia sp!, {r0, r1} +#endif + .endm + .macro pabt_helper @ PABORT handler takes pt_regs in r2, fault address in r4 and psr in r5 #ifdef MULTI_PABORT @@ -89,11 +170,15 @@ * Invalid mode handlers */ .macro inv_entry, reason + + pax_enter_kernel + sub sp, sp, #S_FRAME_SIZE ARM( stmib sp, {r1 - lr} ) THUMB( stmia sp, {r0 - r12} ) THUMB( str sp, [sp, #S_SP] ) THUMB( str lr, [sp, #S_LR] ) + mov r1, #\reason .endm @@ -149,7 +234,11 @@ ENDPROC(__und_invalid) .macro svc_entry, stack_hole=0 UNWIND(.fnstart ) UNWIND(.save {r0 - pc} ) + + pax_enter_kernel + sub sp, sp, #(S_FRAME_SIZE + \stack_hole - 4) + #ifdef CONFIG_THUMB2_KERNEL SPFIX( str r0, [sp] ) @ temporarily saved SPFIX( mov r0, sp ) @@ -164,7 +253,12 @@ ENDPROC(__und_invalid) ldmia r0, {r3 - r5} add r7, sp, #S_SP - 4 @ here for interlock avoidance mov r6, #-1 @ "" "" "" "" +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + @ offset sp by 8 as done in pax_enter_kernel + add r2, sp, #(S_FRAME_SIZE + \stack_hole + 4) +#else add r2, sp, #(S_FRAME_SIZE + \stack_hole - 4) +#endif SPFIX( addeq r2, r2, #4 ) str r3, [sp, #-4]! @ save the "real" r0 copied @ from the exception stack @@ -317,6 +411,9 @@ ENDPROC(__pabt_svc) .macro usr_entry UNWIND(.fnstart ) UNWIND(.cantunwind ) @ don't unwind the user space + + pax_enter_kernel_user + sub sp, sp, #S_FRAME_SIZE ARM( stmib sp, {r1 - r12} ) THUMB( stmia sp, {r0 - r12} ) @@ -421,7 +518,9 @@ __und_usr: tst r3, #PSR_T_BIT @ Thumb mode? bne __und_usr_thumb sub r4, r2, #4 @ ARM instr at LR - 4 + pax_open_userland 1: ldrt r0, [r4] + pax_close_userland ARM_BE8(rev r0, r0) @ little endian instruction @ r0 = 32-bit ARM instruction which caused the exception @@ -455,11 +554,15 @@ __und_usr_thumb: */ .arch armv6t2 #endif + pax_open_userland 2: ldrht r5, [r4] + pax_close_userland ARM_BE8(rev16 r5, r5) @ little endian instruction cmp r5, #0xe800 @ 32bit instruction if xx != 0 blo __und_usr_fault_16 @ 16bit undefined instruction + pax_open_userland 3: ldrht r0, [r2] + pax_close_userland ARM_BE8(rev16 r0, r0) @ little endian instruction add r2, r2, #2 @ r2 is PC + 2, make it PC + 4 str r2, [sp, #S_PC] @ it's a 2x16bit instr, update @@ -489,7 +592,8 @@ ENDPROC(__und_usr) */ .pushsection .fixup, "ax" .align 2 -4: str r4, [sp, #S_PC] @ retry current instruction +4: pax_close_userland + str r4, [sp, #S_PC] @ retry current instruction mov pc, r9 .popsection .pushsection __ex_table,"a" @@ -698,7 +802,7 @@ ENTRY(__switch_to) THUMB( str lr, [ip], #4 ) ldr r4, [r2, #TI_TP_VALUE] ldr r5, [r2, #TI_TP_VALUE + 4] -#ifdef CONFIG_CPU_USE_DOMAINS +#if defined(CONFIG_CPU_USE_DOMAINS) || defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) ldr r6, [r2, #TI_CPU_DOMAIN] #endif switch_tls r1, r4, r5, r3, r7 @@ -707,7 +811,7 @@ ENTRY(__switch_to) ldr r8, =__stack_chk_guard ldr r7, [r7, #TSK_STACK_CANARY] #endif -#ifdef CONFIG_CPU_USE_DOMAINS +#if defined(CONFIG_CPU_USE_DOMAINS) || defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) mcr p15, 0, r6, c3, c0, 0 @ Set domain register #endif mov r5, r0 diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 7139d4a..feaf37f 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -10,18 +10,46 @@ #include #include +#include #include +#include "entry-header.S" + #ifdef CONFIG_NEED_RET_TO_USER #include #else .macro arch_ret_to_user, tmp1, tmp2 +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + @ save regs + stmdb sp!, {r1, r2} + @ read DACR from cpu_domain into r1 + mov r2, sp + @ assume 8K pages, since we have to split the immediate in two + bic r2, r2, #(0x1fc0) + bic r2, r2, #(0x3f) + ldr r1, [r2, #TI_CPU_DOMAIN] +#ifdef CONFIG_PAX_KERNEXEC + @ set type of DOMAIN_KERNEL to DOMAIN_KERNELCLIENT + bic r1, r1, #(domain_val(DOMAIN_KERNEL, 3)) + orr r1, r1, #(domain_val(DOMAIN_KERNEL, DOMAIN_KERNELCLIENT)) +#endif +#ifdef CONFIG_PAX_MEMORY_UDEREF + @ set current DOMAIN_USER to DOMAIN_UDEREF + bic r1, r1, #(domain_val(DOMAIN_USER, 3)) + orr r1, r1, #(domain_val(DOMAIN_USER, DOMAIN_UDEREF)) +#endif + @ write r1 to current_thread_info()->cpu_domain + str r1, [r2, #TI_CPU_DOMAIN] + @ write r1 to DACR + mcr p15, 0, r1, c3, c0, 0 + @ instruction sync + instr_sync + @ restore regs + ldmia sp!, {r1, r2} +#endif .endm #endif -#include "entry-header.S" - - .align 5 /* * This is the fast syscall return path. We do as little as @@ -405,6 +433,12 @@ ENTRY(vector_swi) USER( ldr scno, [lr, #-4] ) @ get SWI instruction #endif + /* + * do this here to avoid a performance hit of wrapping the code above + * that directly dereferences userland to parse the SWI instruction + */ + pax_enter_kernel_user + adr tbl, sys_call_table @ load syscall table pointer #if defined(CONFIG_OABI_COMPAT) diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 0325dbf..e8e47ff 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -188,6 +188,60 @@ msr cpsr_c, \rtemp @ switch back to the SVC mode .endm + .macro pax_enter_kernel_user +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + @ save regs + stmdb sp!, {r0, r1} + @ read DACR from cpu_domain into r1 + mov r0, sp + @ assume 8K pages, since we have to split the immediate in two + bic r0, r0, #(0x1fc0) + bic r0, r0, #(0x3f) + ldr r1, [r0, #TI_CPU_DOMAIN] +#ifdef CONFIG_PAX_MEMORY_UDEREF + @ set current DOMAIN_USER to DOMAIN_NOACCESS + bic r1, r1, #(domain_val(DOMAIN_USER, 3)) +#endif +#ifdef CONFIG_PAX_KERNEXEC + @ set current DOMAIN_KERNEL to DOMAIN_KERNELCLIENT + bic r1, r1, #(domain_val(DOMAIN_KERNEL, 3)) + orr r1, r1, #(domain_val(DOMAIN_KERNEL, DOMAIN_KERNELCLIENT)) +#endif + @ write r1 to current_thread_info()->cpu_domain + str r1, [r0, #TI_CPU_DOMAIN] + @ write r1 to DACR + mcr p15, 0, r1, c3, c0, 0 + @ instruction sync + instr_sync + @ restore regs + ldmia sp!, {r0, r1} +#endif + .endm + + .macro pax_exit_kernel +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + @ save regs + stmdb sp!, {r0, r1} + @ read old DACR from stack into r1 + ldr r1, [sp, #(8 + S_SP)] + sub r1, r1, #8 + ldr r1, [r1] + + @ write r1 to current_thread_info()->cpu_domain + mov r0, sp + @ assume 8K pages, since we have to split the immediate in two + bic r0, r0, #(0x1fc0) + bic r0, r0, #(0x3f) + str r1, [r0, #TI_CPU_DOMAIN] + @ write r1 to DACR + mcr p15, 0, r1, c3, c0, 0 + @ instruction sync + instr_sync + @ restore regs + ldmia sp!, {r0, r1} +#endif + .endm + #ifndef CONFIG_THUMB2_KERNEL .macro svc_exit, rpsr, irq = 0 .if \irq != 0 @@ -207,6 +261,9 @@ blne trace_hardirqs_off #endif .endif + + pax_exit_kernel + msr spsr_cxsf, \rpsr #if defined(CONFIG_CPU_V6) || defined(CONFIG_CPU_32v6K) @ We must avoid clrex due to Cortex-A15 erratum #830321 @@ -260,6 +317,9 @@ blne trace_hardirqs_off #endif .endif + + pax_exit_kernel + ldr lr, [sp, #S_SP] @ top of the stack ldrd r0, r1, [sp, #S_LR] @ calling lr and pc diff --git a/arch/arm/kernel/fiq.c b/arch/arm/kernel/fiq.c index 918875d..cd5fa27 100644 --- a/arch/arm/kernel/fiq.c +++ b/arch/arm/kernel/fiq.c @@ -87,7 +87,10 @@ void set_fiq_handler(void *start, unsigned int length) void *base = vectors_page; unsigned offset = FIQ_OFFSET; + pax_open_kernel(); memcpy(base + offset, start, length); + pax_close_kernel(); + if (!cache_is_vipt_nonaliasing()) flush_icache_range((unsigned long)base + offset, offset + length); diff --git a/arch/arm/kernel/head.S b/arch/arm/kernel/head.S index 2c35f0f..7747ee6 100644 --- a/arch/arm/kernel/head.S +++ b/arch/arm/kernel/head.S @@ -437,7 +437,7 @@ __enable_mmu: mov r5, #(domain_val(DOMAIN_USER, DOMAIN_MANAGER) | \ domain_val(DOMAIN_KERNEL, DOMAIN_MANAGER) | \ domain_val(DOMAIN_TABLE, DOMAIN_MANAGER) | \ - domain_val(DOMAIN_IO, DOMAIN_CLIENT)) + domain_val(DOMAIN_IO, DOMAIN_KERNELCLIENT)) mcr p15, 0, r5, c3, c0, 0 @ load domain access register mcr p15, 0, r4, c2, c0, 0 @ load page table pointer #endif diff --git a/arch/arm/kernel/module.c b/arch/arm/kernel/module.c index 45e4781..8eac93d 100644 --- a/arch/arm/kernel/module.c +++ b/arch/arm/kernel/module.c @@ -38,12 +38,39 @@ #endif #ifdef CONFIG_MMU -void *module_alloc(unsigned long size) +static inline void *__module_alloc(unsigned long size, pgprot_t prot) { + if (!size || PAGE_ALIGN(size) > MODULES_END - MODULES_VADDR) + return NULL; return __vmalloc_node_range(size, 1, MODULES_VADDR, MODULES_END, - GFP_KERNEL, PAGE_KERNEL_EXEC, NUMA_NO_NODE, + GFP_KERNEL, prot, NUMA_NO_NODE, __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 +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_EXEC); +} +EXPORT_SYMBOL(module_alloc_exec); +#endif #endif int diff --git a/arch/arm/kernel/patch.c b/arch/arm/kernel/patch.c index 07314af..c46655c 100644 --- a/arch/arm/kernel/patch.c +++ b/arch/arm/kernel/patch.c @@ -18,6 +18,7 @@ void __kprobes __patch_text(void *addr, unsigned int insn) bool thumb2 = IS_ENABLED(CONFIG_THUMB2_KERNEL); int size; + pax_open_kernel(); if (thumb2 && __opcode_is_thumb16(insn)) { *(u16 *)addr = __opcode_to_mem_thumb16(insn); size = sizeof(u16); @@ -39,6 +40,7 @@ void __kprobes __patch_text(void *addr, unsigned int insn) *(u32 *)addr = insn; size = sizeof(u32); } + pax_close_kernel(); flush_icache_range((uintptr_t)(addr), (uintptr_t)(addr) + size); diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c index a35f6eb..7af43a0 100644 --- a/arch/arm/kernel/process.c +++ b/arch/arm/kernel/process.c @@ -212,6 +212,7 @@ void machine_power_off(void) if (pm_power_off) pm_power_off(); + BUG(); } /* @@ -225,7 +226,7 @@ void machine_power_off(void) * executing pre-reset code, and using RAM that the primary CPU's code wishes * to use. Implementing such co-ordination would be essentially impossible. */ -void machine_restart(char *cmd) +__noreturn void machine_restart(char *cmd) { local_irq_disable(); smp_send_stop(); @@ -248,8 +249,8 @@ void __show_regs(struct pt_regs *regs) show_regs_print_info(KERN_DEFAULT); - print_symbol("PC is at %s\n", instruction_pointer(regs)); - print_symbol("LR is at %s\n", regs->ARM_lr); + printk("PC is at %pA\n", (void *)instruction_pointer(regs)); + printk("LR is at %pA\n", (void *)regs->ARM_lr); printk("pc : [<%08lx>] lr : [<%08lx>] psr: %08lx\n" "sp : %08lx ip : %08lx fp : %08lx\n", regs->ARM_pc, regs->ARM_lr, regs->ARM_cpsr, @@ -427,12 +428,6 @@ 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 #ifdef CONFIG_KUSER_HELPERS /* @@ -448,7 +443,7 @@ static struct vm_area_struct gate_vma = { static int __init gate_vma_init(void) { - gate_vma.vm_page_prot = PAGE_READONLY_EXEC; + gate_vma.vm_page_prot = vm_get_page_prot(gate_vma.vm_flags); return 0; } arch_initcall(gate_vma_init); @@ -474,41 +469,16 @@ int in_gate_area_no_mm(unsigned long addr) const char *arch_vma_name(struct vm_area_struct *vma) { - return is_gate_vma(vma) ? "[vectors]" : - (vma->vm_mm && vma->vm_start == vma->vm_mm->context.sigpage) ? - "[sigpage]" : NULL; + return is_gate_vma(vma) ? "[vectors]" : NULL; } -static struct page *signal_page; -extern struct page *get_signal_page(void); - int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) { struct mm_struct *mm = current->mm; - unsigned long addr; - int ret; - - if (!signal_page) - signal_page = get_signal_page(); - if (!signal_page) - return -ENOMEM; down_write(&mm->mmap_sem); - addr = get_unmapped_area(NULL, 0, PAGE_SIZE, 0, 0); - if (IS_ERR_VALUE(addr)) { - ret = addr; - goto up_fail; - } - - ret = install_special_mapping(mm, addr, PAGE_SIZE, - VM_READ | VM_EXEC | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC, - &signal_page); - - if (ret == 0) - mm->context.sigpage = addr; - - up_fail: + mm->context.sigpage = (PAGE_OFFSET + (get_random_int() % 0x3FFEFFE0)) & 0xFFFFFFFC; up_write(&mm->mmap_sem); - return ret; + return 0; } #endif diff --git a/arch/arm/kernel/psci.c b/arch/arm/kernel/psci.c index f73891b..cf3004e 100644 --- a/arch/arm/kernel/psci.c +++ b/arch/arm/kernel/psci.c @@ -28,7 +28,7 @@ #include #include -struct psci_operations psci_ops; +struct psci_operations psci_ops __read_only; static int (*invoke_psci_fn)(u32, u32, u32, u32); typedef int (*psci_initcall_t)(const struct device_node *); diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index 0c27ed6..b67388e 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -928,10 +928,19 @@ static void tracehook_report_syscall(struct pt_regs *regs, regs->ARM_ip = ip; } +#ifdef CONFIG_GRKERNSEC_SETXID +extern void gr_delayed_cred_worker(void); +#endif + asmlinkage int syscall_trace_enter(struct pt_regs *regs, int scno) { current_thread_info()->syscall = scno; +#ifdef CONFIG_GRKERNSEC_SETXID + if (unlikely(test_and_clear_thread_flag(TIF_GRSEC_SETXID))) + gr_delayed_cred_worker(); +#endif + /* Do the secure computing check first; failures should be fast. */ if (secure_computing(scno) == -1) return -1; diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index 8a16ee5..4f560e5 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -104,21 +104,23 @@ EXPORT_SYMBOL(elf_hwcap); unsigned int elf_hwcap2 __read_mostly; EXPORT_SYMBOL(elf_hwcap2); +pteval_t __supported_pte_mask __read_only; +pmdval_t __supported_pmd_mask __read_only; #ifdef MULTI_CPU -struct processor processor __read_mostly; +struct processor processor __read_only; #endif #ifdef MULTI_TLB -struct cpu_tlb_fns cpu_tlb __read_mostly; +struct cpu_tlb_fns cpu_tlb __read_only; #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 @@ -251,9 +253,13 @@ static int __get_cpu_architecture(void) asm("mrc p15, 0, %0, c0, c1, 4" : "=r" (mmfr0)); if ((mmfr0 & 0x0000000f) >= 0x00000003 || - (mmfr0 & 0x000000f0) >= 0x00000030) + (mmfr0 & 0x000000f0) >= 0x00000030) { cpu_arch = CPU_ARCH_ARMv7; - else if ((mmfr0 & 0x0000000f) == 0x00000002 || + if ((mmfr0 & 0x0000000f) == 0x00000005 || (mmfr0 & 0x0000000f) == 0x00000004) { + __supported_pte_mask |= L_PTE_PXN; + __supported_pmd_mask |= PMD_PXNTABLE; + } + } else if ((mmfr0 & 0x0000000f) == 0x00000002 || (mmfr0 & 0x000000f0) == 0x00000020) cpu_arch = CPU_ARCH_ARMv6; else diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c index bd19834..e4d8c66 100644 --- a/arch/arm/kernel/signal.c +++ b/arch/arm/kernel/signal.c @@ -24,8 +24,6 @@ extern const unsigned long sigreturn_codes[7]; -static unsigned long signal_return_offset; - #ifdef CONFIG_CRUNCH static int preserve_crunch_context(struct crunch_sigframe __user *frame) { @@ -396,8 +394,7 @@ setup_return(struct pt_regs *regs, struct ksignal *ksig, * except when the MPU has protected the vectors * page from PL0 */ - retcode = mm->context.sigpage + signal_return_offset + - (idx << 2) + thumb; + retcode = mm->context.sigpage + (idx << 2) + thumb; } else #endif { @@ -604,33 +601,3 @@ do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall) } while (thread_flags & _TIF_WORK_MASK); return 0; } - -struct page *get_signal_page(void) -{ - unsigned long ptr; - unsigned offset; - struct page *page; - void *addr; - - page = alloc_pages(GFP_KERNEL, 0); - - if (!page) - return NULL; - - addr = page_address(page); - - /* Give the signal return code some randomness */ - offset = 0x200 + (get_random_int() & 0x7fc); - signal_return_offset = offset; - - /* - * Copy signal return handlers into the vector page, and - * set sigreturn to be a pointer to these. - */ - memcpy(addr + offset, sigreturn_codes, sizeof(sigreturn_codes)); - - ptr = (unsigned long)addr + offset; - flush_icache_range(ptr, ptr + sizeof(sigreturn_codes)); - - return page; -} diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c index 7c4fada..8581286 100644 --- a/arch/arm/kernel/smp.c +++ b/arch/arm/kernel/smp.c @@ -73,7 +73,7 @@ enum ipi_msg_type { static DECLARE_COMPLETION(cpu_running); -static struct smp_operations smp_ops; +static struct smp_operations smp_ops __read_only; void __init smp_set_ops(struct smp_operations *ops) { diff --git a/arch/arm/kernel/tcm.c b/arch/arm/kernel/tcm.c index 7a3be1d..b00c7de 100644 --- a/arch/arm/kernel/tcm.c +++ b/arch/arm/kernel/tcm.c @@ -61,7 +61,7 @@ static struct map_desc itcm_iomap[] __initdata = { .virtual = ITCM_OFFSET, .pfn = __phys_to_pfn(ITCM_OFFSET), .length = 0, - .type = MT_MEMORY_RWX_ITCM, + .type = MT_MEMORY_RX_ITCM, } }; @@ -267,7 +267,9 @@ no_dtcm: start = &__sitcm_text; end = &__eitcm_text; ram = &__itcm_start; + pax_open_kernel(); memcpy(start, ram, itcm_code_sz); + pax_close_kernel(); pr_debug("CPU ITCM: copied code from %p - %p\n", start, end); itcm_present = true; diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index da11b28..1e2696e 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -62,7 +62,7 @@ static void dump_mem(const char *, const char *, unsigned long, unsigned long); void dump_backtrace_entry(unsigned long where, unsigned long from, unsigned long frame) { #ifdef CONFIG_KALLSYMS - printk("[<%08lx>] (%ps) from [<%08lx>] (%pS)\n", where, (void *)where, from, (void *)from); + printk("[<%08lx>] (%pA) from [<%08lx>] (%pA)\n", where, (void *)where, from, (void *)from); #else printk("Function entered at [<%08lx>] from [<%08lx>]\n", where, from); #endif @@ -264,6 +264,8 @@ static arch_spinlock_t die_lock = __ARCH_SPIN_LOCK_UNLOCKED; static int die_owner = -1; static unsigned int die_nest_count; +extern void gr_handle_kernel_exploit(void); + static unsigned long oops_begin(void) { int cpu; @@ -306,6 +308,9 @@ static void oops_end(unsigned long flags, struct pt_regs *regs, int signr) panic("Fatal exception in interrupt"); if (panic_on_oops) panic("Fatal exception"); + + gr_handle_kernel_exploit(); + if (signr) do_exit(signr); } @@ -885,7 +890,11 @@ void __init early_trap_init(void *vectors_base) kuser_init(vectors_base); flush_icache_range(vectors, vectors + PAGE_SIZE * 2); - modify_domain(DOMAIN_USER, DOMAIN_CLIENT); + +#ifndef CONFIG_PAX_MEMORY_UDEREF + modify_domain(DOMAIN_USER, DOMAIN_USERCLIENT); +#endif + #else /* ifndef CONFIG_CPU_V7M */ /* * on V7-M there is no need to copy the vector table to a dedicated diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index 7bcee5c..e2f3249 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -8,7 +8,11 @@ #include #include #include - + +#ifdef CONFIG_PAX_KERNEXEC +#include +#endif + #define PROC_INFO \ . = ALIGN(4); \ VMLINUX_SYMBOL(__proc_info_begin) = .; \ @@ -34,7 +38,7 @@ #endif #if (defined(CONFIG_SMP_ON_UP) && !defined(CONFIG_DEBUG_SPINLOCK)) || \ - defined(CONFIG_GENERIC_BUG) + defined(CONFIG_GENERIC_BUG) || defined(CONFIG_PAX_REFCOUNT) #define ARM_EXIT_KEEP(x) x #define ARM_EXIT_DISCARD(x) #else @@ -90,6 +94,11 @@ SECTIONS _text = .; HEAD_TEXT } + +#ifdef CONFIG_PAX_KERNEXEC + . = ALIGN(1<arch.vmid_gen != atomic64_read(&kvm_vmid_gen)); + return unlikely(kvm->arch.vmid_gen != atomic64_read_unchecked(&kvm_vmid_gen)); } /** @@ -442,7 +442,7 @@ static void update_vttbr(struct kvm *kvm) /* First user of a new VMID generation? */ if (unlikely(kvm_next_vmid == 0)) { - atomic64_inc(&kvm_vmid_gen); + atomic64_inc_unchecked(&kvm_vmid_gen); kvm_next_vmid = 1; /* @@ -459,7 +459,7 @@ static void update_vttbr(struct kvm *kvm) kvm_call_hyp(__kvm_flush_vm_context); } - kvm->arch.vmid_gen = atomic64_read(&kvm_vmid_gen); + kvm->arch.vmid_gen = atomic64_read_unchecked(&kvm_vmid_gen); kvm->arch.vmid = kvm_next_vmid; kvm_next_vmid++; @@ -1034,7 +1034,7 @@ static void check_kvm_target_cpu(void *ret) /** * Initialize Hyp-mode and memory mappings on all CPUs. */ -int kvm_arch_init(void *opaque) +int kvm_arch_init(const void *opaque) { int err; int ret, cpu; diff --git a/arch/arm/lib/clear_user.S b/arch/arm/lib/clear_user.S index 14a0d98..7771a7d 100644 --- a/arch/arm/lib/clear_user.S +++ b/arch/arm/lib/clear_user.S @@ -12,14 +12,14 @@ .text -/* Prototype: int __clear_user(void *addr, size_t sz) +/* Prototype: int ___clear_user(void *addr, size_t sz) * Purpose : clear some user memory * Params : addr - user memory address to clear * : sz - number of bytes to clear * Returns : number of bytes NOT cleared */ ENTRY(__clear_user_std) -WEAK(__clear_user) +WEAK(___clear_user) stmfd sp!, {r1, lr} mov r2, #0 cmp r1, #4 @@ -44,7 +44,7 @@ WEAK(__clear_user) USER( strnebt r2, [r0]) mov r0, #0 ldmfd sp!, {r1, pc} -ENDPROC(__clear_user) +ENDPROC(___clear_user) ENDPROC(__clear_user_std) .pushsection .fixup,"ax" 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/csumpartialcopyuser.S b/arch/arm/lib/csumpartialcopyuser.S index 7d08b43..f7ca7ea 100644 --- a/arch/arm/lib/csumpartialcopyuser.S +++ b/arch/arm/lib/csumpartialcopyuser.S @@ -57,8 +57,8 @@ * Returns : r0 = checksum, [[sp, #0], #0] = 0 or -EFAULT */ -#define FN_ENTRY ENTRY(csum_partial_copy_from_user) -#define FN_EXIT ENDPROC(csum_partial_copy_from_user) +#define FN_ENTRY ENTRY(__csum_partial_copy_from_user) +#define FN_EXIT ENDPROC(__csum_partial_copy_from_user) #include "csumpartialcopygeneric.S" diff --git a/arch/arm/lib/delay.c b/arch/arm/lib/delay.c index 5306de3..aed6d03 100644 --- a/arch/arm/lib/delay.c +++ b/arch/arm/lib/delay.c @@ -28,7 +28,7 @@ /* * Default to the loop-based delay implementation. */ -struct arm_delay_ops arm_delay_ops = { +struct arm_delay_ops arm_delay_ops __read_only = { .delay = __loop_delay, .const_udelay = __loop_const_udelay, .udelay = __loop_udelay, diff --git a/arch/arm/lib/uaccess_with_memcpy.c b/arch/arm/lib/uaccess_with_memcpy.c index 3e58d71..029817c 100644 --- a/arch/arm/lib/uaccess_with_memcpy.c +++ b/arch/arm/lib/uaccess_with_memcpy.c @@ -136,7 +136,7 @@ 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 @@ -190,7 +190,7 @@ out: return n; } -unsigned long __clear_user(void __user *addr, unsigned long n) +unsigned long ___clear_user(void __user *addr, unsigned long n) { /* See rational for this in __copy_to_user() above. */ if (n < 64) diff --git a/arch/arm/mach-at91/setup.c b/arch/arm/mach-at91/setup.c index f7a07a5..258e1f7 100644 --- a/arch/arm/mach-at91/setup.c +++ b/arch/arm/mach-at91/setup.c @@ -81,7 +81,7 @@ void __init at91_init_sram(int bank, unsigned long base, unsigned int length) desc->pfn = __phys_to_pfn(base); desc->length = length; - desc->type = MT_MEMORY_RWX_NONCACHED; + desc->type = MT_MEMORY_RW_NONCACHED; pr_info("AT91: sram at 0x%lx of 0x%x mapped at 0x%lx\n", base, length, desc->virtual); diff --git a/arch/arm/mach-kirkwood/common.c b/arch/arm/mach-kirkwood/common.c index 255f33a..507b157 100644 --- a/arch/arm/mach-kirkwood/common.c +++ b/arch/arm/mach-kirkwood/common.c @@ -157,7 +157,16 @@ static void clk_gate_fn_disable(struct clk_hw *hw) clk_gate_ops.disable(hw); } -static struct clk_ops clk_gate_fn_ops; +static int clk_gate_fn_is_enabled(struct clk_hw *hw) +{ + return clk_gate_ops.is_enabled(hw); +} + +static struct clk_ops clk_gate_fn_ops = { + .enable = clk_gate_fn_enable, + .disable = clk_gate_fn_disable, + .is_enabled = clk_gate_fn_is_enabled, +}; static struct clk __init *clk_register_gate_fn(struct device *dev, const char *name, @@ -191,14 +200,6 @@ static struct clk __init *clk_register_gate_fn(struct device *dev, gate_fn->fn_en = fn_en; gate_fn->fn_dis = fn_dis; - /* ops is the gate ops, but with our enable/disable functions */ - if (clk_gate_fn_ops.enable != clk_gate_fn_enable || - clk_gate_fn_ops.disable != clk_gate_fn_disable) { - clk_gate_fn_ops = clk_gate_ops; - clk_gate_fn_ops.enable = clk_gate_fn_enable; - clk_gate_fn_ops.disable = clk_gate_fn_disable; - } - clk = clk_register(dev, &gate_fn->gate.hw); if (IS_ERR(clk)) diff --git a/arch/arm/mach-mvebu/coherency.c b/arch/arm/mach-mvebu/coherency.c index 2bdc323..cf1c607 100644 --- a/arch/arm/mach-mvebu/coherency.c +++ b/arch/arm/mach-mvebu/coherency.c @@ -316,7 +316,7 @@ static void __init armada_370_coherency_init(struct device_node *np) /* * This ioremap hook is used on Armada 375/38x to ensure that PCIe - * memory areas are mapped as MT_UNCACHED instead of MT_DEVICE. This + * memory areas are mapped as MT_UNCACHED_RW instead of MT_DEVICE. This * is needed as a workaround for a deadlock issue between the PCIe * interface and the cache controller. */ @@ -329,7 +329,7 @@ armada_pcie_wa_ioremap_caller(phys_addr_t phys_addr, size_t size, mvebu_mbus_get_pcie_mem_aperture(&pcie_mem); if (pcie_mem.start <= phys_addr && (phys_addr + size) <= pcie_mem.end) - mtype = MT_UNCACHED; + mtype = MT_UNCACHED_RW; return __arm_ioremap_caller(phys_addr, size, mtype, caller); } diff --git a/arch/arm/mach-omap2/board-n8x0.c b/arch/arm/mach-omap2/board-n8x0.c index aead77a..a2253fa 100644 --- a/arch/arm/mach-omap2/board-n8x0.c +++ b/arch/arm/mach-omap2/board-n8x0.c @@ -568,7 +568,7 @@ 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-omap2/gpmc.c b/arch/arm/mach-omap2/gpmc.c index 8bc1338..8b28b69 100644 --- a/arch/arm/mach-omap2/gpmc.c +++ b/arch/arm/mach-omap2/gpmc.c @@ -151,7 +151,6 @@ struct omap3_gpmc_regs { }; static struct gpmc_client_irq gpmc_client_irq[GPMC_NR_IRQ]; -static struct irq_chip gpmc_irq_chip; static int gpmc_irq_start; static struct resource gpmc_mem_root; @@ -736,6 +735,18 @@ static void gpmc_irq_noop(struct irq_data *data) { } static unsigned int gpmc_irq_noop_ret(struct irq_data *data) { return 0; } +static struct irq_chip gpmc_irq_chip = { + .name = "gpmc", + .irq_startup = gpmc_irq_noop_ret, + .irq_enable = gpmc_irq_enable, + .irq_disable = gpmc_irq_disable, + .irq_shutdown = gpmc_irq_noop, + .irq_ack = gpmc_irq_noop, + .irq_mask = gpmc_irq_noop, + .irq_unmask = gpmc_irq_noop, + +}; + static int gpmc_setup_irq(void) { int i; @@ -750,15 +761,6 @@ static int gpmc_setup_irq(void) return gpmc_irq_start; } - gpmc_irq_chip.name = "gpmc"; - gpmc_irq_chip.irq_startup = gpmc_irq_noop_ret; - gpmc_irq_chip.irq_enable = gpmc_irq_enable; - gpmc_irq_chip.irq_disable = gpmc_irq_disable; - gpmc_irq_chip.irq_shutdown = gpmc_irq_noop; - gpmc_irq_chip.irq_ack = gpmc_irq_noop; - gpmc_irq_chip.irq_mask = gpmc_irq_noop; - gpmc_irq_chip.irq_unmask = gpmc_irq_noop; - gpmc_client_irq[0].bitmask = GPMC_IRQ_FIFOEVENTENABLE; gpmc_client_irq[1].bitmask = GPMC_IRQ_COUNT_EVENT; diff --git a/arch/arm/mach-omap2/omap-mpuss-lowpower.c b/arch/arm/mach-omap2/omap-mpuss-lowpower.c index 4001325..b14e2a0 100644 --- a/arch/arm/mach-omap2/omap-mpuss-lowpower.c +++ b/arch/arm/mach-omap2/omap-mpuss-lowpower.c @@ -84,7 +84,7 @@ struct cpu_pm_ops { int (*finish_suspend)(unsigned long cpu_state); void (*resume)(void); void (*scu_prepare)(unsigned int cpu_id, unsigned int cpu_state); -}; +} __no_const; static DEFINE_PER_CPU(struct omap4_cpu_pm_info, omap4_pm_info); static struct powerdomain *mpuss_pd; @@ -102,7 +102,7 @@ static void dummy_cpu_resume(void) static void dummy_scu_prepare(unsigned int cpu_id, unsigned int cpu_state) {} -struct cpu_pm_ops omap_pm_ops = { +static struct cpu_pm_ops omap_pm_ops __read_only = { .finish_suspend = default_finish_suspend, .resume = dummy_cpu_resume, .scu_prepare = dummy_scu_prepare, diff --git a/arch/arm/mach-omap2/omap-wakeupgen.c b/arch/arm/mach-omap2/omap-wakeupgen.c index 37843a7..a98df13 100644 --- a/arch/arm/mach-omap2/omap-wakeupgen.c +++ b/arch/arm/mach-omap2/omap-wakeupgen.c @@ -343,7 +343,7 @@ static int irq_cpu_hotplug_notify(struct notifier_block *self, return NOTIFY_OK; } -static struct notifier_block __refdata irq_hotplug_notifier = { +static struct notifier_block irq_hotplug_notifier = { .notifier_call = irq_cpu_hotplug_notify, }; diff --git a/arch/arm/mach-omap2/omap_device.c b/arch/arm/mach-omap2/omap_device.c index 01ef59d..32ae28a8 100644 --- a/arch/arm/mach-omap2/omap_device.c +++ b/arch/arm/mach-omap2/omap_device.c @@ -510,7 +510,7 @@ void omap_device_delete(struct omap_device *od) struct platform_device __init *omap_device_build(const char *pdev_name, int pdev_id, struct omap_hwmod *oh, - void *pdata, int pdata_len) + const void *pdata, int pdata_len) { struct omap_hwmod *ohs[] = { oh }; @@ -538,7 +538,7 @@ struct platform_device __init *omap_device_build(const char *pdev_name, struct platform_device __init *omap_device_build_ss(const char *pdev_name, int pdev_id, struct omap_hwmod **ohs, - int oh_cnt, void *pdata, + int oh_cnt, const void *pdata, int pdata_len) { int ret = -ENOMEM; diff --git a/arch/arm/mach-omap2/omap_device.h b/arch/arm/mach-omap2/omap_device.h index 78c02b3..c94109a 100644 --- a/arch/arm/mach-omap2/omap_device.h +++ b/arch/arm/mach-omap2/omap_device.h @@ -72,12 +72,12 @@ int omap_device_idle(struct platform_device *pdev); /* Core code interface */ struct platform_device *omap_device_build(const char *pdev_name, int pdev_id, - struct omap_hwmod *oh, void *pdata, + struct omap_hwmod *oh, const void *pdata, int pdata_len); struct platform_device *omap_device_build_ss(const char *pdev_name, int pdev_id, struct omap_hwmod **oh, int oh_cnt, - void *pdata, int pdata_len); + const void *pdata, int pdata_len); struct omap_device *omap_device_alloc(struct platform_device *pdev, struct omap_hwmod **ohs, int oh_cnt); diff --git a/arch/arm/mach-omap2/omap_hwmod.c b/arch/arm/mach-omap2/omap_hwmod.c index 8fd87a3..099ed60 100644 --- a/arch/arm/mach-omap2/omap_hwmod.c +++ b/arch/arm/mach-omap2/omap_hwmod.c @@ -194,10 +194,10 @@ struct omap_hwmod_soc_ops { int (*init_clkdm)(struct omap_hwmod *oh); void (*update_context_lost)(struct omap_hwmod *oh); int (*get_context_lost)(struct omap_hwmod *oh); -}; +} __no_const; /* soc_ops: adapts the omap_hwmod code to the currently-booted SoC */ -static struct omap_hwmod_soc_ops soc_ops; +static struct omap_hwmod_soc_ops soc_ops __read_only; /* omap_hwmod_list contains all registered struct omap_hwmods */ static LIST_HEAD(omap_hwmod_list); diff --git a/arch/arm/mach-omap2/powerdomains43xx_data.c b/arch/arm/mach-omap2/powerdomains43xx_data.c index 95fee54..cfa9cf1 100644 --- a/arch/arm/mach-omap2/powerdomains43xx_data.c +++ b/arch/arm/mach-omap2/powerdomains43xx_data.c @@ -10,6 +10,7 @@ #include #include +#include #include "powerdomain.h" @@ -129,7 +130,9 @@ static int am43xx_check_vcvp(void) void __init am43xx_powerdomains_init(void) { - omap4_pwrdm_operations.pwrdm_has_voltdm = am43xx_check_vcvp; + pax_open_kernel(); + *(void **)&omap4_pwrdm_operations.pwrdm_has_voltdm = am43xx_check_vcvp; + pax_close_kernel(); pwrdm_register_platform_funcs(&omap4_pwrdm_operations); pwrdm_register_pwrdms(powerdomains_am43xx); pwrdm_complete_init(); diff --git a/arch/arm/mach-omap2/wd_timer.c b/arch/arm/mach-omap2/wd_timer.c index 97d6607..8429d14 100644 --- a/arch/arm/mach-omap2/wd_timer.c +++ b/arch/arm/mach-omap2/wd_timer.c @@ -110,7 +110,9 @@ static int __init omap_init_wdt(void) struct omap_hwmod *oh; char *oh_name = "wd_timer2"; char *dev_name = "omap_wdt"; - struct omap_wd_timer_platform_data pdata; + static struct omap_wd_timer_platform_data pdata = { + .read_reset_sources = prm_read_reset_sources + }; if (!cpu_class_is_omap2() || of_have_populated_dt()) return 0; @@ -121,8 +123,6 @@ static int __init omap_init_wdt(void) return -EINVAL; } - pdata.read_reset_sources = prm_read_reset_sources; - pdev = omap_device_build(dev_name, id, oh, &pdata, sizeof(struct omap_wd_timer_platform_data)); WARN(IS_ERR(pdev), "Can't build omap_device for %s:%s.\n", diff --git a/arch/arm/mach-tegra/cpuidle-tegra20.c b/arch/arm/mach-tegra/cpuidle-tegra20.c index b82dcae..44ee5b6 100644 --- a/arch/arm/mach-tegra/cpuidle-tegra20.c +++ b/arch/arm/mach-tegra/cpuidle-tegra20.c @@ -180,7 +180,7 @@ static int tegra20_idle_lp2_coupled(struct cpuidle_device *dev, bool entered_lp2 = false; if (tegra_pending_sgi()) - ACCESS_ONCE(abort_flag) = true; + ACCESS_ONCE_RW(abort_flag) = true; cpuidle_coupled_parallel_barrier(dev, &abort_barrier); diff --git a/arch/arm/mach-ux500/setup.h b/arch/arm/mach-ux500/setup.h index 2dea8b5..6499da2 100644 --- a/arch/arm/mach-ux500/setup.h +++ b/arch/arm/mach-ux500/setup.h @@ -33,13 +33,6 @@ extern void ux500_timer_init(void); .type = MT_DEVICE, \ } -#define __MEM_DEV_DESC(x, sz) { \ - .virtual = IO_ADDRESS(x), \ - .pfn = __phys_to_pfn(x), \ - .length = sz, \ - .type = MT_MEMORY_RWX, \ -} - extern struct smp_operations ux500_smp_ops; extern void ux500_cpu_die(unsigned int cpu); diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index c348eae..456a1a4 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -446,6 +446,7 @@ config CPU_32v5 config CPU_32v6 bool + select CPU_USE_DOMAINS if CPU_V6 && MMU && !PAX_KERNEXEC && !PAX_MEMORY_UDEREF select TLS_REG_EMUL if !CPU_32v6K && !MMU config CPU_32v6K @@ -600,6 +601,7 @@ config CPU_CP15_MPU config CPU_USE_DOMAINS bool + depends on !ARM_LPAE && !PAX_KERNEXEC && !PAX_MEMORY_UDEREF help This option enables or disables the use of domain switching via the set_fs() function. @@ -799,6 +801,7 @@ config NEED_KUSER_HELPERS config KUSER_HELPERS bool "Enable kuser helpers in vector page" if !NEED_KUSER_HELPERS default y + depends on !(CPU_V6 || CPU_V6K || CPU_V7) || GRKERNSEC_OLD_ARM_USERLAND help Warning: disabling this option may break user programs. @@ -811,7 +814,7 @@ config KUSER_HELPERS See Documentation/arm/kernel_user_helpers.txt for details. However, the fixed address nature of these helpers can be used - by ROP (return orientated programming) authors when creating + by ROP (Return Oriented Programming) authors when creating exploits. If all of the binaries and libraries which run on your platform diff --git a/arch/arm/mm/alignment.c b/arch/arm/mm/alignment.c index 33ca980..6b23b44 100644 --- a/arch/arm/mm/alignment.c +++ b/arch/arm/mm/alignment.c @@ -215,10 +215,12 @@ union offset_union { #define __get16_unaligned_check(ins,val,addr) \ do { \ unsigned int err = 0, v, a = addr; \ + pax_open_userland(); \ __get8_unaligned_check(ins,v,a,err); \ val = v << ((BE) ? 8 : 0); \ __get8_unaligned_check(ins,v,a,err); \ val |= v << ((BE) ? 0 : 8); \ + pax_close_userland(); \ if (err) \ goto fault; \ } while (0) @@ -232,6 +234,7 @@ union offset_union { #define __get32_unaligned_check(ins,val,addr) \ do { \ unsigned int err = 0, v, a = addr; \ + pax_open_userland(); \ __get8_unaligned_check(ins,v,a,err); \ val = v << ((BE) ? 24 : 0); \ __get8_unaligned_check(ins,v,a,err); \ @@ -240,6 +243,7 @@ union offset_union { val |= v << ((BE) ? 8 : 16); \ __get8_unaligned_check(ins,v,a,err); \ val |= v << ((BE) ? 0 : 24); \ + pax_close_userland(); \ if (err) \ goto fault; \ } while (0) @@ -253,6 +257,7 @@ union offset_union { #define __put16_unaligned_check(ins,val,addr) \ do { \ unsigned int err = 0, v = val, a = addr; \ + pax_open_userland(); \ __asm__( FIRST_BYTE_16 \ ARM( "1: "ins" %1, [%2], #1\n" ) \ THUMB( "1: "ins" %1, [%2]\n" ) \ @@ -272,6 +277,7 @@ union offset_union { " .popsection\n" \ : "=r" (err), "=&r" (v), "=&r" (a) \ : "0" (err), "1" (v), "2" (a)); \ + pax_close_userland(); \ if (err) \ goto fault; \ } while (0) @@ -285,6 +291,7 @@ union offset_union { #define __put32_unaligned_check(ins,val,addr) \ do { \ unsigned int err = 0, v = val, a = addr; \ + pax_open_userland(); \ __asm__( FIRST_BYTE_32 \ ARM( "1: "ins" %1, [%2], #1\n" ) \ THUMB( "1: "ins" %1, [%2]\n" ) \ @@ -314,6 +321,7 @@ union offset_union { " .popsection\n" \ : "=r" (err), "=&r" (v), "=&r" (a) \ : "0" (err), "1" (v), "2" (a)); \ + pax_close_userland(); \ if (err) \ goto fault; \ } while (0) diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c index 7c3fb41..bfb87d8 100644 --- a/arch/arm/mm/cache-l2x0.c +++ b/arch/arm/mm/cache-l2x0.c @@ -41,7 +41,7 @@ struct l2c_init_data { void (*fixup)(void __iomem *, u32, struct outer_cache_fns *); void (*save)(void __iomem *); struct outer_cache_fns outer_cache; -}; +} __do_const; #define CACHE_LINE_SIZE 32 diff --git a/arch/arm/mm/context.c b/arch/arm/mm/context.c index 6eb97b3..ac509f6 100644 --- a/arch/arm/mm/context.c +++ b/arch/arm/mm/context.c @@ -43,7 +43,7 @@ #define NUM_USER_ASIDS ASID_FIRST_VERSION static DEFINE_RAW_SPINLOCK(cpu_asid_lock); -static atomic64_t asid_generation = ATOMIC64_INIT(ASID_FIRST_VERSION); +static atomic64_unchecked_t asid_generation = ATOMIC64_INIT(ASID_FIRST_VERSION); static DECLARE_BITMAP(asid_map, NUM_USER_ASIDS); static DEFINE_PER_CPU(atomic64_t, active_asids); @@ -182,7 +182,7 @@ static u64 new_context(struct mm_struct *mm, unsigned int cpu) { static u32 cur_idx = 1; u64 asid = atomic64_read(&mm->context.id); - u64 generation = atomic64_read(&asid_generation); + u64 generation = atomic64_read_unchecked(&asid_generation); if (asid != 0 && is_reserved_asid(asid)) { /* @@ -203,7 +203,7 @@ static u64 new_context(struct mm_struct *mm, unsigned int cpu) */ asid = find_next_zero_bit(asid_map, NUM_USER_ASIDS, cur_idx); if (asid == NUM_USER_ASIDS) { - generation = atomic64_add_return(ASID_FIRST_VERSION, + generation = atomic64_add_return_unchecked(ASID_FIRST_VERSION, &asid_generation); flush_context(cpu); asid = find_next_zero_bit(asid_map, NUM_USER_ASIDS, 1); @@ -234,14 +234,14 @@ void check_and_switch_context(struct mm_struct *mm, struct task_struct *tsk) cpu_set_reserved_ttbr0(); asid = atomic64_read(&mm->context.id); - if (!((asid ^ atomic64_read(&asid_generation)) >> ASID_BITS) + if (!((asid ^ atomic64_read_unchecked(&asid_generation)) >> ASID_BITS) && atomic64_xchg(&per_cpu(active_asids, cpu), asid)) goto switch_mm_fastpath; raw_spin_lock_irqsave(&cpu_asid_lock, flags); /* Check that our ASID belongs to the current generation. */ asid = atomic64_read(&mm->context.id); - if ((asid ^ atomic64_read(&asid_generation)) >> ASID_BITS) { + if ((asid ^ atomic64_read_unchecked(&asid_generation)) >> ASID_BITS) { asid = new_context(mm, cpu); atomic64_set(&mm->context.id, asid); } diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c index eb8830a..5360ce7 100644 --- a/arch/arm/mm/fault.c +++ b/arch/arm/mm/fault.c @@ -25,6 +25,7 @@ #include #include #include +#include #include "fault.h" @@ -138,6 +139,31 @@ __do_kernel_fault(struct mm_struct *mm, unsigned long addr, unsigned int fsr, if (fixup_exception(regs)) return; +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (addr < TASK_SIZE) { + if (current->signal->curr_ip) + printk(KERN_ERR "PAX: From %pI4: %s:%d, uid/euid: %u/%u, attempted to access userland memory at %08lx\n", ¤t->signal->curr_ip, current->comm, task_pid_nr(current), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid()), addr); + else + printk(KERN_ERR "PAX: %s:%d, uid/euid: %u/%u, attempted to access userland memory at %08lx\n", current->comm, task_pid_nr(current), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid()), addr); + } +#endif + +#ifdef CONFIG_PAX_KERNEXEC + if ((fsr & FSR_WRITE) && + (((unsigned long)_stext <= addr && addr < init_mm.end_code) || + (MODULES_VADDR <= addr && addr < MODULES_END))) + { + 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), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, 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), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid())); + } +#endif + /* * No handler, we'll have to terminate things with extreme prejudice. */ @@ -174,6 +200,13 @@ __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; @@ -401,6 +434,33 @@ 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 * @@ -548,9 +608,22 @@ do_DataAbort(unsigned long addr, unsigned int fsr, struct pt_regs *regs) const struct fsr_info *inf = fsr_info + fsr_fs(fsr); struct siginfo info; +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (addr < TASK_SIZE && is_domain_fault(fsr)) { + if (current->signal->curr_ip) + printk(KERN_ERR "PAX: From %pI4: %s:%d, uid/euid: %u/%u, attempted to access userland memory at %08lx\n", ¤t->signal->curr_ip, current->comm, task_pid_nr(current), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid()), addr); + else + printk(KERN_ERR "PAX: %s:%d, uid/euid: %u/%u, attempted to access userland memory at %08lx\n", current->comm, task_pid_nr(current), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid()), addr); + goto die; + } +#endif + if (!inf->fn(addr, fsr & ~FSR_LNX_PF, regs)) return; +die: printk(KERN_ALERT "Unhandled fault: %s (0x%03x) at 0x%08lx\n", inf->name, fsr, addr); @@ -574,15 +647,98 @@ hook_ifault_code(int nr, int (*fn)(unsigned long, unsigned int, struct pt_regs * ifsr_info[nr].name = name; } +asmlinkage int sys_sigreturn(struct pt_regs *regs); +asmlinkage int sys_rt_sigreturn(struct pt_regs *regs); + asmlinkage void __exception 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; + unsigned long pc = instruction_pointer(regs); + + if (user_mode(regs)) { + unsigned long sigpage = current->mm->context.sigpage; + + if (sigpage <= pc && pc < sigpage + 7*4) { + if (pc < sigpage + 3*4) + sys_sigreturn(regs); + else + sys_rt_sigreturn(regs); + return; + } + if (pc == 0xffff0f60UL) { + /* + * PaX: __kuser_cmpxchg64 emulation + */ + // TODO + //regs->ARM_pc = regs->ARM_lr; + //return; + } + if (pc == 0xffff0fa0UL) { + /* + * PaX: __kuser_memory_barrier emulation + */ + // dmb(); implied by the exception + regs->ARM_pc = regs->ARM_lr; + return; + } + if (pc == 0xffff0fc0UL) { + /* + * PaX: __kuser_cmpxchg emulation + */ + // TODO + //long new; + //int op; + + //op = FUTEX_OP_SET << 28; + //new = futex_atomic_op_inuser(op, regs->ARM_r2); + //regs->ARM_r0 = old != new; + //regs->ARM_pc = regs->ARM_lr; + //return; + } + if (pc == 0xffff0fe0UL) { + /* + * PaX: __kuser_get_tls emulation + */ + regs->ARM_r0 = current_thread_info()->tp_value[0]; + regs->ARM_pc = regs->ARM_lr; + return; + } + } + +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + else if (is_domain_fault(ifsr) || is_xn_fault(ifsr)) { + if (current->signal->curr_ip) + printk(KERN_ERR "PAX: From %pI4: %s:%d, uid/euid: %u/%u, attempted to execute %s memory at %08lx\n", ¤t->signal->curr_ip, current->comm, task_pid_nr(current), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid()), + pc >= TASK_SIZE ? "non-executable kernel" : "userland", pc); + else + printk(KERN_ERR "PAX: %s:%d, uid/euid: %u/%u, attempted to execute %s memory at %08lx\n", current->comm, task_pid_nr(current), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid()), + pc >= TASK_SIZE ? "non-executable kernel" : "userland", pc); + goto die; + } +#endif + +#ifdef CONFIG_PAX_REFCOUNT + if (fsr_fs(ifsr) == FAULT_CODE_DEBUG) { + unsigned int bkpt; + + if (!probe_kernel_address(pc, bkpt) && cpu_to_le32(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; +die: printk(KERN_ALERT "Unhandled prefetch abort: %s (0x%03x) at 0x%08lx\n", inf->name, ifsr, addr); diff --git a/arch/arm/mm/fault.h b/arch/arm/mm/fault.h index cf08bdf..772656c 100644 --- a/arch/arm/mm/fault.h +++ b/arch/arm/mm/fault.h @@ -3,6 +3,7 @@ /* * Fault status register encodings. We steal bit 31 for our own purposes. + * Set when the FSR value is from an instruction fault. */ #define FSR_LNX_PF (1 << 31) #define FSR_WRITE (1 << 11) @@ -22,6 +23,17 @@ static inline int fsr_fs(unsigned int fsr) } #endif +/* valid for LPAE and !LPAE */ +static inline int is_xn_fault(unsigned int fsr) +{ + return ((fsr_fs(fsr) & 0x3c) == 0xc); +} + +static inline int is_domain_fault(unsigned int fsr) +{ + return ((fsr_fs(fsr) & 0xD) == 0x9); +} + void do_bad_area(unsigned long addr, unsigned int fsr, struct pt_regs *regs); unsigned long search_exception_table(unsigned long addr); diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c index 659c75d..6f8c029 100644 --- a/arch/arm/mm/init.c +++ b/arch/arm/mm/init.c @@ -31,6 +31,8 @@ #include #include #include +#include +#include #include #include @@ -619,7 +621,46 @@ void free_initmem(void) { #ifdef CONFIG_HAVE_TCM extern char __tcm_start, __tcm_end; +#endif +#ifdef CONFIG_PAX_KERNEXEC + unsigned long addr; + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + int cpu_arch = cpu_architecture(); + unsigned int cr = get_cr(); + + if (cpu_arch >= CPU_ARCH_ARMv6 && (cr & CR_XP)) { + /* make pages tables, etc before .text NX */ + for (addr = PAGE_OFFSET; addr < (unsigned long)_stext; addr += SECTION_SIZE) { + pgd = pgd_offset_k(addr); + pud = pud_offset(pgd, addr); + pmd = pmd_offset(pud, addr); + __section_update(pmd, addr, PMD_SECT_XN); + } + /* make init NX */ + for (addr = (unsigned long)__init_begin; addr < (unsigned long)_sdata; addr += SECTION_SIZE) { + pgd = pgd_offset_k(addr); + pud = pud_offset(pgd, addr); + pmd = pmd_offset(pud, addr); + __section_update(pmd, addr, PMD_SECT_XN); + } + /* make kernel code/rodata RX */ + for (addr = (unsigned long)_stext; addr < (unsigned long)__init_begin; addr += SECTION_SIZE) { + pgd = pgd_offset_k(addr); + pud = pud_offset(pgd, addr); + pmd = pmd_offset(pud, addr); +#ifdef CONFIG_ARM_LPAE + __section_update(pmd, addr, PMD_SECT_RDONLY); +#else + __section_update(pmd, addr, PMD_SECT_APX|PMD_SECT_AP_WRITE); +#endif + } + } +#endif + +#ifdef CONFIG_HAVE_TCM poison_init_mem(&__tcm_start, &__tcm_end - &__tcm_start); free_reserved_area(&__tcm_start, &__tcm_end, -1, "TCM link"); #endif diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c index d1e5ad7..84dcbf2 100644 --- a/arch/arm/mm/ioremap.c +++ b/arch/arm/mm/ioremap.c @@ -392,9 +392,9 @@ __arm_ioremap_exec(phys_addr_t phys_addr, size_t size, bool cached) unsigned int mtype; if (cached) - mtype = MT_MEMORY_RWX; + mtype = MT_MEMORY_RX; else - mtype = MT_MEMORY_RWX_NONCACHED; + mtype = MT_MEMORY_RX_NONCACHED; return __arm_ioremap_caller(phys_addr, size, mtype, __builtin_return_address(0)); diff --git a/arch/arm/mm/mmap.c b/arch/arm/mm/mmap.c index 5e85ed3..b10a7ed 100644 --- a/arch/arm/mm/mmap.c +++ b/arch/arm/mm/mmap.c @@ -59,6 +59,7 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, struct vm_area_struct *vma; int do_align = 0; int aliasing = cache_is_vipt_aliasing(); + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); struct vm_unmapped_area_info info; /* @@ -81,6 +82,10 @@ 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); @@ -88,8 +93,7 @@ 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, offset)) return addr; } @@ -99,6 +103,7 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, info.high_limit = TASK_SIZE; info.align_mask = do_align ? (PAGE_MASK & (SHMLBA - 1)) : 0; info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; return vm_unmapped_area(&info); } @@ -112,6 +117,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, unsigned long addr = addr0; int do_align = 0; int aliasing = cache_is_vipt_aliasing(); + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); struct vm_unmapped_area_info info; /* @@ -132,6 +138,10 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, return addr; } +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + /* requesting a specific address */ if (addr) { if (do_align) @@ -139,8 +149,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, else 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, offset)) return addr; } @@ -150,6 +159,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, info.high_limit = mm->mmap_base; info.align_mask = do_align ? (PAGE_MASK & (SHMLBA - 1)) : 0; info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); /* @@ -173,6 +183,10 @@ void arch_pick_mmap_layout(struct mm_struct *mm) { unsigned long random_factor = 0UL; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + /* 8 bits of randomness in 20 address space bits */ if ((current->flags & PF_RANDOMIZE) && !(current->personality & ADDR_NO_RANDOMIZE)) @@ -180,9 +194,21 @@ 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; } 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; } } diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c index 6e3ba8d..9cbb4d7 100644 --- a/arch/arm/mm/mmu.c +++ b/arch/arm/mm/mmu.c @@ -40,6 +40,22 @@ #include "mm.h" #include "tcm.h" +#if defined(CONFIG_CPU_USE_DOMAINS) || defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) +void modify_domain(unsigned int dom, unsigned int type) +{ + struct thread_info *thread = current_thread_info(); + unsigned int domain = thread->cpu_domain; + /* + * DOMAIN_MANAGER might be defined to some other value, + * use the arch-defined constant + */ + domain &= ~domain_val(dom, 3); + thread->cpu_domain = domain | domain_val(dom, type); + set_domain(thread->cpu_domain); +} +EXPORT_SYMBOL(modify_domain); +#endif + /* * empty_zero_page is a special page that is used for * zero-initialized data and COW. @@ -239,7 +255,15 @@ __setup("noalign", noalign_setup); #define PROT_PTE_S2_DEVICE PROT_PTE_DEVICE #define PROT_SECT_DEVICE PMD_TYPE_SECT|PMD_SECT_AP_WRITE -static struct mem_type mem_types[] = { +#ifdef CONFIG_PAX_KERNEXEC +#define L_PTE_KERNEXEC L_PTE_RDONLY +#define PMD_SECT_KERNEXEC PMD_SECT_RDONLY +#else +#define L_PTE_KERNEXEC L_PTE_DIRTY +#define PMD_SECT_KERNEXEC PMD_SECT_AP_WRITE +#endif + +static struct mem_type mem_types[] __read_only = { [MT_DEVICE] = { /* Strongly ordered / ARMv6 shared device */ .prot_pte = PROT_PTE_DEVICE | L_PTE_MT_DEV_SHARED | L_PTE_SHARED, @@ -268,19 +292,19 @@ static struct mem_type mem_types[] = { .prot_sect = PROT_SECT_DEVICE, .domain = DOMAIN_IO, }, - [MT_UNCACHED] = { + [MT_UNCACHED_RW] = { .prot_pte = PROT_PTE_DEVICE, .prot_l1 = PMD_TYPE_TABLE, .prot_sect = PMD_TYPE_SECT | PMD_SECT_XN, .domain = DOMAIN_IO, }, - [MT_CACHECLEAN] = { - .prot_sect = PMD_TYPE_SECT | PMD_SECT_XN, + [MT_CACHECLEAN_RO] = { + .prot_sect = PMD_TYPE_SECT | PMD_SECT_XN | PMD_SECT_RDONLY, .domain = DOMAIN_KERNEL, }, #ifndef CONFIG_ARM_LPAE - [MT_MINICLEAN] = { - .prot_sect = PMD_TYPE_SECT | PMD_SECT_XN | PMD_SECT_MINICACHE, + [MT_MINICLEAN_RO] = { + .prot_sect = PMD_TYPE_SECT | PMD_SECT_MINICACHE | PMD_SECT_XN | PMD_SECT_RDONLY, .domain = DOMAIN_KERNEL, }, #endif @@ -288,15 +312,15 @@ static struct mem_type mem_types[] = { .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | L_PTE_RDONLY, .prot_l1 = PMD_TYPE_TABLE, - .domain = DOMAIN_USER, + .domain = DOMAIN_VECTORS, }, [MT_HIGH_VECTORS] = { .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | L_PTE_USER | L_PTE_RDONLY, .prot_l1 = PMD_TYPE_TABLE, - .domain = DOMAIN_USER, + .domain = DOMAIN_VECTORS, }, - [MT_MEMORY_RWX] = { + [__MT_MEMORY_RWX] = { .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY, .prot_l1 = PMD_TYPE_TABLE, .prot_sect = PMD_TYPE_SECT | PMD_SECT_AP_WRITE, @@ -309,17 +333,30 @@ static struct mem_type mem_types[] = { .prot_sect = PMD_TYPE_SECT | PMD_SECT_AP_WRITE, .domain = DOMAIN_KERNEL, }, - [MT_ROM] = { - .prot_sect = PMD_TYPE_SECT, + [MT_MEMORY_RX] = { + .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_KERNEXEC, + .prot_l1 = PMD_TYPE_TABLE, + .prot_sect = PMD_TYPE_SECT | PMD_SECT_KERNEXEC, + .domain = DOMAIN_KERNEL, + }, + [MT_ROM_RX] = { + .prot_sect = PMD_TYPE_SECT | PMD_SECT_RDONLY, .domain = DOMAIN_KERNEL, }, - [MT_MEMORY_RWX_NONCACHED] = { + [MT_MEMORY_RW_NONCACHED] = { .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | L_PTE_MT_BUFFERABLE, .prot_l1 = PMD_TYPE_TABLE, .prot_sect = PMD_TYPE_SECT | PMD_SECT_AP_WRITE, .domain = DOMAIN_KERNEL, }, + [MT_MEMORY_RX_NONCACHED] = { + .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_KERNEXEC | + L_PTE_MT_BUFFERABLE, + .prot_l1 = PMD_TYPE_TABLE, + .prot_sect = PMD_TYPE_SECT | PMD_SECT_KERNEXEC, + .domain = DOMAIN_KERNEL, + }, [MT_MEMORY_RW_DTCM] = { .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | L_PTE_XN, @@ -327,9 +364,10 @@ static struct mem_type mem_types[] = { .prot_sect = PMD_TYPE_SECT | PMD_SECT_XN, .domain = DOMAIN_KERNEL, }, - [MT_MEMORY_RWX_ITCM] = { - .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY, + [MT_MEMORY_RX_ITCM] = { + .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_KERNEXEC, .prot_l1 = PMD_TYPE_TABLE, + .prot_sect = PMD_TYPE_SECT | PMD_SECT_KERNEXEC, .domain = DOMAIN_KERNEL, }, [MT_MEMORY_RW_SO] = { @@ -547,9 +585,14 @@ static void __init build_mem_type_table(void) * Mark cache clean areas and XIP ROM read only * from SVC mode and no access from userspace. */ - mem_types[MT_ROM].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; - mem_types[MT_MINICLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; - mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; + mem_types[MT_ROM_RX].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; +#ifdef CONFIG_PAX_KERNEXEC + mem_types[MT_MEMORY_RX].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; + mem_types[MT_MEMORY_RX_NONCACHED].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; + mem_types[MT_MEMORY_RX_ITCM].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; +#endif + mem_types[MT_MINICLEAN_RO].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; + mem_types[MT_CACHECLEAN_RO].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; #endif /* @@ -566,13 +609,17 @@ static void __init build_mem_type_table(void) mem_types[MT_DEVICE_WC].prot_pte |= L_PTE_SHARED; mem_types[MT_DEVICE_CACHED].prot_sect |= PMD_SECT_S; mem_types[MT_DEVICE_CACHED].prot_pte |= L_PTE_SHARED; - mem_types[MT_MEMORY_RWX].prot_sect |= PMD_SECT_S; - mem_types[MT_MEMORY_RWX].prot_pte |= L_PTE_SHARED; + mem_types[__MT_MEMORY_RWX].prot_sect |= PMD_SECT_S; + mem_types[__MT_MEMORY_RWX].prot_pte |= L_PTE_SHARED; mem_types[MT_MEMORY_RW].prot_sect |= PMD_SECT_S; mem_types[MT_MEMORY_RW].prot_pte |= L_PTE_SHARED; + mem_types[MT_MEMORY_RX].prot_sect |= PMD_SECT_S; + mem_types[MT_MEMORY_RX].prot_pte |= L_PTE_SHARED; mem_types[MT_MEMORY_DMA_READY].prot_pte |= L_PTE_SHARED; - mem_types[MT_MEMORY_RWX_NONCACHED].prot_sect |= PMD_SECT_S; - mem_types[MT_MEMORY_RWX_NONCACHED].prot_pte |= L_PTE_SHARED; + mem_types[MT_MEMORY_RW_NONCACHED].prot_sect |= PMD_SECT_S; + mem_types[MT_MEMORY_RW_NONCACHED].prot_pte |= L_PTE_SHARED; + mem_types[MT_MEMORY_RX_NONCACHED].prot_sect |= PMD_SECT_S; + mem_types[MT_MEMORY_RX_NONCACHED].prot_pte |= L_PTE_SHARED; } } @@ -583,15 +630,20 @@ static void __init build_mem_type_table(void) if (cpu_arch >= CPU_ARCH_ARMv6) { if (cpu_arch >= CPU_ARCH_ARMv7 && (cr & CR_TRE)) { /* Non-cacheable Normal is XCB = 001 */ - mem_types[MT_MEMORY_RWX_NONCACHED].prot_sect |= + mem_types[MT_MEMORY_RW_NONCACHED].prot_sect |= + PMD_SECT_BUFFERED; + mem_types[MT_MEMORY_RX_NONCACHED].prot_sect |= PMD_SECT_BUFFERED; } else { /* For both ARMv6 and non-TEX-remapping ARMv7 */ - mem_types[MT_MEMORY_RWX_NONCACHED].prot_sect |= + mem_types[MT_MEMORY_RW_NONCACHED].prot_sect |= + PMD_SECT_TEX(1); + mem_types[MT_MEMORY_RX_NONCACHED].prot_sect |= PMD_SECT_TEX(1); } } else { - mem_types[MT_MEMORY_RWX_NONCACHED].prot_sect |= PMD_SECT_BUFFERABLE; + mem_types[MT_MEMORY_RW_NONCACHED].prot_sect |= PMD_SECT_BUFFERABLE; + mem_types[MT_MEMORY_RX_NONCACHED].prot_sect |= PMD_SECT_BUFFERABLE; } #ifdef CONFIG_ARM_LPAE @@ -607,6 +659,8 @@ static void __init build_mem_type_table(void) vecs_pgprot |= PTE_EXT_AF; #endif + user_pgprot |= __supported_pte_mask; + for (i = 0; i < 16; i++) { pteval_t v = pgprot_val(protection_map[i]); protection_map[i] = __pgprot(v | user_pgprot); @@ -624,21 +678,24 @@ static void __init build_mem_type_table(void) mem_types[MT_LOW_VECTORS].prot_l1 |= ecc_mask; mem_types[MT_HIGH_VECTORS].prot_l1 |= ecc_mask; - mem_types[MT_MEMORY_RWX].prot_sect |= ecc_mask | cp->pmd; - mem_types[MT_MEMORY_RWX].prot_pte |= kern_pgprot; + mem_types[__MT_MEMORY_RWX].prot_sect |= ecc_mask | cp->pmd; + mem_types[__MT_MEMORY_RWX].prot_pte |= kern_pgprot; mem_types[MT_MEMORY_RW].prot_sect |= ecc_mask | cp->pmd; mem_types[MT_MEMORY_RW].prot_pte |= kern_pgprot; + mem_types[MT_MEMORY_RX].prot_sect |= ecc_mask | cp->pmd; + mem_types[MT_MEMORY_RX].prot_pte |= kern_pgprot; mem_types[MT_MEMORY_DMA_READY].prot_pte |= kern_pgprot; - mem_types[MT_MEMORY_RWX_NONCACHED].prot_sect |= ecc_mask; - mem_types[MT_ROM].prot_sect |= cp->pmd; + mem_types[MT_MEMORY_RW_NONCACHED].prot_sect |= ecc_mask; + mem_types[MT_MEMORY_RX_NONCACHED].prot_sect |= ecc_mask; + mem_types[MT_ROM_RX].prot_sect |= cp->pmd; switch (cp->pmd) { case PMD_SECT_WT: - mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_WT; + mem_types[MT_CACHECLEAN_RO].prot_sect |= PMD_SECT_WT; break; case PMD_SECT_WB: case PMD_SECT_WBWA: - mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_WB; + mem_types[MT_CACHECLEAN_RO].prot_sect |= PMD_SECT_WB; break; } pr_info("Memory policy: %sData cache %s\n", @@ -856,7 +913,7 @@ static void __init create_mapping(struct map_desc *md) return; } - if ((md->type == MT_DEVICE || md->type == MT_ROM) && + if ((md->type == MT_DEVICE || md->type == MT_ROM_RX) && md->virtual >= PAGE_OFFSET && (md->virtual < VMALLOC_START || md->virtual >= VMALLOC_END)) { printk(KERN_WARNING "BUG: mapping for 0x%08llx" @@ -1224,18 +1281,15 @@ void __init arm_mm_memblock_reserve(void) * called function. This means you can't use any function or debugging * method which may touch any device, otherwise the kernel _will_ crash. */ + +static char vectors[PAGE_SIZE * 2] __read_only __aligned(PAGE_SIZE); + static void __init devicemaps_init(const struct machine_desc *mdesc) { struct map_desc map; unsigned long addr; - void *vectors; - /* - * Allocate the vector page early. - */ - vectors = early_alloc(PAGE_SIZE * 2); - - early_trap_init(vectors); + early_trap_init(&vectors); for (addr = VMALLOC_START; addr; addr += PMD_SIZE) pmd_clear(pmd_off_k(addr)); @@ -1248,7 +1302,7 @@ static void __init devicemaps_init(const struct machine_desc *mdesc) map.pfn = __phys_to_pfn(CONFIG_XIP_PHYS_ADDR & SECTION_MASK); map.virtual = MODULES_VADDR; map.length = ((unsigned long)_etext - map.virtual + ~SECTION_MASK) & SECTION_MASK; - map.type = MT_ROM; + map.type = MT_ROM_RX; create_mapping(&map); #endif @@ -1259,14 +1313,14 @@ static void __init devicemaps_init(const struct machine_desc *mdesc) map.pfn = __phys_to_pfn(FLUSH_BASE_PHYS); map.virtual = FLUSH_BASE; map.length = SZ_1M; - map.type = MT_CACHECLEAN; + map.type = MT_CACHECLEAN_RO; create_mapping(&map); #endif #ifdef FLUSH_BASE_MINICACHE map.pfn = __phys_to_pfn(FLUSH_BASE_PHYS + SZ_1M); map.virtual = FLUSH_BASE_MINICACHE; map.length = SZ_1M; - map.type = MT_MINICLEAN; + map.type = MT_MINICLEAN_RO; create_mapping(&map); #endif @@ -1275,7 +1329,7 @@ static void __init devicemaps_init(const struct machine_desc *mdesc) * location (0xffff0000). If we aren't using high-vectors, also * create a mapping at the low-vectors virtual address. */ - map.pfn = __phys_to_pfn(virt_to_phys(vectors)); + map.pfn = __phys_to_pfn(virt_to_phys(&vectors)); map.virtual = 0xffff0000; map.length = PAGE_SIZE; #ifdef CONFIG_KUSER_HELPERS @@ -1335,8 +1389,10 @@ static void __init kmap_init(void) static void __init map_lowmem(void) { struct memblock_region *reg; +#ifndef CONFIG_PAX_KERNEXEC unsigned long kernel_x_start = round_down(__pa(_stext), SECTION_SIZE); unsigned long kernel_x_end = round_up(__pa(__init_end), SECTION_SIZE); +#endif /* Map all the lowmem memory banks. */ for_each_memblock(memory, reg) { @@ -1349,11 +1405,48 @@ static void __init map_lowmem(void) if (start >= end) break; +#ifdef CONFIG_PAX_KERNEXEC + map.pfn = __phys_to_pfn(start); + map.virtual = __phys_to_virt(start); + map.length = end - start; + + if (map.virtual <= (unsigned long)_stext && ((unsigned long)_end < (map.virtual + map.length))) { + struct map_desc kernel; + struct map_desc initmap; + + /* when freeing initmem we will make this RW */ + initmap.pfn = __phys_to_pfn(__pa(__init_begin)); + initmap.virtual = (unsigned long)__init_begin; + initmap.length = _sdata - __init_begin; + initmap.type = __MT_MEMORY_RWX; + create_mapping(&initmap); + + /* when freeing initmem we will make this RX */ + kernel.pfn = __phys_to_pfn(__pa(_stext)); + kernel.virtual = (unsigned long)_stext; + kernel.length = __init_begin - _stext; + kernel.type = __MT_MEMORY_RWX; + create_mapping(&kernel); + + if (map.virtual < (unsigned long)_stext) { + map.length = (unsigned long)_stext - map.virtual; + map.type = __MT_MEMORY_RWX; + create_mapping(&map); + } + + map.pfn = __phys_to_pfn(__pa(_sdata)); + map.virtual = (unsigned long)_sdata; + map.length = end - __pa(_sdata); + } + + map.type = MT_MEMORY_RW; + create_mapping(&map); +#else if (end < kernel_x_start || start >= kernel_x_end) { map.pfn = __phys_to_pfn(start); map.virtual = __phys_to_virt(start); map.length = end - start; - map.type = MT_MEMORY_RWX; + map.type = __MT_MEMORY_RWX; create_mapping(&map); } else { @@ -1370,7 +1463,7 @@ static void __init map_lowmem(void) map.pfn = __phys_to_pfn(kernel_x_start); map.virtual = __phys_to_virt(kernel_x_start); map.length = kernel_x_end - kernel_x_start; - map.type = MT_MEMORY_RWX; + map.type = __MT_MEMORY_RWX; create_mapping(&map); @@ -1383,6 +1476,7 @@ static void __init map_lowmem(void) create_mapping(&map); } } +#endif } } diff --git a/arch/arm/plat-iop/setup.c b/arch/arm/plat-iop/setup.c index 5b217f4..c23f40e 100644 --- a/arch/arm/plat-iop/setup.c +++ b/arch/arm/plat-iop/setup.c @@ -24,7 +24,7 @@ static struct map_desc iop3xx_std_desc[] __initdata = { .virtual = IOP3XX_PERIPHERAL_VIRT_BASE, .pfn = __phys_to_pfn(IOP3XX_PERIPHERAL_PHYS_BASE), .length = IOP3XX_PERIPHERAL_SIZE, - .type = MT_UNCACHED, + .type = MT_UNCACHED_RW, }, }; diff --git a/arch/arm/plat-omap/sram.c b/arch/arm/plat-omap/sram.c index a5bc92d..0bb4730 100644 --- a/arch/arm/plat-omap/sram.c +++ b/arch/arm/plat-omap/sram.c @@ -93,6 +93,8 @@ void __init omap_map_sram(unsigned long start, unsigned long size, * Looks like we need to preserve some bootloader code at the * beginning of SRAM for jumping to flash for reboot to work... */ + pax_open_kernel(); memset_io(omap_sram_base + omap_sram_skip, 0, omap_sram_size - omap_sram_skip); + pax_close_kernel(); } diff --git a/arch/arm/plat-samsung/include/plat/dma-ops.h b/arch/arm/plat-samsung/include/plat/dma-ops.h index ce6d763..cfea917 100644 --- a/arch/arm/plat-samsung/include/plat/dma-ops.h +++ b/arch/arm/plat-samsung/include/plat/dma-ops.h @@ -47,7 +47,7 @@ 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/arm64/include/asm/barrier.h b/arch/arm64/include/asm/barrier.h index 6389d60..b5d3bdd 100644 --- a/arch/arm64/include/asm/barrier.h +++ b/arch/arm64/include/asm/barrier.h @@ -41,7 +41,7 @@ do { \ compiletime_assert_atomic_type(*p); \ barrier(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ diff --git a/arch/arm64/include/asm/uaccess.h b/arch/arm64/include/asm/uaccess.h index 3bf8f4e..5dd5491 100644 --- a/arch/arm64/include/asm/uaccess.h +++ b/arch/arm64/include/asm/uaccess.h @@ -99,6 +99,7 @@ static inline void set_fs(mm_segment_t fs) flag; \ }) +#define access_ok_noprefault(type, addr, size) access_ok((type), (addr), (size)) #define access_ok(type, addr, size) __range_ok(addr, size) #define user_addr_max get_fs 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 d232888..87c8df1 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 479330b..53717a8 100644 --- a/arch/avr32/include/asm/kmap_types.h +++ b/arch/avr32/include/asm/kmap_types.h @@ -2,9 +2,9 @@ #define __ASM_AVR32_KMAP_TYPES_H #ifdef CONFIG_DEBUG_HIGHMEM -# define KM_TYPE_NR 29 +# define KM_TYPE_NR 30 #else -# define KM_TYPE_NR 14 +# define KM_TYPE_NR 15 #endif #endif /* __ASM_AVR32_KMAP_TYPES_H */ diff --git a/arch/avr32/mm/fault.c b/arch/avr32/mm/fault.c index 0eca933..eb78c7b 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. @@ -176,6 +193,16 @@ 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 7caf25d..ee65ac5 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 f6c3a16..cd422a4 100644 --- a/arch/frv/include/asm/atomic.h +++ b/arch/frv/include/asm/atomic.h @@ -181,6 +181,16 @@ static inline void atomic64_dec(atomic64_t *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 43901f2..0d8b865 100644 --- a/arch/frv/include/asm/kmap_types.h +++ b/arch/frv/include/asm/kmap_types.h @@ -2,6 +2,6 @@ #ifndef _ASM_KMAP_TYPES_H #define _ASM_KMAP_TYPES_H -#define KM_TYPE_NR 17 +#define KM_TYPE_NR 18 #endif diff --git a/arch/frv/mm/elf-fdpic.c b/arch/frv/mm/elf-fdpic.c index 836f147..4cf23f5 100644 --- a/arch/frv/mm/elf-fdpic.c +++ b/arch/frv/mm/elf-fdpic.c @@ -61,6 +61,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi { struct vm_area_struct *vma; struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(current->mm, filp, flags); if (len > TASK_SIZE) return -ENOMEM; @@ -73,8 +74,7 @@ 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, offset)) goto success; } @@ -85,6 +85,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi info.high_limit = (current->mm->start_stack - 0x00200000); info.align_mask = 0; info.align_offset = 0; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); if (!(addr & ~PAGE_MASK)) goto success; diff --git a/arch/hexagon/include/asm/cache.h b/arch/hexagon/include/asm/cache.h index f4ca594..adc72fd6 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/Kconfig b/arch/ia64/Kconfig index 2f3abcf..e63c7fa 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -547,6 +547,7 @@ source "drivers/sn/Kconfig" config KEXEC bool "kexec system call" depends on !IA64_HP_SIM && (!SMP || HOTPLUG_CPU) + depends on !GRKERNSEC_KMEM help kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot diff --git a/arch/ia64/Makefile b/arch/ia64/Makefile index f37238f..810b95f 100644 --- a/arch/ia64/Makefile +++ b/arch/ia64/Makefile @@ -99,5 +99,6 @@ endef archprepare: make_nr_irqs_h FORCE PHONY += make_nr_irqs_h FORCE +make_nr_irqs_h: KBUILD_CFLAGS := $(filter-out $(GCC_PLUGINS_CFLAGS),$(KBUILD_CFLAGS)) make_nr_irqs_h: FORCE $(Q)$(MAKE) $(build)=arch/ia64/kernel include/generated/nr-irqs.h diff --git a/arch/ia64/include/asm/atomic.h b/arch/ia64/include/asm/atomic.h index 0f8bf48..40ea950 100644 --- a/arch/ia64/include/asm/atomic.h +++ b/arch/ia64/include/asm/atomic.h @@ -209,4 +209,14 @@ 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)) + #endif /* _ASM_IA64_ATOMIC_H */ diff --git a/arch/ia64/include/asm/barrier.h b/arch/ia64/include/asm/barrier.h index a48957c..e097b56 100644 --- a/arch/ia64/include/asm/barrier.h +++ b/arch/ia64/include/asm/barrier.h @@ -67,7 +67,7 @@ do { \ compiletime_assert_atomic_type(*p); \ barrier(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ 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 5a83c5c..4d7f553 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 5767cdf..7462574 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 7935115..c0eca6a 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 @@ -142,6 +142,17 @@ #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 45698cd..e8e2dbc 100644 --- a/arch/ia64/include/asm/spinlock.h +++ b/arch/ia64/include/asm/spinlock.h @@ -71,7 +71,7 @@ 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..3d4b1e9 100644 --- a/arch/ia64/include/asm/uaccess.h +++ b/arch/ia64/include/asm/uaccess.h @@ -70,6 +70,7 @@ && ((segment).seg == KERNEL_DS.seg \ || likely(REGION_OFFSET((unsigned long) (addr)) < RGN_MAP_LIMIT))); \ }) +#define access_ok_noprefault(type, addr, size) access_ok((type), (addr), (size)) #define access_ok(type, addr, size) __access_ok((addr), (size), get_fs()) /* @@ -240,12 +241,24 @@ extern unsigned long __must_check __copy_user (void __user *to, const void __use static inline unsigned long __copy_to_user (void __user *to, const void *from, unsigned long count) { + if (count > INT_MAX) + return count; + + if (!__builtin_constant_p(count)) + check_object_size(from, count, true); + return __copy_user(to, (__force void __user *) from, count); } static inline unsigned long __copy_from_user (void *to, const void __user *from, unsigned long count) { + if (count > INT_MAX) + return count; + + if (!__builtin_constant_p(count)) + check_object_size(to, count, false); + return __copy_user((__force void __user *) to, from, count); } @@ -255,10 +268,13 @@ __copy_from_user (void *to, const void __user *from, unsigned long count) ({ \ void __user *__cu_to = (to); \ const void *__cu_from = (from); \ - long __cu_len = (n); \ + unsigned long __cu_len = (n); \ \ - if (__access_ok(__cu_to, __cu_len, get_fs())) \ + if (__cu_len <= INT_MAX && __access_ok(__cu_to, __cu_len, get_fs())) { \ + if (!__builtin_constant_p(n)) \ + check_object_size(__cu_from, __cu_len, true); \ __cu_len = __copy_user(__cu_to, (__force void __user *) __cu_from, __cu_len); \ + } \ __cu_len; \ }) @@ -266,11 +282,14 @@ __copy_from_user (void *to, const void __user *from, unsigned long count) ({ \ void *__cu_to = (to); \ const void __user *__cu_from = (from); \ - long __cu_len = (n); \ + unsigned long __cu_len = (n); \ \ __chk_user_ptr(__cu_from); \ - if (__access_ok(__cu_from, __cu_len, get_fs())) \ + if (__cu_len <= INT_MAX && __access_ok(__cu_from, __cu_len, get_fs())) { \ + if (!__builtin_constant_p(n)) \ + check_object_size(__cu_to, __cu_len, false); \ __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/palinfo.c b/arch/ia64/kernel/palinfo.c index c39c3cd..3c77738 100644 --- a/arch/ia64/kernel/palinfo.c +++ b/arch/ia64/kernel/palinfo.c @@ -980,7 +980,7 @@ static int palinfo_cpu_callback(struct notifier_block *nfb, return NOTIFY_OK; } -static struct notifier_block __refdata palinfo_cpu_notifier = +static struct notifier_block palinfo_cpu_notifier = { .notifier_call = palinfo_cpu_callback, .priority = 0, diff --git a/arch/ia64/kernel/sys_ia64.c b/arch/ia64/kernel/sys_ia64.c index 41e33f8..65180b2a 100644 --- a/arch/ia64/kernel/sys_ia64.c +++ b/arch/ia64/kernel/sys_ia64.c @@ -28,6 +28,7 @@ arch_get_unmapped_area (struct file *filp, unsigned long addr, unsigned long len unsigned long align_mask = 0; struct mm_struct *mm = current->mm; struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); if (len > RGN_MAP_LIMIT) return -ENOMEM; @@ -43,6 +44,13 @@ 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 = TASK_UNMAPPED_BASE; @@ -61,6 +69,7 @@ arch_get_unmapped_area (struct file *filp, unsigned long addr, unsigned long len info.high_limit = TASK_SIZE; info.align_mask = align_mask; info.align_offset = 0; + info.threadstack_offset = offset; return vm_unmapped_area(&info); } diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S index 84f8a52..7c76178 100644 --- a/arch/ia64/kernel/vmlinux.lds.S +++ b/arch/ia64/kernel/vmlinux.lds.S @@ -192,7 +192,7 @@ 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 7225dad..2a7c8256 100644 --- a/arch/ia64/mm/fault.c +++ b/arch/ia64/mm/fault.c @@ -72,6 +72,23 @@ 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 + # define VM_READ_BIT 0 # define VM_WRITE_BIT 1 # define VM_EXEC_BIT 2 @@ -151,8 +168,21 @@ retry: if (((isr >> IA64_ISR_R_BIT) & 1UL) && (!(vma->vm_flags & (VM_READ | VM_WRITE)))) goto bad_area; - 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 diff --git a/arch/ia64/mm/hugetlbpage.c b/arch/ia64/mm/hugetlbpage.c index 76069c1..c2aa816 100644 --- a/arch/ia64/mm/hugetlbpage.c +++ b/arch/ia64/mm/hugetlbpage.c @@ -149,6 +149,7 @@ unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr, u unsigned long pgoff, unsigned long flags) { struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(current->mm, file, flags); if (len > RGN_MAP_LIMIT) return -ENOMEM; @@ -172,6 +173,7 @@ unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr, u info.high_limit = HPAGE_REGION_BASE + RGN_MAP_LIMIT; info.align_mask = PAGE_MASK & (HPAGE_SIZE - 1); info.align_offset = 0; + info.threadstack_offset = offset; return vm_unmapped_area(&info); } diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 25c3502..560dae7 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -120,6 +120,19 @@ 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/metag/include/asm/barrier.h b/arch/metag/include/asm/barrier.h index c7591e8..ecef036 100644 --- a/arch/metag/include/asm/barrier.h +++ b/arch/metag/include/asm/barrier.h @@ -89,7 +89,7 @@ static inline void fence(void) do { \ compiletime_assert_atomic_type(*p); \ smp_mb(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ diff --git a/arch/metag/mm/hugetlbpage.c b/arch/metag/mm/hugetlbpage.c index 3c52fa6..11b2ad8 100644 --- a/arch/metag/mm/hugetlbpage.c +++ b/arch/metag/mm/hugetlbpage.c @@ -200,6 +200,7 @@ hugetlb_get_unmapped_area_new_pmd(unsigned long len) info.high_limit = TASK_SIZE; info.align_mask = PAGE_MASK & HUGEPT_MASK; info.align_offset = 0; + info.threadstack_offset = 0; return vm_unmapped_area(&info); } 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/Kconfig b/arch/mips/Kconfig index 4e238e6..7c9ed92 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -2392,6 +2392,7 @@ source "kernel/Kconfig.preempt" config KEXEC bool "Kexec system call" + depends on !GRKERNSEC_KMEM help kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot diff --git a/arch/mips/cavium-octeon/dma-octeon.c b/arch/mips/cavium-octeon/dma-octeon.c index 02f2444..506969c 100644 --- a/arch/mips/cavium-octeon/dma-octeon.c +++ b/arch/mips/cavium-octeon/dma-octeon.c @@ -199,7 +199,7 @@ static void octeon_dma_free_coherent(struct device *dev, size_t size, if (dma_release_from_coherent(dev, order, vaddr)) return; - swiotlb_free_coherent(dev, size, vaddr, dma_handle); + swiotlb_free_coherent(dev, size, vaddr, dma_handle, attrs); } static dma_addr_t octeon_unity_phys_to_dma(struct device *dev, phys_addr_t paddr) diff --git a/arch/mips/include/asm/atomic.h b/arch/mips/include/asm/atomic.h index 37b2bef..02122b8 100644 --- a/arch/mips/include/asm/atomic.h +++ b/arch/mips/include/asm/atomic.h @@ -21,15 +21,39 @@ #include #include +#ifdef CONFIG_GENERIC_ATOMIC64 +#include +#endif + #define ATOMIC_INIT(i) { (i) } +#ifdef CONFIG_64BIT +#define _ASM_EXTABLE(from, to) \ +" .section __ex_table,\"a\"\n" \ +" .dword " #from ", " #to"\n" \ +" .previous\n" +#else +#define _ASM_EXTABLE(from, to) \ +" .section __ex_table,\"a\"\n" \ +" .word " #from ", " #to"\n" \ +" .previous\n" +#endif + /* * atomic_read - read atomic variable * @v: pointer of type atomic_t * * Atomically reads the value of @v. */ -#define atomic_read(v) (*(volatile int *)&(v)->counter) +static inline int atomic_read(const atomic_t *v) +{ + return (*(volatile const int *) &v->counter); +} + +static inline int atomic_read_unchecked(const atomic_unchecked_t *v) +{ + return (*(volatile const int *) &v->counter); +} /* * atomic_set - set atomic variable @@ -38,7 +62,15 @@ * * Atomically sets the value of @v to @i. */ -#define atomic_set(v, i) ((v)->counter = (i)) +static inline void atomic_set(atomic_t *v, int i) +{ + v->counter = i; +} + +static inline void atomic_set_unchecked(atomic_unchecked_t *v, int i) +{ + v->counter = i; +} /* * atomic_add - add integer to atomic variable @@ -47,7 +79,67 @@ * * Atomically adds @i to @v. */ -static __inline__ void atomic_add(int i, atomic_t * v) +static __inline__ void atomic_add(int i, atomic_t *v) +{ + int temp; + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %0, %1 # atomic_add \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: add %0, %2 \n" +#else + " addu %0, %2 \n" +#endif + " sc %0, %1 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %0, %1 # atomic_add \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: add %0, %2 \n" +#else + " addu %0, %2 \n" +#endif + " sc %0, %1 \n" + " beqz %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: add %0, %1 \n" + "2: \n" + _ASM_EXTABLE(1b, 2b) +#else + " addu %0, %1 \n" +#endif + : "+r" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } +} + +static __inline__ void atomic_add_unchecked(int i, atomic_unchecked_t *v) { if (kernel_uses_llsc && R10000_LLSC_WAR) { int temp; @@ -90,7 +182,67 @@ static __inline__ void atomic_add(int i, atomic_t * v) * * Atomically subtracts @i from @v. */ -static __inline__ void atomic_sub(int i, atomic_t * v) +static __inline__ void atomic_sub(int i, atomic_t *v) +{ + int temp; + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %0, %1 # atomic64_sub \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: sub %0, %2 \n" +#else + " subu %0, %2 \n" +#endif + " sc %0, %1 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %0, %1 # atomic64_sub \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: sub %0, %2 \n" +#else + " subu %0, %2 \n" +#endif + " sc %0, %1 \n" + " beqz %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: sub %0, %1 \n" + "2: \n" + _ASM_EXTABLE(1b, 2b) +#else + " subu %0, %1 \n" +#endif + : "+r" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } +} + +static __inline__ void atomic_sub_unchecked(long i, atomic_unchecked_t *v) { if (kernel_uses_llsc && R10000_LLSC_WAR) { int temp; @@ -129,7 +281,93 @@ static __inline__ void atomic_sub(int i, atomic_t * v) /* * Same as above, but return the result value */ -static __inline__ int atomic_add_return(int i, atomic_t * v) +static __inline__ int atomic_add_return(int i, atomic_t *v) +{ + int result; + int temp; + + smp_mb__before_llsc(); + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %1, %2 # atomic_add_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: add %0, %1, %3 \n" +#else + " addu %0, %1, %3 \n" +#endif + " sc %0, %2 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " b 4f \n" + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: addu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %1, %2 # atomic_add_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: add %0, %1, %3 \n" +#else + " addu %0, %1, %3 \n" +#endif + " sc %0, %2 \n" + " bnez %0, 4f \n" + " b 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: addu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( + " lw %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: add %0, %2 \n" +#else + " addu %0, %2 \n" +#endif + " sw %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Note: Dest reg is not modified on overflow */ + "2: \n" + _ASM_EXTABLE(1b, 2b) +#endif + : "=&r" (result), "+m" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } + + smp_llsc_mb(); + + return result; +} + +static __inline__ int atomic_add_return_unchecked(int i, atomic_unchecked_t *v) { int result; @@ -178,7 +416,93 @@ static __inline__ int atomic_add_return(int i, atomic_t * v) return result; } -static __inline__ int atomic_sub_return(int i, atomic_t * v) +static __inline__ int atomic_sub_return(int i, atomic_t *v) +{ + int result; + int temp; + + smp_mb__before_llsc(); + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %1, %2 # atomic_sub_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: sub %0, %1, %3 \n" +#else + " subu %0, %1, %3 \n" +#endif + " sc %0, %2 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " b 4f \n" + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: subu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "=m" (v->counter) + : "Ir" (i), "m" (v->counter) + : "memory"); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: ll %1, %2 # atomic_sub_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: sub %0, %1, %3 \n" +#else + " subu %0, %1, %3 \n" +#endif + " sc %0, %2 \n" + " bnez %0, 4f \n" + " b 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: subu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( + " lw %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: sub %0, %2 \n" +#else + " subu %0, %2 \n" +#endif + " sw %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Note: Dest reg is not modified on overflow */ + "2: \n" + _ASM_EXTABLE(1b, 2b) +#endif + : "=&r" (result), "+m" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } + + smp_llsc_mb(); + + return result; +} +static __inline__ int atomic_sub_return_unchecked(int i, atomic_unchecked_t *v) { int result; @@ -238,7 +562,7 @@ static __inline__ int atomic_sub_return(int i, atomic_t * v) * Atomically test @v and subtract @i if @v is greater or equal than @i. * The function returns the old value of @v minus @i. */ -static __inline__ int atomic_sub_if_positive(int i, atomic_t * v) +static __inline__ int atomic_sub_if_positive(int i, atomic_t *v) { int result; @@ -295,8 +619,26 @@ static __inline__ int atomic_sub_if_positive(int i, atomic_t * v) return result; } -#define atomic_cmpxchg(v, o, n) (cmpxchg(&((v)->counter), (o), (n))) -#define atomic_xchg(v, new) (xchg(&((v)->counter), (new))) +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 a given value @@ -324,6 +666,10 @@ static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u) #define atomic_dec_return(v) atomic_sub_return(1, (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); +} /* * atomic_sub_and_test - subtract value from variable and test result @@ -345,6 +691,10 @@ static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u) * 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_add_return_unchecked(1, v) == 0; +} /* * atomic_dec_and_test - decrement by 1 and test @@ -369,6 +719,10 @@ static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u) * Atomically increments @v by 1. */ #define atomic_inc(v) atomic_add(1, (v)) +static __inline__ void atomic_inc_unchecked(atomic_unchecked_t *v) +{ + atomic_add_unchecked(1, v); +} /* * atomic_dec - decrement and test @@ -377,6 +731,10 @@ static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u) * Atomically decrements @v by 1. */ #define atomic_dec(v) atomic_sub(1, (v)) +static __inline__ void atomic_dec_unchecked(atomic_unchecked_t *v) +{ + atomic_sub_unchecked(1, v); +} /* * atomic_add_negative - add and test if negative @@ -398,14 +756,30 @@ static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u) * @v: pointer of type atomic64_t * */ -#define atomic64_read(v) (*(volatile long *)&(v)->counter) +static inline long atomic64_read(const atomic64_t *v) +{ + return (*(volatile const long *) &v->counter); +} + +static inline long atomic64_read_unchecked(const atomic64_unchecked_t *v) +{ + return (*(volatile const long *) &v->counter); +} /* * atomic64_set - set atomic variable * @v: pointer of type atomic64_t * @i: required value */ -#define atomic64_set(v, i) ((v)->counter = (i)) +static inline void atomic64_set(atomic64_t *v, long i) +{ + v->counter = i; +} + +static inline void atomic64_set_unchecked(atomic64_unchecked_t *v, long i) +{ + v->counter = i; +} /* * atomic64_add - add integer to atomic variable @@ -414,7 +788,66 @@ static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u) * * Atomically adds @i to @v. */ -static __inline__ void atomic64_add(long i, atomic64_t * v) +static __inline__ void atomic64_add(long i, atomic64_t *v) +{ + long temp; + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %0, %1 # atomic64_add \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: dadd %0, %2 \n" +#else + " daddu %0, %2 \n" +#endif + " scd %0, %1 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %0, %1 # atomic64_add \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: dadd %0, %2 \n" +#else + " daddu %0, %2 \n" +#endif + " scd %0, %1 \n" + " beqz %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: dadd %0, %1 \n" + "2: \n" + _ASM_EXTABLE(1b, 2b) +#else + " daddu %0, %1 \n" +#endif + : "+r" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } +} +static __inline__ void atomic64_add_unchecked(long i, atomic64_unchecked_t *v) { if (kernel_uses_llsc && R10000_LLSC_WAR) { long temp; @@ -457,7 +890,67 @@ static __inline__ void atomic64_add(long i, atomic64_t * v) * * Atomically subtracts @i from @v. */ -static __inline__ void atomic64_sub(long i, atomic64_t * v) +static __inline__ void atomic64_sub(long i, atomic64_t *v) +{ + long temp; + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %0, %1 # atomic64_sub \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: dsub %0, %2 \n" +#else + " dsubu %0, %2 \n" +#endif + " scd %0, %1 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %0, %1 # atomic64_sub \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "2: dsub %0, %2 \n" +#else + " dsubu %0, %2 \n" +#endif + " scd %0, %1 \n" + " beqz %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + "3: \n" + _ASM_EXTABLE(2b, 3b) +#endif + " .set mips0 \n" + : "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: dsub %0, %1 \n" + "2: \n" + _ASM_EXTABLE(1b, 2b) +#else + " dsubu %0, %1 \n" +#endif + : "+r" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } +} + +static __inline__ void atomic64_sub_unchecked(long i, atomic64_unchecked_t *v) { if (kernel_uses_llsc && R10000_LLSC_WAR) { long temp; @@ -496,7 +989,93 @@ static __inline__ void atomic64_sub(long i, atomic64_t * v) /* * Same as above, but return the result value */ -static __inline__ long atomic64_add_return(long i, atomic64_t * v) +static __inline__ long atomic64_add_return(long i, atomic64_t *v) +{ + long result; + long temp; + + smp_mb__before_llsc(); + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %1, %2 # atomic64_add_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: dadd %0, %1, %3 \n" +#else + " daddu %0, %1, %3 \n" +#endif + " scd %0, %2 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " b 4f \n" + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: daddu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "+m" (v->counter) + : "Ir" (i)); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %1, %2 # atomic64_add_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: dadd %0, %1, %3 \n" +#else + " daddu %0, %1, %3 \n" +#endif + " scd %0, %2 \n" + " bnez %0, 4f \n" + " b 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: daddu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "=m" (v->counter) + : "Ir" (i), "m" (v->counter) + : "memory"); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( + " ld %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: dadd %0, %2 \n" +#else + " daddu %0, %2 \n" +#endif + " sd %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Note: Dest reg is not modified on overflow */ + "2: \n" + _ASM_EXTABLE(1b, 2b) +#endif + : "=&r" (result), "+m" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } + + smp_llsc_mb(); + + return result; +} +static __inline__ long atomic64_add_return_unchecked(long i, atomic64_unchecked_t *v) { long result; @@ -546,7 +1125,97 @@ static __inline__ long atomic64_add_return(long i, atomic64_t * v) return result; } -static __inline__ long atomic64_sub_return(long i, atomic64_t * v) +static __inline__ long atomic64_sub_return(long i, atomic64_t *v) +{ + long result; + long temp; + + smp_mb__before_llsc(); + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + long temp; + + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %1, %2 # atomic64_sub_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: dsub %0, %1, %3 \n" +#else + " dsubu %0, %1, %3 \n" +#endif + " scd %0, %2 \n" + " beqzl %0, 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " b 4f \n" + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: dsubu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "=m" (v->counter) + : "Ir" (i), "m" (v->counter) + : "memory"); + } else if (kernel_uses_llsc) { + __asm__ __volatile__( + " .set mips3 \n" + "1: lld %1, %2 # atomic64_sub_return \n" +#ifdef CONFIG_PAX_REFCOUNT + "2: dsub %0, %1, %3 \n" +#else + " dsubu %0, %1, %3 \n" +#endif + " scd %0, %2 \n" + " bnez %0, 4f \n" + " b 1b \n" +#ifdef CONFIG_PAX_REFCOUNT + " .set noreorder \n" + "3: b 5f \n" + " move %0, %1 \n" + " .set reorder \n" + _ASM_EXTABLE(2b, 3b) +#endif + "4: dsubu %0, %1, %3 \n" +#ifdef CONFIG_PAX_REFCOUNT + "5: \n" +#endif + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "=m" (v->counter) + : "Ir" (i), "m" (v->counter) + : "memory"); + } else { + unsigned long flags; + + raw_local_irq_save(flags); + __asm__ __volatile__( + " ld %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Exception on overflow. */ + "1: dsub %0, %2 \n" +#else + " dsubu %0, %2 \n" +#endif + " sd %0, %1 \n" +#ifdef CONFIG_PAX_REFCOUNT + /* Note: Dest reg is not modified on overflow */ + "2: \n" + _ASM_EXTABLE(1b, 2b) +#endif + : "=&r" (result), "+m" (v->counter) : "Ir" (i)); + raw_local_irq_restore(flags); + } + + smp_llsc_mb(); + + return result; +} + +static __inline__ long atomic64_sub_return_unchecked(long i, atomic64_unchecked_t *v) { long result; @@ -605,7 +1274,7 @@ static __inline__ long atomic64_sub_return(long i, atomic64_t * v) * Atomically test @v and subtract @i if @v is greater or equal than @i. * The function returns the old value of @v minus @i. */ -static __inline__ long atomic64_sub_if_positive(long i, atomic64_t * v) +static __inline__ long atomic64_sub_if_positive(long i, atomic64_t *v) { long result; @@ -662,9 +1331,26 @@ static __inline__ long atomic64_sub_if_positive(long i, atomic64_t * v) return result; } -#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_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); +} + +static inline long atomic64_xchg_unchecked(atomic64_unchecked_t *v, long new) +{ + return xchg(&(v->counter), new); +} /** * atomic64_add_unless - add unless the number is a given value @@ -694,6 +1380,7 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) #define atomic64_dec_return(v) atomic64_sub_return(1, (v)) #define atomic64_inc_return(v) atomic64_add_return(1, (v)) +#define atomic64_inc_return_unchecked(v) atomic64_add_return_unchecked(1, (v)) /* * atomic64_sub_and_test - subtract value from variable and test result @@ -715,6 +1402,7 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) * other cases. */ #define atomic64_inc_and_test(v) (atomic64_inc_return(v) == 0) +#define atomic64_inc_and_test_unchecked(v) atomic64_add_return_unchecked(1, (v)) == 0) /* * atomic64_dec_and_test - decrement by 1 and test @@ -739,6 +1427,7 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) * Atomically increments @v by 1. */ #define atomic64_inc(v) atomic64_add(1, (v)) +#define atomic64_inc_unchecked(v) atomic64_add_unchecked(1, (v)) /* * atomic64_dec - decrement and test @@ -747,6 +1436,7 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) * Atomically decrements @v by 1. */ #define atomic64_dec(v) atomic64_sub(1, (v)) +#define atomic64_dec_unchecked(v) atomic64_sub_unchecked(1, (v)) /* * atomic64_add_negative - add and test if negative diff --git a/arch/mips/include/asm/barrier.h b/arch/mips/include/asm/barrier.h index d0101dd..266982c 100644 --- a/arch/mips/include/asm/barrier.h +++ b/arch/mips/include/asm/barrier.h @@ -184,7 +184,7 @@ do { \ compiletime_assert_atomic_type(*p); \ smp_mb(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ 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 d414405..6bb4ba2 100644 --- a/arch/mips/include/asm/elf.h +++ b/arch/mips/include/asm/elf.h @@ -398,13 +398,16 @@ 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/exec.h b/arch/mips/include/asm/exec.h index c1f6afa..38cc6e9 100644 --- a/arch/mips/include/asm/exec.h +++ b/arch/mips/include/asm/exec.h @@ -12,6 +12,6 @@ #ifndef _ASM_EXEC_H #define _ASM_EXEC_H -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) #endif /* _ASM_EXEC_H */ diff --git a/arch/mips/include/asm/hw_irq.h b/arch/mips/include/asm/hw_irq.h index 9e8ef59..1139d6b 100644 --- a/arch/mips/include/asm/hw_irq.h +++ b/arch/mips/include/asm/hw_irq.h @@ -10,7 +10,7 @@ #include -extern atomic_t irq_err_count; +extern atomic_unchecked_t irq_err_count; /* * interrupt-retrigger: NOP for now. This may not be appropriate for all diff --git a/arch/mips/include/asm/local.h b/arch/mips/include/asm/local.h index 46dfc3c..a16b13a 100644 --- a/arch/mips/include/asm/local.h +++ b/arch/mips/include/asm/local.h @@ -12,15 +12,25 @@ typedef struct atomic_long_t a; } local_t; +typedef struct { + atomic_long_unchecked_t a; +} local_unchecked_t; + #define LOCAL_INIT(i) { ATOMIC_LONG_INIT(i) } #define local_read(l) atomic_long_read(&(l)->a) +#define local_read_unchecked(l) atomic_long_read_unchecked(&(l)->a) #define local_set(l, i) atomic_long_set(&(l)->a, (i)) +#define local_set_unchecked(l, i) atomic_long_set_unchecked(&(l)->a, (i)) #define local_add(i, l) atomic_long_add((i), (&(l)->a)) +#define local_add_unchecked(i, l) atomic_long_add_unchecked((i), (&(l)->a)) #define local_sub(i, l) atomic_long_sub((i), (&(l)->a)) +#define local_sub_unchecked(i, l) atomic_long_sub_unchecked((i), (&(l)->a)) #define local_inc(l) atomic_long_inc(&(l)->a) +#define local_inc_unchecked(l) atomic_long_inc_unchecked(&(l)->a) #define local_dec(l) atomic_long_dec(&(l)->a) +#define local_dec_unchecked(l) atomic_long_dec_unchecked(&(l)->a) /* * Same as above, but return the result value @@ -70,6 +80,51 @@ static __inline__ long local_add_return(long i, local_t * l) return result; } +static __inline__ long local_add_return_unchecked(long i, local_unchecked_t * l) +{ + unsigned long result; + + if (kernel_uses_llsc && R10000_LLSC_WAR) { + unsigned long temp; + + __asm__ __volatile__( + " .set mips3 \n" + "1:" __LL "%1, %2 # local_add_return \n" + " addu %0, %1, %3 \n" + __SC "%0, %2 \n" + " beqzl %0, 1b \n" + " addu %0, %1, %3 \n" + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "=m" (l->a.counter) + : "Ir" (i), "m" (l->a.counter) + : "memory"); + } else if (kernel_uses_llsc) { + unsigned long temp; + + __asm__ __volatile__( + " .set mips3 \n" + "1:" __LL "%1, %2 # local_add_return \n" + " addu %0, %1, %3 \n" + __SC "%0, %2 \n" + " beqz %0, 1b \n" + " addu %0, %1, %3 \n" + " .set mips0 \n" + : "=&r" (result), "=&r" (temp), "=m" (l->a.counter) + : "Ir" (i), "m" (l->a.counter) + : "memory"); + } else { + unsigned long flags; + + local_irq_save(flags); + result = l->a.counter; + result += i; + l->a.counter = result; + local_irq_restore(flags); + } + + return result; +} + static __inline__ long local_sub_return(long i, local_t * l) { unsigned long result; @@ -117,6 +172,8 @@ static __inline__ long local_sub_return(long i, local_t * l) #define local_cmpxchg(l, o, n) \ ((long)cmpxchg_local(&((l)->a.counter), (o), (n))) +#define local_cmpxchg_unchecked(l, o, n) \ + ((long)cmpxchg_local(&((l)->a.counter), (o), (n))) #define local_xchg(l, n) (atomic_long_xchg((&(l)->a), (n))) /** diff --git a/arch/mips/include/asm/page.h b/arch/mips/include/asm/page.h index 5699ec3..95def83 100644 --- a/arch/mips/include/asm/page.h +++ b/arch/mips/include/asm/page.h @@ -120,7 +120,7 @@ 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 b336037..5b874cc 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/pgtable.h b/arch/mips/include/asm/pgtable.h index 539ddd1..8783f9a 100644 --- a/arch/mips/include/asm/pgtable.h +++ b/arch/mips/include/asm/pgtable.h @@ -20,6 +20,9 @@ #include #include +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) + struct mm_struct; struct vm_area_struct; diff --git a/arch/mips/include/asm/thread_info.h b/arch/mips/include/asm/thread_info.h index 7de8658..c109224 100644 --- a/arch/mips/include/asm/thread_info.h +++ b/arch/mips/include/asm/thread_info.h @@ -105,6 +105,9 @@ static inline struct thread_info *current_thread_info(void) #define TIF_SECCOMP 4 /* secure computing */ #define TIF_NOTIFY_RESUME 5 /* callback before returning to user */ #define TIF_RESTORE_SIGMASK 9 /* restore signal mask in do_signal() */ +/* li takes a 32bit immediate */ +#define TIF_GRSEC_SETXID 10 /* update credentials on syscall entry/exit */ + #define TIF_USEDFPU 16 /* FPU was used by this task this quantum (SMP) */ #define TIF_MEMDIE 18 /* is terminating due to OOM killer */ #define TIF_NOHZ 19 /* in adaptive nohz mode */ @@ -138,14 +141,16 @@ static inline struct thread_info *current_thread_info(void) #define _TIF_USEDMSA (1< #include #include diff --git a/arch/mips/kernel/binfmt_elfo32.c b/arch/mips/kernel/binfmt_elfo32.c index 71df942..199dd19 100644 --- a/arch/mips/kernel/binfmt_elfo32.c +++ b/arch/mips/kernel/binfmt_elfo32.c @@ -70,6 +70,13 @@ 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 /* These MUST be defined before elf.h gets included */ diff --git a/arch/mips/kernel/i8259.c b/arch/mips/kernel/i8259.c index 50b3648..c2f3cec 100644 --- a/arch/mips/kernel/i8259.c +++ b/arch/mips/kernel/i8259.c @@ -201,7 +201,7 @@ spurious_8259A_irq: printk(KERN_DEBUG "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/mips/kernel/irq-gt641xx.c b/arch/mips/kernel/irq-gt641xx.c index 44a1f79..2bd6aa3 100644 --- a/arch/mips/kernel/irq-gt641xx.c +++ b/arch/mips/kernel/irq-gt641xx.c @@ -110,7 +110,7 @@ void gt641xx_irq_dispatch(void) } } - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); } void __init gt641xx_irq_init(void) diff --git a/arch/mips/kernel/irq.c b/arch/mips/kernel/irq.c index d2bfbc2..a8eacd2 100644 --- a/arch/mips/kernel/irq.c +++ b/arch/mips/kernel/irq.c @@ -76,17 +76,17 @@ void ack_bad_irq(unsigned int irq) printk("unexpected IRQ # %d\n", irq); } -atomic_t irq_err_count; +atomic_unchecked_t irq_err_count; int arch_show_interrupts(struct seq_file *p, int prec) { - 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)); return 0; } asmlinkage void spurious_interrupt(void) { - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); } void __init init_IRQ(void) @@ -109,7 +109,10 @@ void __init init_IRQ(void) #endif } + #ifdef DEBUG_STACKOVERFLOW +extern void gr_handle_kernel_exploit(void); + static inline void check_stack_overflow(void) { unsigned long sp; @@ -125,6 +128,7 @@ static inline void check_stack_overflow(void) printk("do_IRQ: stack overflow: %ld\n", sp - sizeof(struct thread_info)); dump_stack(); + gr_handle_kernel_exploit(); } } #else diff --git a/arch/mips/kernel/pm-cps.c b/arch/mips/kernel/pm-cps.c index c4c2069..bde8051 100644 --- a/arch/mips/kernel/pm-cps.c +++ b/arch/mips/kernel/pm-cps.c @@ -168,7 +168,7 @@ int cps_pm_enter_state(enum cps_pm_state state) nc_core_ready_count = nc_addr; /* Ensure ready_count is zero-initialised before the assembly runs */ - ACCESS_ONCE(*nc_core_ready_count) = 0; + ACCESS_ONCE_RW(*nc_core_ready_count) = 0; coupled_barrier(&per_cpu(pm_barrier, core), online); /* Run the generated entry code */ diff --git a/arch/mips/kernel/process.c b/arch/mips/kernel/process.c index 0a1ec0f..d9e93b6 100644 --- a/arch/mips/kernel/process.c +++ b/arch/mips/kernel/process.c @@ -572,15 +572,3 @@ 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/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index aae7119..8b16434 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -762,6 +762,10 @@ long arch_ptrace(struct task_struct *child, long request, return ret; } +#ifdef CONFIG_GRKERNSEC_SETXID +extern void gr_delayed_cred_worker(void); +#endif + /* * Notification of system call entry/exit * - triggered by current->work.syscall_trace @@ -778,6 +782,11 @@ asmlinkage long syscall_trace_enter(struct pt_regs *regs, long syscall) tracehook_report_syscall_entry(regs)) ret = -1; +#ifdef CONFIG_GRKERNSEC_SETXID + if (unlikely(test_and_clear_thread_flag(TIF_GRSEC_SETXID))) + gr_delayed_cred_worker(); +#endif + if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_enter(regs, regs->regs[2]); diff --git a/arch/mips/kernel/reset.c b/arch/mips/kernel/reset.c index 07fc524..b9d7f28 100644 --- a/arch/mips/kernel/reset.c +++ b/arch/mips/kernel/reset.c @@ -13,6 +13,7 @@ #include #include +#include /* * Urgs ... Too many MIPS machines to handle this in a generic way. @@ -29,16 +30,19 @@ void machine_restart(char *command) { if (_machine_restart) _machine_restart(command); + BUG(); } void machine_halt(void) { if (_machine_halt) _machine_halt(); + BUG(); } void machine_power_off(void) { if (pm_power_off) pm_power_off(); + BUG(); } diff --git a/arch/mips/kernel/sync-r4k.c b/arch/mips/kernel/sync-r4k.c index 2242bdd..b284048 100644 --- a/arch/mips/kernel/sync-r4k.c +++ b/arch/mips/kernel/sync-r4k.c @@ -18,8 +18,8 @@ #include static atomic_t count_start_flag = ATOMIC_INIT(0); -static atomic_t count_count_start = ATOMIC_INIT(0); -static atomic_t count_count_stop = ATOMIC_INIT(0); +static atomic_unchecked_t count_count_start = ATOMIC_INIT(0); +static atomic_unchecked_t count_count_stop = ATOMIC_INIT(0); static atomic_t count_reference = ATOMIC_INIT(0); #define COUNTON 100 @@ -58,13 +58,13 @@ void synchronise_count_master(int cpu) for (i = 0; i < NR_LOOPS; i++) { /* slaves loop on '!= 2' */ - while (atomic_read(&count_count_start) != 1) + while (atomic_read_unchecked(&count_count_start) != 1) mb(); - atomic_set(&count_count_stop, 0); + atomic_set_unchecked(&count_count_stop, 0); smp_wmb(); /* this lets the slaves write their count register */ - atomic_inc(&count_count_start); + atomic_inc_unchecked(&count_count_start); /* * Everyone initialises count in the last loop: @@ -75,11 +75,11 @@ void synchronise_count_master(int cpu) /* * Wait for all slaves to leave the synchronization point: */ - while (atomic_read(&count_count_stop) != 1) + while (atomic_read_unchecked(&count_count_stop) != 1) mb(); - atomic_set(&count_count_start, 0); + atomic_set_unchecked(&count_count_start, 0); smp_wmb(); - atomic_inc(&count_count_stop); + atomic_inc_unchecked(&count_count_stop); } /* Arrange for an interrupt in a short while */ write_c0_compare(read_c0_count() + COUNTON); @@ -112,8 +112,8 @@ void synchronise_count_slave(int cpu) initcount = atomic_read(&count_reference); for (i = 0; i < NR_LOOPS; i++) { - atomic_inc(&count_count_start); - while (atomic_read(&count_count_start) != 2) + atomic_inc_unchecked(&count_count_start); + while (atomic_read_unchecked(&count_count_start) != 2) mb(); /* @@ -122,8 +122,8 @@ void synchronise_count_slave(int cpu) if (i == NR_LOOPS-1) write_c0_count(initcount); - atomic_inc(&count_count_stop); - while (atomic_read(&count_count_stop) != 2) + atomic_inc_unchecked(&count_count_stop); + while (atomic_read_unchecked(&count_count_stop) != 2) mb(); } /* Arrange for an interrupt in a short while */ diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c index 51706d6..ec1178c 100644 --- a/arch/mips/kernel/traps.c +++ b/arch/mips/kernel/traps.c @@ -687,7 +687,18 @@ asmlinkage void do_ov(struct pt_regs *regs) siginfo_t info; prev_state = exception_enter(); - die_if_kernel("Integer overflow", regs); + if (unlikely(!user_mode(regs))) { + +#ifdef CONFIG_PAX_REFCOUNT + if (fixup_exception(regs)) { + pax_report_refcount_overflow(regs); + exception_exit(prev_state); + return; + } +#endif + + die("Integer overflow", regs); + } info.si_code = FPE_INTOVF; info.si_signo = SIGFPE; diff --git a/arch/mips/kvm/kvm_mips.c b/arch/mips/kvm/kvm_mips.c index f3c56a1..6a2f01c 100644 --- a/arch/mips/kvm/kvm_mips.c +++ b/arch/mips/kvm/kvm_mips.c @@ -841,7 +841,7 @@ long kvm_arch_vm_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg) return r; } -int kvm_arch_init(void *opaque) +int kvm_arch_init(const void *opaque) { int ret; diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c index becc42b..9e43d4b 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 @@ -199,6 +216,14 @@ bad_area: bad_area_nosemaphore: /* User mode accesses just cause a SIGSEGV */ if (user_mode(regs)) { + +#ifdef CONFIG_PAX_PAGEEXEC + if (cpu_has_rixi && (mm->pax_flags & MF_PAX_PAGEEXEC) && !write && address == instruction_pointer(regs)) { + pax_report_fault(regs, (void *)address, (void *)user_stack_pointer(regs)); + do_group_exit(SIGKILL); + } +#endif + tsk->thread.cp0_badvaddr = address; tsk->thread.error_code = write; #if 0 diff --git a/arch/mips/mm/mmap.c b/arch/mips/mm/mmap.c index f1baadd..5472dca 100644 --- a/arch/mips/mm/mmap.c +++ b/arch/mips/mm/mmap.c @@ -59,6 +59,7 @@ static unsigned long arch_get_unmapped_area_common(struct file *filp, struct vm_area_struct *vma; unsigned long addr = addr0; int do_color_align; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); struct vm_unmapped_area_info info; if (unlikely(len > TASK_SIZE)) @@ -84,6 +85,11 @@ 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); @@ -91,14 +97,14 @@ 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(vma, addr, len, offset)) return addr; } info.length = len; info.align_mask = do_color_align ? (PAGE_MASK & shm_align_mask) : 0; info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; if (dir == DOWN) { info.flags = VM_UNMAPPED_AREA_TOPDOWN; @@ -146,6 +152,10 @@ void arch_pick_mmap_layout(struct mm_struct *mm) { unsigned long random_factor = 0UL; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (current->flags & PF_RANDOMIZE) { random_factor = get_random_int(); random_factor = random_factor << PAGE_SHIFT; @@ -157,40 +167,25 @@ 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; } 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; } } -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; -} - int __virt_addr_valid(const volatile void *kaddr) { return pfn_valid(PFN_DOWN(virt_to_phys(kaddr))); diff --git a/arch/mips/pci/pci-octeon.c b/arch/mips/pci/pci-octeon.c index 59cccd9..f39ac2f 100644 --- a/arch/mips/pci/pci-octeon.c +++ b/arch/mips/pci/pci-octeon.c @@ -327,8 +327,8 @@ static int octeon_write_config(struct pci_bus *bus, unsigned int devfn, static struct pci_ops octeon_pci_ops = { - octeon_read_config, - octeon_write_config, + .read = octeon_read_config, + .write = octeon_write_config, }; static struct resource octeon_pci_mem_resource = { diff --git a/arch/mips/pci/pcie-octeon.c b/arch/mips/pci/pcie-octeon.c index 5e36c33..eb4a17b 100644 --- a/arch/mips/pci/pcie-octeon.c +++ b/arch/mips/pci/pcie-octeon.c @@ -1792,8 +1792,8 @@ static int octeon_dummy_write_config(struct pci_bus *bus, unsigned int devfn, } static struct pci_ops octeon_pcie0_ops = { - octeon_pcie0_read_config, - octeon_pcie0_write_config, + .read = octeon_pcie0_read_config, + .write = octeon_pcie0_write_config, }; static struct resource octeon_pcie0_mem_resource = { @@ -1813,8 +1813,8 @@ static struct pci_controller octeon_pcie0_controller = { }; static struct pci_ops octeon_pcie1_ops = { - octeon_pcie1_read_config, - octeon_pcie1_write_config, + .read = octeon_pcie1_read_config, + .write = octeon_pcie1_write_config, }; static struct resource octeon_pcie1_mem_resource = { @@ -1834,8 +1834,8 @@ static struct pci_controller octeon_pcie1_controller = { }; static struct pci_ops octeon_dummy_ops = { - octeon_dummy_read_config, - octeon_dummy_write_config, + .read = octeon_dummy_read_config, + .write = octeon_dummy_write_config, }; static struct resource octeon_dummy_mem_resource = { diff --git a/arch/mips/sgi-ip27/ip27-nmi.c b/arch/mips/sgi-ip27/ip27-nmi.c index a2358b4..7cead4f 100644 --- a/arch/mips/sgi-ip27/ip27-nmi.c +++ b/arch/mips/sgi-ip27/ip27-nmi.c @@ -187,9 +187,9 @@ void cont_nmi_dump(void) { #ifndef REAL_NMI_SIGNAL - static atomic_t nmied_cpus = ATOMIC_INIT(0); + static atomic_unchecked_t nmied_cpus = ATOMIC_INIT(0); - atomic_inc(&nmied_cpus); + atomic_inc_unchecked(&nmied_cpus); #endif /* * Only allow 1 cpu to proceed @@ -233,7 +233,7 @@ cont_nmi_dump(void) udelay(10000); } #else - while (atomic_read(&nmied_cpus) != num_online_cpus()); + while (atomic_read_unchecked(&nmied_cpus) != num_online_cpus()); #endif /* diff --git a/arch/mips/sni/rm200.c b/arch/mips/sni/rm200.c index a046b30..6799527 100644 --- a/arch/mips/sni/rm200.c +++ b/arch/mips/sni/rm200.c @@ -270,7 +270,7 @@ spurious_8259A_irq: "spurious RM200 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/mips/vr41xx/common/icu.c b/arch/mips/vr41xx/common/icu.c index 41e873b..34d33a7 100644 --- a/arch/mips/vr41xx/common/icu.c +++ b/arch/mips/vr41xx/common/icu.c @@ -653,7 +653,7 @@ static int icu_get_irq(unsigned int irq) printk(KERN_ERR "spurious ICU interrupt: %04x,%04x\n", pend1, pend2); - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); return -1; } diff --git a/arch/mips/vr41xx/common/irq.c b/arch/mips/vr41xx/common/irq.c index ae0e4ee..e8f0692 100644 --- a/arch/mips/vr41xx/common/irq.c +++ b/arch/mips/vr41xx/common/irq.c @@ -64,7 +64,7 @@ static void irq_dispatch(unsigned int irq) irq_cascade_t *cascade; if (irq >= NR_IRQS) { - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); return; } @@ -84,7 +84,7 @@ static void irq_dispatch(unsigned int irq) ret = cascade->get_irq(irq); irq = ret; if (ret < 0) - atomic_inc(&irq_err_count); + atomic_inc_unchecked(&irq_err_count); else irq_dispatch(irq); if (!irqd_irq_disabled(idata) && chip->irq_unmask) 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 0be2db2..1b0f26d 100644 --- a/arch/parisc/include/asm/atomic.h +++ b/arch/parisc/include/asm/atomic.h @@ -248,6 +248,16 @@ static inline long atomic64_dec_if_positive(atomic64_t *v) return dec; } +#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 3391d06..c23a2cc 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 f213f5b..0af3e8e 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 22b89d1..ce34230 100644 --- a/arch/parisc/include/asm/pgtable.h +++ b/arch/parisc/include/asm/pgtable.h @@ -223,6 +223,17 @@ extern void purge_tlb_entries(struct mm_struct *, unsigned long); #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/include/asm/uaccess.h b/arch/parisc/include/asm/uaccess.h index 4006964..fcb3cc2 100644 --- a/arch/parisc/include/asm/uaccess.h +++ b/arch/parisc/include/asm/uaccess.h @@ -246,10 +246,10 @@ static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, unsigned long n) { - int sz = __compiletime_object_size(to); + size_t sz = __compiletime_object_size(to); int ret = -EFAULT; - if (likely(sz == -1 || !__builtin_constant_p(n) || sz >= n)) + if (likely(sz == (size_t)-1 || !__builtin_constant_p(n) || sz >= n)) ret = __copy_from_user(to, from, n); else copy_from_user_overflow(); diff --git a/arch/parisc/kernel/module.c b/arch/parisc/kernel/module.c index 50dfafc..b9fc230 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) @@ -371,13 +393,13 @@ 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; @@ -395,7 +417,7 @@ 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; @@ -413,7 +435,7 @@ 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); @@ -431,7 +453,7 @@ 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 */ @@ -843,7 +865,7 @@ 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 e1ffea2..46ed66e 100644 --- a/arch/parisc/kernel/sys_parisc.c +++ b/arch/parisc/kernel/sys_parisc.c @@ -89,6 +89,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsigned long task_size = TASK_SIZE; int do_color_align, last_mmap; struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(current->mm, filp, flags); if (len > task_size) return -ENOMEM; @@ -106,6 +107,10 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, goto found_addr; } +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_color_align && last_mmap) addr = COLOR_ALIGN(addr, last_mmap, pgoff); @@ -124,6 +129,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, info.high_limit = mmap_upper_limit(); info.align_mask = last_mmap ? (PAGE_MASK & (SHM_COLOUR - 1)) : 0; info.align_offset = shared_align_offset(last_mmap, pgoff); + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); found_addr: @@ -143,6 +149,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, unsigned long addr = addr0; int do_color_align, last_mmap; struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(current->mm, filp, flags); #ifdef CONFIG_64BIT /* This should only ever run for 32-bit processes. */ @@ -167,6 +174,10 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, } /* requesting a specific address */ +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_color_align && last_mmap) addr = COLOR_ALIGN(addr, last_mmap, pgoff); @@ -184,6 +195,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, info.high_limit = mm->mmap_base; info.align_mask = last_mmap ? (PAGE_MASK & (SHM_COLOUR - 1)) : 0; info.align_offset = shared_align_offset(last_mmap, pgoff); + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); if (!(addr & ~PAGE_MASK)) goto found_addr; @@ -249,6 +261,13 @@ void arch_pick_mmap_layout(struct mm_struct *mm) mm->mmap_legacy_base = mmap_legacy_base(); mm->mmap_base = mmap_upper_limit(); +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) { + mm->mmap_legacy_base += mm->delta_mmap; + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; + } +#endif + if (mmap_is_legacy()) { mm->mmap_base = mm->mmap_legacy_base; mm->get_unmapped_area = arch_get_unmapped_area; diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c index 47ee620..1107387 100644 --- a/arch/parisc/kernel/traps.c +++ b/arch/parisc/kernel/traps.c @@ -726,9 +726,7 @@ 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 3ca9c11..d163ef7 100644 --- a/arch/parisc/mm/fault.c +++ b/arch/parisc/mm/fault.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -50,7 +51,7 @@ int show_unhandled_signals = 1; 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) { @@ -136,6 +137,116 @@ 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; @@ -234,8 +345,33 @@ retry: good_area: - 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/Kconfig b/arch/powerpc/Kconfig index 80b94b0..a3274fb 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -398,6 +398,7 @@ config PPC64_SUPPORTS_MEMORY_FAILURE config KEXEC bool "kexec system call" depends on (PPC_BOOK3S || FSL_BOOKE || (44x && !SMP)) + depends on !GRKERNSEC_KMEM help kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot diff --git a/arch/powerpc/include/asm/atomic.h b/arch/powerpc/include/asm/atomic.h index 28992d0..c797b20 100644 --- a/arch/powerpc/include/asm/atomic.h +++ b/arch/powerpc/include/asm/atomic.h @@ -519,6 +519,16 @@ static __inline__ long atomic64_inc_not_zero(atomic64_t *v) return t1; } +#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/barrier.h b/arch/powerpc/include/asm/barrier.h index bab79a1..4a3eabc 100644 --- a/arch/powerpc/include/asm/barrier.h +++ b/arch/powerpc/include/asm/barrier.h @@ -73,7 +73,7 @@ do { \ compiletime_assert_atomic_type(*p); \ __lwsync(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ diff --git a/arch/powerpc/include/asm/cache.h b/arch/powerpc/include/asm/cache.h index ed0afc1..0332825 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 888d8f3..66f581c 100644 --- a/arch/powerpc/include/asm/elf.h +++ b/arch/powerpc/include/asm/elf.h @@ -28,8 +28,19 @@ 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 #define ELF_CORE_EFLAGS (is_elf2_task() ? 2 : 0) @@ -129,10 +140,6 @@ 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 - - #ifdef CONFIG_SPU_BASE /* Notes used in ET_CORE. Note name is "SPU//". */ #define NT_SPU 1 diff --git a/arch/powerpc/include/asm/exec.h b/arch/powerpc/include/asm/exec.h index 8196e9c..d83a9f3 100644 --- a/arch/powerpc/include/asm/exec.h +++ b/arch/powerpc/include/asm/exec.h @@ -4,6 +4,6 @@ #ifndef _ASM_POWERPC_EXEC_H #define _ASM_POWERPC_EXEC_H -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) #endif /* _ASM_POWERPC_EXEC_H */ diff --git a/arch/powerpc/include/asm/kmap_types.h b/arch/powerpc/include/asm/kmap_types.h index 5acabbd..7ea14fa 100644 --- a/arch/powerpc/include/asm/kmap_types.h +++ b/arch/powerpc/include/asm/kmap_types.h @@ -10,7 +10,7 @@ * 2 of the License, or (at your option) any later version. */ -#define KM_TYPE_NR 16 +#define KM_TYPE_NR 17 #endif /* __KERNEL__ */ #endif /* _ASM_POWERPC_KMAP_TYPES_H */ diff --git a/arch/powerpc/include/asm/local.h b/arch/powerpc/include/asm/local.h index b8da913..60b608a 100644 --- a/arch/powerpc/include/asm/local.h +++ b/arch/powerpc/include/asm/local.h @@ -9,15 +9,26 @@ typedef struct atomic_long_t a; } local_t; +typedef struct +{ + atomic_long_unchecked_t a; +} local_unchecked_t; + #define LOCAL_INIT(i) { ATOMIC_LONG_INIT(i) } #define local_read(l) atomic_long_read(&(l)->a) +#define local_read_unchecked(l) atomic_long_read_unchecked(&(l)->a) #define local_set(l,i) atomic_long_set(&(l)->a, (i)) +#define local_set_unchecked(l,i) atomic_long_set_unchecked(&(l)->a, (i)) #define local_add(i,l) atomic_long_add((i),(&(l)->a)) +#define local_add_unchecked(i,l) atomic_long_add_unchecked((i),(&(l)->a)) #define local_sub(i,l) atomic_long_sub((i),(&(l)->a)) +#define local_sub_unchecked(i,l) atomic_long_sub_unchecked((i),(&(l)->a)) #define local_inc(l) atomic_long_inc(&(l)->a) +#define local_inc_unchecked(l) atomic_long_inc_unchecked(&(l)->a) #define local_dec(l) atomic_long_dec(&(l)->a) +#define local_dec_unchecked(l) atomic_long_dec_unchecked(&(l)->a) static __inline__ long local_add_return(long a, local_t *l) { @@ -35,6 +46,7 @@ static __inline__ long local_add_return(long a, local_t *l) return t; } +#define local_add_return_unchecked(i, l) atomic_long_add_return_unchecked((i), (&(l)->a)) #define local_add_negative(a, l) (local_add_return((a), (l)) < 0) @@ -54,6 +66,7 @@ static __inline__ long local_sub_return(long a, local_t *l) return t; } +#define local_sub_return_unchecked(i, l) atomic_long_sub_return_unchecked((i), (&(l)->a)) static __inline__ long local_inc_return(local_t *l) { @@ -101,6 +114,8 @@ static __inline__ long local_dec_return(local_t *l) #define local_cmpxchg(l, o, n) \ (cmpxchg_local(&((l)->a.counter), (o), (n))) +#define local_cmpxchg_unchecked(l, o, n) \ + (cmpxchg_local(&((l)->a.counter), (o), (n))) #define local_xchg(l, n) (xchg_local(&((l)->a.counter), (n))) /** diff --git a/arch/powerpc/include/asm/mman.h b/arch/powerpc/include/asm/mman.h index 8565c25..2865190 100644 --- a/arch/powerpc/include/asm/mman.h +++ b/arch/powerpc/include/asm/mman.h @@ -24,7 +24,7 @@ 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 32e4e21..62afb12 100644 --- a/arch/powerpc/include/asm/page.h +++ b/arch/powerpc/include/asm/page.h @@ -230,8 +230,9 @@ 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) @@ -259,6 +260,9 @@ 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) + #ifndef CONFIG_PPC_BOOK3S_64 /* * Use the top bit of the higher-level page table entries to indicate whether diff --git a/arch/powerpc/include/asm/page_64.h b/arch/powerpc/include/asm/page_64.h index 88693ce..ac6f9ab 100644 --- a/arch/powerpc/include/asm/page_64.h +++ b/arch/powerpc/include/asm/page_64.h @@ -153,15 +153,18 @@ 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 4b0be20..c15a27d 100644 --- a/arch/powerpc/include/asm/pgalloc-64.h +++ b/arch/powerpc/include/asm/pgalloc-64.h @@ -54,6 +54,7 @@ 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) { @@ -71,6 +72,11 @@ 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)) @@ -173,6 +179,7 @@ extern void __tlb_remove_table(void *_table); #endif #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 d98c1ec..9f61569 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 #include /* For TASK_SIZE */ 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 bffd89d..a6641ed 100644 --- a/arch/powerpc/include/asm/reg.h +++ b/arch/powerpc/include/asm/reg.h @@ -251,6 +251,7 @@ #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/smp.h b/arch/powerpc/include/asm/smp.h index 5a6614a..d89995d1 100644 --- a/arch/powerpc/include/asm/smp.h +++ b/arch/powerpc/include/asm/smp.h @@ -51,7 +51,7 @@ struct smp_ops_t { int (*cpu_disable)(void); void (*cpu_die)(unsigned int nr); int (*cpu_bootable)(unsigned int nr); -}; +} __no_const; extern void smp_send_debugger_break(void); extern void start_secondary_resume(void); diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h index b034ecd..af7e31f 100644 --- a/arch/powerpc/include/asm/thread_info.h +++ b/arch/powerpc/include/asm/thread_info.h @@ -107,6 +107,8 @@ static inline struct thread_info *current_thread_info(void) #if defined(CONFIG_PPC64) #define TIF_ELF2ABI 18 /* function descriptors must die! */ #endif +/* mask must be expressable within 16 bits to satisfy 'andi' instruction reqs */ +#define TIF_GRSEC_SETXID 6 /* update credentials on syscall entry/exit */ /* as above, but as bit values */ #define _TIF_SYSCALL_TRACE (1< 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/Makefile b/arch/powerpc/kernel/Makefile index 670c312..60c2b52 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -27,6 +27,8 @@ CFLAGS_REMOVE_ftrace.o = -pg -mno-sched-epilog CFLAGS_REMOVE_time.o = -pg -mno-sched-epilog endif +CFLAGS_REMOVE_prom_init.o += $(LATENT_ENTROPY_PLUGIN_CFLAGS) + obj-y := cputable.o ptrace.o syscalls.o \ irq.o align.o signal_32.o pmc.o vdso.o \ process.o systbl.o idle.o \ diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S index bb9cac6..5181202 100644 --- a/arch/powerpc/kernel/exceptions-64e.S +++ b/arch/powerpc/kernel/exceptions-64e.S @@ -1010,6 +1010,7 @@ 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) @@ -1018,8 +1019,7 @@ 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 a7d36b1..53af150 100644 --- a/arch/powerpc/kernel/exceptions-64s.S +++ b/arch/powerpc/kernel/exceptions-64s.S @@ -1637,10 +1637,10 @@ 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+ 12f - 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 248ee7e..1eb60dd 100644 --- a/arch/powerpc/kernel/irq.c +++ b/arch/powerpc/kernel/irq.c @@ -447,6 +447,8 @@ void migrate_irqs(void) } #endif +extern void gr_handle_kernel_exploit(void); + static inline void check_stack_overflow(void) { #ifdef CONFIG_DEBUG_STACKOVERFLOW @@ -459,6 +461,7 @@ static inline void check_stack_overflow(void) printk("do_IRQ: stack overflow: %ld\n", sp - sizeof(struct thread_info)); dump_stack(); + gr_handle_kernel_exploit(); } #endif } diff --git a/arch/powerpc/kernel/module_32.c b/arch/powerpc/kernel/module_32.c index 6cff040..74ac5d1b 100644 --- a/arch/powerpc/kernel/module_32.c +++ b/arch/powerpc/kernel/module_32.c @@ -161,7 +161,7 @@ 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; } @@ -191,11 +191,16 @@ 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]) { @@ -299,7 +304,7 @@ int apply_relocate_add(Elf32_Shdr *sechdrs, } #ifdef CONFIG_DYNAMIC_FTRACE module->arch.tramp = - do_plt_call(module->module_core, + do_plt_call(module->module_core_rx, (unsigned long)ftrace_caller, sechdrs, module); #endif diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index be99774..9879c82 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -1039,8 +1039,8 @@ 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)) @@ -1554,10 +1554,10 @@ 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--; } @@ -1577,7 +1577,7 @@ 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; } @@ -1613,58 +1613,3 @@ void notrace __ppc64_runlatch_off(void) mtspr(SPRN_CTRLT, ctrl); } #endif /* CONFIG_PPC64 */ - -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/ptrace.c b/arch/powerpc/kernel/ptrace.c index 2e3d2bf..35df241 100644 --- a/arch/powerpc/kernel/ptrace.c +++ b/arch/powerpc/kernel/ptrace.c @@ -1762,6 +1762,10 @@ long arch_ptrace(struct task_struct *child, long request, return ret; } +#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. @@ -1774,6 +1778,11 @@ long do_syscall_trace_enter(struct pt_regs *regs) secure_computing_strict(regs->gpr[0]); +#ifdef CONFIG_GRKERNSEC_SETXID + if (unlikely(test_and_clear_thread_flag(TIF_GRSEC_SETXID))) + gr_delayed_cred_worker(); +#endif + if (test_thread_flag(TIF_SYSCALL_TRACE) && tracehook_report_syscall_entry(regs)) /* @@ -1808,6 +1817,11 @@ void do_syscall_trace_leave(struct pt_regs *regs) { int 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/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c index 1bc5a17..910d3f3 100644 --- a/arch/powerpc/kernel/signal_32.c +++ b/arch/powerpc/kernel/signal_32.c @@ -1012,7 +1012,7 @@ 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) { sigret = 0; tramp = current->mm->context.vdso_base + vdso32_rt_sigtramp; } else { diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c index 97c1e4b..f427f81 100644 --- a/arch/powerpc/kernel/signal_64.c +++ b/arch/powerpc/kernel/signal_64.c @@ -755,7 +755,7 @@ int handle_rt_signal64(int signr, struct k_sigaction *ka, siginfo_t *info, current->thread.fp_state.fpscr = 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 239f1cd..5359f76 100644 --- a/arch/powerpc/kernel/traps.c +++ b/arch/powerpc/kernel/traps.c @@ -142,6 +142,8 @@ 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) { @@ -191,6 +193,9 @@ 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 ce74c33..0803371 100644 --- a/arch/powerpc/kernel/vdso.c +++ b/arch/powerpc/kernel/vdso.c @@ -35,6 +35,7 @@ #include #include #include +#include #undef DEBUG @@ -221,7 +222,7 @@ 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 @@ -241,7 +242,7 @@ 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/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c index 61c738a..b1092d6 100644 --- a/arch/powerpc/kvm/powerpc.c +++ b/arch/powerpc/kvm/powerpc.c @@ -1195,7 +1195,7 @@ void kvmppc_init_lpid(unsigned long nr_lpids_param) } EXPORT_SYMBOL_GPL(kvmppc_init_lpid); -int kvm_arch_init(void *opaque) +int kvm_arch_init(const void *opaque) { return 0; } 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 51ab9e7..7d3c78b 100644 --- a/arch/powerpc/mm/fault.c +++ b/arch/powerpc/mm/fault.c @@ -33,6 +33,10 @@ #include #include #include +#include +#include +#include +#include #include #include @@ -69,6 +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. @@ -216,7 +247,7 @@ 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 @@ -378,7 +409,7 @@ 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 */ @@ -393,7 +424,7 @@ 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 */ @@ -483,6 +514,23 @@ 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); goto bail; } diff --git a/arch/powerpc/mm/mmap.c b/arch/powerpc/mm/mmap.c index cb8bdbe..cde4bc7 100644 --- a/arch/powerpc/mm/mmap.c +++ b/arch/powerpc/mm/mmap.c @@ -53,10 +53,14 @@ static inline int mmap_is_legacy(void) return sysctl_legacy_va_layout; } -static unsigned long mmap_rnd(void) +static unsigned long mmap_rnd(struct mm_struct *mm) { unsigned long rnd = 0; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (current->flags & PF_RANDOMIZE) { /* 8MB for 32bit, 1GB for 64bit */ if (is_32bit_task()) @@ -67,7 +71,7 @@ static unsigned long mmap_rnd(void) return rnd << PAGE_SHIFT; } -static inline unsigned long mmap_base(void) +static inline unsigned long mmap_base(struct mm_struct *mm) { unsigned long gap = rlimit(RLIMIT_STACK); @@ -76,7 +80,7 @@ static inline unsigned long mmap_base(void) else if (gap > MAX_GAP) gap = MAX_GAP; - return PAGE_ALIGN(TASK_SIZE - gap - mmap_rnd()); + return PAGE_ALIGN(TASK_SIZE - gap - mmap_rnd(mm)); } /* @@ -91,9 +95,21 @@ 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; } 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; } } diff --git a/arch/powerpc/mm/slice.c b/arch/powerpc/mm/slice.c index b0c75cc..ef7fb93 100644 --- a/arch/powerpc/mm/slice.c +++ b/arch/powerpc/mm/slice.c @@ -103,7 +103,7 @@ 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, 0); } static int slice_low_has_vma(struct mm_struct *mm, unsigned long slice) @@ -277,6 +277,12 @@ static unsigned long slice_find_area_bottomup(struct mm_struct *mm, info.align_offset = 0; addr = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + addr += mm->delta_mmap; +#endif + while (addr < TASK_SIZE) { info.low_limit = addr; if (!slice_scan_available(addr, available, 1, &addr)) @@ -410,6 +416,11 @@ unsigned long slice_get_unmapped_area(unsigned long addr, unsigned long len, if (fixed && addr > (mm->task_size - len)) return -ENOMEM; +#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/powerpc/platforms/cell/celleb_scc_pciex.c b/arch/powerpc/platforms/cell/celleb_scc_pciex.c index 4278acf..67fd0e6 100644 --- a/arch/powerpc/platforms/cell/celleb_scc_pciex.c +++ b/arch/powerpc/platforms/cell/celleb_scc_pciex.c @@ -400,8 +400,8 @@ static int scc_pciex_write_config(struct pci_bus *bus, unsigned int devfn, } static struct pci_ops scc_pciex_pci_ops = { - scc_pciex_read_config, - scc_pciex_write_config, + .read = scc_pciex_read_config, + .write = scc_pciex_write_config, }; static void pciex_clear_intr_all(unsigned int __iomem *base) diff --git a/arch/powerpc/platforms/cell/spufs/file.c b/arch/powerpc/platforms/cell/spufs/file.c index 9098692..3d54cd1 100644 --- a/arch/powerpc/platforms/cell/spufs/file.c +++ b/arch/powerpc/platforms/cell/spufs/file.c @@ -280,9 +280,9 @@ spufs_mem_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf) return VM_FAULT_NOPAGE; } -static int spufs_mem_mmap_access(struct vm_area_struct *vma, +static ssize_t spufs_mem_mmap_access(struct vm_area_struct *vma, unsigned long address, - void *buf, int len, int write) + void *buf, size_t len, int write) { struct spu_context *ctx = vma->vm_file->private_data; unsigned long offset = address - vma->vm_start; diff --git a/arch/s390/include/asm/atomic.h b/arch/s390/include/asm/atomic.h index fa934fe..c296056 100644 --- a/arch/s390/include/asm/atomic.h +++ b/arch/s390/include/asm/atomic.h @@ -412,4 +412,14 @@ 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)) + #endif /* __ARCH_S390_ATOMIC__ */ diff --git a/arch/s390/include/asm/barrier.h b/arch/s390/include/asm/barrier.h index 19ff956..8d39cb1 100644 --- a/arch/s390/include/asm/barrier.h +++ b/arch/s390/include/asm/barrier.h @@ -37,7 +37,7 @@ do { \ compiletime_assert_atomic_type(*p); \ barrier(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ diff --git a/arch/s390/include/asm/cache.h b/arch/s390/include/asm/cache.h index 4d7ccac..d03d0ad 100644 --- a/arch/s390/include/asm/cache.h +++ b/arch/s390/include/asm/cache.h @@ -9,8 +9,10 @@ #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 78f4f87..598ce39 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. */ @@ -222,9 +228,6 @@ 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 - void *fill_cpu_elf_notes(void *ptr, struct save_area *sa); #endif diff --git a/arch/s390/include/asm/exec.h b/arch/s390/include/asm/exec.h index c4a93d6..4d2a9b4 100644 --- a/arch/s390/include/asm/exec.h +++ b/arch/s390/include/asm/exec.h @@ -7,6 +7,6 @@ #ifndef __ASM_EXEC_H #define __ASM_EXEC_H -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) #endif /* __ASM_EXEC_H */ diff --git a/arch/s390/include/asm/uaccess.h b/arch/s390/include/asm/uaccess.h index cd4c68e..6764641 100644 --- a/arch/s390/include/asm/uaccess.h +++ b/arch/s390/include/asm/uaccess.h @@ -59,6 +59,7 @@ static inline int __range_ok(unsigned long addr, unsigned long size) __range_ok((unsigned long)(addr), (size)); \ }) +#define access_ok_noprefault(type, addr, size) access_ok((type), (addr), (size)) #define access_ok(type, addr, size) __access_ok(addr, size) /* @@ -275,6 +276,10 @@ 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; + return __copy_to_user(to, from, n); } @@ -303,10 +308,14 @@ __compiletime_warning("copy_from_user() buffer size is not provably correct") static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, unsigned long n) { - unsigned int sz = __compiletime_object_size(to); + size_t sz = __compiletime_object_size(to); might_fault(); - if (unlikely(sz != -1 && sz < n)) { + + if ((long)n < 0) + return n; + + if (unlikely(sz != (size_t)-1 && sz < n)) { copy_from_user_overflow(); return n; } diff --git a/arch/s390/kernel/module.c b/arch/s390/kernel/module.c index b89b591..fd9609d 100644 --- a/arch/s390/kernel/module.c +++ b/arch/s390/kernel/module.c @@ -169,11 +169,11 @@ int 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; } @@ -289,7 +289,7 @@ static int 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; @@ -312,7 +312,7 @@ static int apply_rela(Elf_Rela *rela, Elf_Addr base, Elf_Sym *symtab, rc = apply_rela_bits(loc, val, 0, 64, 0); else if (r_type == R_390_GOTENT || r_type == R_390_GOTPLTENT) { - val += (Elf_Addr) me->module_core - loc; + val += (Elf_Addr) me->module_core_rw - loc; rc = apply_rela_bits(loc, val, 1, 32, 1); } break; @@ -325,7 +325,7 @@ static int 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 */ @@ -350,7 +350,7 @@ static int 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; @@ -372,7 +372,7 @@ static int 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) rc = apply_rela_bits(loc, val, 0, 16, 0); else if (r_type == R_390_GOTOFF32) @@ -382,7 +382,7 @@ static int 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) rc = apply_rela_bits(loc, val, 1, 32, 0); diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index 93b9ca4..4ea1454 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -242,37 +242,3 @@ 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; - - ret = PAGE_ALIGN(mm->brk + brk_rnd()); - return (ret > mm->brk) ? ret : mm->brk; -} - -unsigned long randomize_et_dyn(unsigned long base) -{ - unsigned long ret; - - if (!(current->flags & PF_RANDOMIZE)) - return base; - ret = PAGE_ALIGN(base + brk_rnd()); - return (ret > base) ? ret : base; -} diff --git a/arch/s390/mm/mmap.c b/arch/s390/mm/mmap.c index 9b436c2..54fbf0a 100644 --- a/arch/s390/mm/mmap.c +++ b/arch/s390/mm/mmap.c @@ -95,9 +95,21 @@ void arch_pick_mmap_layout(struct mm_struct *mm) */ if (mmap_is_legacy()) { mm->mmap_base = mmap_base_legacy(); + +#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; } 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; } } @@ -170,9 +182,21 @@ void arch_pick_mmap_layout(struct mm_struct *mm) */ if (mmap_is_legacy()) { mm->mmap_base = mmap_base_legacy(); + +#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; } 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; } } 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/exec.h b/arch/score/include/asm/exec.h index f9f3cd5..58ff438 100644 --- a/arch/score/include/asm/exec.h +++ b/arch/score/include/asm/exec.h @@ -1,6 +1,6 @@ #ifndef _ASM_SCORE_EXEC_H #define _ASM_SCORE_EXEC_H -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) (x) #endif /* _ASM_SCORE_EXEC_H */ diff --git a/arch/score/kernel/process.c b/arch/score/kernel/process.c index a1519ad3..e8ac1ff 100644 --- a/arch/score/kernel/process.c +++ b/arch/score/kernel/process.c @@ -116,8 +116,3 @@ 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 6777177..cb5e44f 100644 --- a/arch/sh/mm/mmap.c +++ b/arch/sh/mm/mmap.c @@ -36,6 +36,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, struct mm_struct *mm = current->mm; struct vm_area_struct *vma; int do_colour_align; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); struct vm_unmapped_area_info info; if (flags & MAP_FIXED) { @@ -55,6 +56,10 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, if (filp || (flags & MAP_SHARED)) do_colour_align = 1; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_colour_align) addr = COLOUR_ALIGN(addr, pgoff); @@ -62,14 +67,13 @@ 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, offset)) return addr; } info.flags = 0; info.length = len; - info.low_limit = TASK_UNMAPPED_BASE; + info.low_limit = mm->mmap_base; info.high_limit = TASK_SIZE; info.align_mask = do_colour_align ? (PAGE_MASK & shm_align_mask) : 0; info.align_offset = pgoff << PAGE_SHIFT; @@ -85,6 +89,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, struct mm_struct *mm = current->mm; unsigned long addr = addr0; int do_colour_align; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); struct vm_unmapped_area_info info; if (flags & MAP_FIXED) { @@ -104,6 +109,10 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (filp || (flags & MAP_SHARED)) do_colour_align = 1; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + /* requesting a specific address */ if (addr) { if (do_colour_align) @@ -112,8 +121,7 @@ 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, offset)) return addr; } @@ -135,6 +143,12 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, VM_BUG_ON(addr != -ENOMEM); info.flags = 0; info.low_limit = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + info.low_limit += mm->delta_mmap; +#endif + info.high_limit = TASK_SIZE; addr = vm_unmapped_area(&info); } diff --git a/arch/sparc/include/asm/atomic_64.h b/arch/sparc/include/asm/atomic_64.h index bb894c8..8141d5c 100644 --- a/arch/sparc/include/asm/atomic_64.h +++ b/arch/sparc/include/asm/atomic_64.h @@ -15,18 +15,40 @@ #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; +} void atomic_add(int, atomic_t *); +void atomic_add_unchecked(int, atomic_unchecked_t *); void atomic64_add(long, atomic64_t *); +void atomic64_add_unchecked(long, atomic64_unchecked_t *); void atomic_sub(int, atomic_t *); +void atomic_sub_unchecked(int, atomic_unchecked_t *); void atomic64_sub(long, atomic64_t *); +void atomic64_sub_unchecked(long, atomic64_unchecked_t *); int atomic_add_ret(int, atomic_t *); +int atomic_add_ret_unchecked(int, atomic_unchecked_t *); long atomic64_add_ret(long, atomic64_t *); +long atomic64_add_ret_unchecked(long, atomic64_unchecked_t *); int atomic_sub_ret(int, atomic_t *); long atomic64_sub_ret(long, atomic64_t *); @@ -34,13 +56,29 @@ 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 @@ -51,6 +89,10 @@ 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) @@ -60,25 +102,60 @@ 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/barrier_64.h b/arch/sparc/include/asm/barrier_64.h index 305dcc3..7835030 100644 --- a/arch/sparc/include/asm/barrier_64.h +++ b/arch/sparc/include/asm/barrier_64.h @@ -57,7 +57,7 @@ do { __asm__ __volatile__("ba,pt %%xcc, 1f\n\t" \ do { \ compiletime_assert_atomic_type(*p); \ barrier(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ diff --git a/arch/sparc/include/asm/cache.h b/arch/sparc/include/asm/cache.h index 5bb6991..5c2132e 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 a24e41f..47677ff 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 370ca1e..d4f4a98 100644 --- a/arch/sparc/include/asm/elf_64.h +++ b/arch/sparc/include/asm/elf_64.h @@ -189,6 +189,13 @@ 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 a3890da..f6a408e 100644 --- a/arch/sparc/include/asm/pgalloc_32.h +++ b/arch/sparc/include/asm/pgalloc_32.h @@ -35,6 +35,7 @@ static inline void pgd_set(pgd_t * pgdp, pmd_t * pmdp) } #define pgd_populate(MM, PGD, PMD) pgd_set(PGD, PMD) +#define pgd_populate_kernel(MM, PGD, PMD) pgd_populate((MM), (PGD), (PMD)) static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long address) diff --git a/arch/sparc/include/asm/pgalloc_64.h b/arch/sparc/include/asm/pgalloc_64.h index 39a7ac4..2c9b586 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.h b/arch/sparc/include/asm/pgtable.h index 59ba6f6..4518128 100644 --- a/arch/sparc/include/asm/pgtable.h +++ b/arch/sparc/include/asm/pgtable.h @@ -5,4 +5,8 @@ #else #include #endif + +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) + #endif diff --git a/arch/sparc/include/asm/pgtable_32.h b/arch/sparc/include/asm/pgtable_32.h index b9b91ae..950b91e 100644 --- a/arch/sparc/include/asm/pgtable_32.h +++ b/arch/sparc/include/asm/pgtable_32.h @@ -51,6 +51,9 @@ unsigned long __init bootmem_init(unsigned long *pages_avail); #define PAGE_SHARED SRMMU_PAGE_SHARED #define PAGE_COPY SRMMU_PAGE_COPY #define PAGE_READONLY SRMMU_PAGE_RDONLY +#define PAGE_SHARED_NOEXEC SRMMU_PAGE_SHARED_NOEXEC +#define PAGE_COPY_NOEXEC SRMMU_PAGE_COPY_NOEXEC +#define PAGE_READONLY_NOEXEC SRMMU_PAGE_RDONLY_NOEXEC #define PAGE_KERNEL SRMMU_PAGE_KERNEL /* Top-level page directory - dummy used by init-mm. @@ -63,18 +66,18 @@ extern unsigned long ptr_in_current_pgd; /* xwr */ #define __P000 PAGE_NONE -#define __P001 PAGE_READONLY -#define __P010 PAGE_COPY -#define __P011 PAGE_COPY +#define __P001 PAGE_READONLY_NOEXEC +#define __P010 PAGE_COPY_NOEXEC +#define __P011 PAGE_COPY_NOEXEC #define __P100 PAGE_READONLY #define __P101 PAGE_READONLY #define __P110 PAGE_COPY #define __P111 PAGE_COPY #define __S000 PAGE_NONE -#define __S001 PAGE_READONLY -#define __S010 PAGE_SHARED -#define __S011 PAGE_SHARED +#define __S001 PAGE_READONLY_NOEXEC +#define __S010 PAGE_SHARED_NOEXEC +#define __S011 PAGE_SHARED_NOEXEC #define __S100 PAGE_READONLY #define __S101 PAGE_READONLY #define __S110 PAGE_SHARED diff --git a/arch/sparc/include/asm/pgtsrmmu.h b/arch/sparc/include/asm/pgtsrmmu.h index 79da178..c2eede8 100644 --- a/arch/sparc/include/asm/pgtsrmmu.h +++ b/arch/sparc/include/asm/pgtsrmmu.h @@ -115,6 +115,11 @@ SRMMU_EXEC | SRMMU_REF) #define SRMMU_PAGE_RDONLY __pgprot(SRMMU_VALID | SRMMU_CACHE | \ SRMMU_EXEC | SRMMU_REF) + +#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) + #define SRMMU_PAGE_KERNEL __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_PRIV | \ SRMMU_DIRTY | SRMMU_REF) diff --git a/arch/sparc/include/asm/setup.h b/arch/sparc/include/asm/setup.h index f5fffd8..a0669f0 100644 --- a/arch/sparc/include/asm/setup.h +++ b/arch/sparc/include/asm/setup.h @@ -53,8 +53,8 @@ int handle_ldf_stq(u32 insn, struct pt_regs *regs); void handle_ld_nf(u32 insn, struct pt_regs *regs); /* init_64.c */ -extern atomic_t dcpage_flushes; -extern atomic_t dcpage_flushes_xcall; +extern atomic_unchecked_t dcpage_flushes; +extern atomic_unchecked_t dcpage_flushes_xcall; extern int sysctl_tsb_ratio; #endif 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 96efa7a..16858bf 100644 --- a/arch/sparc/include/asm/thread_info_32.h +++ b/arch/sparc/include/asm/thread_info_32.h @@ -49,6 +49,8 @@ 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 a5f01ac..703b554 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))); }; @@ -188,12 +190,13 @@ register struct thread_info *current_thread_info_reg asm("g6"); #define TIF_NEED_RESCHED 3 /* rescheduling necessary */ /* flag bit 4 is available */ #define TIF_UNALIGNED 5 /* allowed to do unaligned accesses */ -/* flag bit 6 is available */ +#define TIF_GRSEC_SETXID 6 /* update credentials on syscall entry/exit */ #define TIF_32BIT 7 /* 32-bit binary */ #define TIF_NOHZ 8 /* in adaptive nohz mode */ #define TIF_SECCOMP 9 /* secure computing */ #define TIF_SYSCALL_AUDIT 10 /* syscall auditing active */ #define TIF_SYSCALL_TRACEPOINT 11 /* syscall tracepoint instrumentation */ + /* NOTE: Thread flags >= 12 should be ones we have no interest * in using in assembly, else we can't use the mask as * an immediate value in instructions such as andcc. @@ -213,12 +216,18 @@ register struct thread_info *current_thread_info_reg asm("g6"); #define _TIF_SYSCALL_AUDIT (1< #else diff --git a/arch/sparc/include/asm/uaccess_32.h b/arch/sparc/include/asm/uaccess_32.h index 9634d08..f55fe4f 100644 --- a/arch/sparc/include/asm/uaccess_32.h +++ b/arch/sparc/include/asm/uaccess_32.h @@ -250,27 +250,46 @@ unsigned long __copy_user(void __user *to, const void __user *from, unsigned lon 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 c990a5e..f17b9c1 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 @@ -214,8 +215,15 @@ 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); @@ -231,8 +239,15 @@ 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 7cf9c6e..6206648 100644 --- a/arch/sparc/kernel/Makefile +++ b/arch/sparc/kernel/Makefile @@ -4,7 +4,7 @@ # asflags-y := -ansi -ccflags-y := -Werror +#ccflags-y := -Werror extra-y := head_$(BITS).o diff --git a/arch/sparc/kernel/process_32.c b/arch/sparc/kernel/process_32.c index 50e7b62..79fae35 100644 --- a/arch/sparc/kernel/process_32.c +++ b/arch/sparc/kernel/process_32.c @@ -123,14 +123,14 @@ 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], @@ -167,7 +167,7 @@ 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 027e099..6d4178f 100644 --- a/arch/sparc/kernel/process_64.c +++ b/arch/sparc/kernel/process_64.c @@ -161,7 +161,7 @@ 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) @@ -170,7 +170,7 @@ 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]); @@ -183,7 +183,7 @@ 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]); } @@ -278,7 +278,7 @@ void arch_trigger_all_cpu_backtrace(bool include_self) ((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/prom_common.c b/arch/sparc/kernel/prom_common.c index 79cc0d1..ec62734 100644 --- a/arch/sparc/kernel/prom_common.c +++ b/arch/sparc/kernel/prom_common.c @@ -144,7 +144,7 @@ static int __init prom_common_nextprop(phandle node, char *prev, char *buf) unsigned int prom_early_allocated __initdata; -static struct of_pdt_ops prom_sparc_ops __initdata = { +static struct of_pdt_ops prom_sparc_ops __initconst = { .nextprop = prom_common_nextprop, .getproplen = prom_getproplen, .getproperty = prom_getproperty, diff --git a/arch/sparc/kernel/ptrace_64.c b/arch/sparc/kernel/ptrace_64.c index c13c9f2..d572c34 100644 --- a/arch/sparc/kernel/ptrace_64.c +++ b/arch/sparc/kernel/ptrace_64.c @@ -1060,6 +1060,10 @@ long arch_ptrace(struct task_struct *child, long request, return ret; } +#ifdef CONFIG_GRKERNSEC_SETXID +extern void gr_delayed_cred_worker(void); +#endif + asmlinkage int syscall_trace_enter(struct pt_regs *regs) { int ret = 0; @@ -1070,6 +1074,11 @@ asmlinkage int syscall_trace_enter(struct pt_regs *regs) if (test_thread_flag(TIF_NOHZ)) user_exit(); +#ifdef CONFIG_GRKERNSEC_SETXID + if (unlikely(test_and_clear_thread_flag(TIF_GRSEC_SETXID))) + gr_delayed_cred_worker(); +#endif + if (test_thread_flag(TIF_SYSCALL_TRACE)) ret = tracehook_report_syscall_entry(regs); @@ -1093,6 +1102,11 @@ asmlinkage void syscall_trace_leave(struct pt_regs *regs) if (test_thread_flag(TIF_NOHZ)) user_exit(); +#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/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index 41aa247..eadfb74 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -883,7 +883,7 @@ void smp_flush_dcache_page_impl(struct page *page, int cpu) return; #ifdef CONFIG_DEBUG_DCFLUSH - atomic_inc(&dcpage_flushes); + atomic_inc_unchecked(&dcpage_flushes); #endif this_cpu = get_cpu(); @@ -907,7 +907,7 @@ void smp_flush_dcache_page_impl(struct page *page, int cpu) xcall_deliver(data0, __pa(pg_addr), (u64) pg_addr, cpumask_of(cpu)); #ifdef CONFIG_DEBUG_DCFLUSH - atomic_inc(&dcpage_flushes_xcall); + atomic_inc_unchecked(&dcpage_flushes_xcall); #endif } } @@ -926,7 +926,7 @@ void flush_dcache_page_all(struct mm_struct *mm, struct page *page) preempt_disable(); #ifdef CONFIG_DEBUG_DCFLUSH - atomic_inc(&dcpage_flushes); + atomic_inc_unchecked(&dcpage_flushes); #endif data0 = 0; pg_addr = page_address(page); @@ -943,7 +943,7 @@ void flush_dcache_page_all(struct mm_struct *mm, struct page *page) xcall_deliver(data0, __pa(pg_addr), (u64) pg_addr, cpu_online_mask); #ifdef CONFIG_DEBUG_DCFLUSH - atomic_inc(&dcpage_flushes_xcall); + atomic_inc_unchecked(&dcpage_flushes_xcall); #endif } __local_flush_dcache_page(page); diff --git a/arch/sparc/kernel/sys_sparc_32.c b/arch/sparc/kernel/sys_sparc_32.c index 646988d..b88905f 100644 --- a/arch/sparc/kernel/sys_sparc_32.c +++ b/arch/sparc/kernel/sys_sparc_32.c @@ -54,7 +54,7 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi if (len > TASK_SIZE - PAGE_SIZE) return -ENOMEM; if (!addr) - addr = TASK_UNMAPPED_BASE; + addr = current->mm->mmap_base; info.flags = 0; info.length = len; diff --git a/arch/sparc/kernel/sys_sparc_64.c b/arch/sparc/kernel/sys_sparc_64.c index c85403d..6af95c9 100644 --- a/arch/sparc/kernel/sys_sparc_64.c +++ b/arch/sparc/kernel/sys_sparc_64.c @@ -89,13 +89,14 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, unsi struct vm_area_struct * vma; unsigned long task_size = TASK_SIZE; int do_color_align; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); struct vm_unmapped_area_info info; if (flags & MAP_FIXED) { /* 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; @@ -110,6 +111,10 @@ 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 = COLOR_ALIGN(addr, pgoff); @@ -117,22 +122,28 @@ 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, offset)) return addr; } info.flags = 0; info.length = len; - info.low_limit = TASK_UNMAPPED_BASE; + info.low_limit = mm->mmap_base; info.high_limit = min(task_size, VA_EXCLUDE_START); info.align_mask = do_color_align ? (PAGE_MASK & (SHMLBA - 1)) : 0; info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); if ((addr & ~PAGE_MASK) && task_size > VA_EXCLUDE_END) { VM_BUG_ON(addr != -ENOMEM); info.low_limit = VA_EXCLUDE_END; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + info.low_limit += mm->delta_mmap; +#endif + info.high_limit = task_size; addr = vm_unmapped_area(&info); } @@ -150,6 +161,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, unsigned long task_size = STACK_TOP32; unsigned long addr = addr0; int do_color_align; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); struct vm_unmapped_area_info info; /* This should only ever run for 32-bit processes. */ @@ -159,7 +171,7 @@ 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; @@ -172,6 +184,10 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (filp || (flags & MAP_SHARED)) do_color_align = 1; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + /* requesting a specific address */ if (addr) { if (do_color_align) @@ -180,8 +196,7 @@ 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, offset)) return addr; } @@ -191,6 +206,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, info.high_limit = mm->mmap_base; info.align_mask = do_color_align ? (PAGE_MASK & (SHMLBA - 1)) : 0; info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); /* @@ -203,6 +219,12 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, VM_BUG_ON(addr != -ENOMEM); info.flags = 0; info.low_limit = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + info.low_limit += mm->delta_mmap; +#endif + info.high_limit = STACK_TOP32; addr = vm_unmapped_area(&info); } @@ -259,10 +281,14 @@ unsigned long get_fb_unmapped_area(struct file *filp, unsigned long orig_addr, u EXPORT_SYMBOL(get_fb_unmapped_area); /* Essentially the same as PowerPC. */ -static unsigned long mmap_rnd(void) +static unsigned long mmap_rnd(struct mm_struct *mm) { unsigned long rnd = 0UL; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (current->flags & PF_RANDOMIZE) { unsigned long val = get_random_int(); if (test_thread_flag(TIF_32BIT)) @@ -275,7 +301,7 @@ static unsigned long mmap_rnd(void) void arch_pick_mmap_layout(struct mm_struct *mm) { - unsigned long random_factor = mmap_rnd(); + unsigned long random_factor = mmap_rnd(mm); unsigned long gap; /* @@ -288,6 +314,12 @@ 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; } else { /* We know it's 32-bit */ @@ -299,6 +331,12 @@ 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; } } diff --git a/arch/sparc/kernel/syscalls.S b/arch/sparc/kernel/syscalls.S index 33a17e7..d87fb1f 100644 --- a/arch/sparc/kernel/syscalls.S +++ b/arch/sparc/kernel/syscalls.S @@ -52,7 +52,7 @@ sys32_rt_sigreturn: #endif .align 32 1: ldx [%g6 + TI_FLAGS], %l5 - andcc %l5, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0 + andcc %l5, _TIF_WORK_SYSCALL, %g0 be,pt %icc, rtrap nop call syscall_trace_leave @@ -184,7 +184,7 @@ linux_sparc_syscall32: srl %i3, 0, %o3 ! IEU0 srl %i2, 0, %o2 ! IEU0 Group - andcc %l0, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0 + andcc %l0, _TIF_WORK_SYSCALL, %g0 bne,pn %icc, linux_syscall_trace32 ! CTI mov %i0, %l5 ! IEU1 5: call %l7 ! CTI Group brk forced @@ -208,7 +208,7 @@ linux_sparc_syscall: mov %i3, %o3 ! IEU1 mov %i4, %o4 ! IEU0 Group - andcc %l0, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0 + andcc %l0, _TIF_WORK_SYSCALL, %g0 bne,pn %icc, linux_syscall_trace ! CTI Group mov %i0, %l5 ! IEU0 2: call %l7 ! CTI Group brk forced @@ -223,7 +223,7 @@ ret_sys_call: cmp %o0, -ERESTART_RESTARTBLOCK bgeu,pn %xcc, 1f - andcc %l0, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0 + andcc %l0, _TIF_WORK_SYSCALL, %g0 ldx [%sp + PTREGS_OFF + PT_V9_TNPC], %l1 ! pc = npc 2: diff --git a/arch/sparc/kernel/traps_32.c b/arch/sparc/kernel/traps_32.c index 6fd386c5..6907d81 100644 --- a/arch/sparc/kernel/traps_32.c +++ b/arch/sparc/kernel/traps_32.c @@ -44,6 +44,8 @@ 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 __noreturn die_if_kernel(char *str, struct pt_regs *regs) { static int die_counter; @@ -76,15 +78,17 @@ void __noreturn 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 fb6640e..2daada8 100644 --- a/arch/sparc/kernel/traps_64.c +++ b/arch/sparc/kernel/traps_64.c @@ -79,7 +79,7 @@ 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); } } @@ -99,6 +99,12 @@ 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); } @@ -117,11 +123,16 @@ 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); @@ -1151,7 +1162,7 @@ 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, @@ -1758,7 +1769,7 @@ 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."); } @@ -1766,7 +1777,7 @@ 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 { @@ -1839,8 +1850,8 @@ struct sun4v_error_entry { /*0x38*/u64 reserved_5; }; -static atomic_t sun4v_resum_oflow_cnt = ATOMIC_INIT(0); -static atomic_t sun4v_nonresum_oflow_cnt = ATOMIC_INIT(0); +static atomic_unchecked_t sun4v_resum_oflow_cnt = ATOMIC_INIT(0); +static atomic_unchecked_t sun4v_nonresum_oflow_cnt = ATOMIC_INIT(0); static const char *sun4v_err_type_to_str(u8 type) { @@ -1932,7 +1943,7 @@ static void sun4v_report_real_raddr(const char *pfx, struct pt_regs *regs) } static void sun4v_log_error(struct pt_regs *regs, struct sun4v_error_entry *ent, - int cpu, const char *pfx, atomic_t *ocnt) + int cpu, const char *pfx, atomic_unchecked_t *ocnt) { u64 *raw_ptr = (u64 *) ent; u32 attrs; @@ -1990,8 +2001,8 @@ static void sun4v_log_error(struct pt_regs *regs, struct sun4v_error_entry *ent, show_regs(regs); - if ((cnt = atomic_read(ocnt)) != 0) { - atomic_set(ocnt, 0); + if ((cnt = atomic_read_unchecked(ocnt)) != 0) { + atomic_set_unchecked(ocnt, 0); wmb(); printk("%s: Queue overflowed %d times.\n", pfx, cnt); @@ -2048,7 +2059,7 @@ out: */ void sun4v_resum_overflow(struct pt_regs *regs) { - atomic_inc(&sun4v_resum_oflow_cnt); + atomic_inc_unchecked(&sun4v_resum_oflow_cnt); } /* We run with %pil set to PIL_NORMAL_MAX and PSTATE_IE enabled in %pstate. @@ -2101,7 +2112,7 @@ void sun4v_nonresum_overflow(struct pt_regs *regs) /* XXX Actually even this can make not that much sense. Perhaps * XXX we should just pull the plug and panic directly from here? */ - atomic_inc(&sun4v_nonresum_oflow_cnt); + atomic_inc_unchecked(&sun4v_nonresum_oflow_cnt); } unsigned long sun4v_err_itlb_vaddr; @@ -2116,9 +2127,9 @@ 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", @@ -2140,9 +2151,9 @@ 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", @@ -2359,13 +2370,13 @@ 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++; } } @@ -2383,6 +2394,8 @@ 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 __noreturn die_if_kernel(char *str, struct pt_regs *regs) { static int die_counter; @@ -2411,7 +2424,7 @@ void __noreturn 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); @@ -2424,8 +2437,10 @@ void __noreturn 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 62098a8..547ab2c 100644 --- a/arch/sparc/kernel/unaligned_64.c +++ b/arch/sparc/kernel/unaligned_64.c @@ -297,7 +297,7 @@ 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 3269b02..64f5231 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-$(CONFIG_SPARC32) += ashrdi3.o lib-$(CONFIG_SPARC32) += memcpy.o memset.o diff --git a/arch/sparc/lib/atomic_64.S b/arch/sparc/lib/atomic_64.S index 85c233d..68500e0 100644 --- a/arch/sparc/lib/atomic_64.S +++ b/arch/sparc/lib/atomic_64.S @@ -17,7 +17,12 @@ ENTRY(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) @@ -27,10 +32,28 @@ ENTRY(atomic_add) /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) ENDPROC(atomic_add) +ENTRY(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) +ENDPROC(atomic_add_unchecked) + ENTRY(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) @@ -40,10 +63,28 @@ ENTRY(atomic_sub) /* %o0 = decrement, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) ENDPROC(atomic_sub) +ENTRY(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) +ENDPROC(atomic_sub_unchecked) + ENTRY(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) @@ -53,10 +94,29 @@ ENTRY(atomic_add_ret) /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) ENDPROC(atomic_add_ret) +ENTRY(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) +ENDPROC(atomic_add_ret_unchecked) + ENTRY(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) @@ -69,7 +129,12 @@ ENDPROC(atomic_sub_ret) ENTRY(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) @@ -79,10 +144,28 @@ ENTRY(atomic64_add) /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) ENDPROC(atomic64_add) +ENTRY(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) +ENDPROC(atomic64_add_unchecked) + ENTRY(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) @@ -92,10 +175,28 @@ ENTRY(atomic64_sub) /* %o0 = decrement, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) ENDPROC(atomic64_sub) +ENTRY(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) +ENDPROC(atomic64_sub_unchecked) + ENTRY(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) @@ -105,10 +206,29 @@ ENTRY(atomic64_add_ret) /* %o0 = increment, %o1 = atomic_ptr */ 2: BACKOFF_SPIN(%o2, %o3, 1b) ENDPROC(atomic64_add_ret) +ENTRY(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) +ENDPROC(atomic64_add_ret_unchecked) + ENTRY(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 323335b..ed85ea2 100644 --- a/arch/sparc/lib/ksyms.c +++ b/arch/sparc/lib/ksyms.c @@ -100,12 +100,18 @@ EXPORT_SYMBOL(__clear_user); /* 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); EXPORT_SYMBOL(atomic64_dec_if_positive); diff --git a/arch/sparc/mm/Makefile b/arch/sparc/mm/Makefile index 30c3ecc..736f015 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 908e8c1..1524793 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 @@ -156,6 +159,277 @@ 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); + + do { /* PaX: patched PLT emulation #2 */ + unsigned int ba; + + err = get_user(ba, (unsigned int *)regs->pc); + + if (err) + break; + + if ((ba & 0xFFC00000U) == 0x30800000U || (ba & 0xFFF80000U) == 0x30480000U) { + unsigned int addr; + + if ((ba & 0xFFC00000U) == 0x30800000U) + addr = regs->pc + ((((ba | 0xFFC00000U) ^ 0x00200000U) + 0x00200000U) << 2); + else + addr = regs->pc + ((((ba | 0xFFF80000U) ^ 0x00040000U) + 0x00040000U) << 2); + regs->pc = addr; + regs->npc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #3 */ + unsigned int sethi, bajmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->pc); + err |= get_user(bajmpl, (unsigned int *)(regs->pc+4)); + err |= get_user(nop, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + ((bajmpl & 0xFFFFE000U) == 0x81C06000U || (bajmpl & 0xFFF80000U) == 0x30480000U) && + nop == 0x01000000U) + { + unsigned int addr; + + addr = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] = addr; + if ((bajmpl & 0xFFFFE000U) == 0x81C06000U) + addr += (((bajmpl | 0xFFFFE000U) ^ 0x00001000U) + 0x00001000U); + else + addr = regs->pc + ((((bajmpl | 0xFFF80000U) ^ 0x00040000U) + 0x00040000U) << 2); + 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) { @@ -226,6 +500,24 @@ 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 587cd05..fbdf17a 100644 --- a/arch/sparc/mm/fault_64.c +++ b/arch/sparc/mm/fault_64.c @@ -22,6 +22,9 @@ #include #include #include +#include +#include +#include #include #include @@ -76,7 +79,7 @@ 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); @@ -279,6 +282,466 @@ static void noinline __kprobes bogus_32bit_fault_tpc(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); + + do { /* PaX: patched PLT emulation #2 */ + unsigned int ba; + + err = get_user(ba, (unsigned int *)regs->tpc); + + if (err) + break; + + if ((ba & 0xFFC00000U) == 0x30800000U || (ba & 0xFFF80000U) == 0x30480000U) { + unsigned long addr; + + if ((ba & 0xFFC00000U) == 0x30800000U) + addr = regs->tpc + ((((ba | 0xFFFFFFFFFFC00000UL) ^ 0x00200000UL) + 0x00200000UL) << 2); + else + 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); + + do { /* PaX: patched PLT emulation #3 */ + unsigned int sethi, bajmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(bajmpl, (unsigned int *)(regs->tpc+4)); + err |= get_user(nop, (unsigned int *)(regs->tpc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + ((bajmpl & 0xFFFFE000U) == 0x81C06000U || (bajmpl & 0xFFF80000U) == 0x30480000U) && + nop == 0x01000000U) + { + unsigned long addr; + + addr = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] = addr; + if ((bajmpl & 0xFFFFE000U) == 0x81C06000U) + addr += (((bajmpl | 0xFFFFFFFFFFFFE000UL) ^ 0x00001000UL) + 0x00001000UL); + else + addr = regs->tpc + ((((bajmpl | 0xFFFFFFFFFFF80000UL) ^ 0x00040000UL) + 0x00040000UL) << 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 #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) { enum ctx_state prev_state = exception_enter(); @@ -350,6 +813,29 @@ retry: 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 d329537..2c3746a 100644 --- a/arch/sparc/mm/hugetlbpage.c +++ b/arch/sparc/mm/hugetlbpage.c @@ -25,8 +25,10 @@ static unsigned long hugetlb_get_unmapped_area_bottomup(struct file *filp, unsigned long addr, unsigned long len, unsigned long pgoff, - unsigned long flags) + unsigned long flags, + unsigned long offset) { + struct mm_struct *mm = current->mm; unsigned long task_size = TASK_SIZE; struct vm_unmapped_area_info info; @@ -35,15 +37,22 @@ static unsigned long hugetlb_get_unmapped_area_bottomup(struct file *filp, info.flags = 0; info.length = len; - info.low_limit = TASK_UNMAPPED_BASE; + info.low_limit = mm->mmap_base; info.high_limit = min(task_size, VA_EXCLUDE_START); info.align_mask = PAGE_MASK & ~HPAGE_MASK; info.align_offset = 0; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); if ((addr & ~PAGE_MASK) && task_size > VA_EXCLUDE_END) { VM_BUG_ON(addr != -ENOMEM); info.low_limit = VA_EXCLUDE_END; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + info.low_limit += mm->delta_mmap; +#endif + info.high_limit = task_size; addr = vm_unmapped_area(&info); } @@ -55,7 +64,8 @@ static unsigned long hugetlb_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, const unsigned long len, const unsigned long pgoff, - const unsigned long flags) + const unsigned long flags, + const unsigned long offset) { struct mm_struct *mm = current->mm; unsigned long addr = addr0; @@ -70,6 +80,7 @@ hugetlb_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, info.high_limit = mm->mmap_base; info.align_mask = PAGE_MASK & ~HPAGE_MASK; info.align_offset = 0; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); /* @@ -82,6 +93,12 @@ hugetlb_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, VM_BUG_ON(addr != -ENOMEM); info.flags = 0; info.low_limit = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + info.low_limit += mm->delta_mmap; +#endif + info.high_limit = STACK_TOP32; addr = vm_unmapped_area(&info); } @@ -96,6 +113,7 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, struct mm_struct *mm = current->mm; struct vm_area_struct *vma; unsigned long task_size = TASK_SIZE; + unsigned long offset = gr_rand_threadstack_offset(mm, file, flags); if (test_thread_flag(TIF_32BIT)) task_size = STACK_TOP32; @@ -111,19 +129,22 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, return addr; } +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + 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, offset)) return addr; } if (mm->get_unmapped_area == arch_get_unmapped_area) return hugetlb_get_unmapped_area_bottomup(file, addr, len, - pgoff, flags); + pgoff, flags, offset); else return hugetlb_get_unmapped_area_topdown(file, addr, len, - pgoff, flags); + pgoff, flags, offset); } pte_t *huge_pte_alloc(struct mm_struct *mm, diff --git a/arch/sparc/mm/init_64.c b/arch/sparc/mm/init_64.c index 2cfb0f2..e917d9f 100644 --- a/arch/sparc/mm/init_64.c +++ b/arch/sparc/mm/init_64.c @@ -189,9 +189,9 @@ unsigned long sparc64_kern_sec_context __read_mostly; int num_kernel_image_mappings; #ifdef CONFIG_DEBUG_DCFLUSH -atomic_t dcpage_flushes = ATOMIC_INIT(0); +atomic_unchecked_t dcpage_flushes = ATOMIC_INIT(0); #ifdef CONFIG_SMP -atomic_t dcpage_flushes_xcall = ATOMIC_INIT(0); +atomic_unchecked_t dcpage_flushes_xcall = ATOMIC_INIT(0); #endif #endif @@ -199,7 +199,7 @@ inline void flush_dcache_page_impl(struct page *page) { BUG_ON(tlb_type == hypervisor); #ifdef CONFIG_DEBUG_DCFLUSH - atomic_inc(&dcpage_flushes); + atomic_inc_unchecked(&dcpage_flushes); #endif #ifdef DCACHE_ALIASING_POSSIBLE @@ -471,10 +471,10 @@ void mmu_info(struct seq_file *m) #ifdef CONFIG_DEBUG_DCFLUSH seq_printf(m, "DCPageFlushes\t: %d\n", - atomic_read(&dcpage_flushes)); + atomic_read_unchecked(&dcpage_flushes)); #ifdef CONFIG_SMP seq_printf(m, "DCPageFlushesXC\t: %d\n", - atomic_read(&dcpage_flushes_xcall)); + atomic_read_unchecked(&dcpage_flushes_xcall)); #endif /* CONFIG_SMP */ #endif /* CONFIG_DEBUG_DCFLUSH */ } diff --git a/arch/tile/Kconfig b/arch/tile/Kconfig index 4f3006b..453f625f 100644 --- a/arch/tile/Kconfig +++ b/arch/tile/Kconfig @@ -192,6 +192,7 @@ source "kernel/Kconfig.hz" config KEXEC bool "kexec system call" + depends on !GRKERNSEC_KMEM ---help--- kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot diff --git a/arch/tile/include/asm/atomic_64.h b/arch/tile/include/asm/atomic_64.h index 7b11c5f..755a026 100644 --- a/arch/tile/include/asm/atomic_64.h +++ b/arch/tile/include/asm/atomic_64.h @@ -105,6 +105,16 @@ 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)) + /* Define this to indicate that cmpxchg is an efficient operation. */ #define __HAVE_ARCH_CMPXCHG diff --git a/arch/tile/include/asm/cache.h b/arch/tile/include/asm/cache.h index 6160761..00cac88 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/tile/include/asm/uaccess.h b/arch/tile/include/asm/uaccess.h index b6cde32..c0cb736 100644 --- a/arch/tile/include/asm/uaccess.h +++ b/arch/tile/include/asm/uaccess.h @@ -414,9 +414,9 @@ static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, unsigned long n) { - int sz = __compiletime_object_size(to); + size_t sz = __compiletime_object_size(to); - if (likely(sz == -1 || sz >= n)) + if (likely(sz == (size_t)-1 || sz >= n)) n = _copy_from_user(to, from, n); else copy_from_user_overflow(); diff --git a/arch/tile/mm/hugetlbpage.c b/arch/tile/mm/hugetlbpage.c index e514899..f8743c4 100644 --- a/arch/tile/mm/hugetlbpage.c +++ b/arch/tile/mm/hugetlbpage.c @@ -207,6 +207,7 @@ static unsigned long hugetlb_get_unmapped_area_bottomup(struct file *file, info.high_limit = TASK_SIZE; info.align_mask = PAGE_MASK & ~huge_page_mask(h); info.align_offset = 0; + info.threadstack_offset = 0; return vm_unmapped_area(&info); } @@ -224,6 +225,7 @@ static unsigned long hugetlb_get_unmapped_area_topdown(struct file *file, info.high_limit = current->mm->mmap_base; info.align_mask = PAGE_MASK & ~huge_page_mask(h); info.align_offset = 0; + info.threadstack_offset = 0; addr = vm_unmapped_area(&info); /* diff --git a/arch/um/Makefile b/arch/um/Makefile index e4b1a96..16162f8 100644 --- a/arch/um/Makefile +++ b/arch/um/Makefile @@ -72,6 +72,10 @@ 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 += -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 2e0a6b1..a64d0f5 100644 --- a/arch/um/include/asm/kmap_types.h +++ b/arch/um/include/asm/kmap_types.h @@ -8,6 +8,6 @@ /* No more #include "asm/arch/kmap_types.h" ! */ -#define KM_TYPE_NR 14 +#define KM_TYPE_NR 15 #endif diff --git a/arch/um/include/asm/page.h b/arch/um/include/asm/page.h index 5ff53d9..5850cdf 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 f17bca8..48adb87 100644 --- a/arch/um/kernel/process.c +++ b/arch/um/kernel/process.c @@ -356,22 +356,6 @@ 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/exec.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 27adfd9..2362ac6 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -22,6 +22,7 @@ config X86_64 config X86 def_bool y select ARCH_HAS_DEBUG_STRICT_USER_COPY_CHECKS + select ARCH_HAS_FAST_MULTIPLIER select ARCH_MIGHT_HAVE_PC_PARPORT select ARCH_MIGHT_HAVE_PC_SERIO select HAVE_AOUT if X86_32 @@ -128,7 +129,7 @@ config X86 select RTC_LIB select HAVE_DEBUG_STACKOVERFLOW select HAVE_IRQ_EXIT_ON_IRQ_STACK if X86_64 - select HAVE_CC_STACKPROTECTOR + select HAVE_CC_STACKPROTECTOR if X86_64 || !PAX_MEMORY_UDEREF select GENERIC_CPU_AUTOPROBE select HAVE_ARCH_AUDITSYSCALL select ARCH_SUPPORTS_ATOMIC_RMW @@ -253,7 +254,7 @@ 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 @@ -549,6 +550,7 @@ config SCHED_OMIT_FRAME_POINTER menuconfig HYPERVISOR_GUEST bool "Linux guest support" + depends on !GRKERNSEC_CONFIG_AUTO || GRKERNSEC_CONFIG_VIRT_GUEST || (GRKERNSEC_CONFIG_VIRT_HOST && GRKERNSEC_CONFIG_VIRT_XEN) ---help--- Say Y here to enable options for running Linux under various hyper- visors. This option enables basic hypervisor detection and platform @@ -1076,6 +1078,7 @@ choice config NOHIGHMEM bool "off" + depends on !(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 @@ -1112,6 +1115,7 @@ config NOHIGHMEM config HIGHMEM4G bool "4GB" + depends on !(PAX_PAGEEXEC && PAX_ENABLE_PAE) ---help--- Select this if you have a 32-bit processor and between 1 and 4 gigabytes of physical RAM. @@ -1164,7 +1168,7 @@ 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 @@ -1578,6 +1582,7 @@ source kernel/Kconfig.hz config KEXEC bool "kexec system call" + depends on !GRKERNSEC_KMEM ---help--- kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot @@ -1728,7 +1733,9 @@ config X86_NEED_RELOCS config PHYSICAL_ALIGN hex "Alignment value to which kernel should be aligned" - default "0x200000" + default "0x1000000" + range 0x200000 0x1000000 if PAX_KERNEXEC && X86_PAE + range 0x400000 0x1000000 if PAX_KERNEXEC && !X86_PAE range 0x2000 0x1000000 if X86_32 range 0x200000 0x1000000 if X86_64 ---help--- @@ -1811,6 +1818,7 @@ config COMPAT_VDSO def_bool n prompt "Disable the 32-bit vDSO (needed for glibc 2.3.3)" depends on X86_32 || IA32_EMULATION + depends on !PAX_PAGEEXEC && !PAX_SEGMEXEC && !PAX_KERNEXEC && !PAX_MEMORY_UDEREF ---help--- Certain buggy versions of glibc will crash if they are presented with a 32-bit vDSO that is not mapped at the address diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index 6983314..54ad7e8 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -319,7 +319,7 @@ config X86_PPRO_FENCE config X86_F00F_BUG def_bool y - depends on M586MMX || M586TSC || M586 || M486 + depends on (M586MMX || M586TSC || M586 || M486) && !PAX_KERNEXEC config X86_INVD_BUG def_bool y @@ -327,7 +327,7 @@ config X86_INVD_BUG 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 @@ -369,7 +369,7 @@ 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 61bd2ad..50b625d 100644 --- a/arch/x86/Kconfig.debug +++ b/arch/x86/Kconfig.debug @@ -93,7 +93,7 @@ config EFI_PGT_DUMP 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 @@ -111,7 +111,7 @@ 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 33f71b0..c2cefa2 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -71,9 +71,6 @@ ifeq ($(CONFIG_X86_32),y) # CPU-specific tuning. Anything which can be shared with UML should go here. include $(srctree)/arch/x86/Makefile_32.cpu KBUILD_CFLAGS += $(cflags-y) - - # temporary until string.h is fixed - KBUILD_CFLAGS += -ffreestanding else BITS := 64 UTS_MACHINE := x86_64 @@ -114,6 +111,9 @@ else KBUILD_CFLAGS += $(call cc-option,-maccumulate-outgoing-args) endif +# temporary until string.h is fixed +KBUILD_CFLAGS += -ffreestanding + # Make sure compiler does not have buggy stack-protector support. ifdef CONFIG_CC_STACKPROTECTOR cc_has_sp := $(srctree)/scripts/gcc-x86_$(BITS)-has-stack-protector.sh @@ -271,3 +271,12 @@ define archhelp echo ' FDINITRD=file initrd for the booted kernel' echo ' kvmconfig - Enable additional options for guest kernel support' 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 dbe8dd2..2f0a98f 100644 --- a/arch/x86/boot/Makefile +++ b/arch/x86/boot/Makefile @@ -52,6 +52,9 @@ $(obj)/cpustr.h: $(obj)/mkcpustr FORCE # --------------------------------------------------------------------------- KBUILD_CFLAGS := $(USERINCLUDE) $(REALMODE_CFLAGS) -D_SETUP +ifdef CONSTIFY_PLUGIN +KBUILD_CFLAGS += -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 bd49ec6..94c7f58 100644 --- a/arch/x86/boot/boot.h +++ b/arch/x86/boot/boot.h @@ -84,7 +84,7 @@ 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; } diff --git a/arch/x86/boot/compressed/Makefile b/arch/x86/boot/compressed/Makefile index 0fcd913..3bb5c42 100644 --- a/arch/x86/boot/compressed/Makefile +++ b/arch/x86/boot/compressed/Makefile @@ -16,6 +16,9 @@ KBUILD_CFLAGS += $(cflags-y) KBUILD_CFLAGS += -mno-mmx -mno-sse KBUILD_CFLAGS += $(call cc-option,-ffreestanding) KBUILD_CFLAGS += $(call cc-option,-fno-stack-protector) +ifdef CONSTIFY_PLUGIN +KBUILD_CFLAGS += -fplugin-arg-constify_plugin-no-constify +endif KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ GCOV_PROFILE := n diff --git a/arch/x86/boot/compressed/efi_stub_32.S b/arch/x86/boot/compressed/efi_stub_32.S index a53440e..c3dbf1e 100644 --- a/arch/x86/boot/compressed/efi_stub_32.S +++ b/arch/x86/boot/compressed/efi_stub_32.S @@ -46,16 +46,13 @@ 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 %ecx - movl %ecx, saved_return_addr(%edx) - /* get the function pointer into ECX*/ - popl %ecx - movl %ecx, efi_rt_function_ptr(%edx) + popl saved_return_addr(%edx) + popl efi_rt_function_ptr(%edx) /* * 3. Call the physical function. */ - call *%ecx + call *efi_rt_function_ptr(%edx) /* * 4. Balance the stack. And because EAX contain the return value, @@ -67,15 +64,12 @@ ENTRY(efi_call_phys) 1: popl %edx subl $1b, %edx - movl efi_rt_function_ptr(%edx), %ecx - pushl %ecx + pushl efi_rt_function_ptr(%edx) /* * 10. Push the saved return address onto the stack and return. */ - movl saved_return_addr(%edx), %ecx - pushl %ecx - ret + jmpl *saved_return_addr(%edx) ENDPROC(efi_call_phys) .previous diff --git a/arch/x86/boot/compressed/head_32.S b/arch/x86/boot/compressed/head_32.S index cbed140..5f2ca57 100644 --- a/arch/x86/boot/compressed/head_32.S +++ b/arch/x86/boot/compressed/head_32.S @@ -140,10 +140,10 @@ preferred_addr: addl %eax, %ebx notl %eax andl %eax, %ebx - cmpl $LOAD_PHYSICAL_ADDR, %ebx + cmpl $____LOAD_PHYSICAL_ADDR, %ebx jge 1f #endif - movl $LOAD_PHYSICAL_ADDR, %ebx + movl $____LOAD_PHYSICAL_ADDR, %ebx 1: /* Target address to relocate to for decompression */ diff --git a/arch/x86/boot/compressed/head_64.S b/arch/x86/boot/compressed/head_64.S index 2884e0c..904a2f7 100644 --- a/arch/x86/boot/compressed/head_64.S +++ b/arch/x86/boot/compressed/head_64.S @@ -94,10 +94,10 @@ ENTRY(startup_32) addl %eax, %ebx notl %eax andl %eax, %ebx - cmpl $LOAD_PHYSICAL_ADDR, %ebx + cmpl $____LOAD_PHYSICAL_ADDR, %ebx jge 1f #endif - movl $LOAD_PHYSICAL_ADDR, %ebx + movl $____LOAD_PHYSICAL_ADDR, %ebx 1: /* Target address to relocate to for decompression */ @@ -322,10 +322,10 @@ preferred_addr: addq %rax, %rbp notq %rax andq %rax, %rbp - cmpq $LOAD_PHYSICAL_ADDR, %rbp + cmpq $____LOAD_PHYSICAL_ADDR, %rbp jge 1f #endif - movq $LOAD_PHYSICAL_ADDR, %rbp + movq $____LOAD_PHYSICAL_ADDR, %rbp 1: /* Target address to relocate to for decompression */ @@ -431,8 +431,8 @@ gdt: .long gdt .word 0 .quad 0x0000000000000000 /* NULL descriptor */ - .quad 0x00af9a000000ffff /* __KERNEL_CS */ - .quad 0x00cf92000000ffff /* __KERNEL_DS */ + .quad 0x00af9b000000ffff /* __KERNEL_CS */ + .quad 0x00cf93000000ffff /* __KERNEL_DS */ .quad 0x0080890000000000 /* TS descriptor */ .quad 0x0000000000000000 /* TS continued */ gdt_end: diff --git a/arch/x86/boot/compressed/misc.c b/arch/x86/boot/compressed/misc.c index 57ab74d..7c52182 100644 --- a/arch/x86/boot/compressed/misc.c +++ b/arch/x86/boot/compressed/misc.c @@ -242,7 +242,7 @@ static void handle_relocations(void *output, unsigned long output_len) * Calculate the delta between where vmlinux was linked to load * and where it was actually loaded. */ - delta = min_addr - LOAD_PHYSICAL_ADDR; + delta = min_addr - ____LOAD_PHYSICAL_ADDR; if (!delta) { debug_putstr("No relocation needed... "); return; @@ -312,7 +312,7 @@ static void parse_elf(void *output) Elf32_Ehdr ehdr; Elf32_Phdr *phdrs, *phdr; #endif - void *dest; + void *dest, *prev; int i; memcpy(&ehdr, output, sizeof(ehdr)); @@ -339,13 +339,16 @@ 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 memcpy(dest, output + phdr->p_offset, phdr->p_filesz); + if (i) + memset(prev, 0xff, dest - prev); + prev = dest + phdr->p_filesz; break; default: /* Ignore other PT_* */ break; } @@ -395,7 +398,7 @@ asmlinkage __visible 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/cpucheck.c b/arch/x86/boot/cpucheck.c index 1fd7d57..0f7d096 100644 --- a/arch/x86/boot/cpucheck.c +++ b/arch/x86/boot/cpucheck.c @@ -125,9 +125,9 @@ 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_cpuflags(); /* Make sure it really did something */ err = check_cpuflags(); @@ -140,9 +140,9 @@ 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_cpuflags(); @@ -153,12 +153,12 @@ 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_cpuflags(); } else if (err == 0x01 && diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S index 7a6d43a..edf6e40 100644 --- a/arch/x86/boot/header.S +++ b/arch/x86/boot/header.S @@ -438,10 +438,14 @@ 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) +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) +#define VO_INIT_SIZE (VO__end - VO__text - __PAGE_OFFSET - ____LOAD_PHYSICAL_ADDR) +#else #define VO_INIT_SIZE (VO__end - VO__text) +#endif #if ZO_INIT_SIZE > VO_INIT_SIZE #define INIT_SIZE ZO_INIT_SIZE #else 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 ba3e100..6501b8f 100644 --- a/arch/x86/boot/video-vesa.c +++ b/arch/x86/boot/video-vesa.c @@ -201,6 +201,7 @@ 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 9105655..41779c1 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 @@ -70,6 +72,8 @@ je B192; \ leaq 32(r9),r9; +#define ret pax_force_retaddr; ret + #define epilogue(FUNC,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 477e9d7..c92c7d8 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 @@ -205,7 +206,7 @@ enc: .octa 0x2 * num_initial_blocks = b mod 4 * encrypt the initial num_initial_blocks blocks and apply ghash on * the ciphertext -* %r10, %r11, %r12, %rax, %xmm5, %xmm6, %xmm7, %xmm8, %xmm9 registers +* %r10, %r11, %r15, %rax, %xmm5, %xmm6, %xmm7, %xmm8, %xmm9 registers * are clobbered * arg1, %arg2, %arg3, %r14 are used as a pointer only, not modified */ @@ -214,8 +215,8 @@ enc: .octa 0x2 .macro INITIAL_BLOCKS_DEC num_initial_blocks TMP1 TMP2 TMP3 TMP4 TMP5 XMM0 XMM1 \ XMM2 XMM3 XMM4 XMMDst TMP6 TMP7 i i_seq operation mov arg7, %r10 # %r10 = AAD - mov arg8, %r12 # %r12 = aadLen - mov %r12, %r11 + mov arg8, %r15 # %r15 = aadLen + mov %r15, %r11 pxor %xmm\i, %xmm\i _get_AAD_loop\num_initial_blocks\operation: movd (%r10), \TMP1 @@ -223,15 +224,15 @@ _get_AAD_loop\num_initial_blocks\operation: psrldq $4, %xmm\i pxor \TMP1, %xmm\i add $4, %r10 - sub $4, %r12 + sub $4, %r15 jne _get_AAD_loop\num_initial_blocks\operation cmp $16, %r11 je _get_AAD_loop2_done\num_initial_blocks\operation - mov $16, %r12 + mov $16, %r15 _get_AAD_loop2\num_initial_blocks\operation: psrldq $4, %xmm\i - sub $4, %r12 - cmp %r11, %r12 + sub $4, %r15 + cmp %r11, %r15 jne _get_AAD_loop2\num_initial_blocks\operation _get_AAD_loop2_done\num_initial_blocks\operation: movdqa SHUF_MASK(%rip), %xmm14 @@ -443,7 +444,7 @@ _initial_blocks_done\num_initial_blocks\operation: * num_initial_blocks = b mod 4 * encrypt the initial num_initial_blocks blocks and apply ghash on * the ciphertext -* %r10, %r11, %r12, %rax, %xmm5, %xmm6, %xmm7, %xmm8, %xmm9 registers +* %r10, %r11, %r15, %rax, %xmm5, %xmm6, %xmm7, %xmm8, %xmm9 registers * are clobbered * arg1, %arg2, %arg3, %r14 are used as a pointer only, not modified */ @@ -452,8 +453,8 @@ _initial_blocks_done\num_initial_blocks\operation: .macro INITIAL_BLOCKS_ENC num_initial_blocks TMP1 TMP2 TMP3 TMP4 TMP5 XMM0 XMM1 \ XMM2 XMM3 XMM4 XMMDst TMP6 TMP7 i i_seq operation mov arg7, %r10 # %r10 = AAD - mov arg8, %r12 # %r12 = aadLen - mov %r12, %r11 + mov arg8, %r15 # %r15 = aadLen + mov %r15, %r11 pxor %xmm\i, %xmm\i _get_AAD_loop\num_initial_blocks\operation: movd (%r10), \TMP1 @@ -461,15 +462,15 @@ _get_AAD_loop\num_initial_blocks\operation: psrldq $4, %xmm\i pxor \TMP1, %xmm\i add $4, %r10 - sub $4, %r12 + sub $4, %r15 jne _get_AAD_loop\num_initial_blocks\operation cmp $16, %r11 je _get_AAD_loop2_done\num_initial_blocks\operation - mov $16, %r12 + mov $16, %r15 _get_AAD_loop2\num_initial_blocks\operation: psrldq $4, %xmm\i - sub $4, %r12 - cmp %r11, %r12 + sub $4, %r15 + cmp %r11, %r15 jne _get_AAD_loop2\num_initial_blocks\operation _get_AAD_loop2_done\num_initial_blocks\operation: movdqa SHUF_MASK(%rip), %xmm14 @@ -1269,7 +1270,7 @@ TMP7 XMM1 XMM2 XMM3 XMM4 XMMDst * *****************************************************************************/ ENTRY(aesni_gcm_dec) - push %r12 + push %r15 push %r13 push %r14 mov %rsp, %r14 @@ -1279,8 +1280,8 @@ ENTRY(aesni_gcm_dec) */ sub $VARIABLE_OFFSET, %rsp and $~63, %rsp # align rsp to 64 bytes - mov %arg6, %r12 - movdqu (%r12), %xmm13 # %xmm13 = HashKey + mov %arg6, %r15 + movdqu (%r15), %xmm13 # %xmm13 = HashKey movdqa SHUF_MASK(%rip), %xmm2 PSHUFB_XMM %xmm2, %xmm13 @@ -1308,10 +1309,10 @@ ENTRY(aesni_gcm_dec) movdqa %xmm13, HashKey(%rsp) # store HashKey<<1 (mod poly) mov %arg4, %r13 # save the number of bytes of plaintext/ciphertext and $-16, %r13 # %r13 = %r13 - (%r13 mod 16) - mov %r13, %r12 - and $(3<<4), %r12 + mov %r13, %r15 + and $(3<<4), %r15 jz _initial_num_blocks_is_0_decrypt - cmp $(2<<4), %r12 + cmp $(2<<4), %r15 jb _initial_num_blocks_is_1_decrypt je _initial_num_blocks_is_2_decrypt _initial_num_blocks_is_3_decrypt: @@ -1361,16 +1362,16 @@ _zero_cipher_left_decrypt: sub $16, %r11 add %r13, %r11 movdqu (%arg3,%r11,1), %xmm1 # receive the last <16 byte block - lea SHIFT_MASK+16(%rip), %r12 - sub %r13, %r12 + lea SHIFT_MASK+16(%rip), %r15 + sub %r13, %r15 # adjust the shuffle mask pointer to be able to shift 16-%r13 bytes # (%r13 is the number of bytes in plaintext mod 16) - movdqu (%r12), %xmm2 # get the appropriate shuffle mask + movdqu (%r15), %xmm2 # get the appropriate shuffle mask PSHUFB_XMM %xmm2, %xmm1 # right shift 16-%r13 butes movdqa %xmm1, %xmm2 pxor %xmm1, %xmm0 # Ciphertext XOR E(K, Yn) - movdqu ALL_F-SHIFT_MASK(%r12), %xmm1 + movdqu ALL_F-SHIFT_MASK(%r15), %xmm1 # get the appropriate mask to mask out top 16-%r13 bytes of %xmm0 pand %xmm1, %xmm0 # mask out top 16-%r13 bytes of %xmm0 pand %xmm1, %xmm2 @@ -1399,9 +1400,9 @@ _less_than_8_bytes_left_decrypt: sub $1, %r13 jne _less_than_8_bytes_left_decrypt _multiple_of_16_bytes_decrypt: - mov arg8, %r12 # %r13 = aadLen (number of bytes) - shl $3, %r12 # convert into number of bits - movd %r12d, %xmm15 # len(A) in %xmm15 + mov arg8, %r15 # %r13 = aadLen (number of bytes) + shl $3, %r15 # convert into number of bits + movd %r15d, %xmm15 # len(A) in %xmm15 shl $3, %arg4 # len(C) in bits (*128) MOVQ_R64_XMM %arg4, %xmm1 pslldq $8, %xmm15 # %xmm15 = len(A)||0x0000000000000000 @@ -1440,7 +1441,8 @@ _return_T_done_decrypt: mov %r14, %rsp pop %r14 pop %r13 - pop %r12 + pop %r15 + pax_force_retaddr ret ENDPROC(aesni_gcm_dec) @@ -1529,7 +1531,7 @@ ENDPROC(aesni_gcm_dec) * poly = x^128 + x^127 + x^126 + x^121 + 1 ***************************************************************************/ ENTRY(aesni_gcm_enc) - push %r12 + push %r15 push %r13 push %r14 mov %rsp, %r14 @@ -1539,8 +1541,8 @@ ENTRY(aesni_gcm_enc) # sub $VARIABLE_OFFSET, %rsp and $~63, %rsp - mov %arg6, %r12 - movdqu (%r12), %xmm13 + mov %arg6, %r15 + movdqu (%r15), %xmm13 movdqa SHUF_MASK(%rip), %xmm2 PSHUFB_XMM %xmm2, %xmm13 @@ -1564,13 +1566,13 @@ ENTRY(aesni_gcm_enc) movdqa %xmm13, HashKey(%rsp) mov %arg4, %r13 # %xmm13 holds HashKey<<1 (mod poly) and $-16, %r13 - mov %r13, %r12 + mov %r13, %r15 # Encrypt first few blocks - and $(3<<4), %r12 + and $(3<<4), %r15 jz _initial_num_blocks_is_0_encrypt - cmp $(2<<4), %r12 + cmp $(2<<4), %r15 jb _initial_num_blocks_is_1_encrypt je _initial_num_blocks_is_2_encrypt _initial_num_blocks_is_3_encrypt: @@ -1623,14 +1625,14 @@ _zero_cipher_left_encrypt: sub $16, %r11 add %r13, %r11 movdqu (%arg3,%r11,1), %xmm1 # receive the last <16 byte blocks - lea SHIFT_MASK+16(%rip), %r12 - sub %r13, %r12 + lea SHIFT_MASK+16(%rip), %r15 + sub %r13, %r15 # adjust the shuffle mask pointer to be able to shift 16-r13 bytes # (%r13 is the number of bytes in plaintext mod 16) - movdqu (%r12), %xmm2 # get the appropriate shuffle mask + movdqu (%r15), %xmm2 # get the appropriate shuffle mask PSHUFB_XMM %xmm2, %xmm1 # shift right 16-r13 byte pxor %xmm1, %xmm0 # Plaintext XOR Encrypt(K, Yn) - movdqu ALL_F-SHIFT_MASK(%r12), %xmm1 + movdqu ALL_F-SHIFT_MASK(%r15), %xmm1 # get the appropriate mask to mask out top 16-r13 bytes of xmm0 pand %xmm1, %xmm0 # mask out top 16-r13 bytes of xmm0 movdqa SHUF_MASK(%rip), %xmm10 @@ -1663,9 +1665,9 @@ _less_than_8_bytes_left_encrypt: sub $1, %r13 jne _less_than_8_bytes_left_encrypt _multiple_of_16_bytes_encrypt: - mov arg8, %r12 # %r12 = addLen (number of bytes) - shl $3, %r12 - movd %r12d, %xmm15 # len(A) in %xmm15 + mov arg8, %r15 # %r15 = addLen (number of bytes) + shl $3, %r15 + movd %r15d, %xmm15 # len(A) in %xmm15 shl $3, %arg4 # len(C) in bits (*128) MOVQ_R64_XMM %arg4, %xmm1 pslldq $8, %xmm15 # %xmm15 = len(A)||0x0000000000000000 @@ -1704,7 +1706,8 @@ _return_T_done_encrypt: mov %r14, %rsp pop %r14 pop %r13 - pop %r12 + pop %r15 + pax_force_retaddr ret ENDPROC(aesni_gcm_enc) @@ -1722,6 +1725,7 @@ _key_expansion_256a: pxor %xmm1, %xmm0 movaps %xmm0, (TKEYP) add $0x10, TKEYP + pax_force_retaddr ret ENDPROC(_key_expansion_128) ENDPROC(_key_expansion_256a) @@ -1748,6 +1752,7 @@ _key_expansion_192a: shufps $0b01001110, %xmm2, %xmm1 movaps %xmm1, 0x10(TKEYP) add $0x20, TKEYP + pax_force_retaddr ret ENDPROC(_key_expansion_192a) @@ -1768,6 +1773,7 @@ _key_expansion_192b: movaps %xmm0, (TKEYP) add $0x10, TKEYP + pax_force_retaddr ret ENDPROC(_key_expansion_192b) @@ -1781,6 +1787,7 @@ _key_expansion_256b: pxor %xmm1, %xmm2 movaps %xmm2, (TKEYP) add $0x10, TKEYP + pax_force_retaddr ret ENDPROC(_key_expansion_256b) @@ -1894,6 +1901,7 @@ ENTRY(aesni_set_key) #ifndef __x86_64__ popl KEYP #endif + pax_force_retaddr ret ENDPROC(aesni_set_key) @@ -1916,6 +1924,7 @@ ENTRY(aesni_enc) popl KLEN popl KEYP #endif + pax_force_retaddr ret ENDPROC(aesni_enc) @@ -1974,6 +1983,7 @@ _aesni_enc1: AESENC KEY STATE movaps 0x70(TKEYP), KEY AESENCLAST KEY STATE + pax_force_retaddr ret ENDPROC(_aesni_enc1) @@ -2083,6 +2093,7 @@ _aesni_enc4: AESENCLAST KEY STATE2 AESENCLAST KEY STATE3 AESENCLAST KEY STATE4 + pax_force_retaddr ret ENDPROC(_aesni_enc4) @@ -2106,6 +2117,7 @@ ENTRY(aesni_dec) popl KLEN popl KEYP #endif + pax_force_retaddr ret ENDPROC(aesni_dec) @@ -2164,6 +2176,7 @@ _aesni_dec1: AESDEC KEY STATE movaps 0x70(TKEYP), KEY AESDECLAST KEY STATE + pax_force_retaddr ret ENDPROC(_aesni_dec1) @@ -2273,6 +2286,7 @@ _aesni_dec4: AESDECLAST KEY STATE2 AESDECLAST KEY STATE3 AESDECLAST KEY STATE4 + pax_force_retaddr ret ENDPROC(_aesni_dec4) @@ -2331,6 +2345,7 @@ ENTRY(aesni_ecb_enc) popl KEYP popl LEN #endif + pax_force_retaddr ret ENDPROC(aesni_ecb_enc) @@ -2390,6 +2405,7 @@ ENTRY(aesni_ecb_dec) popl KEYP popl LEN #endif + pax_force_retaddr ret ENDPROC(aesni_ecb_dec) @@ -2432,6 +2448,7 @@ ENTRY(aesni_cbc_enc) popl LEN popl IVP #endif + pax_force_retaddr ret ENDPROC(aesni_cbc_enc) @@ -2523,6 +2540,7 @@ ENTRY(aesni_cbc_dec) popl LEN popl IVP #endif + pax_force_retaddr ret ENDPROC(aesni_cbc_dec) @@ -2550,6 +2568,7 @@ _aesni_inc_init: mov $1, TCTR_LOW MOVQ_R64_XMM TCTR_LOW INC MOVQ_R64_XMM CTR TCTR_LOW + pax_force_retaddr ret ENDPROC(_aesni_inc_init) @@ -2579,6 +2598,7 @@ _aesni_inc: .Linc_low: movaps CTR, IV PSHUFB_XMM BSWAP_MASK IV + pax_force_retaddr ret ENDPROC(_aesni_inc) @@ -2640,6 +2660,7 @@ ENTRY(aesni_ctr_enc) .Lctr_enc_ret: movups IV, (IVP) .Lctr_enc_just_ret: + pax_force_retaddr ret ENDPROC(aesni_ctr_enc) @@ -2766,6 +2787,7 @@ ENTRY(aesni_xts_crypt8) pxor INC, STATE4 movdqu STATE4, 0x70(OUTP) + pax_force_retaddr ret ENDPROC(aesni_xts_crypt8) diff --git a/arch/x86/crypto/blowfish-x86_64-asm_64.S b/arch/x86/crypto/blowfish-x86_64-asm_64.S index 246c670..466e2d6 100644 --- a/arch/x86/crypto/blowfish-x86_64-asm_64.S +++ b/arch/x86/crypto/blowfish-x86_64-asm_64.S @@ -21,6 +21,7 @@ */ #include +#include .file "blowfish-x86_64-asm.S" .text @@ -149,9 +150,11 @@ ENTRY(__blowfish_enc_blk) jnz .L__enc_xor; write_block(); + pax_force_retaddr ret; .L__enc_xor: xor_block(); + pax_force_retaddr ret; ENDPROC(__blowfish_enc_blk) @@ -183,6 +186,7 @@ ENTRY(blowfish_dec_blk) movq %r11, %rbp; + pax_force_retaddr ret; ENDPROC(blowfish_dec_blk) @@ -334,6 +338,7 @@ ENTRY(__blowfish_enc_blk_4way) popq %rbx; popq %rbp; + pax_force_retaddr ret; .L__enc_xor4: @@ -341,6 +346,7 @@ ENTRY(__blowfish_enc_blk_4way) popq %rbx; popq %rbp; + pax_force_retaddr ret; ENDPROC(__blowfish_enc_blk_4way) @@ -375,5 +381,6 @@ ENTRY(blowfish_dec_blk_4way) popq %rbx; popq %rbp; + pax_force_retaddr ret; ENDPROC(blowfish_dec_blk_4way) diff --git a/arch/x86/crypto/camellia-aesni-avx-asm_64.S b/arch/x86/crypto/camellia-aesni-avx-asm_64.S index ce71f92..1dce7ec 100644 --- a/arch/x86/crypto/camellia-aesni-avx-asm_64.S +++ b/arch/x86/crypto/camellia-aesni-avx-asm_64.S @@ -16,6 +16,7 @@ */ #include +#include #define CAMELLIA_TABLE_BYTE_LEN 272 @@ -191,6 +192,7 @@ roundsm16_x0_x1_x2_x3_x4_x5_x6_x7_y0_y1_y2_y3_y4_y5_y6_y7_cd: roundsm16(%xmm0, %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7, %xmm8, %xmm9, %xmm10, %xmm11, %xmm12, %xmm13, %xmm14, %xmm15, %rcx, (%r9)); + pax_force_retaddr ret; ENDPROC(roundsm16_x0_x1_x2_x3_x4_x5_x6_x7_y0_y1_y2_y3_y4_y5_y6_y7_cd) @@ -199,6 +201,7 @@ roundsm16_x4_x5_x6_x7_x0_x1_x2_x3_y4_y5_y6_y7_y0_y1_y2_y3_ab: roundsm16(%xmm4, %xmm5, %xmm6, %xmm7, %xmm0, %xmm1, %xmm2, %xmm3, %xmm12, %xmm13, %xmm14, %xmm15, %xmm8, %xmm9, %xmm10, %xmm11, %rax, (%r9)); + pax_force_retaddr ret; ENDPROC(roundsm16_x4_x5_x6_x7_x0_x1_x2_x3_y4_y5_y6_y7_y0_y1_y2_y3_ab) @@ -780,6 +783,7 @@ __camellia_enc_blk16: %xmm8, %xmm9, %xmm10, %xmm11, %xmm12, %xmm13, %xmm14, %xmm15, (key_table)(CTX, %r8, 8), (%rax), 1 * 16(%rax)); + pax_force_retaddr ret; .align 8 @@ -865,6 +869,7 @@ __camellia_dec_blk16: %xmm8, %xmm9, %xmm10, %xmm11, %xmm12, %xmm13, %xmm14, %xmm15, (key_table)(CTX), (%rax), 1 * 16(%rax)); + pax_force_retaddr ret; .align 8 @@ -904,6 +909,7 @@ ENTRY(camellia_ecb_enc_16way) %xmm15, %xmm14, %xmm13, %xmm12, %xmm11, %xmm10, %xmm9, %xmm8, %rsi); + pax_force_retaddr ret; ENDPROC(camellia_ecb_enc_16way) @@ -932,6 +938,7 @@ ENTRY(camellia_ecb_dec_16way) %xmm15, %xmm14, %xmm13, %xmm12, %xmm11, %xmm10, %xmm9, %xmm8, %rsi); + pax_force_retaddr ret; ENDPROC(camellia_ecb_dec_16way) @@ -981,6 +988,7 @@ ENTRY(camellia_cbc_dec_16way) %xmm15, %xmm14, %xmm13, %xmm12, %xmm11, %xmm10, %xmm9, %xmm8, %rsi); + pax_force_retaddr ret; ENDPROC(camellia_cbc_dec_16way) @@ -1092,6 +1100,7 @@ ENTRY(camellia_ctr_16way) %xmm15, %xmm14, %xmm13, %xmm12, %xmm11, %xmm10, %xmm9, %xmm8, %rsi); + pax_force_retaddr ret; ENDPROC(camellia_ctr_16way) @@ -1234,6 +1243,7 @@ camellia_xts_crypt_16way: %xmm15, %xmm14, %xmm13, %xmm12, %xmm11, %xmm10, %xmm9, %xmm8, %rsi); + pax_force_retaddr ret; ENDPROC(camellia_xts_crypt_16way) diff --git a/arch/x86/crypto/camellia-aesni-avx2-asm_64.S b/arch/x86/crypto/camellia-aesni-avx2-asm_64.S index 0e0b886..5a3123c 100644 --- a/arch/x86/crypto/camellia-aesni-avx2-asm_64.S +++ b/arch/x86/crypto/camellia-aesni-avx2-asm_64.S @@ -11,6 +11,7 @@ */ #include +#include #define CAMELLIA_TABLE_BYTE_LEN 272 @@ -230,6 +231,7 @@ roundsm32_x0_x1_x2_x3_x4_x5_x6_x7_y0_y1_y2_y3_y4_y5_y6_y7_cd: roundsm32(%ymm0, %ymm1, %ymm2, %ymm3, %ymm4, %ymm5, %ymm6, %ymm7, %ymm8, %ymm9, %ymm10, %ymm11, %ymm12, %ymm13, %ymm14, %ymm15, %rcx, (%r9)); + pax_force_retaddr ret; ENDPROC(roundsm32_x0_x1_x2_x3_x4_x5_x6_x7_y0_y1_y2_y3_y4_y5_y6_y7_cd) @@ -238,6 +240,7 @@ roundsm32_x4_x5_x6_x7_x0_x1_x2_x3_y4_y5_y6_y7_y0_y1_y2_y3_ab: roundsm32(%ymm4, %ymm5, %ymm6, %ymm7, %ymm0, %ymm1, %ymm2, %ymm3, %ymm12, %ymm13, %ymm14, %ymm15, %ymm8, %ymm9, %ymm10, %ymm11, %rax, (%r9)); + pax_force_retaddr ret; ENDPROC(roundsm32_x4_x5_x6_x7_x0_x1_x2_x3_y4_y5_y6_y7_y0_y1_y2_y3_ab) @@ -820,6 +823,7 @@ __camellia_enc_blk32: %ymm8, %ymm9, %ymm10, %ymm11, %ymm12, %ymm13, %ymm14, %ymm15, (key_table)(CTX, %r8, 8), (%rax), 1 * 32(%rax)); + pax_force_retaddr ret; .align 8 @@ -905,6 +909,7 @@ __camellia_dec_blk32: %ymm8, %ymm9, %ymm10, %ymm11, %ymm12, %ymm13, %ymm14, %ymm15, (key_table)(CTX), (%rax), 1 * 32(%rax)); + pax_force_retaddr ret; .align 8 @@ -948,6 +953,7 @@ ENTRY(camellia_ecb_enc_32way) vzeroupper; + pax_force_retaddr ret; ENDPROC(camellia_ecb_enc_32way) @@ -980,6 +986,7 @@ ENTRY(camellia_ecb_dec_32way) vzeroupper; + pax_force_retaddr ret; ENDPROC(camellia_ecb_dec_32way) @@ -1046,6 +1053,7 @@ ENTRY(camellia_cbc_dec_32way) vzeroupper; + pax_force_retaddr ret; ENDPROC(camellia_cbc_dec_32way) @@ -1184,6 +1192,7 @@ ENTRY(camellia_ctr_32way) vzeroupper; + pax_force_retaddr ret; ENDPROC(camellia_ctr_32way) @@ -1349,6 +1358,7 @@ camellia_xts_crypt_32way: vzeroupper; + pax_force_retaddr ret; ENDPROC(camellia_xts_crypt_32way) diff --git a/arch/x86/crypto/camellia-x86_64-asm_64.S b/arch/x86/crypto/camellia-x86_64-asm_64.S index 310319c..db3d7b5 100644 --- a/arch/x86/crypto/camellia-x86_64-asm_64.S +++ b/arch/x86/crypto/camellia-x86_64-asm_64.S @@ -21,6 +21,7 @@ */ #include +#include .file "camellia-x86_64-asm_64.S" .text @@ -228,12 +229,14 @@ ENTRY(__camellia_enc_blk) enc_outunpack(mov, RT1); movq RRBP, %rbp; + pax_force_retaddr ret; .L__enc_xor: enc_outunpack(xor, RT1); movq RRBP, %rbp; + pax_force_retaddr ret; ENDPROC(__camellia_enc_blk) @@ -272,6 +275,7 @@ ENTRY(camellia_dec_blk) dec_outunpack(); movq RRBP, %rbp; + pax_force_retaddr ret; ENDPROC(camellia_dec_blk) @@ -463,6 +467,7 @@ ENTRY(__camellia_enc_blk_2way) movq RRBP, %rbp; popq %rbx; + pax_force_retaddr ret; .L__enc2_xor: @@ -470,6 +475,7 @@ ENTRY(__camellia_enc_blk_2way) movq RRBP, %rbp; popq %rbx; + pax_force_retaddr ret; ENDPROC(__camellia_enc_blk_2way) @@ -510,5 +516,6 @@ ENTRY(camellia_dec_blk_2way) movq RRBP, %rbp; movq RXOR, %rbx; + pax_force_retaddr ret; ENDPROC(camellia_dec_blk_2way) diff --git a/arch/x86/crypto/cast5-avx-x86_64-asm_64.S b/arch/x86/crypto/cast5-avx-x86_64-asm_64.S index c35fd5d..2d8c7db 100644 --- a/arch/x86/crypto/cast5-avx-x86_64-asm_64.S +++ b/arch/x86/crypto/cast5-avx-x86_64-asm_64.S @@ -24,6 +24,7 @@ */ #include +#include .file "cast5-avx-x86_64-asm_64.S" @@ -281,6 +282,7 @@ __cast5_enc_blk16: outunpack_blocks(RR3, RL3, RTMP, RX, RKM); outunpack_blocks(RR4, RL4, RTMP, RX, RKM); + pax_force_retaddr ret; ENDPROC(__cast5_enc_blk16) @@ -352,6 +354,7 @@ __cast5_dec_blk16: outunpack_blocks(RR3, RL3, RTMP, RX, RKM); outunpack_blocks(RR4, RL4, RTMP, RX, RKM); + pax_force_retaddr ret; .L__skip_dec: @@ -388,6 +391,7 @@ ENTRY(cast5_ecb_enc_16way) vmovdqu RR4, (6*4*4)(%r11); vmovdqu RL4, (7*4*4)(%r11); + pax_force_retaddr ret; ENDPROC(cast5_ecb_enc_16way) @@ -420,6 +424,7 @@ ENTRY(cast5_ecb_dec_16way) vmovdqu RR4, (6*4*4)(%r11); vmovdqu RL4, (7*4*4)(%r11); + pax_force_retaddr ret; ENDPROC(cast5_ecb_dec_16way) @@ -430,10 +435,10 @@ ENTRY(cast5_cbc_dec_16way) * %rdx: src */ - pushq %r12; + pushq %r14; movq %rsi, %r11; - movq %rdx, %r12; + movq %rdx, %r14; vmovdqu (0*16)(%rdx), RL1; vmovdqu (1*16)(%rdx), RR1; @@ -447,16 +452,16 @@ ENTRY(cast5_cbc_dec_16way) call __cast5_dec_blk16; /* xor with src */ - vmovq (%r12), RX; + vmovq (%r14), RX; vpshufd $0x4f, RX, RX; vpxor RX, RR1, RR1; - vpxor 0*16+8(%r12), RL1, RL1; - vpxor 1*16+8(%r12), RR2, RR2; - vpxor 2*16+8(%r12), RL2, RL2; - vpxor 3*16+8(%r12), RR3, RR3; - vpxor 4*16+8(%r12), RL3, RL3; - vpxor 5*16+8(%r12), RR4, RR4; - vpxor 6*16+8(%r12), RL4, RL4; + vpxor 0*16+8(%r14), RL1, RL1; + vpxor 1*16+8(%r14), RR2, RR2; + vpxor 2*16+8(%r14), RL2, RL2; + vpxor 3*16+8(%r14), RR3, RR3; + vpxor 4*16+8(%r14), RL3, RL3; + vpxor 5*16+8(%r14), RR4, RR4; + vpxor 6*16+8(%r14), RL4, RL4; vmovdqu RR1, (0*16)(%r11); vmovdqu RL1, (1*16)(%r11); @@ -467,8 +472,9 @@ ENTRY(cast5_cbc_dec_16way) vmovdqu RR4, (6*16)(%r11); vmovdqu RL4, (7*16)(%r11); - popq %r12; + popq %r14; + pax_force_retaddr ret; ENDPROC(cast5_cbc_dec_16way) @@ -480,10 +486,10 @@ ENTRY(cast5_ctr_16way) * %rcx: iv (big endian, 64bit) */ - pushq %r12; + pushq %r14; movq %rsi, %r11; - movq %rdx, %r12; + movq %rdx, %r14; vpcmpeqd RTMP, RTMP, RTMP; vpsrldq $8, RTMP, RTMP; /* low: -1, high: 0 */ @@ -523,14 +529,14 @@ ENTRY(cast5_ctr_16way) call __cast5_enc_blk16; /* dst = src ^ iv */ - vpxor (0*16)(%r12), RR1, RR1; - vpxor (1*16)(%r12), RL1, RL1; - vpxor (2*16)(%r12), RR2, RR2; - vpxor (3*16)(%r12), RL2, RL2; - vpxor (4*16)(%r12), RR3, RR3; - vpxor (5*16)(%r12), RL3, RL3; - vpxor (6*16)(%r12), RR4, RR4; - vpxor (7*16)(%r12), RL4, RL4; + vpxor (0*16)(%r14), RR1, RR1; + vpxor (1*16)(%r14), RL1, RL1; + vpxor (2*16)(%r14), RR2, RR2; + vpxor (3*16)(%r14), RL2, RL2; + vpxor (4*16)(%r14), RR3, RR3; + vpxor (5*16)(%r14), RL3, RL3; + vpxor (6*16)(%r14), RR4, RR4; + vpxor (7*16)(%r14), RL4, RL4; vmovdqu RR1, (0*16)(%r11); vmovdqu RL1, (1*16)(%r11); vmovdqu RR2, (2*16)(%r11); @@ -540,7 +546,8 @@ ENTRY(cast5_ctr_16way) vmovdqu RR4, (6*16)(%r11); vmovdqu RL4, (7*16)(%r11); - popq %r12; + popq %r14; + pax_force_retaddr ret; ENDPROC(cast5_ctr_16way) diff --git a/arch/x86/crypto/cast6-avx-x86_64-asm_64.S b/arch/x86/crypto/cast6-avx-x86_64-asm_64.S index e3531f8..e123f35 100644 --- a/arch/x86/crypto/cast6-avx-x86_64-asm_64.S +++ b/arch/x86/crypto/cast6-avx-x86_64-asm_64.S @@ -24,6 +24,7 @@ */ #include +#include #include "glue_helper-asm-avx.S" .file "cast6-avx-x86_64-asm_64.S" @@ -295,6 +296,7 @@ __cast6_enc_blk8: outunpack_blocks(RA1, RB1, RC1, RD1, RTMP, RX, RKRF, RKM); outunpack_blocks(RA2, RB2, RC2, RD2, RTMP, RX, RKRF, RKM); + pax_force_retaddr ret; ENDPROC(__cast6_enc_blk8) @@ -340,6 +342,7 @@ __cast6_dec_blk8: outunpack_blocks(RA1, RB1, RC1, RD1, RTMP, RX, RKRF, RKM); outunpack_blocks(RA2, RB2, RC2, RD2, RTMP, RX, RKRF, RKM); + pax_force_retaddr ret; ENDPROC(__cast6_dec_blk8) @@ -358,6 +361,7 @@ ENTRY(cast6_ecb_enc_8way) store_8way(%r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(cast6_ecb_enc_8way) @@ -376,6 +380,7 @@ ENTRY(cast6_ecb_dec_8way) store_8way(%r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(cast6_ecb_dec_8way) @@ -386,19 +391,20 @@ ENTRY(cast6_cbc_dec_8way) * %rdx: src */ - pushq %r12; + pushq %r14; movq %rsi, %r11; - movq %rdx, %r12; + movq %rdx, %r14; load_8way(%rdx, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); call __cast6_dec_blk8; - store_cbc_8way(%r12, %r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + store_cbc_8way(%r14, %r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); - popq %r12; + popq %r14; + pax_force_retaddr ret; ENDPROC(cast6_cbc_dec_8way) @@ -410,20 +416,21 @@ ENTRY(cast6_ctr_8way) * %rcx: iv (little endian, 128bit) */ - pushq %r12; + pushq %r14; movq %rsi, %r11; - movq %rdx, %r12; + movq %rdx, %r14; load_ctr_8way(%rcx, .Lbswap128_mask, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2, RX, RKR, RKM); call __cast6_enc_blk8; - store_ctr_8way(%r12, %r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + store_ctr_8way(%r14, %r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); - popq %r12; + popq %r14; + pax_force_retaddr ret; ENDPROC(cast6_ctr_8way) @@ -446,6 +453,7 @@ ENTRY(cast6_xts_enc_8way) /* dst <= regs xor IVs(in dst) */ store_xts_8way(%r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(cast6_xts_enc_8way) @@ -468,5 +476,6 @@ ENTRY(cast6_xts_dec_8way) /* dst <= regs xor IVs(in dst) */ store_xts_8way(%r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(cast6_xts_dec_8way) diff --git a/arch/x86/crypto/crc32c-pcl-intel-asm_64.S b/arch/x86/crypto/crc32c-pcl-intel-asm_64.S index dbc4339..de6e120 100644 --- a/arch/x86/crypto/crc32c-pcl-intel-asm_64.S +++ b/arch/x86/crypto/crc32c-pcl-intel-asm_64.S @@ -45,6 +45,7 @@ #include #include +#include ## ISCSI CRC 32 Implementation with crc32 and pclmulqdq Instruction @@ -312,6 +313,7 @@ do_return: popq %rsi popq %rdi popq %rbx + pax_force_retaddr ret ################################################################ diff --git a/arch/x86/crypto/ghash-clmulni-intel_asm.S b/arch/x86/crypto/ghash-clmulni-intel_asm.S index 5d1e007..098cb4f 100644 --- a/arch/x86/crypto/ghash-clmulni-intel_asm.S +++ b/arch/x86/crypto/ghash-clmulni-intel_asm.S @@ -18,6 +18,7 @@ #include #include +#include .data @@ -89,6 +90,7 @@ __clmul_gf128mul_ble: psrlq $1, T2 pxor T2, T1 pxor T1, DATA + pax_force_retaddr ret ENDPROC(__clmul_gf128mul_ble) @@ -101,6 +103,7 @@ ENTRY(clmul_ghash_mul) call __clmul_gf128mul_ble PSHUFB_XMM BSWAP DATA movups DATA, (%rdi) + pax_force_retaddr ret ENDPROC(clmul_ghash_mul) @@ -128,5 +131,6 @@ ENTRY(clmul_ghash_update) PSHUFB_XMM BSWAP DATA movups DATA, (%rdi) .Lupdate_just_ret: + pax_force_retaddr ret ENDPROC(clmul_ghash_update) diff --git a/arch/x86/crypto/salsa20-x86_64-asm_64.S b/arch/x86/crypto/salsa20-x86_64-asm_64.S index 9279e0b..c4b3d2c 100644 --- a/arch/x86/crypto/salsa20-x86_64-asm_64.S +++ b/arch/x86/crypto/salsa20-x86_64-asm_64.S @@ -1,4 +1,5 @@ #include +#include # enter salsa20_encrypt_bytes ENTRY(salsa20_encrypt_bytes) @@ -789,6 +790,7 @@ ENTRY(salsa20_encrypt_bytes) add %r11,%rsp mov %rdi,%rax mov %rsi,%rdx + pax_force_retaddr ret # bytesatleast65: ._bytesatleast65: @@ -889,6 +891,7 @@ ENTRY(salsa20_keysetup) add %r11,%rsp mov %rdi,%rax mov %rsi,%rdx + pax_force_retaddr ret ENDPROC(salsa20_keysetup) @@ -914,5 +917,6 @@ ENTRY(salsa20_ivsetup) add %r11,%rsp mov %rdi,%rax mov %rsi,%rdx + pax_force_retaddr ret ENDPROC(salsa20_ivsetup) diff --git a/arch/x86/crypto/serpent-avx-x86_64-asm_64.S b/arch/x86/crypto/serpent-avx-x86_64-asm_64.S index 2f202f4..d9164d6 100644 --- a/arch/x86/crypto/serpent-avx-x86_64-asm_64.S +++ b/arch/x86/crypto/serpent-avx-x86_64-asm_64.S @@ -24,6 +24,7 @@ */ #include +#include #include "glue_helper-asm-avx.S" .file "serpent-avx-x86_64-asm_64.S" @@ -618,6 +619,7 @@ __serpent_enc_blk8_avx: write_blocks(RA1, RB1, RC1, RD1, RK0, RK1, RK2); write_blocks(RA2, RB2, RC2, RD2, RK0, RK1, RK2); + pax_force_retaddr ret; ENDPROC(__serpent_enc_blk8_avx) @@ -672,6 +674,7 @@ __serpent_dec_blk8_avx: write_blocks(RC1, RD1, RB1, RE1, RK0, RK1, RK2); write_blocks(RC2, RD2, RB2, RE2, RK0, RK1, RK2); + pax_force_retaddr ret; ENDPROC(__serpent_dec_blk8_avx) @@ -688,6 +691,7 @@ ENTRY(serpent_ecb_enc_8way_avx) store_8way(%rsi, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(serpent_ecb_enc_8way_avx) @@ -704,6 +708,7 @@ ENTRY(serpent_ecb_dec_8way_avx) store_8way(%rsi, RC1, RD1, RB1, RE1, RC2, RD2, RB2, RE2); + pax_force_retaddr ret; ENDPROC(serpent_ecb_dec_8way_avx) @@ -720,6 +725,7 @@ ENTRY(serpent_cbc_dec_8way_avx) store_cbc_8way(%rdx, %rsi, RC1, RD1, RB1, RE1, RC2, RD2, RB2, RE2); + pax_force_retaddr ret; ENDPROC(serpent_cbc_dec_8way_avx) @@ -738,6 +744,7 @@ ENTRY(serpent_ctr_8way_avx) store_ctr_8way(%rdx, %rsi, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(serpent_ctr_8way_avx) @@ -758,6 +765,7 @@ ENTRY(serpent_xts_enc_8way_avx) /* dst <= regs xor IVs(in dst) */ store_xts_8way(%rsi, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(serpent_xts_enc_8way_avx) @@ -778,5 +786,6 @@ ENTRY(serpent_xts_dec_8way_avx) /* dst <= regs xor IVs(in dst) */ store_xts_8way(%rsi, RC1, RD1, RB1, RE1, RC2, RD2, RB2, RE2); + pax_force_retaddr ret; ENDPROC(serpent_xts_dec_8way_avx) diff --git a/arch/x86/crypto/serpent-avx2-asm_64.S b/arch/x86/crypto/serpent-avx2-asm_64.S index b222085..abd483c 100644 --- a/arch/x86/crypto/serpent-avx2-asm_64.S +++ b/arch/x86/crypto/serpent-avx2-asm_64.S @@ -15,6 +15,7 @@ */ #include +#include #include "glue_helper-asm-avx2.S" .file "serpent-avx2-asm_64.S" @@ -610,6 +611,7 @@ __serpent_enc_blk16: write_blocks(RA1, RB1, RC1, RD1, RK0, RK1, RK2); write_blocks(RA2, RB2, RC2, RD2, RK0, RK1, RK2); + pax_force_retaddr ret; ENDPROC(__serpent_enc_blk16) @@ -664,6 +666,7 @@ __serpent_dec_blk16: write_blocks(RC1, RD1, RB1, RE1, RK0, RK1, RK2); write_blocks(RC2, RD2, RB2, RE2, RK0, RK1, RK2); + pax_force_retaddr ret; ENDPROC(__serpent_dec_blk16) @@ -684,6 +687,7 @@ ENTRY(serpent_ecb_enc_16way) vzeroupper; + pax_force_retaddr ret; ENDPROC(serpent_ecb_enc_16way) @@ -704,6 +708,7 @@ ENTRY(serpent_ecb_dec_16way) vzeroupper; + pax_force_retaddr ret; ENDPROC(serpent_ecb_dec_16way) @@ -725,6 +730,7 @@ ENTRY(serpent_cbc_dec_16way) vzeroupper; + pax_force_retaddr ret; ENDPROC(serpent_cbc_dec_16way) @@ -748,6 +754,7 @@ ENTRY(serpent_ctr_16way) vzeroupper; + pax_force_retaddr ret; ENDPROC(serpent_ctr_16way) @@ -772,6 +779,7 @@ ENTRY(serpent_xts_enc_16way) vzeroupper; + pax_force_retaddr ret; ENDPROC(serpent_xts_enc_16way) @@ -796,5 +804,6 @@ ENTRY(serpent_xts_dec_16way) vzeroupper; + pax_force_retaddr ret; ENDPROC(serpent_xts_dec_16way) 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 acc066c..1559cc4 100644 --- a/arch/x86/crypto/serpent-sse2-x86_64-asm_64.S +++ b/arch/x86/crypto/serpent-sse2-x86_64-asm_64.S @@ -25,6 +25,7 @@ */ #include +#include .file "serpent-sse2-x86_64-asm_64.S" .text @@ -690,12 +691,14 @@ ENTRY(__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; .L__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; ENDPROC(__serpent_enc_blk_8way) @@ -750,5 +753,6 @@ ENTRY(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; ENDPROC(serpent_dec_blk_8way) diff --git a/arch/x86/crypto/sha1_ssse3_asm.S b/arch/x86/crypto/sha1_ssse3_asm.S index a410950..9dfe7ad 100644 --- a/arch/x86/crypto/sha1_ssse3_asm.S +++ b/arch/x86/crypto/sha1_ssse3_asm.S @@ -29,6 +29,7 @@ */ #include +#include #define CTX %rdi // arg1 #define BUF %rsi // arg2 @@ -75,9 +76,9 @@ push %rbx push %rbp - push %r12 + push %r14 - mov %rsp, %r12 + mov %rsp, %r14 sub $64, %rsp # allocate workspace and $~15, %rsp # align stack @@ -99,11 +100,12 @@ xor %rax, %rax rep stosq - mov %r12, %rsp # deallocate workspace + mov %r14, %rsp # deallocate workspace - pop %r12 + pop %r14 pop %rbp pop %rbx + pax_force_retaddr ret ENDPROC(\name) diff --git a/arch/x86/crypto/sha256-avx-asm.S b/arch/x86/crypto/sha256-avx-asm.S index 642f156..51a513c 100644 --- a/arch/x86/crypto/sha256-avx-asm.S +++ b/arch/x86/crypto/sha256-avx-asm.S @@ -49,6 +49,7 @@ #ifdef CONFIG_AS_AVX #include +#include ## assume buffers not aligned #define VMOVDQ vmovdqu @@ -460,6 +461,7 @@ done_hash: popq %r13 popq %rbp popq %rbx + pax_force_retaddr ret ENDPROC(sha256_transform_avx) diff --git a/arch/x86/crypto/sha256-avx2-asm.S b/arch/x86/crypto/sha256-avx2-asm.S index 9e86944..3795e6a 100644 --- a/arch/x86/crypto/sha256-avx2-asm.S +++ b/arch/x86/crypto/sha256-avx2-asm.S @@ -50,6 +50,7 @@ #ifdef CONFIG_AS_AVX2 #include +#include ## assume buffers not aligned #define VMOVDQ vmovdqu @@ -720,6 +721,7 @@ done_hash: popq %r12 popq %rbp popq %rbx + pax_force_retaddr ret ENDPROC(sha256_transform_rorx) diff --git a/arch/x86/crypto/sha256-ssse3-asm.S b/arch/x86/crypto/sha256-ssse3-asm.S index f833b74..8c62a9e 100644 --- a/arch/x86/crypto/sha256-ssse3-asm.S +++ b/arch/x86/crypto/sha256-ssse3-asm.S @@ -47,6 +47,7 @@ ######################################################################## #include +#include ## assume buffers not aligned #define MOVDQ movdqu @@ -471,6 +472,7 @@ done_hash: popq %rbp popq %rbx + pax_force_retaddr ret ENDPROC(sha256_transform_ssse3) diff --git a/arch/x86/crypto/sha512-avx-asm.S b/arch/x86/crypto/sha512-avx-asm.S index 974dde9..a823ff9 100644 --- a/arch/x86/crypto/sha512-avx-asm.S +++ b/arch/x86/crypto/sha512-avx-asm.S @@ -49,6 +49,7 @@ #ifdef CONFIG_AS_AVX #include +#include .text @@ -364,6 +365,7 @@ updateblock: mov frame_RSPSAVE(%rsp), %rsp nowork: + pax_force_retaddr ret ENDPROC(sha512_transform_avx) diff --git a/arch/x86/crypto/sha512-avx2-asm.S b/arch/x86/crypto/sha512-avx2-asm.S index 568b961..ed20c37 100644 --- a/arch/x86/crypto/sha512-avx2-asm.S +++ b/arch/x86/crypto/sha512-avx2-asm.S @@ -51,6 +51,7 @@ #ifdef CONFIG_AS_AVX2 #include +#include .text @@ -678,6 +679,7 @@ done_hash: # Restore Stack Pointer mov frame_RSPSAVE(%rsp), %rsp + pax_force_retaddr ret ENDPROC(sha512_transform_rorx) diff --git a/arch/x86/crypto/sha512-ssse3-asm.S b/arch/x86/crypto/sha512-ssse3-asm.S index fb56855..6edd768 100644 --- a/arch/x86/crypto/sha512-ssse3-asm.S +++ b/arch/x86/crypto/sha512-ssse3-asm.S @@ -48,6 +48,7 @@ ######################################################################## #include +#include .text @@ -363,6 +364,7 @@ updateblock: mov frame_RSPSAVE(%rsp), %rsp nowork: + pax_force_retaddr ret ENDPROC(sha512_transform_ssse3) diff --git a/arch/x86/crypto/twofish-avx-x86_64-asm_64.S b/arch/x86/crypto/twofish-avx-x86_64-asm_64.S index 0505813..b067311 100644 --- a/arch/x86/crypto/twofish-avx-x86_64-asm_64.S +++ b/arch/x86/crypto/twofish-avx-x86_64-asm_64.S @@ -24,6 +24,7 @@ */ #include +#include #include "glue_helper-asm-avx.S" .file "twofish-avx-x86_64-asm_64.S" @@ -284,6 +285,7 @@ __twofish_enc_blk8: outunpack_blocks(RC1, RD1, RA1, RB1, RK1, RX0, RY0, RK2); outunpack_blocks(RC2, RD2, RA2, RB2, RK1, RX0, RY0, RK2); + pax_force_retaddr ret; ENDPROC(__twofish_enc_blk8) @@ -324,6 +326,7 @@ __twofish_dec_blk8: outunpack_blocks(RA1, RB1, RC1, RD1, RK1, RX0, RY0, RK2); outunpack_blocks(RA2, RB2, RC2, RD2, RK1, RX0, RY0, RK2); + pax_force_retaddr ret; ENDPROC(__twofish_dec_blk8) @@ -342,6 +345,7 @@ ENTRY(twofish_ecb_enc_8way) store_8way(%r11, RC1, RD1, RA1, RB1, RC2, RD2, RA2, RB2); + pax_force_retaddr ret; ENDPROC(twofish_ecb_enc_8way) @@ -360,6 +364,7 @@ ENTRY(twofish_ecb_dec_8way) store_8way(%r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(twofish_ecb_dec_8way) @@ -370,19 +375,20 @@ ENTRY(twofish_cbc_dec_8way) * %rdx: src */ - pushq %r12; + pushq %r14; movq %rsi, %r11; - movq %rdx, %r12; + movq %rdx, %r14; load_8way(%rdx, RC1, RD1, RA1, RB1, RC2, RD2, RA2, RB2); call __twofish_dec_blk8; - store_cbc_8way(%r12, %r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + store_cbc_8way(%r14, %r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); - popq %r12; + popq %r14; + pax_force_retaddr ret; ENDPROC(twofish_cbc_dec_8way) @@ -394,20 +400,21 @@ ENTRY(twofish_ctr_8way) * %rcx: iv (little endian, 128bit) */ - pushq %r12; + pushq %r14; movq %rsi, %r11; - movq %rdx, %r12; + movq %rdx, %r14; load_ctr_8way(%rcx, .Lbswap128_mask, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2, RX0, RX1, RY0); call __twofish_enc_blk8; - store_ctr_8way(%r12, %r11, RC1, RD1, RA1, RB1, RC2, RD2, RA2, RB2); + store_ctr_8way(%r14, %r11, RC1, RD1, RA1, RB1, RC2, RD2, RA2, RB2); - popq %r12; + popq %r14; + pax_force_retaddr ret; ENDPROC(twofish_ctr_8way) @@ -430,6 +437,7 @@ ENTRY(twofish_xts_enc_8way) /* dst <= regs xor IVs(in dst) */ store_xts_8way(%r11, RC1, RD1, RA1, RB1, RC2, RD2, RA2, RB2); + pax_force_retaddr ret; ENDPROC(twofish_xts_enc_8way) @@ -452,5 +460,6 @@ ENTRY(twofish_xts_dec_8way) /* dst <= regs xor IVs(in dst) */ store_xts_8way(%r11, RA1, RB1, RC1, RD1, RA2, RB2, RC2, RD2); + pax_force_retaddr ret; ENDPROC(twofish_xts_dec_8way) 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 1c3b7ce..02f578d 100644 --- a/arch/x86/crypto/twofish-x86_64-asm_64-3way.S +++ b/arch/x86/crypto/twofish-x86_64-asm_64-3way.S @@ -21,6 +21,7 @@ */ #include +#include .file "twofish-x86_64-asm-3way.S" .text @@ -258,6 +259,7 @@ ENTRY(__twofish_enc_blk_3way) popq %r13; popq %r14; popq %r15; + pax_force_retaddr ret; .L__enc_xor3: @@ -269,6 +271,7 @@ ENTRY(__twofish_enc_blk_3way) popq %r13; popq %r14; popq %r15; + pax_force_retaddr ret; ENDPROC(__twofish_enc_blk_3way) @@ -308,5 +311,6 @@ ENTRY(twofish_dec_blk_3way) popq %r13; popq %r14; popq %r15; + pax_force_retaddr ret; ENDPROC(twofish_dec_blk_3way) diff --git a/arch/x86/crypto/twofish-x86_64-asm_64.S b/arch/x86/crypto/twofish-x86_64-asm_64.S index a039d21..524b8b2 100644 --- a/arch/x86/crypto/twofish-x86_64-asm_64.S +++ b/arch/x86/crypto/twofish-x86_64-asm_64.S @@ -22,6 +22,7 @@ #include #include +#include #define a_offset 0 #define b_offset 4 @@ -265,6 +266,7 @@ ENTRY(twofish_enc_blk) popq R1 movq $1,%rax + pax_force_retaddr ret ENDPROC(twofish_enc_blk) @@ -317,5 +319,6 @@ ENTRY(twofish_dec_blk) popq R1 movq $1,%rax + pax_force_retaddr ret ENDPROC(twofish_dec_blk) diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c index d21ff89..6da8e6e 100644 --- a/arch/x86/ia32/ia32_aout.c +++ b/arch/x86/ia32/ia32_aout.c @@ -153,6 +153,8 @@ static int aout_core_dump(struct coredump_params *cprm) 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 f9e181a..300544c 100644 --- a/arch/x86/ia32/ia32_signal.c +++ b/arch/x86/ia32/ia32_signal.c @@ -218,7 +218,7 @@ asmlinkage long sys32_sigreturn(void) if (__get_user(set.sig[0], &frame->sc.oldmask) || (_COMPAT_NSIG_WORDS > 1 && __copy_from_user((((char *) &set.sig) + 4), - &frame->extramask, + frame->extramask, sizeof(frame->extramask)))) goto badframe; @@ -338,7 +338,7 @@ static void __user *get_sigframe(struct ksignal *ksig, 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; } @@ -383,10 +383,10 @@ int ia32_setup_frame(int sig, struct ksignal *ksig, } else { /* Return stub is in 32bit vsyscall page */ if (current->mm->context.vdso) - restorer = current->mm->context.vdso + - selected_vdso32->sym___kernel_sigreturn; + restorer = (void __force_user *)(current->mm->context.vdso + + selected_vdso32->sym___kernel_sigreturn); else - restorer = &frame->retcode; + restorer = frame->retcode; } put_user_try { @@ -396,7 +396,7 @@ int ia32_setup_frame(int sig, struct ksignal *ksig, * These are actually not used anymore, but left because some * gdb versions depend on them as a marker. */ - put_user_ex(*((u64 *)&code), (u64 __user *)frame->retcode); + put_user_ex(*((const u64 *)&code), (u64 __user *)frame->retcode); } put_user_catch(err); if (err) @@ -438,7 +438,7 @@ int ia32_setup_rt_frame(int sig, struct ksignal *ksig, 0xb8, __NR_ia32_rt_sigreturn, 0x80cd, - 0, + 0 }; frame = get_sigframe(ksig, regs, sizeof(*frame), &fpstate); @@ -461,16 +461,19 @@ int ia32_setup_rt_frame(int sig, struct ksignal *ksig, if (ksig->ka.sa.sa_flags & SA_RESTORER) restorer = ksig->ka.sa.sa_restorer; + else if (current->mm->context.vdso) + /* Return stub is in 32bit vsyscall page */ + restorer = (void __force_user *)(current->mm->context.vdso + + selected_vdso32->sym___kernel_rt_sigreturn); else - restorer = current->mm->context.vdso + - selected_vdso32->sym___kernel_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 __user *)frame->retcode); + put_user_ex(*((const u64 *)&code), (u64 __user *)frame->retcode); } put_user_catch(err); err |= copy_siginfo_to_user32(&frame->info, &ksig->info); diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index 4299eb0..fefe70e 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -15,8 +15,10 @@ #include #include #include +#include #include #include +#include /* Avoid __ASSEMBLER__'ifying just for this. */ #include @@ -62,12 +64,12 @@ */ .macro LOAD_ARGS32 offset, _r9=0 .if \_r9 - movl \offset+16(%rsp),%r9d + movl \offset+R9(%rsp),%r9d .endif - movl \offset+40(%rsp),%ecx - movl \offset+48(%rsp),%edx - movl \offset+56(%rsp),%esi - movl \offset+64(%rsp),%edi + movl \offset+RCX(%rsp),%ecx + movl \offset+RDX(%rsp),%edx + movl \offset+RSI(%rsp),%esi + movl \offset+RDI(%rsp),%edi movl %eax,%eax /* zero extension */ .endm @@ -96,6 +98,32 @@ 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. * @@ -122,12 +150,6 @@ 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*/ @@ -135,24 +157,49 @@ 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 + +#ifdef CONFIG_PAX_RANDKSTACK + pax_erase_kstack +#endif + + /* + * 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 + addq pax_user_shadow_base,%rbp + ASM_PAX_OPEN_USERLAND +#endif + ASM_STAC 1: movl (%rbp),%ebp _ASM_EXTABLE(1b,ia32_badarg) ASM_CLAC - orl $TS_COMPAT,TI_status+THREAD_INFO(%rsp,RIP-ARGOFFSET) - testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + +#ifdef CONFIG_PAX_MEMORY_UDEREF + ASM_PAX_CLOSE_USERLAND +#endif + + 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 @@ -162,15 +209,18 @@ 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 */ + andl $~X86_EFLAGS_IF,EFLAGS(%rsp) + movl RIP(%rsp),%edx /* User %eip */ CFI_REGISTER rip,rdx RESTORE_ARGS 0,24,0,0,0,0 xorq %r8,%r8 @@ -193,6 +243,9 @@ 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 @@ -204,7 +257,7 @@ 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 ENABLE_INTERRUPTS(CLBR_NONE) @@ -215,11 +268,12 @@ 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 DISABLE_INTERRUPTS(CLBR_NONE) 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 @@ -237,7 +291,7 @@ 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 @@ -249,6 +303,9 @@ sysenter_tracesys: RESTORE_REST cmpq $(IA32_NR_syscalls-1),%rax ja int_ret_from_sys_call /* sysenter_tracesys has set RAX(%rsp) */ + + pax_erase_kstack + jmp sysenter_do_call CFI_ENDPROC ENDPROC(ia32_sysenter_target) @@ -276,19 +333,25 @@ 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 + +#ifdef CONFIG_PAX_RANDKSTACK + pax_erase_kstack +#endif + /* * 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) @@ -304,12 +367,25 @@ 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 + ASM_PAX_OPEN_USERLAND + movq pax_user_shadow_base,%r8 + addq RSP-ARGOFFSET(%rsp),%r8 +#endif + ASM_STAC 1: movl (%r8),%r9d _ASM_EXTABLE(1b,ia32_badarg) ASM_CLAC - orl $TS_COMPAT,TI_status+THREAD_INFO(%rsp,RIP-ARGOFFSET) - testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + +#ifdef CONFIG_PAX_MEMORY_UDEREF + ASM_PAX_CLOSE_USERLAND +#endif + + 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 @@ -319,13 +395,16 @@ 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) - RESTORE_ARGS 0,-ARG_SKIP,0,0,0 + pax_exit_kernel_user + pax_erase_kstack + andl $~TS_COMPAT,TI_status(%r11) + RESTORE_ARGS 0,-ORIG_RAX,0,0,0 movl RIP-ARGOFFSET(%rsp),%ecx CFI_REGISTER rip,rcx movl EFLAGS-ARGOFFSET(%rsp),%r11d @@ -352,7 +431,7 @@ 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 @@ -366,11 +445,19 @@ cstar_tracesys: xchgl %ebp,%r9d cmpq $(IA32_NR_syscalls-1),%rax ja int_ret_from_sys_call /* cstar_tracesys has set RAX(%rsp) */ + + pax_erase_kstack + jmp cstar_do_call END(ia32_cstar_target) ia32_badarg: ASM_CLAC + +#ifdef CONFIG_PAX_MEMORY_UDEREF + ASM_PAX_CLOSE_USERLAND +#endif + movq $-EFAULT,%rax jmp ia32_sysret CFI_ENDPROC @@ -407,19 +494,26 @@ 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 + +#ifdef CONFIG_PAX_RANDKSTACK + pax_erase_kstack +#endif + + /* + * 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 @@ -442,6 +536,9 @@ ia32_tracesys: RESTORE_REST cmpq $(IA32_NR_syscalls-1),%rax ja int_ret_from_sys_call /* ia32_tracesys has set RAX(%rsp) */ + + pax_erase_kstack + jmp ia32_do_call END(ia32_syscall) diff --git a/arch/x86/ia32/sys_ia32.c b/arch/x86/ia32/sys_ia32.c index 8e0ceec..af13504 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, from_kuid_munged(current_user_ns(), stat->uid)); SET_GID(gid, from_kgid_munged(current_user_ns(), stat->gid)); if (!access_ok(VERIFY_WRITE, ubuf, sizeof(struct stat64)) || diff --git a/arch/x86/include/asm/alternative-asm.h b/arch/x86/include/asm/alternative-asm.h index 372231c..51b537d 100644 --- a/arch/x86/include/asm/alternative-asm.h +++ b/arch/x86/include/asm/alternative-asm.h @@ -18,6 +18,45 @@ .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 %r12,\rip(%rsp) + .endm + .macro pax_force_fptr ptr + orq %r12,\ptr + .endm + .macro pax_set_fptr_mask + movabs $0x8000000000000000,%r12 + .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 0a3f9c9..c9d081d 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -106,7 +106,7 @@ static inline int alternatives_text_reserved(void *start, void *end) ".pushsection .discard,\"aw\",@progbits\n" \ DISCARD_ENTRY(1) \ ".popsection\n" \ - ".pushsection .altinstr_replacement, \"ax\"\n" \ + ".pushsection .altinstr_replacement, \"a\"\n" \ ALTINSTR_REPLACEMENT(newinstr, feature, 1) \ ".popsection" @@ -120,7 +120,7 @@ static inline int alternatives_text_reserved(void *start, void *end) DISCARD_ENTRY(1) \ DISCARD_ENTRY(2) \ ".popsection\n" \ - ".pushsection .altinstr_replacement, \"ax\"\n" \ + ".pushsection .altinstr_replacement, \"a\"\n" \ ALTINSTR_REPLACEMENT(newinstr1, feature1, 1) \ ALTINSTR_REPLACEMENT(newinstr2, feature2, 2) \ ".popsection" diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index 19b0eba..12254cd 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 6dd1c7dd..2edd216 100644 --- a/arch/x86/include/asm/atomic.h +++ b/arch/x86/include/asm/atomic.h @@ -24,7 +24,18 @@ */ 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 __intentional_overflow(-1) atomic_read_unchecked(const atomic_unchecked_t *v) +{ + return (*(volatile const int *)&(v)->counter); } /** @@ -40,6 +51,18 @@ 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 @@ -48,7 +71,29 @@ 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)); } @@ -62,7 +107,29 @@ 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 +145,7 @@ static inline void atomic_sub(int i, atomic_t *v) */ static inline int atomic_sub_and_test(int i, atomic_t *v) { - GEN_BINARY_RMWcc(LOCK_PREFIX "subl", v->counter, "er", i, "%0", "e"); + GEN_BINARY_RMWcc(LOCK_PREFIX "subl", LOCK_PREFIX "addl", v->counter, "er", i, "%0", "e"); } /** @@ -89,7 +156,27 @@ 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)); } @@ -101,7 +188,27 @@ 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)); } @@ -115,7 +222,7 @@ static inline void atomic_dec(atomic_t *v) */ static inline int atomic_dec_and_test(atomic_t *v) { - GEN_UNARY_RMWcc(LOCK_PREFIX "decl", v->counter, "%0", "e"); + GEN_UNARY_RMWcc(LOCK_PREFIX "decl", LOCK_PREFIX "incl", v->counter, "%0", "e"); } /** @@ -128,7 +235,20 @@ static inline int atomic_dec_and_test(atomic_t *v) */ static inline int atomic_inc_and_test(atomic_t *v) { - GEN_UNARY_RMWcc(LOCK_PREFIX "incl", v->counter, "%0", "e"); + GEN_UNARY_RMWcc(LOCK_PREFIX "incl", LOCK_PREFIX "decl", v->counter, "%0", "e"); +} + +/** + * 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) +{ + GEN_UNARY_RMWcc_unchecked(LOCK_PREFIX "incl", v->counter, "%0", "e"); } /** @@ -142,7 +262,7 @@ static inline int atomic_inc_and_test(atomic_t *v) */ static inline int atomic_add_negative(int i, atomic_t *v) { - GEN_BINARY_RMWcc(LOCK_PREFIX "addl", v->counter, "er", i, "%0", "s"); + GEN_BINARY_RMWcc(LOCK_PREFIX "addl", LOCK_PREFIX "subl", v->counter, "er", i, "%0", "s"); } /** @@ -154,6 +274,18 @@ static inline int atomic_add_negative(int i, atomic_t *v) */ static inline int atomic_add_return(int i, atomic_t *v) { + return i + xadd_check_overflow(&v->counter, i); +} + +/** + * 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) +{ return i + xadd(&v->counter, i); } @@ -170,9 +302,18 @@ 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) +static inline int __intentional_overflow(-1) 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); } @@ -182,6 +323,11 @@ 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 @@ -191,14 +337,27 @@ static inline int atomic_xchg(atomic_t *v, int new) * Atomically adds @a to @v, so long as @v was not already @u. * Returns the old value of @v. */ -static inline int __atomic_add_unless(atomic_t *v, int a, int u) +static inline int __intentional_overflow(-1) __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; @@ -207,6 +366,49 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) } /** + * 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_inc_short - increment of a short integer * @v: pointer to type int * @@ -235,14 +437,37 @@ static inline void atomic_or_long(unsigned long *v1, unsigned long v2) #endif /* These are x86-specific, used by some header files */ -#define atomic_clear_mask(mask, addr) \ - asm volatile(LOCK_PREFIX "andl %0,%1" \ - : : "r" (~(mask)), "m" (*(addr)) : "memory") +static inline void atomic_clear_mask(unsigned int mask, atomic_t *v) +{ + asm volatile(LOCK_PREFIX "andl %1,%0" + : "+m" (v->counter) + : "r" (~(mask)) + : "memory"); +} -#define atomic_set_mask(mask, addr) \ - asm volatile(LOCK_PREFIX "orl %0,%1" \ - : : "r" ((unsigned)(mask)), "m" (*(addr)) \ - : "memory") +static inline void atomic_clear_mask_unchecked(unsigned int mask, atomic_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "andl %1,%0" + : "+m" (v->counter) + : "r" (~(mask)) + : "memory"); +} + +static inline void atomic_set_mask(unsigned int mask, atomic_t *v) +{ + asm volatile(LOCK_PREFIX "orl %1,%0" + : "+m" (v->counter) + : "r" (mask) + : "memory"); +} + +static inline void atomic_set_mask_unchecked(unsigned int mask, atomic_unchecked_t *v) +{ + asm volatile(LOCK_PREFIX "orl %1,%0" + : "+m" (v->counter) + : "r" (mask) + : "memory"); +} #ifdef CONFIG_X86_32 # include diff --git a/arch/x86/include/asm/atomic64_32.h b/arch/x86/include/asm/atomic64_32.h index b154de7..bf18a5a 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) } #define __ATOMIC64_DECL(sym) void atomic64_##sym(atomic64_t *, ...) @@ -37,21 +45,31 @@ typedef struct { ATOMIC64_DECL_ONE(sym##_386) ATOMIC64_DECL_ONE(add_386); +ATOMIC64_DECL_ONE(add_unchecked_386); ATOMIC64_DECL_ONE(sub_386); +ATOMIC64_DECL_ONE(sub_unchecked_386); ATOMIC64_DECL_ONE(inc_386); +ATOMIC64_DECL_ONE(inc_unchecked_386); ATOMIC64_DECL_ONE(dec_386); +ATOMIC64_DECL_ONE(dec_unchecked_386); #endif #define alternative_atomic64(f, out, in...) \ __alternative_atomic64(f, f, ASM_OUTPUT2(out), ## in) ATOMIC64_DECL(read); +ATOMIC64_DECL(read_unchecked); ATOMIC64_DECL(set); +ATOMIC64_DECL(set_unchecked); ATOMIC64_DECL(xchg); ATOMIC64_DECL(add_return); +ATOMIC64_DECL(add_return_unchecked); ATOMIC64_DECL(sub_return); +ATOMIC64_DECL(sub_return_unchecked); ATOMIC64_DECL(inc_return); +ATOMIC64_DECL(inc_return_unchecked); ATOMIC64_DECL(dec_return); +ATOMIC64_DECL(dec_return_unchecked); ATOMIC64_DECL(dec_if_positive); ATOMIC64_DECL(inc_not_zero); ATOMIC64_DECL(add_unless); @@ -77,6 +95,21 @@ 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 @@ -112,6 +145,22 @@ 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; + alternative_atomic64(set, /* no output */, + "S" (v), "b" (low), "c" (high) + : "eax", "edx", "memory"); +} + +/** * atomic64_read - read atomic64 variable * @v: pointer to type atomic64_t * @@ -125,6 +174,19 @@ 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 __intentional_overflow(-1) atomic64_read_unchecked(atomic64_unchecked_t *v) +{ + long long r; + alternative_atomic64(read, "=&A" (r), "c" (v) : "memory"); + return r; + } + +/** * atomic64_add_return - add and return * @i: integer value to add * @v: pointer to type atomic64_t @@ -139,6 +201,21 @@ 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) +{ + alternative_atomic64(add_return_unchecked, + ASM_OUTPUT2("+A" (i), "+c" (v)), + ASM_NO_INPUT_CLOBBER("memory")); + return i; +} + /* * Other variants with different arithmetic operators: */ @@ -158,6 +235,14 @@ 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; + alternative_atomic64(inc_return_unchecked, "=&A" (a), + "S" (v) : "memory", "ecx"); + return a; +} + static inline long long atomic64_dec_return(atomic64_t *v) { long long a; @@ -182,6 +267,21 @@ 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) +{ + __alternative_atomic64(add_unchecked, add_return_unchecked, + ASM_OUTPUT2("+A" (i), "+c" (v)), + ASM_NO_INPUT_CLOBBER("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 46e9052..ae45136 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 __intentional_overflow(-1) 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)); } @@ -72,7 +140,7 @@ static inline void atomic64_sub(long i, atomic64_t *v) */ static inline int atomic64_sub_and_test(long i, atomic64_t *v) { - GEN_BINARY_RMWcc(LOCK_PREFIX "subq", v->counter, "er", i, "%0", "e"); + GEN_BINARY_RMWcc(LOCK_PREFIX "subq", LOCK_PREFIX "addq", v->counter, "er", i, "%0", "e"); } /** @@ -83,6 +151,27 @@ 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)); @@ -96,7 +185,28 @@ 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)); } @@ -111,7 +221,7 @@ static inline void atomic64_dec(atomic64_t *v) */ static inline int atomic64_dec_and_test(atomic64_t *v) { - GEN_UNARY_RMWcc(LOCK_PREFIX "decq", v->counter, "%0", "e"); + GEN_UNARY_RMWcc(LOCK_PREFIX "decq", LOCK_PREFIX "incq", v->counter, "%0", "e"); } /** @@ -124,7 +234,7 @@ static inline int atomic64_dec_and_test(atomic64_t *v) */ static inline int atomic64_inc_and_test(atomic64_t *v) { - GEN_UNARY_RMWcc(LOCK_PREFIX "incq", v->counter, "%0", "e"); + GEN_UNARY_RMWcc(LOCK_PREFIX "incq", LOCK_PREFIX "decq", v->counter, "%0", "e"); } /** @@ -138,7 +248,7 @@ static inline int atomic64_inc_and_test(atomic64_t *v) */ static inline int atomic64_add_negative(long i, atomic64_t *v) { - GEN_BINARY_RMWcc(LOCK_PREFIX "addq", v->counter, "er", i, "%0", "s"); + GEN_BINARY_RMWcc(LOCK_PREFIX "addq", LOCK_PREFIX "subq", v->counter, "er", i, "%0", "s"); } /** @@ -150,6 +260,18 @@ 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); } @@ -159,6 +281,10 @@ 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) @@ -166,6 +292,11 @@ 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); @@ -182,17 +313,30 @@ 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/barrier.h b/arch/x86/include/asm/barrier.h index 5c7198c..44180b5 100644 --- a/arch/x86/include/asm/barrier.h +++ b/arch/x86/include/asm/barrier.h @@ -107,7 +107,7 @@ do { \ compiletime_assert_atomic_type(*p); \ smp_mb(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ @@ -124,7 +124,7 @@ do { \ do { \ compiletime_assert_atomic_type(*p); \ barrier(); \ - ACCESS_ONCE(*p) = (v); \ + ACCESS_ONCE_RW(*p) = (v); \ } while (0) #define smp_load_acquire(p) \ diff --git a/arch/x86/include/asm/bitops.h b/arch/x86/include/asm/bitops.h index afcd35d..d01b118 100644 --- a/arch/x86/include/asm/bitops.h +++ b/arch/x86/include/asm/bitops.h @@ -50,7 +50,7 @@ * 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)) /** @@ -203,7 +203,7 @@ static inline void change_bit(long nr, volatile unsigned long *addr) */ static inline int test_and_set_bit(long nr, volatile unsigned long *addr) { - GEN_BINARY_RMWcc(LOCK_PREFIX "bts", *addr, "Ir", nr, "%0", "c"); + GEN_BINARY_RMWcc_unchecked(LOCK_PREFIX "bts", *addr, "Ir", nr, "%0", "c"); } /** @@ -249,7 +249,7 @@ static inline int __test_and_set_bit(long nr, volatile unsigned long *addr) */ static inline int test_and_clear_bit(long nr, volatile unsigned long *addr) { - GEN_BINARY_RMWcc(LOCK_PREFIX "btr", *addr, "Ir", nr, "%0", "c"); + GEN_BINARY_RMWcc_unchecked(LOCK_PREFIX "btr", *addr, "Ir", nr, "%0", "c"); } /** @@ -302,7 +302,7 @@ static inline int __test_and_change_bit(long nr, volatile unsigned long *addr) */ static inline int test_and_change_bit(long nr, volatile unsigned long *addr) { - GEN_BINARY_RMWcc(LOCK_PREFIX "btc", *addr, "Ir", nr, "%0", "c"); + GEN_BINARY_RMWcc_unchecked(LOCK_PREFIX "btc", *addr, "Ir", nr, "%0", "c"); } static __always_inline int constant_test_bit(long nr, const volatile unsigned long *addr) @@ -343,7 +343,7 @@ static int test_bit(int nr, const volatile unsigned long *addr); * * Undefined if no bit exists, so code should check against 0 first. */ -static inline unsigned long __ffs(unsigned long word) +static inline unsigned long __intentional_overflow(-1) __ffs(unsigned long word) { asm("rep; bsf %1,%0" : "=r" (word) @@ -357,7 +357,7 @@ static inline unsigned long __ffs(unsigned long word) * * Undefined if no zero exists, so code should check against ~0UL first. */ -static inline unsigned long ffz(unsigned long word) +static inline unsigned long __intentional_overflow(-1) ffz(unsigned long word) { asm("rep; bsf %1,%0" : "=r" (word) @@ -371,7 +371,7 @@ static inline unsigned long ffz(unsigned long word) * * Undefined if no set bit exists, so code should check against 0 first. */ -static inline unsigned long __fls(unsigned long word) +static inline unsigned long __intentional_overflow(-1) __fls(unsigned long word) { asm("bsr %1,%0" : "=r" (word) @@ -434,7 +434,7 @@ static inline int ffs(int x) * set bit if value is nonzero. The last (most significant) bit is * at position 32. */ -static inline int fls(int x) +static inline int __intentional_overflow(-1) fls(int x) { int r; @@ -476,7 +476,7 @@ static inline int fls(int x) * at position 64. */ #ifdef CONFIG_X86_64 -static __always_inline int fls64(__u64 x) +static __always_inline __intentional_overflow(-1) int fls64(__u64 x) { int bitpos = -1; /* @@ -497,8 +497,6 @@ static __always_inline int fls64(__u64 x) #include -#define ARCH_HAS_FAST_MULTIPLIER 1 - #include #include diff --git a/arch/x86/include/asm/boot.h b/arch/x86/include/asm/boot.h index 4fa687a..60f2d39 100644 --- a/arch/x86/include/asm/boot.h +++ b/arch/x86/include/asm/boot.h @@ -6,10 +6,15 @@ #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 9863ee3..4a1f8e1 100644 --- a/arch/x86/include/asm/cacheflush.h +++ b/arch/x86/include/asm/cacheflush.h @@ -27,7 +27,7 @@ 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/calling.h b/arch/x86/include/asm/calling.h index cb4c73b..c473c29 100644 --- a/arch/x86/include/asm/calling.h +++ b/arch/x86/include/asm/calling.h @@ -82,103 +82,113 @@ For 32-bit we have the following conventions - kernel is built with #define RSP 152 #define SS 160 -#define ARGOFFSET R11 -#define SWFRAME ORIG_RAX +#define ARGOFFSET R15 .macro SAVE_ARGS addskip=0, save_rcx=1, save_r891011=1 - subq $9*8+\addskip, %rsp - CFI_ADJUST_CFA_OFFSET 9*8+\addskip - movq_cfi rdi, 8*8 - movq_cfi rsi, 7*8 - movq_cfi rdx, 6*8 + subq $ORIG_RAX-ARGOFFSET+\addskip, %rsp + CFI_ADJUST_CFA_OFFSET ORIG_RAX-ARGOFFSET+\addskip + movq_cfi rdi, RDI + movq_cfi rsi, RSI + movq_cfi rdx, RDX .if \save_rcx - movq_cfi rcx, 5*8 + movq_cfi rcx, RCX .endif - movq_cfi rax, 4*8 + movq_cfi rax, RAX .if \save_r891011 - movq_cfi r8, 3*8 - movq_cfi r9, 2*8 - movq_cfi r10, 1*8 - movq_cfi r11, 0*8 + movq_cfi r8, R8 + movq_cfi r9, R9 + movq_cfi r10, R10 + movq_cfi r11, R11 .endif +#ifdef CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_OR + movq_cfi r12, R12 +#endif + .endm -#define ARG_SKIP (9*8) +#define ARG_SKIP ORIG_RAX .macro RESTORE_ARGS rstor_rax=1, addskip=0, rstor_rcx=1, rstor_r11=1, \ rstor_r8910=1, rstor_rdx=1 + +#ifdef CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_OR + movq_cfi_restore R12, r12 +#endif + .if \rstor_r11 - movq_cfi_restore 0*8, r11 + movq_cfi_restore R11, r11 .endif .if \rstor_r8910 - movq_cfi_restore 1*8, r10 - movq_cfi_restore 2*8, r9 - movq_cfi_restore 3*8, r8 + movq_cfi_restore R10, r10 + movq_cfi_restore R9, r9 + movq_cfi_restore R8, r8 .endif .if \rstor_rax - movq_cfi_restore 4*8, rax + movq_cfi_restore RAX, rax .endif .if \rstor_rcx - movq_cfi_restore 5*8, rcx + movq_cfi_restore RCX, rcx .endif .if \rstor_rdx - movq_cfi_restore 6*8, rdx + movq_cfi_restore RDX, rdx .endif - movq_cfi_restore 7*8, rsi - movq_cfi_restore 8*8, rdi + movq_cfi_restore RSI, rsi + movq_cfi_restore RDI, rdi - .if ARG_SKIP+\addskip > 0 - addq $ARG_SKIP+\addskip, %rsp - CFI_ADJUST_CFA_OFFSET -(ARG_SKIP+\addskip) + .if ORIG_RAX+\addskip > 0 + addq $ORIG_RAX+\addskip, %rsp + CFI_ADJUST_CFA_OFFSET -(ORIG_RAX+\addskip) .endif .endm - .macro LOAD_ARGS offset, skiprax=0 - movq \offset(%rsp), %r11 - movq \offset+8(%rsp), %r10 - movq \offset+16(%rsp), %r9 - movq \offset+24(%rsp), %r8 - movq \offset+40(%rsp), %rcx - movq \offset+48(%rsp), %rdx - movq \offset+56(%rsp), %rsi - movq \offset+64(%rsp), %rdi + .macro LOAD_ARGS skiprax=0 + movq R11(%rsp), %r11 + movq R10(%rsp), %r10 + movq R9(%rsp), %r9 + movq R8(%rsp), %r8 + movq RCX(%rsp), %rcx + movq RDX(%rsp), %rdx + movq RSI(%rsp), %rsi + movq RDI(%rsp), %rdi .if \skiprax .else - movq \offset+72(%rsp), %rax + movq RAX(%rsp), %rax .endif .endm -#define REST_SKIP (6*8) - .macro SAVE_REST - subq $REST_SKIP, %rsp - CFI_ADJUST_CFA_OFFSET REST_SKIP - movq_cfi rbx, 5*8 - movq_cfi rbp, 4*8 - movq_cfi r12, 3*8 - movq_cfi r13, 2*8 - movq_cfi r14, 1*8 - movq_cfi r15, 0*8 + movq_cfi rbx, RBX + movq_cfi rbp, RBP + +#ifndef CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_OR + movq_cfi r12, R12 +#endif + + movq_cfi r13, R13 + movq_cfi r14, R14 + movq_cfi r15, R15 .endm .macro RESTORE_REST - movq_cfi_restore 0*8, r15 - movq_cfi_restore 1*8, r14 - movq_cfi_restore 2*8, r13 - movq_cfi_restore 3*8, r12 - movq_cfi_restore 4*8, rbp - movq_cfi_restore 5*8, rbx - addq $REST_SKIP, %rsp - CFI_ADJUST_CFA_OFFSET -(REST_SKIP) + movq_cfi_restore R15, r15 + movq_cfi_restore R14, r14 + movq_cfi_restore R13, r13 + +#ifndef CONFIG_PAX_KERNEXEC_PLUGIN_METHOD_OR + movq_cfi_restore R12, r12 +#endif + + movq_cfi_restore RBP, rbp + movq_cfi_restore RBX, rbx .endm .macro SAVE_ALL diff --git a/arch/x86/include/asm/checksum_32.h b/arch/x86/include/asm/checksum_32.h index f50de69..2b0a458 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. @@ -53,7 +61,7 @@ static inline __wsum csum_partial_copy_from_user(const void __user *src, might_sleep(); stac(); - ret = csum_partial_copy_generic((__force void *)src, dst, + ret = csum_partial_copy_generic_from_user((__force void *)src, dst, len, sum, err_ptr, NULL); clac(); @@ -187,7 +195,7 @@ static inline __wsum csum_and_copy_to_user(const void *src, might_sleep(); if (access_ok(VERIFY_WRITE, dst, len)) { stac(); - ret = csum_partial_copy_generic(src, (__force void *)dst, + ret = csum_partial_copy_generic_to_user(src, (__force void *)dst, len, sum, NULL, err_ptr); clac(); return ret; diff --git a/arch/x86/include/asm/cmpxchg.h b/arch/x86/include/asm/cmpxchg.h index d47786a..2d8883e 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 +71,38 @@ extern void __add_wrong_size(void) __ret; \ }) +#ifdef CONFIG_PAX_REFCOUNT +#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; \ + }) +#else +#define __xchg_op_check_overflow(ptr, arg, op, lock) __xchg_op(ptr, arg, op, lock) +#endif + /* * 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 +203,9 @@ 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/compat.h b/arch/x86/include/asm/compat.h index 59c6c40..5e0b22c 100644 --- a/arch/x86/include/asm/compat.h +++ b/arch/x86/include/asm/compat.h @@ -41,7 +41,7 @@ typedef s64 __attribute__((aligned(4))) compat_s64; typedef u32 compat_uint_t; typedef u32 compat_ulong_t; typedef u64 __attribute__((aligned(4))) compat_u64; -typedef u32 compat_uptr_t; +typedef u32 __user compat_uptr_t; struct compat_timespec { compat_time_t tv_sec; diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index e265ff9..72c253b 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -203,7 +203,7 @@ #define X86_FEATURE_DECODEASSISTS (8*32+12) /* AMD Decode Assists support */ #define X86_FEATURE_PAUSEFILTER (8*32+13) /* AMD filtered pause intercept */ #define X86_FEATURE_PFTHRESHOLD (8*32+14) /* AMD pause filter threshold */ - +#define X86_FEATURE_STRONGUDEREF (8*32+31) /* PaX PCID based strong UDEREF */ /* Intel-defined CPU features, CPUID level 0x00000007:0 (ebx), word 9 */ #define X86_FEATURE_FSGSBASE (9*32+ 0) /* {RD/WR}{FS/GS}BASE instructions*/ @@ -211,7 +211,7 @@ #define X86_FEATURE_BMI1 (9*32+ 3) /* 1st group bit manipulation extensions */ #define X86_FEATURE_HLE (9*32+ 4) /* Hardware Lock Elision */ #define X86_FEATURE_AVX2 (9*32+ 5) /* AVX2 instructions */ -#define X86_FEATURE_SMEP (9*32+ 7) /* Supervisor Mode Execution Protection */ +#define X86_FEATURE_SMEP (9*32+ 7) /* Supervisor Mode Execution Prevention */ #define X86_FEATURE_BMI2 (9*32+ 8) /* 2nd group bit manipulation extensions */ #define X86_FEATURE_ERMS (9*32+ 9) /* Enhanced REP MOVSB/STOSB */ #define X86_FEATURE_INVPCID (9*32+10) /* Invalidate Processor Context ID */ @@ -359,6 +359,7 @@ extern const char * const x86_power_flags[32]; #undef cpu_has_centaur_mcr #define cpu_has_centaur_mcr 0 +#define cpu_has_pcid boot_cpu_has(X86_FEATURE_PCID) #endif /* CONFIG_X86_64 */ #if __GNUC__ >= 4 @@ -411,7 +412,8 @@ static __always_inline __pure bool __static_cpu_has(u16 bit) #ifdef CONFIG_X86_DEBUG_STATIC_CPU_HAS t_warn: - warn_pre_alternatives(); + if (bit != X86_FEATURE_PCID && bit != X86_FEATURE_INVPCID) + warn_pre_alternatives(); return false; #endif @@ -431,7 +433,7 @@ 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" @@ -468,7 +470,7 @@ static __always_inline __pure bool _static_cpu_has_safe(u16 bit) " .byte 2b - 1b\n" /* src len */ " .byte 4f - 3f\n" /* repl len */ ".previous\n" - ".section .altinstr_replacement,\"ax\"\n" + ".section .altinstr_replacement,\"a\"\n" "3: .byte 0xe9\n .long %l[t_no] - 2b\n" "4:\n" ".previous\n" @@ -501,7 +503,7 @@ static __always_inline __pure bool _static_cpu_has_safe(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 $0,%0\n" "4:\n" ".previous\n" @@ -515,7 +517,7 @@ static __always_inline __pure bool _static_cpu_has_safe(u16 bit) ".section .discard,\"aw\",@progbits\n" " .byte 0xff + (6f-5f) - (4b-3b)\n" /* size check */ ".previous\n" - ".section .altinstr_replacement,\"ax\"\n" + ".section .altinstr_replacement,\"a\"\n" "5: movb $1,%0\n" "6:\n" ".previous\n" diff --git a/arch/x86/include/asm/desc.h b/arch/x86/include/asm/desc.h index 50d033a..37deb26 100644 --- a/arch/x86/include/asm/desc.h +++ b/arch/x86/include/asm/desc.h @@ -4,6 +4,7 @@ #include #include #include +#include #include #include @@ -17,6 +18,7 @@ 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; @@ -35,19 +37,14 @@ 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 debug_idt_descr; -extern gate_desc debug_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[IDT_ENTRIES]; +extern const struct desc_ptr debug_idt_descr; +extern gate_desc debug_idt_table[IDT_ENTRIES]; +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 @@ -72,8 +69,14 @@ 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 @@ -118,12 +121,16 @@ 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 @@ -137,7 +144,9 @@ 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, @@ -210,7 +219,9 @@ 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) @@ -247,8 +258,10 @@ 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) \ @@ -287,7 +300,7 @@ static inline void load_LDT(mm_context_t *pc) preempt_enable(); } -static inline unsigned long get_desc_base(const struct desc_struct *desc) +static inline unsigned long __intentional_overflow(-1) get_desc_base(const struct desc_struct *desc) { return (unsigned)(desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24)); } @@ -311,7 +324,7 @@ 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; @@ -321,14 +334,14 @@ static inline void set_nmi_gate(int gate, void *addr) #endif #ifdef CONFIG_TRACING -extern struct desc_ptr trace_idt_descr; -extern gate_desc trace_idt_table[]; +extern const struct desc_ptr trace_idt_descr; +extern gate_desc trace_idt_table[IDT_ENTRIES]; static inline void write_trace_idt_entry(int entry, const gate_desc *gate) { write_idt_entry(trace_idt_table, entry, gate); } -static inline void _trace_set_gate(int gate, unsigned type, void *addr, +static inline void _trace_set_gate(int gate, unsigned type, const void *addr, unsigned dpl, unsigned ist, unsigned seg) { gate_desc s; @@ -348,7 +361,7 @@ static inline void write_trace_idt_entry(int entry, const gate_desc *gate) #define _trace_set_gate(gate, type, addr, dpl, ist, seg) #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; @@ -371,9 +384,9 @@ static inline void _set_gate(int gate, unsigned type, void *addr, #define set_intr_gate(n, addr) \ do { \ BUG_ON((unsigned)n > 0xFF); \ - _set_gate(n, GATE_INTERRUPT, (void *)addr, 0, 0, \ + _set_gate(n, GATE_INTERRUPT, (const void *)addr, 0, 0, \ __KERNEL_CS); \ - _trace_set_gate(n, GATE_INTERRUPT, (void *)trace_##addr,\ + _trace_set_gate(n, GATE_INTERRUPT, (const void *)trace_##addr,\ 0, 0, __KERNEL_CS); \ } while (0) @@ -401,19 +414,19 @@ static inline void alloc_system_vector(int vector) /* * 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); @@ -422,16 +435,16 @@ 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); @@ -503,4 +516,17 @@ static inline void load_current_idt(void) else load_idt((const struct desc_ptr *)&idt_descr); } + +#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/div64.h b/arch/x86/include/asm/div64.h index ced283a..ffe04cc 100644 --- a/arch/x86/include/asm/div64.h +++ b/arch/x86/include/asm/div64.h @@ -39,7 +39,7 @@ __mod; \ }) -static inline u64 div_u64_rem(u64 dividend, u32 divisor, u32 *remainder) +static inline u64 __intentional_overflow(-1) div_u64_rem(u64 dividend, u32 divisor, u32 *remainder) { union { u64 v64; diff --git a/arch/x86/include/asm/elf.h b/arch/x86/include/asm/elf.h index 1a055c8..a1701de 100644 --- a/arch/x86/include/asm/elf.h +++ b/arch/x86/include/asm/elf.h @@ -75,9 +75,6 @@ typedef struct user_fxsr_struct elf_fpxregset_t; #include -#ifdef CONFIG_X86_64 -extern unsigned int vdso64_enabled; -#endif #if defined(CONFIG_X86_32) || defined(CONFIG_COMPAT) extern unsigned int vdso32_enabled; #endif @@ -248,7 +245,25 @@ 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_ADDR32)) ? 16 : TASK_SIZE_MAX_SHIFT - PAGE_SHIFT - 3) +#define PAX_DELTA_STACK_LEN ((test_thread_flag(TIF_ADDR32)) ? 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, @@ -297,17 +312,13 @@ do { \ #define ARCH_DLINFO \ do { \ - if (vdso64_enabled) \ - NEW_AUX_ENT(AT_SYSINFO_EHDR, \ - (unsigned long __force)current->mm->context.vdso); \ + NEW_AUX_ENT(AT_SYSINFO_EHDR, current->mm->context.vdso); \ } while (0) /* As a historical oddity, the x32 and x86_64 vDSOs are controlled together. */ #define ARCH_DLINFO_X32 \ do { \ - if (vdso64_enabled) \ - NEW_AUX_ENT(AT_SYSINFO_EHDR, \ - (unsigned long __force)current->mm->context.vdso); \ + NEW_AUX_ENT(AT_SYSINFO_EHDR, current->mm->context.vdso); \ } while (0) #define AT_SYSINFO 32 @@ -322,10 +333,10 @@ else \ #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)current->mm->context.vdso + \ + (current->mm->context.vdso + \ selected_vdso32->sym___kernel_vsyscall) struct linux_binprm; @@ -337,9 +348,6 @@ extern int compat_arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp); #define compat_arch_setup_additional_pages compat_arch_setup_additional_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 77a99ac..39ff7f5 100644 --- a/arch/x86/include/asm/emergency-restart.h +++ b/arch/x86/include/asm/emergency-restart.h @@ -1,6 +1,6 @@ #ifndef _ASM_X86_EMERGENCY_RESTART_H #define _ASM_X86_EMERGENCY_RESTART_H -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 1c7eefe..d0e4702 100644 --- a/arch/x86/include/asm/floppy.h +++ b/arch/x86/include/asm/floppy.h @@ -229,18 +229,18 @@ static struct fd_routine_l { int (*_dma_setup)(char *addr, unsigned long size, int mode, int io); } fd_routine[] = { { - request_dma, - free_dma, - get_dma_residue, - dma_mem_alloc, - hard_dma_setup + ._request_dma = request_dma, + ._free_dma = free_dma, + ._get_dma_residue = get_dma_residue, + ._dma_mem_alloc = dma_mem_alloc, + ._dma_setup = hard_dma_setup }, { - vdma_request_dma, - vdma_nop, - vdma_get_dma_residue, - vdma_mem_alloc, - vdma_dma_setup + ._request_dma = vdma_request_dma, + ._free_dma = vdma_nop, + ._get_dma_residue = vdma_get_dma_residue, + ._dma_mem_alloc = vdma_mem_alloc, + ._dma_setup = vdma_dma_setup } }; diff --git a/arch/x86/include/asm/fpu-internal.h b/arch/x86/include/asm/fpu-internal.h index 115e368..76ecf6c 100644 --- a/arch/x86/include/asm/fpu-internal.h +++ b/arch/x86/include/asm/fpu-internal.h @@ -124,8 +124,11 @@ static inline void sanitize_i387_state(struct task_struct *tsk) #define user_insn(insn, output, input...) \ ({ \ int err; \ + pax_open_userland(); \ asm volatile(ASM_STAC "\n" \ - "1:" #insn "\n\t" \ + "1:" \ + __copyuser_seg \ + #insn "\n\t" \ "2: " ASM_CLAC "\n" \ ".section .fixup,\"ax\"\n" \ "3: movl $-1,%[err]\n" \ @@ -134,6 +137,7 @@ static inline void sanitize_i387_state(struct task_struct *tsk) _ASM_EXTABLE(1b, 3b) \ : [err] "=r" (err), output \ : "0"(0), input); \ + pax_close_userland(); \ err; \ }) @@ -298,7 +302,7 @@ static inline int restore_fpu_checking(struct task_struct *tsk) "fnclex\n\t" "emms\n\t" "fildl %P[addr]" /* set F?P to defined value */ - : : [addr] "m" (tsk->thread.fpu.has_fpu)); + : : [addr] "m" (init_tss[raw_smp_processor_id()].x86_tss.sp0)); } return fpu_restore_checking(&tsk->thread.fpu); diff --git a/arch/x86/include/asm/futex.h b/arch/x86/include/asm/futex.h index b4c1f54..e290c08 100644 --- a/arch/x86/include/asm/futex.h +++ b/arch/x86/include/asm/futex.h @@ -12,6 +12,7 @@ #include #define __futex_atomic_op1(insn, ret, oldval, uaddr, oparg) \ + typecheck(u32 __user *, uaddr); \ asm volatile("\t" ASM_STAC "\n" \ "1:\t" insn "\n" \ "2:\t" ASM_CLAC "\n" \ @@ -20,15 +21,16 @@ "\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("\t" ASM_STAC "\n" \ "1:\tmovl %2, %0\n" \ "\tmovl\t%0, %3\n" \ "\t" insn "\n" \ - "2:\t" LOCK_PREFIX "cmpxchgl %3, %2\n" \ + "2:\t" LOCK_PREFIX __copyuser_seg"cmpxchgl %3, %2\n" \ "\tjnz\t1b\n" \ "3:\t" ASM_CLAC "\n" \ "\t.section .fixup,\"ax\"\n" \ @@ -38,7 +40,7 @@ _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) @@ -57,12 +59,13 @@ static inline int futex_atomic_op_inuser(int encoded_op, u32 __user *uaddr) pagefault_disable(); + pax_open_userland(); 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: @@ -77,6 +80,7 @@ static inline int futex_atomic_op_inuser(int encoded_op, u32 __user *uaddr) default: ret = -ENOSYS; } + pax_close_userland(); pagefault_enable(); diff --git a/arch/x86/include/asm/hw_irq.h b/arch/x86/include/asm/hw_irq.h index 4615906..788c817 100644 --- a/arch/x86/include/asm/hw_irq.h +++ b/arch/x86/include/asm/hw_irq.h @@ -164,8 +164,8 @@ 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/i8259.h b/arch/x86/include/asm/i8259.h index a203659..9889f1c 100644 --- a/arch/x86/include/asm/i8259.h +++ b/arch/x86/include/asm/i8259.h @@ -62,7 +62,7 @@ struct legacy_pic { void (*init)(int auto_eoi); int (*irq_pending)(unsigned int irq); void (*make_irq)(unsigned int irq); -}; +} __do_const; extern struct legacy_pic *legacy_pic; extern struct legacy_pic null_legacy_pic; diff --git a/arch/x86/include/asm/io.h b/arch/x86/include/asm/io.h index b8237d8..3e8864e 100644 --- a/arch/x86/include/asm/io.h +++ b/arch/x86/include/asm/io.h @@ -52,12 +52,12 @@ static inline void name(type val, volatile void __iomem *addr) \ "m" (*(volatile type __force *)addr) barrier); } build_mmio_read(readb, "b", unsigned char, "=q", :"memory") -build_mmio_read(readw, "w", unsigned short, "=r", :"memory") -build_mmio_read(readl, "l", unsigned int, "=r", :"memory") +build_mmio_read(__intentional_overflow(-1) readw, "w", unsigned short, "=r", :"memory") +build_mmio_read(__intentional_overflow(-1) readl, "l", unsigned int, "=r", :"memory") build_mmio_read(__readb, "b", unsigned char, "=q", ) -build_mmio_read(__readw, "w", unsigned short, "=r", ) -build_mmio_read(__readl, "l", unsigned int, "=r", ) +build_mmio_read(__intentional_overflow(-1) __readw, "w", unsigned short, "=r", ) +build_mmio_read(__intentional_overflow(-1) __readl, "l", unsigned int, "=r", ) build_mmio_write(writeb, "b", unsigned char, "q", :"memory") build_mmio_write(writew, "w", unsigned short, "r", :"memory") @@ -109,7 +109,7 @@ build_mmio_write(writeq, "q", unsigned long, "r", :"memory") * this function */ -static inline phys_addr_t virt_to_phys(volatile void *address) +static inline phys_addr_t __intentional_overflow(-1) virt_to_phys(volatile void *address) { return __pa(address); } @@ -185,7 +185,7 @@ static inline void __iomem *ioremap(resource_size_t offset, unsigned long size) return ioremap_nocache(offset, size); } -extern void iounmap(volatile void __iomem *addr); +extern void iounmap(const volatile void __iomem *addr); extern void set_iounmap_nonlazy(void); @@ -195,6 +195,17 @@ 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 0a8b519..80e7d5b 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 53cdfb2..d1369e6 100644 --- a/arch/x86/include/asm/kprobes.h +++ b/arch/x86/include/asm/kprobes.h @@ -38,13 +38,8 @@ 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/local.h b/arch/x86/include/asm/local.h index 4ad6560..75c7bdd 100644 --- a/arch/x86/include/asm/local.h +++ b/arch/x86/include/asm/local.h @@ -10,33 +10,97 @@ typedef struct { atomic_long_t a; } local_t; +typedef struct { + atomic_long_unchecked_t a; +} local_unchecked_t; + #define LOCAL_INIT(i) { ATOMIC_LONG_INIT(i) } #define local_read(l) atomic_long_read(&(l)->a) +#define local_read_unchecked(l) atomic_long_read_unchecked(&(l)->a) #define local_set(l, i) atomic_long_set(&(l)->a, (i)) +#define local_set_unchecked(l, i) atomic_long_set_unchecked(&(l)->a, (i)) 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_inc_unchecked(local_unchecked_t *l) +{ + asm volatile(_ASM_INC "%0\n" : "+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_dec_unchecked(local_unchecked_t *l) +{ + asm volatile(_ASM_DEC "%0\n" : "+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_add_unchecked(long i, local_unchecked_t *l) +{ + asm volatile(_ASM_ADD "%1,%0\n" : "+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)); +} + +static inline void local_sub_unchecked(long i, local_unchecked_t *l) +{ + asm volatile(_ASM_SUB "%1,%0\n" : "+m" (l->a.counter) : "ir" (i)); } @@ -52,7 +116,7 @@ static inline void local_sub(long i, local_t *l) */ static inline int local_sub_and_test(long i, local_t *l) { - GEN_BINARY_RMWcc(_ASM_SUB, l->a.counter, "er", i, "%0", "e"); + GEN_BINARY_RMWcc(_ASM_SUB, _ASM_ADD, l->a.counter, "er", i, "%0", "e"); } /** @@ -65,7 +129,7 @@ static inline int local_sub_and_test(long i, local_t *l) */ static inline int local_dec_and_test(local_t *l) { - GEN_UNARY_RMWcc(_ASM_DEC, l->a.counter, "%0", "e"); + GEN_UNARY_RMWcc(_ASM_DEC, _ASM_INC, l->a.counter, "%0", "e"); } /** @@ -78,7 +142,7 @@ static inline int local_dec_and_test(local_t *l) */ static inline int local_inc_and_test(local_t *l) { - GEN_UNARY_RMWcc(_ASM_INC, l->a.counter, "%0", "e"); + GEN_UNARY_RMWcc(_ASM_INC, _ASM_DEC, l->a.counter, "%0", "e"); } /** @@ -92,7 +156,7 @@ static inline int local_inc_and_test(local_t *l) */ static inline int local_add_negative(long i, local_t *l) { - GEN_BINARY_RMWcc(_ASM_ADD, l->a.counter, "er", i, "%0", "s"); + GEN_BINARY_RMWcc(_ASM_ADD, _ASM_SUB, l->a.counter, "er", i, "%0", "s"); } /** @@ -105,6 +169,30 @@ static inline int local_add_negative(long i, local_t *l) static inline long local_add_return(long i, local_t *l) { long __i = i; + 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; +} + +/** + * local_add_return_unchecked - add and return + * @i: integer value to add + * @l: pointer to type local_unchecked_t + * + * Atomically adds @i to @l and returns @i + @l + */ +static inline long local_add_return_unchecked(long i, local_unchecked_t *l) +{ + long __i = i; asm volatile(_ASM_XADD "%0, %1;" : "+r" (i), "+m" (l->a.counter) : : "memory"); @@ -121,6 +209,8 @@ static inline long local_sub_return(long i, local_t *l) #define local_cmpxchg(l, o, n) \ (cmpxchg_local(&((l)->a.counter), (o), (n))) +#define local_cmpxchg_unchecked(l, o, n) \ + (cmpxchg_local(&((l)->a.counter), (o), (n))) /* Always has a lock prefix */ #define local_xchg(l, n) (xchg(&((l)->a.counter), (n))) diff --git a/arch/x86/include/asm/mman.h b/arch/x86/include/asm/mman.h new file mode 100644 index 0000000..2bfd3ba --- /dev/null +++ b/arch/x86/include/asm/mman.h @@ -0,0 +1,15 @@ +#ifndef _X86_MMAN_H +#define _X86_MMAN_H + +#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 /* X86_MMAN_H */ diff --git a/arch/x86/include/asm/mmu.h b/arch/x86/include/asm/mmu.h index 876e74e..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 __user *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 be12c53..07fd3ca 100644 --- a/arch/x86/include/asm/mmu_context.h +++ b/arch/x86/include/asm/mmu_context.h @@ -24,6 +24,20 @@ 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) + if (!(static_cpu_has(X86_FEATURE_PCID))) { + unsigned int i; + pgd_t *pgd; + + pax_open_kernel(); + pgd = get_cpu_pgd(smp_processor_id(), kernel); + 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 (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) this_cpu_write(cpu_tlbstate.state, TLBSTATE_LAZY); @@ -34,16 +48,59 @@ 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 = this_cpu_read(cpu_tlbstate.state); +#endif this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK); this_cpu_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(); + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (static_cpu_has(X86_FEATURE_PCID)) + __clone_user_pgds(get_cpu_pgd(cpu, user), next->pgd); + else +#endif + + __clone_user_pgds(get_cpu_pgd(cpu, kernel), next->pgd); + __shadow_user_pgds(get_cpu_pgd(cpu, kernel) + USER_PGD_PTRS, next->pgd); + pax_close_kernel(); + BUG_ON((__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL) != (read_cr3() & __PHYSICAL_MASK) && (__pa(get_cpu_pgd(cpu, user)) | PCID_USER) != (read_cr3() & __PHYSICAL_MASK)); + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (static_cpu_has(X86_FEATURE_PCID)) { + if (static_cpu_has(X86_FEATURE_INVPCID)) { + u64 descriptor[2]; + descriptor[0] = PCID_USER; + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_SINGLE_CONTEXT) : "memory"); + if (!static_cpu_has(X86_FEATURE_STRONGUDEREF)) { + descriptor[0] = PCID_KERNEL; + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_SINGLE_CONTEXT) : "memory"); + } + } else { + write_cr3(__pa(get_cpu_pgd(cpu, user)) | PCID_USER); + if (static_cpu_has(X86_FEATURE_STRONGUDEREF)) + write_cr3(__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL | PCID_NOFLUSH); + else + write_cr3(__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL); + } + } else +#endif + + load_cr3(get_cpu_pgd(cpu, kernel)); +#else load_cr3(next->pgd); +#endif /* Stop flush ipis for the previous mm */ cpumask_clear_cpu(cpu, mm_cpumask(prev)); @@ -51,9 +108,67 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, /* Load the LDT, if the LDT is different: */ 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_atomic(); + cpu_clear(cpu, prev->context.cpu_user_cs_mask); + smp_mb__after_atomic(); + 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(); + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (static_cpu_has(X86_FEATURE_PCID)) + __clone_user_pgds(get_cpu_pgd(cpu, user), next->pgd); + else +#endif + + __clone_user_pgds(get_cpu_pgd(cpu, kernel), next->pgd); + __shadow_user_pgds(get_cpu_pgd(cpu, kernel) + USER_PGD_PTRS, next->pgd); + pax_close_kernel(); + BUG_ON((__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL) != (read_cr3() & __PHYSICAL_MASK) && (__pa(get_cpu_pgd(cpu, user)) | PCID_USER) != (read_cr3() & __PHYSICAL_MASK)); + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (static_cpu_has(X86_FEATURE_PCID)) { + if (static_cpu_has(X86_FEATURE_INVPCID)) { + u64 descriptor[2]; + descriptor[0] = PCID_USER; + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_SINGLE_CONTEXT) : "memory"); + if (!static_cpu_has(X86_FEATURE_STRONGUDEREF)) { + descriptor[0] = PCID_KERNEL; + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_SINGLE_CONTEXT) : "memory"); + } + } else { + write_cr3(__pa(get_cpu_pgd(cpu, user)) | PCID_USER); + if (static_cpu_has(X86_FEATURE_STRONGUDEREF)) + write_cr3(__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL | PCID_NOFLUSH); + else + write_cr3(__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL); + } + } else +#endif + + load_cr3(get_cpu_pgd(cpu, kernel)); +#endif + #ifdef CONFIG_SMP - else { this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK); BUG_ON(this_cpu_read(cpu_tlbstate.active_mm) != next); @@ -70,11 +185,28 @@ 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 e3b7819..b257c64 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_M486 #define MODULE_PROC_FAMILY "486 " #elif defined CONFIG_M586 @@ -57,8 +58,20 @@ #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/nmi.h b/arch/x86/include/asm/nmi.h index 5f2fc44..106caa6 100644 --- a/arch/x86/include/asm/nmi.h +++ b/arch/x86/include/asm/nmi.h @@ -36,26 +36,35 @@ enum { typedef int (*nmi_handler_t)(unsigned int, struct pt_regs *); +struct nmiaction; + +struct nmiwork { + const struct nmiaction *action; + u64 max_duration; + struct irq_work irq_work; +}; + struct nmiaction { struct list_head list; nmi_handler_t handler; - u64 max_duration; - struct irq_work irq_work; unsigned long flags; const char *name; -}; + struct nmiwork *work; +} __do_const; #define register_nmi_handler(t, fn, fg, n, init...) \ ({ \ - static struct nmiaction init fn##_na = { \ + static struct nmiwork fn##_nw; \ + static const struct nmiaction init fn##_na = { \ .handler = (fn), \ .name = (n), \ .flags = (fg), \ + .work = &fn##_nw, \ }; \ __register_nmi_handler((t), &fn##_na); \ }) -int __register_nmi_handler(unsigned int, struct nmiaction *); +int __register_nmi_handler(unsigned int, const struct nmiaction *); void unregister_nmi_handler(unsigned int, const char *); diff --git a/arch/x86/include/asm/page.h b/arch/x86/include/asm/page.h index 775873d..04cd306 100644 --- a/arch/x86/include/asm/page.h +++ b/arch/x86/include/asm/page.h @@ -52,6 +52,7 @@ static inline void copy_user_page(void *to, void *from, unsigned long vaddr, __phys_addr_symbol(__phys_reloc_hide((unsigned long)(x))) #define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET)) +#define __early_va(x) ((void *)((unsigned long)(x)+__START_KERNEL_map - phys_base)) #define __boot_va(x) __va(x) #define __boot_pa(x) __pa(x) @@ -60,11 +61,21 @@ static inline void copy_user_page(void *to, void *from, unsigned long vaddr, * virt_to_page(kaddr) returns a valid pointer if and only if * virt_addr_valid(kaddr) returns true. */ -#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT) #define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT) extern bool __virt_addr_valid(unsigned long kaddr); #define virt_addr_valid(kaddr) __virt_addr_valid((unsigned long) (kaddr)) +#ifdef CONFIG_GRKERNSEC_KSTACKOVERFLOW +#define virt_to_page(kaddr) \ + ({ \ + const void *__kaddr = (const void *)(kaddr); \ + BUG_ON(!virt_addr_valid(__kaddr)); \ + pfn_to_page(__pa(__kaddr) >> PAGE_SHIFT); \ + }) +#else +#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT) +#endif + #endif /* __ASSEMBLY__ */ #include diff --git a/arch/x86/include/asm/page_64.h b/arch/x86/include/asm/page_64.h index 0f1ddee..e2fc3d1 100644 --- a/arch/x86/include/asm/page_64.h +++ b/arch/x86/include/asm/page_64.h @@ -7,9 +7,9 @@ /* duplicated to the one in bootmem.h */ extern unsigned long max_pfn; -extern unsigned long phys_base; +extern const unsigned long phys_base; -static inline unsigned long __phys_addr_nodebug(unsigned long x) +static inline unsigned long __intentional_overflow(-1) __phys_addr_nodebug(unsigned long x) { unsigned long y = x - __START_KERNEL_map; diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index cd6e1610..70f4418 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -560,7 +560,7 @@ static inline pmd_t __pmd(pmdval_t val) return (pmd_t) { ret }; } -static inline pmdval_t pmd_val(pmd_t pmd) +static inline __intentional_overflow(-1) pmdval_t pmd_val(pmd_t pmd) { pmdval_t ret; @@ -626,6 +626,18 @@ 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)); @@ -710,6 +722,21 @@ 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 __always_inline void __ticket_lock_spinning(struct arch_spinlock *lock, @@ -906,7 +933,7 @@ 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 \ @@ -981,6 +1008,21 @@ 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 7549b8b..f0edfda 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -84,7 +84,7 @@ struct pv_init_ops { */ unsigned (*patch)(u8 type, u16 clobber, void *insnbuf, unsigned long addr, unsigned len); -}; +} __no_const __no_randomize_layout; struct pv_lazy_ops { @@ -92,13 +92,13 @@ struct pv_lazy_ops { void (*enter)(void); void (*leave)(void); void (*flush)(void); -}; +} __no_randomize_layout; 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 __no_randomize_layout; struct pv_cpu_ops { /* hooks for various privileged instructions */ @@ -192,7 +192,7 @@ struct pv_cpu_ops { void (*start_context_switch)(struct task_struct *prev); void (*end_context_switch)(struct task_struct *next); -}; +} __no_const __no_randomize_layout; struct pv_irq_ops { /* @@ -215,7 +215,7 @@ struct pv_irq_ops { #ifdef CONFIG_X86_64 void (*adjust_exception_frame)(void); #endif -}; +} __no_randomize_layout; struct pv_apic_ops { #ifdef CONFIG_X86_LOCAL_APIC @@ -223,7 +223,7 @@ struct pv_apic_ops { unsigned long start_eip, unsigned long start_esp); #endif -}; +} __no_const __no_randomize_layout; 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 */ @@ -324,7 +325,13 @@ 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 + +} __no_randomize_layout; struct arch_spinlock; #ifdef CONFIG_SMP @@ -336,11 +343,14 @@ typedef u16 __ticket_t; struct pv_lock_ops { struct paravirt_callee_save lock_spinning; void (*unlock_kick)(struct arch_spinlock *lock, __ticket_t ticket); -}; +} __no_randomize_layout; /* This contains all the paravirt structures: we get a convenient * number for each function using the offset which we use to indicate - * what to patch. */ + * what to patch. + * shouldn't be randomized due to the "NEAT TRICK" in paravirt.c + */ + struct paravirt_patch_template { struct pv_init_ops pv_init_ops; struct pv_time_ops pv_time_ops; @@ -349,7 +359,7 @@ struct paravirt_patch_template { struct pv_apic_ops pv_apic_ops; struct pv_mmu_ops pv_mmu_ops; struct pv_lock_ops pv_lock_ops; -}; +} __no_randomize_layout; extern struct pv_info pv_info; extern struct pv_init_ops pv_init_ops; diff --git a/arch/x86/include/asm/pgalloc.h b/arch/x86/include/asm/pgalloc.h index c4412e9..90e88c5 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)); } @@ -108,12 +115,22 @@ 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 @@ -123,6 +140,12 @@ 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 206a87f..1623b06 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 81bb91b..9392125 100644 --- a/arch/x86/include/asm/pgtable-3level.h +++ b/arch/x86/include/asm/pgtable-3level.h @@ -92,12 +92,16 @@ 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 aa97a07..5c53c32 100644 --- a/arch/x86/include/asm/pgtable.h +++ b/arch/x86/include/asm/pgtable.h @@ -46,6 +46,7 @@ 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 @@ -83,12 +84,53 @@ 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(cr0 & X86_CR0_WP); + write_cr0(cr0); + barrier(); + return cr0 ^ X86_CR0_WP; +} + +static inline unsigned long native_pax_close_kernel(void) +{ + unsigned long cr0; + + barrier(); + cr0 = read_cr0() ^ X86_CR0_WP; + BUG_ON(!(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; @@ -155,6 +197,11 @@ static inline unsigned long pud_pfn(pud_t pud) return (pud_val(pud) & PTE_PFN_MASK) >> PAGE_SHIFT; } +static inline unsigned long pgd_pfn(pgd_t pgd) +{ + return (pgd_val(pgd) & PTE_PFN_MASK) >> PAGE_SHIFT; +} + #define pte_page(pte) pfn_to_page(pte_pfn(pte)) static inline int pmd_large(pmd_t pte) @@ -208,9 +255,29 @@ 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) @@ -440,6 +507,16 @@ pte_t *populate_extra_pte(unsigned long vaddr); #endif #ifndef __ASSEMBLY__ + +#ifdef CONFIG_PAX_PER_CPU_PGD +extern pgd_t cpu_pgd[NR_CPUS][2][PTRS_PER_PGD]; +enum cpu_pgd_type {kernel = 0, user = 1}; +static inline pgd_t *get_cpu_pgd(unsigned int cpu, enum cpu_pgd_type type) +{ + return cpu_pgd[cpu][type]; +} +#endif + #include #include #include @@ -586,7 +663,7 @@ static inline unsigned long pud_page_vaddr(pud_t pud) * Currently stuck as a macro due to indirect forward reference to * linux/mmzone.h's __section_mem_map_addr() definition: */ -#define pud_page(pud) pfn_to_page(pud_val(pud) >> PAGE_SHIFT) +#define pud_page(pud) pfn_to_page((pud_val(pud) & PTE_PFN_MASK) >> PAGE_SHIFT) /* Find an entry in the second-level page table.. */ static inline pmd_t *pmd_offset(pud_t *pud, unsigned long address) @@ -626,7 +703,7 @@ static inline unsigned long pgd_page_vaddr(pgd_t pgd) * Currently stuck as a macro due to indirect forward reference to * linux/mmzone.h's __section_mem_map_addr() definition: */ -#define pgd_page(pgd) pfn_to_page(pgd_val(pgd) >> PAGE_SHIFT) +#define pgd_page(pgd) pfn_to_page((pgd_val(pgd) & PTE_PFN_MASK) >> PAGE_SHIFT) /* to find an entry in a page-table-directory. */ static inline unsigned long pud_index(unsigned long address) @@ -641,7 +718,7 @@ 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) @@ -664,7 +741,12 @@ 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, type, address) (get_cpu_pgd(cpu, type) + pgd_index(address)) +#endif + /* * a shortcut which implies the use of the kernel's pgd, instead * of a process's @@ -675,6 +757,23 @@ 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 +#ifdef __ASSEMBLY__ +#define pax_user_shadow_base pax_user_shadow_base(%rip) +#else +extern unsigned long pax_user_shadow_base; +extern pgdval_t clone_pgd_mask; +#endif +#endif + +#endif + #ifndef __ASSEMBLY__ extern int direct_gbpages; @@ -841,11 +940,24 @@ 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); +#endif + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +extern void __shadow_user_pgds(pgd_t *dst, const pgd_t *src); +#else +static inline void __shadow_user_pgds(pgd_t *dst, const pgd_t *src) {} +#endif + #define PTE_SHIFT ilog2(PTRS_PER_PTE) static inline int page_level_shift(enum pg_level level) { diff --git a/arch/x86/include/asm/pgtable_32.h b/arch/x86/include/asm/pgtable_32.h index 9ee3221..b979c6b 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))) + \ @@ -62,12 +65,17 @@ 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) #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 3874693..d7906ac 100644 --- a/arch/x86/include/asm/pgtable_64.h +++ b/arch/x86/include/asm/pgtable_64.h @@ -16,11 +16,16 @@ 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 pmd_t level2_ident_pgt[512*2]; extern pte_t level1_fixmap_pgt[512]; -extern pgd_t init_level4_pgt[]; +extern pte_t level1_vsyscall_pgt[512]; +extern pgd_t init_level4_pgt[512]; #define swapper_pg_dir init_level4_pgt @@ -62,7 +67,9 @@ 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) @@ -98,7 +105,9 @@ 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) @@ -108,6 +117,13 @@ 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 7166e25..baaa6fe 100644 --- a/arch/x86/include/asm/pgtable_64_types.h +++ b/arch/x86/include/asm/pgtable_64_types.h @@ -61,9 +61,14 @@ typedef struct { pteval_t pte; } pte_t; #define MODULES_VADDR (__START_KERNEL_map + KERNEL_IMAGE_SIZE) #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 ESPFIX_PGD_ENTRY _AC(-2, UL) #define ESPFIX_BASE_ADDR (ESPFIX_PGD_ENTRY << PGDIR_SHIFT) +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) + #define EARLY_DYNAMIC_PAGE_TABLES 64 #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 f216963..6bd7c21 100644 --- a/arch/x86/include/asm/pgtable_types.h +++ b/arch/x86/include/asm/pgtable_types.h @@ -111,8 +111,10 @@ #if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) #define _PAGE_NX (_AT(pteval_t, 1) << _PAGE_BIT_NX) -#else +#elif defined(CONFIG_KMEMCHECK) || defined(CONFIG_MEM_SOFT_DIRTY) #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) @@ -151,6 +153,9 @@ #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) @@ -161,7 +166,7 @@ #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) @@ -218,7 +223,7 @@ #ifdef CONFIG_X86_64 #define __PAGE_KERNEL_IDENT_LARGE_EXEC __PAGE_KERNEL_LARGE_EXEC #else -#define PTE_IDENT_ATTR 0x003 /* PRESENT+RW */ +#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 @@ -257,7 +262,17 @@ 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; @@ -271,8 +286,6 @@ 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); @@ -292,8 +305,6 @@ 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); @@ -333,7 +344,6 @@ 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/preempt.h b/arch/x86/include/asm/preempt.h index 7024c12..71c46b9 100644 --- a/arch/x86/include/asm/preempt.h +++ b/arch/x86/include/asm/preempt.h @@ -87,7 +87,7 @@ static __always_inline void __preempt_count_sub(int val) */ static __always_inline bool __preempt_count_dec_and_test(void) { - GEN_UNARY_RMWcc("decl", __preempt_count, __percpu_arg(0), "e"); + GEN_UNARY_RMWcc("decl", "incl", __preempt_count, __percpu_arg(0), "e"); } /* diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index a4ea023..2ff3cb8 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -128,7 +128,7 @@ struct cpuinfo_x86 { /* Index into per_cpu list: */ u16 cpu_index; u32 microcode; -} __attribute__((__aligned__(SMP_CACHE_BYTES))); +} __attribute__((__aligned__(SMP_CACHE_BYTES))) __randomize_layout; #define X86_VENDOR_INTEL 0 #define X86_VENDOR_CYRIX 1 @@ -199,9 +199,21 @@ static inline void native_cpuid(unsigned int *eax, unsigned int *ebx, : "memory"); } +/* invpcid (%rdx),%rax */ +#define __ASM_INVPCID ".byte 0x66,0x0f,0x38,0x82,0x02" + +#define INVPCID_SINGLE_ADDRESS 0UL +#define INVPCID_SINGLE_CONTEXT 1UL +#define INVPCID_ALL_GLOBAL 2UL +#define INVPCID_ALL_NONGLOBAL 3UL + +#define PCID_KERNEL 0UL +#define PCID_USER 1UL +#define PCID_NOFLUSH (1UL << 63) + static inline void load_cr3(pgd_t *pgdir) { - write_cr3(__pa(pgdir)); + write_cr3(__pa(pgdir) | PCID_KERNEL); } #ifdef CONFIG_X86_32 @@ -283,7 +295,7 @@ 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 @@ -479,6 +491,7 @@ struct thread_struct { unsigned short ds; unsigned short fsindex; unsigned short gsindex; + unsigned short ss; #endif #ifdef CONFIG_X86_32 unsigned long ip; @@ -588,29 +601,8 @@ static inline void load_sp0(struct tss_struct *tss, extern unsigned long mmu_cr4_features; extern u32 *trampoline_cr4_features; -static inline void set_in_cr4(unsigned long mask) -{ - unsigned long cr4; - - mmu_cr4_features |= mask; - if (trampoline_cr4_features) - *trampoline_cr4_features = mmu_cr4_features; - cr4 = read_cr4(); - cr4 |= mask; - write_cr4(cr4); -} - -static inline void clear_in_cr4(unsigned long mask) -{ - unsigned long cr4; - - mmu_cr4_features &= ~mask; - if (trampoline_cr4_features) - *trampoline_cr4_features = mmu_cr4_features; - cr4 = read_cr4(); - cr4 &= ~mask; - write_cr4(cr4); -} +extern void set_in_cr4(unsigned long mask); +extern void clear_in_cr4(unsigned long mask); typedef struct { unsigned long seg; @@ -836,11 +828,18 @@ 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, \ @@ -854,7 +853,7 @@ 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, \ @@ -865,11 +864,7 @@ 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. @@ -884,7 +879,7 @@ 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; \ }) @@ -894,13 +889,13 @@ 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_ADDR32) ? \ IA32_PAGE_OFFSET : TASK_SIZE_MAX) @@ -911,11 +906,11 @@ 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 \ } /* @@ -943,6 +938,10 @@ 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 */ @@ -969,7 +968,7 @@ static inline uint32_t hypervisor_cpuid_base(const char *sig, uint32_t leaves) return 0; } -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); @@ -979,6 +978,6 @@ bool xen_set_default_idle(void); #define xen_set_default_idle 0 #endif -void stop_this_cpu(void *dummy); +void stop_this_cpu(void *dummy) __noreturn; void df_debug(struct pt_regs *regs, long error_code); #endif /* _ASM_X86_PROCESSOR_H */ diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h index 6205f0c..688a3a9 100644 --- a/arch/x86/include/asm/ptrace.h +++ b/arch/x86/include/asm/ptrace.h @@ -84,28 +84,29 @@ 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 } @@ -121,15 +122,16 @@ 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 } @@ -180,9 +182,11 @@ static inline unsigned long regs_get_register(struct pt_regs *regs, * Traps from the kernel do not save sp and ss. * Use the helper function to retrieve sp. */ - if (offset == offsetof(struct pt_regs, sp) && - regs->cs == __KERNEL_CS) - return kernel_stack_pointer(regs); + if (offset == offsetof(struct pt_regs, sp)) { + unsigned long cs = regs->cs & 0xffff; + if (cs == __KERNEL_CS || cs == __KERNEXEC_KERNEL_CS) + return kernel_stack_pointer(regs); + } #endif return *(unsigned long *)((unsigned long)regs + offset); } diff --git a/arch/x86/include/asm/qrwlock.h b/arch/x86/include/asm/qrwlock.h index 70f46f0..adfbdb4 100644 --- a/arch/x86/include/asm/qrwlock.h +++ b/arch/x86/include/asm/qrwlock.h @@ -7,8 +7,8 @@ #define queue_write_unlock queue_write_unlock static inline void queue_write_unlock(struct qrwlock *lock) { - barrier(); - ACCESS_ONCE(*(u8 *)&lock->cnts) = 0; + barrier(); + ACCESS_ONCE_RW(*(u8 *)&lock->cnts) = 0; } #endif diff --git a/arch/x86/include/asm/realmode.h b/arch/x86/include/asm/realmode.h index 9c6b890..5305f53 100644 --- a/arch/x86/include/asm/realmode.h +++ b/arch/x86/include/asm/realmode.h @@ -22,16 +22,14 @@ struct real_mode_header { #endif /* APM/BIOS reboot */ u32 machine_real_restart_asm; -#ifdef CONFIG_X86_64 u32 machine_real_restart_seg; -#endif }; /* This must match data at trampoline_32/64.S */ struct trampoline_header { #ifdef CONFIG_X86_32 u32 start; - u16 gdt_pad; + u16 boot_cs; u16 gdt_limit; u32 gdt_base; #else diff --git a/arch/x86/include/asm/reboot.h b/arch/x86/include/asm/reboot.h index a82c4f1..ac45053 100644 --- a/arch/x86/include/asm/reboot.h +++ b/arch/x86/include/asm/reboot.h @@ -6,13 +6,13 @@ 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; diff --git a/arch/x86/include/asm/rmwcc.h b/arch/x86/include/asm/rmwcc.h index 8f7866a..e442f20 100644 --- a/arch/x86/include/asm/rmwcc.h +++ b/arch/x86/include/asm/rmwcc.h @@ -3,7 +3,34 @@ #ifdef CC_HAVE_ASM_GOTO -#define __GEN_RMWcc(fullop, var, cc, ...) \ +#ifdef CONFIG_PAX_REFCOUNT +#define __GEN_RMWcc(fullop, fullantiop, var, cc, ...) \ +do { \ + asm_volatile_goto (fullop \ + ";jno 0f\n" \ + fullantiop \ + ";int $4\n0:\n" \ + _ASM_EXTABLE(0b, 0b) \ + ";j" cc " %l[cc_label]" \ + : : "m" (var), ## __VA_ARGS__ \ + : "memory" : cc_label); \ + return 0; \ +cc_label: \ + return 1; \ +} while (0) +#else +#define __GEN_RMWcc(fullop, fullantiop, var, cc, ...) \ +do { \ + asm_volatile_goto (fullop ";j" cc " %l[cc_label]" \ + : : "m" (var), ## __VA_ARGS__ \ + : "memory" : cc_label); \ + return 0; \ +cc_label: \ + return 1; \ +} while (0) +#endif + +#define __GEN_RMWcc_unchecked(fullop, var, cc, ...) \ do { \ asm_volatile_goto (fullop "; j" cc " %l[cc_label]" \ : : "m" (var), ## __VA_ARGS__ \ @@ -13,15 +40,46 @@ cc_label: \ return 1; \ } while (0) -#define GEN_UNARY_RMWcc(op, var, arg0, cc) \ - __GEN_RMWcc(op " " arg0, var, cc) +#define GEN_UNARY_RMWcc(op, antiop, var, arg0, cc) \ + __GEN_RMWcc(op " " arg0, antiop " " arg0, var, cc) -#define GEN_BINARY_RMWcc(op, var, vcon, val, arg0, cc) \ - __GEN_RMWcc(op " %1, " arg0, var, cc, vcon (val)) +#define GEN_UNARY_RMWcc_unchecked(op, var, arg0, cc) \ + __GEN_RMWcc_unchecked(op " " arg0, var, cc) + +#define GEN_BINARY_RMWcc(op, antiop, var, vcon, val, arg0, cc) \ + __GEN_RMWcc(op " %1, " arg0, antiop " %1, " arg0, var, cc, vcon (val)) + +#define GEN_BINARY_RMWcc_unchecked(op, var, vcon, val, arg0, cc) \ + __GEN_RMWcc_unchecked(op " %1, " arg0, var, cc, vcon (val)) #else /* !CC_HAVE_ASM_GOTO */ -#define __GEN_RMWcc(fullop, var, cc, ...) \ +#ifdef CONFIG_PAX_REFCOUNT +#define __GEN_RMWcc(fullop, fullantiop, var, cc, ...) \ +do { \ + char c; \ + asm volatile (fullop \ + ";jno 0f\n" \ + fullantiop \ + ";int $4\n0:\n" \ + _ASM_EXTABLE(0b, 0b) \ + "; set" cc " %1" \ + : "+m" (var), "=qm" (c) \ + : __VA_ARGS__ : "memory"); \ + return c != 0; \ +} while (0) +#else +#define __GEN_RMWcc(fullop, fullantiop, var, cc, ...) \ +do { \ + char c; \ + asm volatile (fullop "; set" cc " %1" \ + : "+m" (var), "=qm" (c) \ + : __VA_ARGS__ : "memory"); \ + return c != 0; \ +} while (0) +#endif + +#define __GEN_RMWcc_unchecked(fullop, var, cc, ...) \ do { \ char c; \ asm volatile (fullop "; set" cc " %1" \ @@ -30,11 +88,17 @@ do { \ return c != 0; \ } while (0) -#define GEN_UNARY_RMWcc(op, var, arg0, cc) \ - __GEN_RMWcc(op " " arg0, var, cc) +#define GEN_UNARY_RMWcc(op, antiop, var, arg0, cc) \ + __GEN_RMWcc(op " " arg0, antiop " " arg0, var, cc) + +#define GEN_UNARY_RMWcc_unchecked(op, var, arg0, cc) \ + __GEN_RMWcc_unchecked(op " " arg0, var, cc) + +#define GEN_BINARY_RMWcc(op, antiop, var, vcon, val, arg0, cc) \ + __GEN_RMWcc(op " %2, " arg0, antiop " %2, " arg0, var, cc, vcon (val)) -#define GEN_BINARY_RMWcc(op, var, vcon, val, arg0, cc) \ - __GEN_RMWcc(op " %2, " arg0, var, cc, vcon (val)) +#define GEN_BINARY_RMWcc_unchecked(op, var, vcon, val, arg0, cc) \ + __GEN_RMWcc_unchecked(op " %2, " arg0, var, cc, vcon (val)) #endif /* CC_HAVE_ASM_GOTO */ diff --git a/arch/x86/include/asm/rwsem.h b/arch/x86/include/asm/rwsem.h index cad82c9..2e5c5c1 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 " __ASM_SEL(%w1,%k1) "," __ASM_SEL(%w1,%k1) "\n\t" /* was the active mask 0 before? */ @@ -155,6 +179,14 @@ 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 */ @@ -173,6 +205,14 @@ 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 */ @@ -190,6 +230,14 @@ 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) @@ -208,7 +256,15 @@ 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)); } @@ -218,7 +274,7 @@ 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 6f1c3a8..7744f19 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 @@ -173,6 +188,8 @@ #define GDT_ENTRY_PER_CPU 15 /* Abused to load per CPU data from limit */ #define __PER_CPU_SEG (GDT_ENTRY_PER_CPU * 8 + 3) +#define GDT_ENTRY_UDEREF_KERNEL_DS 16 + /* TLS indexes for 64bit - hardcoded in arch_prctl */ #define FS_TLS 0 #define GS_TLS 1 @@ -180,12 +197,14 @@ #define GS_TLS_SEL ((GDT_ENTRY_TLS_MIN+GS_TLS)*8 + 3) #define FS_TLS_SEL ((GDT_ENTRY_TLS_MIN+FS_TLS)*8 + 3) -#define GDT_ENTRIES 16 +#define GDT_ENTRIES 17 #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 __UDEREF_KERNEL_DS (GDT_ENTRY_UDEREF_KERNEL_DS*8) #define __USER_DS (GDT_ENTRY_DEFAULT_USER_DS*8+3) #define __USER_CS (GDT_ENTRY_DEFAULT_USER_CS*8+3) #ifndef CONFIG_PARAVIRT @@ -268,7 +287,7 @@ 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; } #endif /* !__ASSEMBLY__ */ diff --git a/arch/x86/include/asm/smap.h b/arch/x86/include/asm/smap.h index 8d3120f..352b440 100644 --- a/arch/x86/include/asm/smap.h +++ b/arch/x86/include/asm/smap.h @@ -25,11 +25,40 @@ #include +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +#define ASM_PAX_OPEN_USERLAND \ + 661: jmp 663f; \ + .pushsection .altinstr_replacement, "a" ; \ + 662: pushq %rax; nop; \ + .popsection ; \ + .pushsection .altinstructions, "a" ; \ + altinstruction_entry 661b, 662b, X86_FEATURE_STRONGUDEREF, 2, 2;\ + .popsection ; \ + call __pax_open_userland; \ + popq %rax; \ + 663: + +#define ASM_PAX_CLOSE_USERLAND \ + 661: jmp 663f; \ + .pushsection .altinstr_replacement, "a" ; \ + 662: pushq %rax; nop; \ + .popsection; \ + .pushsection .altinstructions, "a" ; \ + altinstruction_entry 661b, 662b, X86_FEATURE_STRONGUDEREF, 2, 2;\ + .popsection; \ + call __pax_close_userland; \ + popq %rax; \ + 663: +#else +#define ASM_PAX_OPEN_USERLAND +#define ASM_PAX_CLOSE_USERLAND +#endif + #ifdef CONFIG_X86_SMAP #define ASM_CLAC \ 661: ASM_NOP3 ; \ - .pushsection .altinstr_replacement, "ax" ; \ + .pushsection .altinstr_replacement, "a" ; \ 662: __ASM_CLAC ; \ .popsection ; \ .pushsection .altinstructions, "a" ; \ @@ -38,7 +67,7 @@ #define ASM_STAC \ 661: ASM_NOP3 ; \ - .pushsection .altinstr_replacement, "ax" ; \ + .pushsection .altinstr_replacement, "a" ; \ 662: __ASM_STAC ; \ .popsection ; \ .pushsection .altinstructions, "a" ; \ @@ -56,6 +85,37 @@ #include +#define __HAVE_ARCH_PAX_OPEN_USERLAND +#define __HAVE_ARCH_PAX_CLOSE_USERLAND + +extern void __pax_open_userland(void); +static __always_inline unsigned long pax_open_userland(void) +{ + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + asm volatile(ALTERNATIVE(ASM_NOP5, "call %P[open]", X86_FEATURE_STRONGUDEREF) + : + : [open] "i" (__pax_open_userland) + : "memory", "rax"); +#endif + + return 0; +} + +extern void __pax_close_userland(void); +static __always_inline unsigned long pax_close_userland(void) +{ + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + asm volatile(ALTERNATIVE(ASM_NOP5, "call %P[close]", X86_FEATURE_STRONGUDEREF) + : + : [close] "i" (__pax_close_userland) + : "memory", "rax"); +#endif + + return 0; +} + #ifdef CONFIG_X86_SMAP static __always_inline void clac(void) diff --git a/arch/x86/include/asm/smp.h b/arch/x86/include/asm/smp.h index 8cd27e0..7f05ec8 100644 --- a/arch/x86/include/asm/smp.h +++ b/arch/x86/include/asm/smp.h @@ -35,7 +35,7 @@ DECLARE_PER_CPU_READ_MOSTLY(cpumask_var_t, cpu_core_map); /* cpus sharing the last level cache: */ DECLARE_PER_CPU_READ_MOSTLY(cpumask_var_t, cpu_llc_shared_map); DECLARE_PER_CPU_READ_MOSTLY(u16, cpu_llc_id); -DECLARE_PER_CPU_READ_MOSTLY(int, cpu_number); +DECLARE_PER_CPU_READ_MOSTLY(unsigned int, cpu_number); static inline struct cpumask *cpu_sibling_mask(int cpu) { @@ -78,7 +78,7 @@ 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); @@ -190,14 +190,8 @@ extern unsigned disabled_cpus; extern int safe_smp_processor_id(void); #elif defined(CONFIG_X86_64_SMP) -#define raw_smp_processor_id() (this_cpu_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() (this_cpu_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 54f1c80..39362a5 100644 --- a/arch/x86/include/asm/spinlock.h +++ b/arch/x86/include/asm/spinlock.h @@ -223,6 +223,14 @@ 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" @@ -232,6 +240,14 @@ 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" @@ -261,13 +277,29 @@ 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"); } #else diff --git a/arch/x86/include/asm/stackprotector.h b/arch/x86/include/asm/stackprotector.h index 6a99859..03cb807 100644 --- a/arch/x86/include/asm/stackprotector.h +++ b/arch/x86/include/asm/stackprotector.h @@ -47,7 +47,7 @@ * 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. @@ -112,7 +112,7 @@ 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/switch_to.h b/arch/x86/include/asm/switch_to.h index d7f3b3b..3cc39f1 100644 --- a/arch/x86/include/asm/switch_to.h +++ b/arch/x86/include/asm/switch_to.h @@ -108,7 +108,7 @@ 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" \ @@ -119,7 +119,7 @@ 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) diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index 8540538..4b0b5e9 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -24,7 +24,6 @@ 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 */ @@ -33,13 +32,13 @@ struct thread_info { mm_segment_t addr_limit; struct restart_block restart_block; void __user *sysenter_return; + 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, \ @@ -50,7 +49,7 @@ 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__ */ @@ -91,6 +90,7 @@ struct thread_info { #define TIF_SYSCALL_TRACEPOINT 28 /* syscall tracepoint instrumentation */ #define TIF_ADDR32 29 /* 32-bit address space on 64 bits */ #define TIF_X32 30 /* 32-bit native x86-64 binary */ +#define TIF_GRSEC_SETXID 31 /* update credentials on syscall entry/exit */ #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) @@ -115,17 +115,18 @@ struct thread_info { #define _TIF_SYSCALL_TRACEPOINT (1 << TIF_SYSCALL_TRACEPOINT) #define _TIF_ADDR32 (1 << TIF_ADDR32) #define _TIF_X32 (1 << TIF_X32) +#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_NOHZ) + _TIF_NOHZ | _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_NOHZ) + _TIF_SYSCALL_TRACEPOINT | _TIF_NOHZ | _TIF_GRSEC_SETXID) /* work to do on interrupt/exception return */ #define _TIF_WORK_MASK \ @@ -136,7 +137,7 @@ struct thread_info { /* work to do on any return to user space */ #define _TIF_ALLWORK_MASK \ ((0x0000FFFF & ~_TIF_SECCOMP) | _TIF_SYSCALL_TRACEPOINT | \ - _TIF_NOHZ) + _TIF_NOHZ | _TIF_GRSEC_SETXID) /* Only used for 64 bit */ #define _TIF_DO_NOTIFY_MASK \ @@ -151,7 +152,6 @@ struct thread_info { #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) #define STACK_WARN (THREAD_SIZE/8) -#define KERNEL_STACK_OFFSET (5*(BITS_PER_LONG/8)) /* * macros/functions for gaining access to the thread information structure @@ -162,26 +162,18 @@ struct thread_info { DECLARE_PER_CPU(unsigned long, kernel_stack); +DECLARE_PER_CPU(struct thread_info *, current_tinfo); + static inline struct thread_info *current_thread_info(void) { - struct thread_info *ti; - ti = (void *)(this_cpu_read_stable(kernel_stack) + - KERNEL_STACK_OFFSET - THREAD_SIZE); - return ti; + return this_cpu_read_stable(current_tinfo); } #else /* !__ASSEMBLY__ */ /* how to get the thread information struct from ASM */ #define GET_THREAD_INFO(reg) \ - _ASM_MOV PER_CPU_VAR(kernel_stack),reg ; \ - _ASM_SUB $(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) + _ASM_MOV PER_CPU_VAR(current_tinfo),reg ; #endif @@ -237,5 +229,12 @@ static inline bool is_ia32_task(void) extern void arch_task_cache_init(void); extern int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src); extern void arch_release_task_struct(struct task_struct *tsk); + +#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) + #endif #endif /* _ASM_X86_THREAD_INFO_H */ diff --git a/arch/x86/include/asm/tlbflush.h b/arch/x86/include/asm/tlbflush.h index 04905bf..1178cdf 100644 --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -17,18 +17,44 @@ static inline void __native_flush_tlb(void) { + if (static_cpu_has(X86_FEATURE_INVPCID)) { + u64 descriptor[2]; + + descriptor[0] = PCID_KERNEL; + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_ALL_NONGLOBAL) : "memory"); + return; + } + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (static_cpu_has(X86_FEATURE_PCID)) { + unsigned int cpu = raw_get_cpu(); + + native_write_cr3(__pa(get_cpu_pgd(cpu, user)) | PCID_USER); + native_write_cr3(__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL); + raw_put_cpu_no_resched(); + return; + } +#endif + native_write_cr3(native_read_cr3()); } static inline void __native_flush_tlb_global_irq_disabled(void) { - unsigned long cr4; + if (static_cpu_has(X86_FEATURE_INVPCID)) { + u64 descriptor[2]; - cr4 = native_read_cr4(); - /* clear PGE */ - native_write_cr4(cr4 & ~X86_CR4_PGE); - /* write old PGE again and flush TLBs */ - native_write_cr4(cr4); + descriptor[0] = PCID_KERNEL; + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_ALL_GLOBAL) : "memory"); + } else { + unsigned long cr4; + + cr4 = native_read_cr4(); + /* clear PGE */ + native_write_cr4(cr4 & ~X86_CR4_PGE); + /* write old PGE again and flush TLBs */ + native_write_cr4(cr4); + } } static inline void __native_flush_tlb_global(void) @@ -49,6 +75,41 @@ static inline void __native_flush_tlb_global(void) static inline void __native_flush_tlb_single(unsigned long addr) { + if (static_cpu_has(X86_FEATURE_INVPCID)) { + u64 descriptor[2]; + + descriptor[0] = PCID_KERNEL; + descriptor[1] = addr; + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (!static_cpu_has(X86_FEATURE_STRONGUDEREF) || addr >= TASK_SIZE_MAX) { + if (addr < TASK_SIZE_MAX) + descriptor[1] += pax_user_shadow_base; + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_SINGLE_ADDRESS) : "memory"); + } + + descriptor[0] = PCID_USER; + descriptor[1] = addr; +#endif + + asm volatile(__ASM_INVPCID : : "d"(&descriptor), "a"(INVPCID_SINGLE_ADDRESS) : "memory"); + return; + } + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + if (static_cpu_has(X86_FEATURE_PCID)) { + unsigned int cpu = raw_get_cpu(); + + native_write_cr3(__pa(get_cpu_pgd(cpu, user)) | PCID_USER | PCID_NOFLUSH); + asm volatile("invlpg (%0)" ::"r" (addr) : "memory"); + native_write_cr3(__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL | PCID_NOFLUSH); + raw_put_cpu_no_resched(); + + if (!static_cpu_has(X86_FEATURE_STRONGUDEREF) && addr < TASK_SIZE_MAX) + addr += pax_user_shadow_base; + } +#endif + asm volatile("invlpg (%0)" ::"r" (addr) : "memory"); } diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h index 0d592e0..526f797 100644 --- a/arch/x86/include/asm/uaccess.h +++ b/arch/x86/include/asm/uaccess.h @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include @@ -29,7 +30,12 @@ #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) @@ -85,8 +91,36 @@ static inline bool __chk_range_not_ok(unsigned long addr, unsigned long size, un * 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, user_addr_max())) +extern int _cond_resched(void); +#define access_ok_noprefault(type, addr, size) (likely(!__range_not_ok(addr, size, user_addr_max()))) +#define access_ok(type, addr, size) \ +({ \ + unsigned long __size = size; \ + unsigned long __addr = (unsigned long)addr; \ + bool __ret_ao = __range_not_ok(__addr, __size, user_addr_max()) == 0;\ + if (__ret_ao && __size) { \ + unsigned long __addr_ao = __addr & PAGE_MASK; \ + unsigned long __end_ao = __addr + __size - 1; \ + if (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 relative to the @@ -176,10 +210,12 @@ __typeof__(__builtin_choose_expr(sizeof(x) > sizeof(0UL), 0ULL, 0UL)) register __inttype(*(ptr)) __val_gu asm("%"_ASM_DX); \ __chk_user_ptr(ptr); \ might_fault(); \ + pax_open_userland(); \ asm volatile("call __get_user_%P3" \ : "=a" (__ret_gu), "=r" (__val_gu) \ : "0" (ptr), "i" (sizeof(*(ptr)))); \ (x) = (__typeof__(*(ptr))) __val_gu; \ + pax_close_userland(); \ __ret_gu; \ }) @@ -187,13 +223,21 @@ __typeof__(__builtin_choose_expr(sizeof(x) > sizeof(0UL), 0ULL, 0UL)) 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(ASM_STAC "\n" \ - "1: movl %%eax,0(%2)\n" \ - "2: movl %%edx,4(%2)\n" \ + "1: "__copyuser_seg"movl %%eax,0(%2)\n" \ + "2: "__copyuser_seg"movl %%edx,4(%2)\n" \ "3: " ASM_CLAC "\n" \ ".section .fixup,\"ax\"\n" \ "4: movl %3,%0\n" \ @@ -206,8 +250,8 @@ __typeof__(__builtin_choose_expr(sizeof(x) > sizeof(0UL), 0ULL, 0UL)) #define __put_user_asm_ex_u64(x, addr) \ asm volatile(ASM_STAC "\n" \ - "1: movl %%eax,0(%1)\n" \ - "2: movl %%edx,4(%1)\n" \ + "1: "__copyuser_seg"movl %%eax,0(%1)\n" \ + "2: "__copyuser_seg"movl %%edx,4(%1)\n" \ "3: " ASM_CLAC "\n" \ _ASM_EXTABLE_EX(1b, 2b) \ _ASM_EXTABLE_EX(2b, 3b) \ @@ -257,7 +301,8 @@ extern void __put_user_8(void); __typeof__(*(ptr)) __pu_val; \ __chk_user_ptr(ptr); \ might_fault(); \ - __pu_val = x; \ + __pu_val = (x); \ + pax_open_userland(); \ switch (sizeof(*(ptr))) { \ case 1: \ __put_user_x(1, __pu_val, ptr, __ret_pu); \ @@ -275,6 +320,7 @@ extern void __put_user_8(void); __put_user_x(X, __pu_val, ptr, __ret_pu); \ break; \ } \ + pax_close_userland(); \ __ret_pu; \ }) @@ -355,8 +401,10 @@ do { \ } while (0) #define __get_user_asm(x, addr, err, itype, rtype, ltype, errret) \ +do { \ + pax_open_userland(); \ asm volatile(ASM_STAC "\n" \ - "1: mov"itype" %2,%"rtype"1\n" \ + "1: "__copyuser_seg"mov"itype" %2,%"rtype"1\n"\ "2: " ASM_CLAC "\n" \ ".section .fixup,\"ax\"\n" \ "3: mov %3,%0\n" \ @@ -364,8 +412,10 @@ do { \ " jmp 2b\n" \ ".previous\n" \ _ASM_EXTABLE(1b, 3b) \ - : "=r" (err), ltype(x) \ - : "m" (__m(addr)), "i" (errret), "0" (err)) + : "=r" (err), ltype (x) \ + : "m" (__m(addr)), "i" (errret), "0" (err)); \ + pax_close_userland(); \ +} while (0) #define __get_user_size_ex(x, ptr, size) \ do { \ @@ -389,7 +439,7 @@ 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_EX(1b, 2b) \ : ltype(x) : "m" (__m(addr))) @@ -406,13 +456,24 @@ 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; \ + (typeof(x))____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 @@ -420,8 +481,10 @@ struct __large_struct { unsigned long buf[100]; }; * aliasing issues. */ #define __put_user_asm(x, addr, err, itype, rtype, ltype, errret) \ +do { \ + pax_open_userland(); \ asm volatile(ASM_STAC "\n" \ - "1: mov"itype" %"rtype"1,%2\n" \ + "1: "__copyuser_seg"mov"itype" %"rtype"1,%2\n"\ "2: " ASM_CLAC "\n" \ ".section .fixup,\"ax\"\n" \ "3: mov %3,%0\n" \ @@ -429,10 +492,12 @@ 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));\ + pax_close_userland(); \ +} while (0) #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_EX(1b, 2b) \ : : ltype(x), "m" (__m(addr))) @@ -442,11 +507,13 @@ struct __large_struct { unsigned long buf[100]; }; */ #define uaccess_try do { \ current_thread_info()->uaccess_err = 0; \ + pax_open_userland(); \ stac(); \ barrier(); #define uaccess_catch(err) \ clac(); \ + pax_close_userland(); \ (err) |= (current_thread_info()->uaccess_err ? -EFAULT : 0); \ } while (0) @@ -471,8 +538,12 @@ 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. @@ -494,8 +565,12 @@ 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 @@ -513,7 +588,7 @@ 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) #define put_user_try uaccess_try @@ -542,18 +617,19 @@ extern void __cmpxchg_wrong_size(void) __typeof__(ptr) __uval = (uval); \ __typeof__(*(ptr)) __old = (old); \ __typeof__(*(ptr)) __new = (new); \ + pax_open_userland(); \ switch (size) { \ case 1: \ { \ asm volatile("\t" ASM_STAC "\n" \ - "1:\t" LOCK_PREFIX "cmpxchgb %4, %2\n" \ + "1:\t" LOCK_PREFIX __copyuser_seg"cmpxchgb %4, %2\n"\ "2:\t" ASM_CLAC "\n" \ "\t.section .fixup, \"ax\"\n" \ "3:\tmov %3, %0\n" \ "\tjmp 2b\n" \ "\t.previous\n" \ _ASM_EXTABLE(1b, 3b) \ - : "+r" (__ret), "=a" (__old), "+m" (*(ptr)) \ + : "+r" (__ret), "=a" (__old), "+m" (*____m(ptr))\ : "i" (-EFAULT), "q" (__new), "1" (__old) \ : "memory" \ ); \ @@ -562,14 +638,14 @@ extern void __cmpxchg_wrong_size(void) case 2: \ { \ asm volatile("\t" ASM_STAC "\n" \ - "1:\t" LOCK_PREFIX "cmpxchgw %4, %2\n" \ + "1:\t" LOCK_PREFIX __copyuser_seg"cmpxchgw %4, %2\n"\ "2:\t" ASM_CLAC "\n" \ "\t.section .fixup, \"ax\"\n" \ "3:\tmov %3, %0\n" \ "\tjmp 2b\n" \ "\t.previous\n" \ _ASM_EXTABLE(1b, 3b) \ - : "+r" (__ret), "=a" (__old), "+m" (*(ptr)) \ + : "+r" (__ret), "=a" (__old), "+m" (*____m(ptr))\ : "i" (-EFAULT), "r" (__new), "1" (__old) \ : "memory" \ ); \ @@ -578,14 +654,14 @@ extern void __cmpxchg_wrong_size(void) case 4: \ { \ asm volatile("\t" ASM_STAC "\n" \ - "1:\t" LOCK_PREFIX "cmpxchgl %4, %2\n" \ + "1:\t" LOCK_PREFIX __copyuser_seg"cmpxchgl %4, %2\n"\ "2:\t" ASM_CLAC "\n" \ "\t.section .fixup, \"ax\"\n" \ "3:\tmov %3, %0\n" \ "\tjmp 2b\n" \ "\t.previous\n" \ _ASM_EXTABLE(1b, 3b) \ - : "+r" (__ret), "=a" (__old), "+m" (*(ptr)) \ + : "+r" (__ret), "=a" (__old), "+m" (*____m(ptr))\ : "i" (-EFAULT), "r" (__new), "1" (__old) \ : "memory" \ ); \ @@ -597,14 +673,14 @@ extern void __cmpxchg_wrong_size(void) __cmpxchg_wrong_size(); \ \ asm volatile("\t" ASM_STAC "\n" \ - "1:\t" LOCK_PREFIX "cmpxchgq %4, %2\n" \ + "1:\t" LOCK_PREFIX __copyuser_seg"cmpxchgq %4, %2\n"\ "2:\t" ASM_CLAC "\n" \ "\t.section .fixup, \"ax\"\n" \ "3:\tmov %3, %0\n" \ "\tjmp 2b\n" \ "\t.previous\n" \ _ASM_EXTABLE(1b, 3b) \ - : "+r" (__ret), "=a" (__old), "+m" (*(ptr)) \ + : "+r" (__ret), "=a" (__old), "+m" (*____m(ptr))\ : "i" (-EFAULT), "r" (__new), "1" (__old) \ : "memory" \ ); \ @@ -613,6 +689,7 @@ extern void __cmpxchg_wrong_size(void) default: \ __cmpxchg_wrong_size(); \ } \ + pax_close_userland(); \ *__uval = __old; \ __ret; \ }) @@ -636,17 +713,6 @@ extern struct movsl_mask { #define ARCH_HAS_NOCACHE_UACCESS 1 -#ifdef CONFIG_X86_32 -# include -#else -# include -#endif - -unsigned long __must_check _copy_from_user(void *to, const void __user *from, - unsigned n); -unsigned long __must_check _copy_to_user(void __user *to, const void *from, - unsigned n); - #ifdef CONFIG_DEBUG_STRICT_USER_COPY_CHECKS # define copy_user_diag __compiletime_error #else @@ -656,7 +722,7 @@ unsigned long __must_check _copy_to_user(void __user *to, const void *from, extern void copy_user_diag("copy_from_user() buffer size is too small") copy_from_user_overflow(void); extern void copy_user_diag("copy_to_user() buffer size is too small") -copy_to_user_overflow(void) __asm__("copy_from_user_overflow"); +copy_to_user_overflow(void); #undef copy_user_diag @@ -669,7 +735,7 @@ __copy_from_user_overflow(void) __asm__("copy_from_user_overflow"); extern void __compiletime_warning("copy_to_user() buffer size is not provably correct") -__copy_to_user_overflow(void) __asm__("copy_from_user_overflow"); +__copy_to_user_overflow(void) __asm__("copy_to_user_overflow"); #define __copy_to_user_overflow(size, count) __copy_to_user_overflow() #else @@ -684,10 +750,16 @@ __copy_from_user_overflow(int size, unsigned long count) #endif +#ifdef CONFIG_X86_32 +# include +#else +# include +#endif + static inline unsigned long __must_check copy_from_user(void *to, const void __user *from, unsigned long n) { - int sz = __compiletime_object_size(to); + size_t sz = __compiletime_object_size(to); might_fault(); @@ -709,12 +781,15 @@ copy_from_user(void *to, const void __user *from, unsigned long n) * case, and do only runtime checking for non-constant sizes. */ - if (likely(sz < 0 || sz >= n)) - n = _copy_from_user(to, from, n); - else if(__builtin_constant_p(n)) - copy_from_user_overflow(); - else - __copy_from_user_overflow(sz, n); + if (likely(sz != (size_t)-1 && sz < n)) { + if(__builtin_constant_p(n)) + copy_from_user_overflow(); + else + __copy_from_user_overflow(sz, n); + } else if (access_ok(VERIFY_READ, from, n)) + n = __copy_from_user(to, from, n); + else if ((long)n > 0) + memset(to, 0, n); return n; } @@ -722,17 +797,18 @@ copy_from_user(void *to, const void __user *from, unsigned long n) static inline unsigned long __must_check copy_to_user(void __user *to, const void *from, unsigned long n) { - int sz = __compiletime_object_size(from); + size_t sz = __compiletime_object_size(from); might_fault(); /* See the comment in copy_from_user() above. */ - if (likely(sz < 0 || sz >= n)) - n = _copy_to_user(to, from, n); - else if(__builtin_constant_p(n)) - copy_to_user_overflow(); - else - __copy_to_user_overflow(sz, n); + if (likely(sz != (size_t)-1 && sz < n)) { + if(__builtin_constant_p(n)) + copy_to_user_overflow(); + else + __copy_to_user_overflow(sz, n); + } else if (access_ok(VERIFY_WRITE, to, n)) + n = __copy_to_user(to, from, n); return n; } diff --git a/arch/x86/include/asm/uaccess_32.h b/arch/x86/include/asm/uaccess_32.h index 3c03a5d..1071638 100644 --- a/arch/x86/include/asm/uaccess_32.h +++ b/arch/x86/include/asm/uaccess_32.h @@ -43,6 +43,11 @@ 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) { + if ((long)n < 0) + return n; + + check_object_size(from, n, true); + if (__builtin_constant_p(n)) { unsigned long ret; @@ -82,12 +87,16 @@ 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) { + 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 @@ -137,6 +146,12 @@ 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; + + check_object_size(to, n, false); + if (__builtin_constant_p(n)) { unsigned long ret; @@ -159,6 +174,10 @@ 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; @@ -181,7 +200,10 @@ 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); } #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 12a26b9..206c200 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 @@ -17,14 +20,14 @@ /* Handles exceptions in both to and from, but doesn't do access_ok */ __must_check unsigned long -copy_user_enhanced_fast_string(void *to, const void *from, unsigned len); +copy_user_enhanced_fast_string(void *to, const void *from, unsigned len) __size_overflow(3); __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 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 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) { unsigned ret; @@ -46,121 +49,170 @@ copy_user_generic(void *to, const void *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); static __always_inline __must_check -int __copy_from_user_nocheck(void *dst, const void __user *src, unsigned size) +unsigned long __copy_from_user_nocheck(void *dst, const void __user *src, unsigned long size) { - int ret = 0; + size_t sz = __compiletime_object_size(dst); + unsigned ret = 0; + + if (size > INT_MAX) + return size; + + check_object_size(dst, size, false); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!access_ok_noprefault(VERIFY_READ, src, size)) + return size; +#endif + + if (unlikely(sz != (size_t)-1 && sz < size)) { + if(__builtin_constant_p(size)) + copy_from_user_overflow(); + else + __copy_from_user_overflow(sz, size); + return size; + } if (!__builtin_constant_p(size)) - return copy_user_generic(dst, (__force void *)src, size); + return copy_user_generic(dst, (__force_kernel const void *)____m(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); + return copy_user_generic(dst, (__force_kernel const void *)____m(src), 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) { might_fault(); return __copy_from_user_nocheck(dst, src, size); } static __always_inline __must_check -int __copy_to_user_nocheck(void __user *dst, const void *src, unsigned size) +unsigned long __copy_to_user_nocheck(void __user *dst, const void *src, unsigned long size) { - int ret = 0; + size_t sz = __compiletime_object_size(src); + unsigned ret = 0; + + if (size > INT_MAX) + return size; + + check_object_size(src, size, true); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!access_ok_noprefault(VERIFY_WRITE, dst, size)) + return size; +#endif + + if (unlikely(sz != (size_t)-1 && sz < size)) { + if(__builtin_constant_p(size)) + copy_to_user_overflow(); + else + __copy_to_user_overflow(sz, size); + return size; + } if (!__builtin_constant_p(size)) - return copy_user_generic((__force void *)dst, src, size); + return copy_user_generic((__force_kernel void *)____m(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); + return copy_user_generic((__force_kernel void *)____m(dst), 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) { might_fault(); return __copy_to_user_nocheck(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 size) { - int ret = 0; + unsigned ret = 0; might_fault(); + + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!access_ok_noprefault(VERIFY_READ, src, size)) + return size; + if (!access_ok_noprefault(VERIFY_WRITE, dst, size)) + return size; +#endif + if (!__builtin_constant_p(size)) - return copy_user_generic((__force void *)dst, - (__force void *)src, size); + return copy_user_generic((__force_kernel void *)____m(dst), + (__force_kernel const void *)____m(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, @@ -169,7 +221,7 @@ 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, @@ -179,7 +231,7 @@ 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, @@ -188,7 +240,7 @@ 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, @@ -196,41 +248,58 @@ 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); + return copy_user_generic((__force_kernel void *)____m(dst), + (__force_kernel const void *)____m(src), size); } } -static __must_check __always_inline int -__copy_from_user_inatomic(void *dst, const void __user *src, unsigned size) +static __must_check __always_inline unsigned long +__copy_from_user_inatomic(void *dst, const void __user *src, unsigned long size) { return __copy_from_user_nocheck(dst, 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) { return __copy_to_user_nocheck(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); -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) { might_fault(); + + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!access_ok_noprefault(VERIFY_READ, src, size)) + return size; +#endif + return __copy_user_nocache(dst, src, size, 1); } -static inline int +static inline unsigned long __copy_from_user_inatomic_nocache(void *dst, const void __user *src, - unsigned size) + unsigned long size) { + if (size > INT_MAX) + return size; + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!access_ok_noprefault(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); +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/word-at-a-time.h b/arch/x86/include/asm/word-at-a-time.h index 5b238981..77fdd78 100644 --- a/arch/x86/include/asm/word-at-a-time.h +++ b/arch/x86/include/asm/word-at-a-time.h @@ -11,7 +11,7 @@ * and shift, for example. */ struct word_at_a_time { - const unsigned long one_bits, high_bits; + unsigned long one_bits, high_bits; }; #define WORD_AT_A_TIME_CONSTANTS { REPEAT_BYTE(0x01), REPEAT_BYTE(0x80) } diff --git a/arch/x86/include/asm/x86_init.h b/arch/x86/include/asm/x86_init.h index e45e4da..44e8572 100644 --- a/arch/x86/include/asm/x86_init.h +++ b/arch/x86/include/asm/x86_init.h @@ -129,7 +129,7 @@ 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 @@ -140,7 +140,7 @@ struct x86_cpuinit_ops { void (*setup_percpu_clockev)(void); void (*early_percpu_clock_init)(void); void (*fixup_cpu_id)(struct cpuinfo_x86 *c, int node); -}; +} __no_const; struct timespec; @@ -168,7 +168,7 @@ struct x86_platform_ops { void (*save_sched_clock_state)(void); void (*restore_sched_clock_state)(void); void (*apic_post_init)(void); -}; +} __no_const; struct pci_dev; struct msi_msg; @@ -185,7 +185,7 @@ struct x86_msi_ops { int (*setup_hpet_msi)(unsigned int irq, unsigned int id); u32 (*msi_mask_irq)(struct msi_desc *desc, u32 mask, u32 flag); u32 (*msix_mask_irq)(struct msi_desc *desc, u32 flag); -}; +} __no_const; struct IO_APIC_route_entry; struct io_apic_irq_attr; @@ -206,7 +206,7 @@ struct x86_io_apic_ops { unsigned int destination, int vector, struct io_apic_irq_attr *attr); void (*eoi_ioapic_pin)(int apic, int pin, int vector); -}; +} __no_const; extern struct x86_init_ops x86_init; extern struct x86_cpuinit_ops x86_cpuinit; diff --git a/arch/x86/include/asm/xen/page.h b/arch/x86/include/asm/xen/page.h index c949923..c22bfa4 100644 --- a/arch/x86/include/asm/xen/page.h +++ b/arch/x86/include/asm/xen/page.h @@ -63,7 +63,7 @@ extern int m2p_remove_override(struct page *page, extern struct page *m2p_find_override(unsigned long mfn); extern unsigned long m2p_find_override_pfn(unsigned long mfn, unsigned long pfn); -static inline unsigned long pfn_to_mfn(unsigned long pfn) +static inline unsigned long __intentional_overflow(-1) pfn_to_mfn(unsigned long pfn) { unsigned long mfn; diff --git a/arch/x86/include/asm/xsave.h b/arch/x86/include/asm/xsave.h index d949ef2..479b9d1 100644 --- a/arch/x86/include/asm/xsave.h +++ b/arch/x86/include/asm/xsave.h @@ -82,8 +82,11 @@ static inline int xsave_user(struct xsave_struct __user *buf) if (unlikely(err)) return -EFAULT; + pax_open_userland(); __asm__ __volatile__(ASM_STAC "\n" - "1: .byte " REX_PREFIX "0x0f,0xae,0x27\n" + "1:" + __copyuser_seg + ".byte " REX_PREFIX "0x0f,0xae,0x27\n" "2: " ASM_CLAC "\n" ".section .fixup,\"ax\"\n" "3: movl $-1,%[err]\n" @@ -93,18 +96,22 @@ static inline int xsave_user(struct xsave_struct __user *buf) : [err] "=r" (err) : "D" (buf), "a" (-1), "d" (-1), "0" (0) : "memory"); + pax_close_userland(); return err; } 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; + pax_open_userland(); __asm__ __volatile__(ASM_STAC "\n" - "1: .byte " REX_PREFIX "0x0f,0xae,0x2f\n" + "1:" + __copyuser_seg + ".byte " REX_PREFIX "0x0f,0xae,0x2f\n" "2: " ASM_CLAC "\n" ".section .fixup,\"ax\"\n" "3: movl $-1,%[err]\n" @@ -114,6 +121,7 @@ static inline int xrestore_user(struct xsave_struct __user *buf, u64 mask) : [err] "=r" (err) : "D" (xstate), "a" (lmask), "d" (hmask), "0" (0) : "memory"); /* memory required? */ + pax_close_userland(); return err; } diff --git a/arch/x86/include/uapi/asm/e820.h b/arch/x86/include/uapi/asm/e820.h index bbae024..e1528f9 100644 --- a/arch/x86/include/uapi/asm/e820.h +++ b/arch/x86/include/uapi/asm/e820.h @@ -63,7 +63,7 @@ struct e820map { #define ISA_START_ADDRESS 0xa0000 #define ISA_END_ADDRESS 0x100000 -#define BIOS_BEGIN 0x000a0000 +#define BIOS_BEGIN 0x000c0000 #define BIOS_END 0x00100000 #define BIOS_ROM_BASE 0xffe00000 diff --git a/arch/x86/include/uapi/asm/ptrace-abi.h b/arch/x86/include/uapi/asm/ptrace-abi.h index 7b0a55a..ad115bf 100644 --- a/arch/x86/include/uapi/asm/ptrace-abi.h +++ b/arch/x86/include/uapi/asm/ptrace-abi.h @@ -49,7 +49,6 @@ #define EFLAGS 144 #define RSP 152 #define SS 160 -#define ARGOFFSET R11 #endif /* __ASSEMBLY__ */ /* top of stack page */ diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 047f9ff..4ba5ea6 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -24,7 +24,7 @@ obj-y += time.o ioport.o ldt.o dumpstack.o nmi.o obj-y += setup.o x86_init.o i8259.o irqinit.o jump_label.o obj-$(CONFIG_IRQ_WORK) += irq_work.o obj-y += probe_roms.o -obj-$(CONFIG_X86_32) += i386_ksyms_32.o +obj-$(CONFIG_X86_32) += sys_i386_32.o i386_ksyms_32.o obj-$(CONFIG_X86_64) += sys_x86_64.o x8664_ksyms_64.o obj-$(CONFIG_X86_64) += mcount_64.o obj-y += syscall_$(BITS).o vsyscall_gtod.o diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 86281ff..e046fc2 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -1296,7 +1296,7 @@ static int __init dmi_ignore_irq0_timer_override(const struct dmi_system_id *d) * If your system is blacklisted here, but you find that acpi=force * works for you, please contact linux-acpi@vger.kernel.org */ -static struct dmi_system_id __initdata acpi_dmi_table[] = { +static const struct dmi_system_id __initconst acpi_dmi_table[] = { /* * Boxes that need ACPI disabled */ @@ -1371,7 +1371,7 @@ static struct dmi_system_id __initdata acpi_dmi_table[] = { }; /* second table for DMI checks that should run after early-quirks */ -static struct dmi_system_id __initdata acpi_dmi_table_late[] = { +static const struct dmi_system_id __initconst acpi_dmi_table_late[] = { /* * HP laptops which use a DSDT reporting as HP/SB400/10000, * which includes some code which overrides all temperature diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index 3136820..e2c6577 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -99,8 +99,12 @@ int x86_acpi_suspend_lowlevel(void) #else /* CONFIG_64BIT */ #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 665c6b7..eae4d56 100644 --- a/arch/x86/kernel/acpi/wakeup_32.S +++ b/arch/x86/kernel/acpi/wakeup_32.S @@ -29,13 +29,11 @@ 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 703130f..27a155d 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -268,6 +268,13 @@ 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)); @@ -284,6 +291,11 @@ void __init_or_module apply_alternatives(struct alt_instr *start, add_nops(insnbuf + a->replacementlen, a->instrlen - a->replacementlen); +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + if (instr < (u8 *)_text || (u8 *)_einittext <= instr) + instr = ktva_ktla(instr); +#endif + text_poke_early(instr, insnbuf, a->instrlen); } } @@ -299,10 +311,16 @@ 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); @@ -317,10 +335,16 @@ 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); @@ -457,7 +481,7 @@ 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); @@ -504,7 +528,7 @@ void __init alternative_instructions(void) if (!uniproc_patched || num_possible_cpus() == 1) free_init_pages("SMP alternatives", (unsigned long)__smp_locks, - (unsigned long)__smp_locks_end); + PAGE_ALIGN((unsigned long)__smp_locks_end)); #endif apply_paravirt(__parainstructions, __parainstructions_end); @@ -524,13 +548,17 @@ 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. */ @@ -552,36 +580,22 @@ void *__init_or_module text_poke_early(void *addr, const void *opcode, */ void *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; } @@ -601,7 +615,7 @@ int poke_int3_handler(struct pt_regs *regs) if (likely(!bp_patching_in_progress)) return 0; - if (user_mode_vm(regs) || regs->ip != (unsigned long)bp_int3_addr) + if (user_mode(regs) || regs->ip != (unsigned long)bp_int3_addr) return 0; /* set up the specified breakpoint handler */ @@ -635,7 +649,7 @@ int poke_int3_handler(struct pt_regs *regs) */ void *text_poke_bp(void *addr, const void *opcode, size_t len, void *handler) { - unsigned char int3 = 0xcc; + const unsigned char int3 = 0xcc; bp_int3_handler = handler; bp_int3_addr = (u8 *)addr + sizeof(int3); diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index ad28db7..c538b2c 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -201,7 +201,7 @@ int first_system_vector = 0xfe; /* * Debug level, exported for io_apic.c */ -unsigned int apic_verbosity; +int apic_verbosity; int pic_mode; @@ -2000,7 +2000,7 @@ static inline void __smp_error_interrupt(struct pt_regs *regs) apic_write(APIC_ESR, 0); v = 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", smp_processor_id(), v); diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index 7c1b294..e71d27f 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -154,7 +154,7 @@ static int flat_probe(void) return 1; } -static struct apic apic_flat = { +static struct apic apic_flat __read_only = { .name = "flat", .probe = flat_probe, .acpi_madt_oem_check = flat_acpi_madt_oem_check, @@ -268,7 +268,7 @@ static int physflat_probe(void) return 0; } -static struct apic apic_physflat = { +static struct apic apic_physflat __read_only = { .name = "physical flat", .probe = physflat_probe, diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c index 8c7c982..a225910 100644 --- a/arch/x86/kernel/apic/apic_noop.c +++ b/arch/x86/kernel/apic/apic_noop.c @@ -118,7 +118,7 @@ static void noop_apic_write(u32 reg, u32 v) WARN_ON_ONCE(cpu_has_apic && !disable_apic); } -struct apic apic_noop = { +struct apic apic_noop __read_only = { .name = "noop", .probe = noop_probe, .acpi_madt_oem_check = NULL, diff --git a/arch/x86/kernel/apic/bigsmp_32.c b/arch/x86/kernel/apic/bigsmp_32.c index e4840aa..e7d9dac 100644 --- a/arch/x86/kernel/apic/bigsmp_32.c +++ b/arch/x86/kernel/apic/bigsmp_32.c @@ -152,7 +152,7 @@ static int probe_bigsmp(void) return dmi_bigsmp; } -static struct apic apic_bigsmp = { +static struct apic apic_bigsmp __read_only = { .name = "bigsmp", .probe = probe_bigsmp, diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 81e08ef..abc77e5 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -1042,7 +1042,7 @@ 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. @@ -1050,7 +1050,7 @@ 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); } @@ -2349,7 +2349,7 @@ static void ack_apic_edge(struct irq_data *data) ack_APIC_irq(); } -atomic_t irq_mis_count; +atomic_unchecked_t irq_mis_count; #ifdef CONFIG_GENERIC_PENDING_IRQ static bool io_apic_level_ack_pending(struct irq_cfg *cfg) @@ -2490,7 +2490,7 @@ 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/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c index cceb352..a635fd8 100644 --- a/arch/x86/kernel/apic/probe_32.c +++ b/arch/x86/kernel/apic/probe_32.c @@ -72,7 +72,7 @@ static int probe_default(void) return 1; } -static struct apic apic_default = { +static struct apic apic_default __read_only = { .name = "default", .probe = probe_default, diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index e66766b..1c008ba 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -182,7 +182,7 @@ update_clusterinfo(struct notifier_block *nfb, unsigned long action, void *hcpu) return notifier_from_errno(err); } -static struct notifier_block __refdata x2apic_cpu_notifier = { +static struct notifier_block x2apic_cpu_notifier = { .notifier_call = update_clusterinfo, }; @@ -234,7 +234,7 @@ static void cluster_vector_allocation_domain(int cpu, struct cpumask *retmask, cpumask_and(retmask, mask, per_cpu(cpus_in_cluster, cpu)); } -static struct apic apic_x2apic_cluster = { +static struct apic apic_x2apic_cluster __read_only = { .name = "cluster x2apic", .probe = x2apic_cluster_probe, diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c index 6d600eb..0300c00 100644 --- a/arch/x86/kernel/apic/x2apic_phys.c +++ b/arch/x86/kernel/apic/x2apic_phys.c @@ -88,7 +88,7 @@ static int x2apic_phys_probe(void) return apic == &apic_x2apic_phys; } -static struct apic apic_x2apic_phys = { +static struct apic apic_x2apic_phys __read_only = { .name = "physical x2apic", .probe = x2apic_phys_probe, diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 293b41d..4df25fd 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -350,7 +350,7 @@ static int uv_probe(void) return apic == &apic_x2apic_uv_x; } -static struct apic __refdata apic_x2apic_uv_x = { +static struct apic apic_x2apic_uv_x __read_only = { .name = "UV large system", .probe = uv_probe, diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index 5848744..56cb598 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -433,7 +433,7 @@ 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 */ @@ -611,7 +611,10 @@ 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; @@ -620,7 +623,11 @@ 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; @@ -687,7 +694,10 @@ 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; @@ -695,7 +705,11 @@ 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; } @@ -2350,12 +2364,15 @@ 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 9f6b934..cf5ffb3 100644 --- a/arch/x86/kernel/asm-offsets.c +++ b/arch/x86/kernel/asm-offsets.c @@ -32,6 +32,8 @@ void common(void) { OFFSET(TI_flags, thread_info, flags); OFFSET(TI_status, thread_info, status); OFFSET(TI_addr_limit, thread_info, addr_limit); + 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); @@ -52,8 +54,26 @@ 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 e7c798b..2b2019b 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -77,6 +77,7 @@ 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 7fd54f0..0691410 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 += rdrand.o diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index ce8b8ff..d7d8851 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -728,7 +728,7 @@ static void init_amd(struct cpuinfo_x86 *c) static unsigned int 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 ef1b93f..150db65 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -90,60 +90,6 @@ static const struct cpu_dev default_cpu = { static const struct cpu_dev *this_cpu = &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); @@ -295,6 +241,59 @@ static __always_inline void setup_smap(struct cpuinfo_x86 *c) } } +#ifdef CONFIG_X86_64 +static __init int setup_disable_pcid(char *arg) +{ + setup_clear_cpu_cap(X86_FEATURE_PCID); + setup_clear_cpu_cap(X86_FEATURE_INVPCID); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (clone_pgd_mask != ~(pgdval_t)0UL) + pax_user_shadow_base = 1UL << TASK_SIZE_MAX_SHIFT; +#endif + + return 1; +} +__setup("nopcid", setup_disable_pcid); + +static void setup_pcid(struct cpuinfo_x86 *c) +{ + if (!cpu_has(c, X86_FEATURE_PCID)) { + clear_cpu_cap(c, X86_FEATURE_INVPCID); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (clone_pgd_mask != ~(pgdval_t)0UL) { + pax_open_kernel(); + pax_user_shadow_base = 1UL << TASK_SIZE_MAX_SHIFT; + pax_close_kernel(); + printk("PAX: slow and weak UDEREF enabled\n"); + } else + printk("PAX: UDEREF disabled\n"); +#endif + + return; + } + + printk("PAX: PCID detected\n"); + set_in_cr4(X86_CR4_PCIDE); + +#ifdef CONFIG_PAX_MEMORY_UDEREF + pax_open_kernel(); + clone_pgd_mask = ~(pgdval_t)0UL; + pax_close_kernel(); + if (pax_user_shadow_base) + printk("PAX: weak UDEREF enabled\n"); + else { + set_cpu_cap(c, X86_FEATURE_STRONGUDEREF); + printk("PAX: strong UDEREF enabled\n"); + } +#endif + + if (cpu_has(c, X86_FEATURE_INVPCID)) + printk("PAX: INVPCID detected\n"); +} +#endif + /* * Some CPU features depend on higher CPUID levels, which may not always * be available due to CPUID level capping or broken virtualization @@ -395,7 +394,7 @@ 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 */ @@ -885,6 +884,10 @@ static void identify_cpu(struct cpuinfo_x86 *c) setup_smep(c); setup_smap(c); +#ifdef CONFIG_X86_64 + setup_pcid(c); +#endif + /* * The vendor-specific functions might have changed features. * Now we do "generic changes." @@ -893,6 +896,10 @@ static void 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; @@ -973,7 +980,7 @@ static void syscall32_cpu_init(void) 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(); @@ -1113,14 +1120,16 @@ 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); + 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); #ifdef CONFIG_X86_64 -struct desc_ptr idt_descr = { NR_VECTORS * 16 - 1, (unsigned long) idt_table }; -struct desc_ptr debug_idt_descr = { NR_VECTORS * 16 - 1, - (unsigned long) debug_idt_table }; +struct desc_ptr idt_descr __read_only = { NR_VECTORS * 16 - 1, (unsigned long) idt_table }; +const struct desc_ptr debug_idt_descr = { NR_VECTORS * 16 - 1, (unsigned long) debug_idt_table }; DEFINE_PER_CPU_FIRST(union irq_stack_union, irq_stack_union) __aligned(PAGE_SIZE) __visible; @@ -1283,7 +1292,7 @@ void cpu_init(void) load_ucode_ap(); cpu = stack_smp_processor_id(); - t = &per_cpu(init_tss, cpu); + t = init_tss + cpu; oist = &per_cpu(orig_ist, cpu); #ifdef CONFIG_NUMA @@ -1318,7 +1327,6 @@ void cpu_init(void) wrmsrl(MSR_KERNEL_GS_BASE, 0); barrier(); - x86_configure_nx(); enable_x2apic(); /* @@ -1370,7 +1378,7 @@ void 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; show_ucode_info_early(); diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 9c8f739..902a9c5 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -1026,6 +1026,22 @@ static struct attribute *default_attrs[] = { }; #ifdef CONFIG_AMD_NB +static struct attribute *default_attrs_amd_nb[] = { + &type.attr, + &level.attr, + &coherency_line_size.attr, + &physical_line_partition.attr, + &ways_of_associativity.attr, + &number_of_sets.attr, + &size.attr, + &shared_cpu_map.attr, + &shared_cpu_list.attr, + NULL, + NULL, + NULL, + NULL +}; + static struct attribute **amd_l3_attrs(void) { static struct attribute **attrs; @@ -1036,18 +1052,7 @@ static struct attribute **amd_l3_attrs(void) n = ARRAY_SIZE(default_attrs); - if (amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE)) - n += 2; - - if (amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) - n += 1; - - attrs = kzalloc(n * sizeof (struct attribute *), GFP_KERNEL); - if (attrs == NULL) - return attrs = default_attrs; - - for (n = 0; default_attrs[n]; n++) - attrs[n] = default_attrs[n]; + attrs = default_attrs_amd_nb; if (amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE)) { attrs[n++] = &cache_disable_0.attr; @@ -1098,6 +1103,13 @@ static struct kobj_type ktype_cache = { .default_attrs = default_attrs, }; +#ifdef CONFIG_AMD_NB +static struct kobj_type ktype_cache_amd_nb = { + .sysfs_ops = &sysfs_ops, + .default_attrs = default_attrs_amd_nb, +}; +#endif + static struct kobj_type ktype_percpu_entry = { .sysfs_ops = &sysfs_ops, }; @@ -1163,20 +1175,26 @@ static int cache_add_dev(struct device *dev) return retval; } +#ifdef CONFIG_AMD_NB + amd_l3_attrs(); +#endif + for (i = 0; i < num_cache_leaves; i++) { + struct kobj_type *ktype; + this_object = INDEX_KOBJECT_PTR(cpu, i); this_object->cpu = cpu; this_object->index = i; this_leaf = CPUID4_INFO_IDX(cpu, i); - ktype_cache.default_attrs = default_attrs; + ktype = &ktype_cache; #ifdef CONFIG_AMD_NB if (this_leaf->base.nb) - ktype_cache.default_attrs = amd_l3_attrs(); + ktype = &ktype_cache_amd_nb; #endif retval = kobject_init_and_add(&(this_object->kobj), - &ktype_cache, + ktype, per_cpu(ici_cache_kobject, cpu), "index%1lu", i); if (unlikely(retval)) { diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index 9a79c8d..158c2f1 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -45,6 +45,7 @@ #include #include #include +#include #include "mce-internal.h" @@ -259,7 +260,7 @@ 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"); } @@ -292,10 +293,10 @@ 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) @@ -319,7 +320,7 @@ 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(); @@ -327,7 +328,7 @@ 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 */ @@ -366,7 +367,7 @@ static void mce_panic(char *msg, struct mce *final, char *exp) if (!fake_panic) { if (panic_timeout == 0) panic_timeout = mca_cfg.panic_timeout; - panic(msg); + panic("%s", msg); } else pr_emerg(HW_ERR "Fake kernel panic: %s\n", msg); } @@ -697,7 +698,7 @@ 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 (!mca_cfg.monarch_timeout) goto out; @@ -1674,7 +1675,7 @@ 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; /* @@ -1697,7 +1698,9 @@ void 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); @@ -1711,7 +1714,7 @@ void 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) @@ -1719,7 +1722,7 @@ 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; @@ -1727,7 +1730,7 @@ 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); @@ -1738,7 +1741,7 @@ 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); @@ -2414,7 +2417,7 @@ static __init void mce_init_banks(void) for (i = 0; i < mca_cfg.banks; i++) { struct mce_bank *b = &mce_banks[i]; - struct device_attribute *a = &b->attr; + device_attribute_no_const *a = &b->attr; sysfs_attr_init(&a->attr); a->attr.name = b->attrname; @@ -2521,7 +2524,7 @@ 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 a304298..49b6d06 100644 --- a/arch/x86/kernel/cpu/mcheck/p5.c +++ b/arch/x86/kernel/cpu/mcheck/p5.c @@ -10,6 +10,7 @@ #include #include #include +#include /* By default disabled */ int mce_p5_enabled __read_mostly; @@ -48,7 +49,9 @@ 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 7dc5564..1273569 100644 --- a/arch/x86/kernel/cpu/mcheck/winchip.c +++ b/arch/x86/kernel/cpu/mcheck/winchip.c @@ -9,6 +9,7 @@ #include #include #include +#include /* Machine check handler for WinChip C6: */ static void winchip_machine_check(struct pt_regs *regs, long error_code) @@ -22,7 +23,9 @@ 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/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c index dd9d619..86e1d81 100644 --- a/arch/x86/kernel/cpu/microcode/core.c +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -516,7 +516,7 @@ mc_cpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) return NOTIFY_OK; } -static struct notifier_block __refdata mc_cpu_notifier = { +static struct notifier_block mc_cpu_notifier = { .notifier_call = mc_cpu_callback, }; diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c index a276fa7..e66810f 100644 --- a/arch/x86/kernel/cpu/microcode/intel.c +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -293,13 +293,13 @@ static enum ucode_state request_microcode_fw(int cpu, struct device *device, 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/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c index f961de9..8a9d332 100644 --- a/arch/x86/kernel/cpu/mtrr/main.c +++ b/arch/x86/kernel/cpu/mtrr/main.c @@ -66,7 +66,7 @@ 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 2879ecd..bb8c80b 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1372,7 +1372,7 @@ static void __init pmu_check_apic(void) } -static struct attribute_group x86_pmu_format_group = { +static attribute_group_no_const x86_pmu_format_group = { .name = "format", .attrs = NULL, }; @@ -1471,7 +1471,7 @@ static struct attribute *events_attr[] = { NULL, }; -static struct attribute_group x86_pmu_events_group = { +static attribute_group_no_const x86_pmu_events_group = { .name = "events", .attrs = events_attr, }; @@ -1995,7 +1995,7 @@ static unsigned long get_segment_base(unsigned int segment) if (idx > GDT_ENTRIES) return 0; - desc = __this_cpu_ptr(&gdt_page.gdt[0]); + desc = get_cpu_gdt_table(smp_processor_id()); } return get_desc_base(desc + idx); @@ -2085,7 +2085,7 @@ 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/cpu/perf_event_amd_iommu.c b/arch/x86/kernel/cpu/perf_event_amd_iommu.c index 639d128..e92d7e5 100644 --- a/arch/x86/kernel/cpu/perf_event_amd_iommu.c +++ b/arch/x86/kernel/cpu/perf_event_amd_iommu.c @@ -405,7 +405,7 @@ static void perf_iommu_del(struct perf_event *event, int flags) static __init int _init_events_attrs(struct perf_amd_iommu *perf_iommu) { struct attribute **attrs; - struct attribute_group *attr_group; + attribute_group_no_const *attr_group; int i = 0, j; while (amd_iommu_v2_event_descs[i].attr.attr.name) diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 2502d0d..e5cc05c 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -2353,10 +2353,10 @@ __init int intel_pmu_init(void) x86_pmu.num_counters_fixed = max((int)edx.split.num_counters_fixed, 3); if (boot_cpu_has(X86_FEATURE_PDCM)) { - u64 capabilities; + u64 capabilities = x86_pmu.intel_cap.capabilities; - rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); - x86_pmu.intel_cap.capabilities = capabilities; + if (rdmsrl_safe(MSR_IA32_PERF_CAPABILITIES, &x86_pmu.intel_cap.capabilities)) + x86_pmu.intel_cap.capabilities = capabilities; } intel_ds_init(); diff --git a/arch/x86/kernel/cpu/perf_event_intel_rapl.c b/arch/x86/kernel/cpu/perf_event_intel_rapl.c index 619f769..d510008 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_rapl.c +++ b/arch/x86/kernel/cpu/perf_event_intel_rapl.c @@ -449,7 +449,7 @@ static struct attribute *rapl_events_hsw_attr[] = { NULL, }; -static struct attribute_group rapl_pmu_events_group = { +static attribute_group_no_const rapl_pmu_events_group __read_only = { .name = "events", .attrs = NULL, /* patched at runtime */ }; diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c index ae6552a..b5be2d3 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_uncore.c +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.c @@ -3694,7 +3694,7 @@ static void __init uncore_types_exit(struct intel_uncore_type **types) static int __init uncore_type_init(struct intel_uncore_type *type) { struct intel_uncore_pmu *pmus; - struct attribute_group *attr_group; + attribute_group_no_const *attr_group; struct attribute **attrs; int i, j; diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.h b/arch/x86/kernel/cpu/perf_event_intel_uncore.h index 90236f0..54cb20d 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_uncore.h +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.h @@ -503,7 +503,7 @@ struct intel_uncore_box { struct uncore_event_desc { struct kobj_attribute attr; const char *config; -}; +} __do_const; #define INTEL_UNCORE_EVENT_DESC(_name, _config) \ { \ diff --git a/arch/x86/kernel/cpuid.c b/arch/x86/kernel/cpuid.c index 3225ae6c..ee3c6db 100644 --- a/arch/x86/kernel/cpuid.c +++ b/arch/x86/kernel/cpuid.c @@ -170,7 +170,7 @@ static int cpuid_class_cpu_callback(struct notifier_block *nfb, return notifier_from_errno(err); } -static struct notifier_block __refdata cpuid_class_cpu_notifier = +static struct notifier_block cpuid_class_cpu_notifier = { .notifier_call = cpuid_class_cpu_callback, }; diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index 507de80..ebaae2a 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -58,7 +58,7 @@ static void kdump_nmi_callback(int cpu, struct pt_regs *regs) #ifdef CONFIG_X86_32 struct pt_regs fixed_regs; - 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/crash_dump_64.c b/arch/x86/kernel/crash_dump_64.c index afa64ad..dce67dd 100644 --- a/arch/x86/kernel/crash_dump_64.c +++ b/arch/x86/kernel/crash_dump_64.c @@ -36,7 +36,7 @@ ssize_t copy_oldmem_page(unsigned long pfn, char *buf, return -ENOMEM; if (userbuf) { - if (copy_to_user(buf, vaddr + offset, csize)) { + if (copy_to_user((char __force_user *)buf, vaddr + offset, csize)) { iounmap(vaddr); return -EFAULT; } diff --git a/arch/x86/kernel/doublefault.c b/arch/x86/kernel/doublefault.c index f6dfd93..892ade4 100644 --- a/arch/x86/kernel/doublefault.c +++ b/arch/x86/kernel/doublefault.c @@ -12,7 +12,7 @@ #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) @@ -22,7 +22,7 @@ static void doublefault_fn(void) unsigned long gdt, tss; native_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); @@ -59,10 +59,10 @@ 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 b74ebc7..2c95874 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 @@ -33,23 +36,21 @@ static void printk_stack_address(unsigned long address, int reliable) void printk_address(unsigned long address) { - pr_cont(" [<%p>] %pS\n", (void *)address, (void *)address); + pr_cont(" [<%p>] %pA\n", (void *)address, (void *)address); } #ifdef CONFIG_FUNCTION_GRAPH_TRACER 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; unsigned long ret_addr; int index; if (addr != (unsigned long)return_to_handler) return; - task = tinfo->task; index = task->curr_ret_stack; if (!task->ret_stack || index < *graph) @@ -66,7 +67,7 @@ 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 @@ -77,10 +78,8 @@ 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; @@ -91,14 +90,14 @@ 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; @@ -110,7 +109,7 @@ 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++; } @@ -119,7 +118,7 @@ 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) @@ -127,7 +126,7 @@ 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)) @@ -136,7 +135,7 @@ 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; @@ -155,7 +154,7 @@ static int print_trace_stack(void *data, char *name) static void print_trace_address(void *data, unsigned long addr, int reliable) { touch_nmi_watchdog(); - printk(data); + printk("%s", (char *)data); printk_stack_address(addr, reliable); } @@ -225,6 +224,8 @@ unsigned long oops_begin(void) EXPORT_SYMBOL_GPL(oops_begin); NOKPROBE_SYMBOL(oops_begin); +extern void gr_handle_kernel_exploit(void); + void oops_end(unsigned long flags, struct pt_regs *regs, int signr) { if (regs && kexec_should_crash(current)) @@ -246,7 +247,10 @@ void 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); } NOKPROBE_SYMBOL(oops_end); @@ -275,7 +279,7 @@ int __die(const char *str, struct pt_regs *regs, long err) print_modules(); show_regs(regs); #ifdef CONFIG_X86_32 - if (user_mode_vm(regs)) { + if (user_mode(regs)) { sp = regs->sp; ss = regs->ss & 0xffff; } else { @@ -304,7 +308,7 @@ 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 5abd4cd..c65733b 100644 --- a/arch/x86/kernel/dumpstack_32.c +++ b/arch/x86/kernel/dumpstack_32.c @@ -61,15 +61,14 @@ 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)); void *end_stack; end_stack = is_hardirq_stack(stack, cpu); if (!end_stack) end_stack = is_softirq_stack(stack, cpu); - context = task_thread_info(task); - bp = ops->walk_stack(context, stack, bp, ops, data, + bp = ops->walk_stack(task, stack_start, stack, bp, ops, data, end_stack, &graph); /* Stop if not on irq stack */ @@ -123,27 +122,28 @@ void show_regs(struct pt_regs *regs) int i; show_regs_print_info(KERN_EMERG); - __show_regs(regs, !user_mode_vm(regs)); + __show_regs(regs, !user_mode(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(0)[(0xffff & regs->cs) >> 3]); pr_emerg("Stack:\n"); show_stack_log_lvl(NULL, regs, ®s->sp, 0, KERN_EMERG); pr_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++) { @@ -152,7 +152,7 @@ void show_regs(struct pt_regs *regs) pr_cont(" Bad EIP value."); break; } - if (ip == (u8 *)regs->ip) + if (ip == (u8 *)regs->ip + cs_base) pr_cont(" <%02x>", c); else pr_cont(" %02x", c); @@ -165,6 +165,7 @@ 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)) @@ -172,3 +173,15 @@ int is_valid_bugaddr(unsigned long ip) return ud2 == 0x0b0f; } + +#if defined(CONFIG_PAX_MEMORY_STACKLEAK) || defined(CONFIG_PAX_USERCOPY) +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 1abcb50..6c8d702 100644 --- a/arch/x86/kernel/dumpstack_64.c +++ b/arch/x86/kernel/dumpstack_64.c @@ -154,12 +154,12 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, const struct stacktrace_ops *ops, void *data) { const unsigned cpu = get_cpu(); - struct thread_info *tinfo; unsigned long *irq_stack = (unsigned long *)per_cpu(irq_stack_ptr, cpu); unsigned long dummy; unsigned used = 0; int graph = 0; int done = 0; + void *stack_start; if (!task) task = current; @@ -180,7 +180,6 @@ 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); while (!done) { unsigned long *stack_end; enum stack_type stype; @@ -203,7 +202,7 @@ 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, stack_end - EXCEPTION_STKSZ, stack, bp, ops, data, stack_end, &graph); ops->stack(data, ""); /* @@ -211,6 +210,8 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, * second-to-last pointer (index -2 to end) in the * exception stack: */ + if ((u16)stack_end[-1] != __KERNEL_DS) + goto out; stack = (unsigned long *) stack_end[-2]; done = 0; break; @@ -219,7 +220,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, 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, stack_end, &graph); /* * We link to the next stack (which would be @@ -241,7 +242,9 @@ 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); +out: put_cpu(); } EXPORT_SYMBOL(dump_trace); @@ -350,3 +353,50 @@ int is_valid_bugaddr(unsigned long ip) return ud2 == 0x0b0f; } + +#if defined(CONFIG_PAX_MEMORY_STACKLEAK) || defined(CONFIG_PAX_USERCOPY) +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/e820.c b/arch/x86/kernel/e820.c index 988c00a..4f673b6 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -803,8 +803,8 @@ unsigned long __init e820_end_of_low_ram_pfn(void) static void early_panic(char *msg) { - early_printk(msg); - panic(msg); + early_printk("%s", msg); + panic("%s", msg); } static int userdef __initdata; diff --git a/arch/x86/kernel/early_printk.c b/arch/x86/kernel/early_printk.c index 01d1c18..8073693 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 0d0c9d4..f65b4f6 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -177,13 +177,153 @@ /*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 + */ +ENTRY(pax_erase_kstack) + pushl %edi + pushl %ecx + 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 + + cmp $THREAD_SIZE_asm, %ecx + jb 3f + ud2 +3: + + shr $2, %ecx + rep stosl + + mov TI_task_thread_sp0(%ebp), %edi + sub $128, %edi + mov %edi, TI_lowest_stack(%ebp) + + popl %eax + popl %ecx + popl %edi + ret +ENDPROC(pax_erase_kstack) +#endif + +.macro __SAVE_ALL _DS cld PUSH_GS pushl_cfi %fs @@ -206,7 +346,7 @@ 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 @@ -214,6 +354,15 @@ 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 @@ -297,7 +446,7 @@ ENTRY(ret_from_fork) popfl_cfi jmp syscall_exit CFI_ENDPROC -END(ret_from_fork) +ENDPROC(ret_from_fork) ENTRY(ret_from_kernel_thread) CFI_STARTPROC @@ -340,7 +489,15 @@ ret_from_intr: 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 @@ -352,8 +509,8 @@ 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) @@ -365,7 +522,7 @@ need_resched: jz restore_all call preempt_schedule_irq jmp need_resched -END(resume_kernel) +ENDPROC(resume_kernel) #endif CFI_ENDPROC @@ -395,30 +552,45 @@ 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 ASM_STAC 1: movl (%ebp),%ebp ASM_CLAC +#endif + movl %ebp,PT_EBP(%esp) _ASM_EXTABLE(1b,syscall_fault) GET_THREAD_INFO(%ebp) +#ifdef CONFIG_PAX_RANDKSTACK + pax_erase_kstack +#endif + testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp) jnz sysenter_audit sysenter_do_call: @@ -434,12 +606,24 @@ sysenter_after_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 @@ -456,6 +640,9 @@ 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 @@ -481,10 +668,16 @@ 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 .popsection - _ASM_EXTABLE(1b,2b) + _ASM_EXTABLE(1b,4b) + _ASM_EXTABLE(2b,5b) + _ASM_EXTABLE(3b,6b) PTGS_TO_GS_EX ENDPROC(ia32_sysenter_target) @@ -495,6 +688,11 @@ ENTRY(system_call) pushl_cfi %eax # save orig_eax SAVE_ALL GET_THREAD_INFO(%ebp) + +#ifdef CONFIG_PAX_RANDKSTACK + pax_erase_kstack +#endif + # system call tracing in operation / emulation testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp) jnz syscall_trace_entry @@ -514,6 +712,15 @@ 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: @@ -568,14 +775,34 @@ 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 @@ -605,20 +832,18 @@ 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 # vm86-space 1: -#else - movl %esp, %eax #endif TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_NONE) @@ -639,7 +864,7 @@ work_notifysig_v86: movl %eax, %esp jmp 1b #endif -END(work_pending) +ENDPROC(work_pending) # perform syscall exit tracing ALIGN @@ -647,11 +872,14 @@ 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 @@ -664,26 +892,30 @@ 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 ASM_CLAC GET_THREAD_INFO(%ebp) movl $-EFAULT,PT_EAX(%esp) jmp resume_userspace -END(syscall_fault) +ENDPROC(syscall_fault) syscall_badsys: movl $-ENOSYS,%eax jmp syscall_after_call -END(syscall_badsys) +ENDPROC(syscall_badsys) sysenter_badsys: movl $-ENOSYS,%eax jmp sysenter_after_call -END(syscall_badsys) +ENDPROC(sysenter_badsys) CFI_ENDPROC .macro FIXUP_ESPFIX_STACK @@ -696,8 +928,15 @@ END(syscall_badsys) */ #ifdef CONFIG_X86_ESPFIX32 /* 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 @@ -753,7 +992,7 @@ vector=vector+1 .endr 2: jmp common_interrupt .endr -END(irq_entries_start) +ENDPROC(irq_entries_start) .previous END(interrupt) @@ -810,7 +1049,7 @@ 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 @@ -823,7 +1062,7 @@ ENTRY(simd_coprocessor_error) .section .altinstructions,"a" altinstruction_entry 661b, 663f, X86_FEATURE_XMM, 662b-661b, 664f-663f .previous -.section .altinstr_replacement,"ax" +.section .altinstr_replacement,"a" 663: pushl $do_simd_coprocessor_error 664: .previous @@ -832,7 +1071,7 @@ 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 @@ -841,18 +1080,18 @@ 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) iret _ASM_EXTABLE(native_iret, iret_exc) -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) @@ -862,7 +1101,7 @@ ENTRY(overflow) pushl_cfi $do_overflow jmp error_code CFI_ENDPROC -END(overflow) +ENDPROC(overflow) ENTRY(bounds) RING0_INT_FRAME @@ -871,7 +1110,7 @@ ENTRY(bounds) pushl_cfi $do_bounds jmp error_code CFI_ENDPROC -END(bounds) +ENDPROC(bounds) ENTRY(invalid_op) RING0_INT_FRAME @@ -880,7 +1119,7 @@ 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 @@ -889,7 +1128,7 @@ 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 @@ -897,7 +1136,7 @@ ENTRY(invalid_TSS) pushl_cfi $do_invalid_TSS jmp error_code CFI_ENDPROC -END(invalid_TSS) +ENDPROC(invalid_TSS) ENTRY(segment_not_present) RING0_EC_FRAME @@ -905,7 +1144,7 @@ ENTRY(segment_not_present) 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 @@ -913,7 +1152,7 @@ ENTRY(stack_segment) pushl_cfi $do_stack_segment jmp error_code CFI_ENDPROC -END(stack_segment) +ENDPROC(stack_segment) ENTRY(alignment_check) RING0_EC_FRAME @@ -921,7 +1160,7 @@ ENTRY(alignment_check) pushl_cfi $do_alignment_check jmp error_code CFI_ENDPROC -END(alignment_check) +ENDPROC(alignment_check) ENTRY(divide_error) RING0_INT_FRAME @@ -930,7 +1169,7 @@ 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) @@ -940,7 +1179,7 @@ ENTRY(machine_check) pushl_cfi machine_check_vector jmp error_code CFI_ENDPROC -END(machine_check) +ENDPROC(machine_check) #endif ENTRY(spurious_interrupt_bug) @@ -950,7 +1189,7 @@ ENTRY(spurious_interrupt_bug) pushl_cfi $do_spurious_interrupt_bug jmp error_code CFI_ENDPROC -END(spurious_interrupt_bug) +ENDPROC(spurious_interrupt_bug) #ifdef CONFIG_XEN /* Xen doesn't set %esp to be precisely what the normal sysenter @@ -1056,7 +1295,7 @@ BUILD_INTERRUPT3(hyperv_callback_vector, HYPERVISOR_CALLBACK_VECTOR, ENTRY(mcount) ret -END(mcount) +ENDPROC(mcount) ENTRY(ftrace_caller) cmpl $0, function_trace_stop @@ -1089,7 +1328,7 @@ ftrace_graph_call: .globl ftrace_stub ftrace_stub: ret -END(ftrace_caller) +ENDPROC(ftrace_caller) ENTRY(ftrace_regs_caller) pushf /* push flags before compare (in cs location) */ @@ -1193,7 +1432,7 @@ trace: popl %ecx popl %eax jmp ftrace_stub -END(mcount) +ENDPROC(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_TRACER */ @@ -1211,7 +1450,7 @@ ENTRY(ftrace_graph_caller) popl %ecx popl %eax ret -END(ftrace_graph_caller) +ENDPROC(ftrace_graph_caller) .globl return_to_handler return_to_handler: @@ -1272,15 +1511,18 @@ 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 @@ -1323,7 +1565,7 @@ 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 @@ -1363,6 +1605,9 @@ 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 @@ -1400,13 +1645,16 @@ 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 #endif CFI_ENDPROC -END(nmi) +ENDPROC(nmi) ENTRY(int3) RING0_INT_FRAME @@ -1419,14 +1667,14 @@ 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) @@ -1435,6 +1683,6 @@ 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 c844f08..966a50e 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -59,6 +59,8 @@ #include #include #include +#include +#include /* Avoid __ASSEMBLER__'ifying just for this. */ #include @@ -81,6 +83,430 @@ 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 +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + call pax_enter_kernel +#endif + .endm + + .macro pax_exit_kernel +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) + call pax_exit_kernel +#endif + + .endm + +#if defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) +ENTRY(pax_enter_kernel) + pushq %rdi + +#ifdef CONFIG_PARAVIRT + PV_SAVE_REGS(CLBR_RDI) +#endif + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_RDI + bts $16,%rdi + jnc 3f + mov %cs,%edi + cmp $__KERNEL_CS,%edi + jnz 2f +1: +#endif + +#ifdef CONFIG_PAX_MEMORY_UDEREF + 661: jmp 111f + .pushsection .altinstr_replacement, "a" + 662: ASM_NOP2 + .popsection + .pushsection .altinstructions, "a" + altinstruction_entry 661b, 662b, X86_FEATURE_PCID, 2, 2 + .popsection + GET_CR3_INTO_RDI + cmp $0,%dil + jnz 112f + mov $__KERNEL_DS,%edi + mov %edi,%ss + jmp 111f +112: cmp $1,%dil + jz 113f + ud2 +113: sub $4097,%rdi + bts $63,%rdi + SET_RDI_INTO_CR3 + mov $__UDEREF_KERNEL_DS,%edi + mov %edi,%ss +111: +#endif + +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI) +#endif + + popq %rdi + pax_force_retaddr + retq + +#ifdef CONFIG_PAX_KERNEXEC +2: ljmpq __KERNEL_CS,1b +3: ljmpq __KERNEXEC_KERNEL_CS,4f +4: SET_RDI_INTO_CR0 + jmp 1b +#endif +ENDPROC(pax_enter_kernel) + +ENTRY(pax_exit_kernel) + pushq %rdi + +#ifdef CONFIG_PARAVIRT + PV_SAVE_REGS(CLBR_RDI) +#endif + +#ifdef CONFIG_PAX_KERNEXEC + mov %cs,%rdi + cmp $__KERNEXEC_KERNEL_CS,%edi + jz 2f + GET_CR0_INTO_RDI + bts $16,%rdi + jnc 4f +1: +#endif + +#ifdef CONFIG_PAX_MEMORY_UDEREF + 661: jmp 111f + .pushsection .altinstr_replacement, "a" + 662: ASM_NOP2 + .popsection + .pushsection .altinstructions, "a" + altinstruction_entry 661b, 662b, X86_FEATURE_PCID, 2, 2 + .popsection + mov %ss,%edi + cmp $__UDEREF_KERNEL_DS,%edi + jnz 111f + GET_CR3_INTO_RDI + cmp $0,%dil + jz 112f + ud2 +112: add $4097,%rdi + bts $63,%rdi + SET_RDI_INTO_CR3 + mov $__KERNEL_DS,%edi + mov %edi,%ss +111: +#endif + +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI); +#endif + + popq %rdi + pax_force_retaddr + retq + +#ifdef CONFIG_PAX_KERNEXEC +2: GET_CR0_INTO_RDI + btr $16,%rdi + jnc 4f + ljmpq __KERNEL_CS,3f +3: SET_RDI_INTO_CR0 + jmp 1b +4: ud2 + jmp 4b +#endif +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 + pushq %r11 + call pax_randomize_kstack + popq %r11 + 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 + + 661: jmp 111f + .pushsection .altinstr_replacement, "a" + 662: ASM_NOP2 + .popsection + .pushsection .altinstructions, "a" + altinstruction_entry 661b, 662b, X86_FEATURE_PCID, 2, 2 + .popsection + GET_CR3_INTO_RDI + cmp $1,%dil + jnz 4f + sub $4097,%rdi + bts $63,%rdi + SET_RDI_INTO_CR3 + jmp 3f +111: + + GET_CR3_INTO_RDI + mov %rdi,%rbx + add $__START_KERNEL_map,%rbx + sub phys_base(%rip),%rbx + +#ifdef CONFIG_PARAVIRT + cmpl $0, pv_info+PARAVIRT_enabled + jz 1f + pushq %rdi + 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 + popq %rdi + jmp 2f +1: +#endif + + i = 0 + .rept USER_PGD_PTRS + movb $0,i*8(%rbx) + i = i + 1 + .endr + +2: SET_RDI_INTO_CR3 + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_RDI + bts $16,%rdi + SET_RDI_INTO_CR0 +#endif + +3: + +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI) +#endif + + popq %rbx + popq %rdi + pax_force_retaddr + retq +4: ud2 +ENDPROC(pax_enter_kernel_user) + +ENTRY(pax_exit_kernel_user) + pushq %rdi + pushq %rbx + +#ifdef CONFIG_PARAVIRT + PV_SAVE_REGS(CLBR_RDI) +#endif + + GET_CR3_INTO_RDI + 661: jmp 1f + .pushsection .altinstr_replacement, "a" + 662: ASM_NOP2 + .popsection + .pushsection .altinstructions, "a" + altinstruction_entry 661b, 662b, X86_FEATURE_PCID, 2, 2 + .popsection + cmp $0,%dil + jnz 3f + add $4097,%rdi + bts $63,%rdi + SET_RDI_INTO_CR3 + jmp 2f +1: + + mov %rdi,%rbx + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_RDI + btr $16,%rdi + jnc 3f + SET_RDI_INTO_CR0 +#endif + + add $__START_KERNEL_map,%rbx + sub phys_base(%rip),%rbx + +#ifdef CONFIG_PARAVIRT + cmpl $0, pv_info+PARAVIRT_enabled + jz 1f + 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(%rbx) + i = i + 1 + .endr +2: + +#ifdef CONFIG_PARAVIRT + PV_RESTORE_REGS(CLBR_RDI) +#endif + + popq %rbx + popq %rdi + pax_force_retaddr + retq +3: ud2 +ENDPROC(pax_exit_kernel_user) +#endif + + .macro pax_enter_kernel_nmi + pax_set_fptr_mask + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_RDI + bts $16,%rdi + jc 110f + SET_RDI_INTO_CR0 + or $2,%ebx +110: +#endif + +#ifdef CONFIG_PAX_MEMORY_UDEREF + 661: jmp 111f + .pushsection .altinstr_replacement, "a" + 662: ASM_NOP2 + .popsection + .pushsection .altinstructions, "a" + altinstruction_entry 661b, 662b, X86_FEATURE_PCID, 2, 2 + .popsection + GET_CR3_INTO_RDI + cmp $0,%dil + jz 111f + sub $4097,%rdi + or $4,%ebx + bts $63,%rdi + SET_RDI_INTO_CR3 + mov $__UDEREF_KERNEL_DS,%edi + mov %edi,%ss +111: +#endif + .endm + + .macro pax_exit_kernel_nmi +#ifdef CONFIG_PAX_KERNEXEC + btr $1,%ebx + jnc 110f + GET_CR0_INTO_RDI + btr $16,%rdi + SET_RDI_INTO_CR0 +110: +#endif + +#ifdef CONFIG_PAX_MEMORY_UDEREF + btr $2,%ebx + jnc 111f + GET_CR3_INTO_RDI + add $4097,%rdi + bts $63,%rdi + SET_RDI_INTO_CR3 + mov $__KERNEL_DS,%edi + mov %edi,%ss +111: +#endif + .endm + + .macro pax_erase_kstack +#ifdef CONFIG_PAX_MEMORY_STACKLEAK + call pax_erase_kstack +#endif + .endm + +#ifdef CONFIG_PAX_MEMORY_STACKLEAK +ENTRY(pax_erase_kstack) + pushq %rdi + pushq %rcx + 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 %rcx + popq %rdi + pax_force_retaddr + ret +ENDPROC(pax_erase_kstack) +#endif .macro TRACE_IRQS_IRETQ offset=ARGOFFSET #ifdef CONFIG_TRACE_IRQFLAGS @@ -117,7 +543,7 @@ ENDPROC(native_usergs_sysret64) .endm .macro TRACE_IRQS_IRETQ_DEBUG offset=ARGOFFSET - bt $9,EFLAGS-\offset(%rsp) /* interrupts off? */ + bt $X86_EFLAGS_IF_BIT,EFLAGS-\offset(%rsp) /* interrupts off? */ jnc 1f TRACE_IRQS_ON_DEBUG 1: @@ -155,27 +581,6 @@ ENDPROC(native_usergs_sysret64) movq \tmp,R11+\offset(%rsp) .endm - .macro FAKE_STACK_FRAME child_rip - /* push in order ss, rsp, eflags, cs, rip */ - xorl %eax, %eax - pushq_cfi $__KERNEL_DS /* ss */ - /*CFI_REL_OFFSET ss,0*/ - pushq_cfi %rax /* rsp */ - CFI_REL_OFFSET rsp,0 - pushq_cfi $(X86_EFLAGS_IF|X86_EFLAGS_FIXED) /* eflags - interrupts on */ - /*CFI_REL_OFFSET rflags,0*/ - pushq_cfi $__KERNEL_CS /* cs */ - /*CFI_REL_OFFSET cs,0*/ - pushq_cfi \child_rip /* rip */ - CFI_REL_OFFSET rip,0 - pushq_cfi %rax /* orig rax */ - .endm - - .macro UNFAKE_STACK_FRAME - addq $8*6, %rsp - CFI_ADJUST_CFA_OFFSET -(6*8) - .endm - /* * initial frame state for interrupts (and exceptions without error code) */ @@ -242,25 +647,26 @@ ENDPROC(native_usergs_sysret64) /* save partial stack frame */ .macro SAVE_ARGS_IRQ cld - /* start from rbp in pt_regs and jump over */ - movq_cfi rdi, (RDI-RBP) - movq_cfi rsi, (RSI-RBP) - movq_cfi rdx, (RDX-RBP) - movq_cfi rcx, (RCX-RBP) - movq_cfi rax, (RAX-RBP) - movq_cfi r8, (R8-RBP) - movq_cfi r9, (R9-RBP) - movq_cfi r10, (R10-RBP) - movq_cfi r11, (R11-RBP) + /* start from r15 in pt_regs and jump over */ + movq_cfi rdi, RDI + movq_cfi rsi, RSI + movq_cfi rdx, RDX + movq_cfi rcx, RCX + movq_cfi rax, RAX + movq_cfi r8, R8 + movq_cfi r9, R9 + movq_cfi r10, R10 + movq_cfi r11, R11 + movq_cfi r12, R12 /* Save rbp so that we can unwind from get_irq_regs() */ - movq_cfi rbp, 0 + movq_cfi rbp, RBP /* Save previous stack value */ movq %rsp, %rsi - leaq -RBP(%rsp),%rdi /* arg1 for handler */ - testl $3, CS-RBP(%rsi) + movq %rsp,%rdi /* arg1 for handler */ + testb $3, CS(%rsi) je 1f SWAPGS /* @@ -280,6 +686,18 @@ ENDPROC(native_usergs_sysret64) 0x06 /* DW_OP_deref */, \ 0x08 /* DW_OP_const1u */, SS+8-RBP, \ 0x22 /* DW_OP_plus */ + +#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 + /* We entered an interrupt context - irqs are off: */ TRACE_IRQS_OFF .endm @@ -309,9 +727,52 @@ ENTRY(save_paranoid) js 1f /* negative -> in kernel */ SWAPGS xorl %ebx,%ebx -1: ret +1: +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS+8(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif + pax_force_retaddr + ret CFI_ENDPROC -END(save_paranoid) +ENDPROC(save_paranoid) + +ENTRY(save_paranoid_nmi) + XCPT_FRAME 1 RDI+8 + cld + movq_cfi rdi, RDI+8 + movq_cfi rsi, RSI+8 + movq_cfi rdx, RDX+8 + movq_cfi rcx, RCX+8 + movq_cfi rax, RAX+8 + movq_cfi r8, R8+8 + movq_cfi r9, R9+8 + movq_cfi r10, R10+8 + movq_cfi r11, R11+8 + movq_cfi rbx, RBX+8 + movq_cfi rbp, RBP+8 + movq_cfi r12, R12+8 + movq_cfi r13, R13+8 + movq_cfi r14, R14+8 + movq_cfi r15, R15+8 + movl $1,%ebx + movl $MSR_GS_BASE,%ecx + rdmsr + testl %edx,%edx + js 1f /* negative -> in kernel */ + SWAPGS + xorl %ebx,%ebx +1: pax_enter_kernel_nmi + pax_force_retaddr + ret + CFI_ENDPROC +ENDPROC(save_paranoid_nmi) /* * A newly forked process directly context switches into this address. @@ -332,7 +793,7 @@ ENTRY(ret_from_fork) RESTORE_REST - testl $3, CS-ARGOFFSET(%rsp) # from kernel_thread? + testb $3, CS-ARGOFFSET(%rsp) # from kernel_thread? jz 1f testl $_TIF_IA32, TI_flags(%rcx) # 32-bit compat task needs IRET @@ -342,15 +803,13 @@ ENTRY(ret_from_fork) jmp ret_from_sys_call # go to the SYSRET fastpath 1: - subq $REST_SKIP, %rsp # leave space for volatiles - CFI_ADJUST_CFA_OFFSET REST_SKIP movq %rbp, %rdi call *%rbx movl $0, RAX(%rsp) RESTORE_REST jmp int_ret_from_sys_call CFI_ENDPROC -END(ret_from_fork) +ENDPROC(ret_from_fork) /* * System call entry. Up to 6 arguments in registers are supported. @@ -387,7 +846,7 @@ 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 @@ -400,16 +859,23 @@ 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 + +#ifdef CONFIG_PAX_RANDKSTACK + pax_erase_kstack +#endif + /* * 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: #if __SYSCALL_MASK == ~0 @@ -433,10 +899,13 @@ 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: */ @@ -495,6 +964,9 @@ auditsys: 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 */ jmp system_call_fastpath @@ -516,7 +988,7 @@ 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 @@ -524,12 +996,15 @@ 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 + LOAD_ARGS 1 RESTORE_REST #if __SYSCALL_MASK == ~0 cmpq $__NR_syscall_max,%rax @@ -559,7 +1034,9 @@ GLOBAL(int_with_check) andl %edi,%edx jnz int_careful andl $~TS_COMPAT,TI_status(%rcx) - jmp retint_swapgs + pax_exit_kernel_user + pax_erase_kstack + jmp retint_swapgs_pax /* Either reschedule or signal or syscall exit tracking needed. */ /* First do a reschedule test. */ @@ -605,7 +1082,7 @@ int_restore_rest: TRACE_IRQS_OFF jmp int_with_check CFI_ENDPROC -END(system_call) +ENDPROC(system_call) .macro FORK_LIKE func ENTRY(stub_\func) @@ -618,9 +1095,10 @@ ENTRY(stub_\func) DEFAULT_FRAME 0 8 /* offset 8: return address */ call sys_\func RESTORE_TOP_OF_STACK %r11, 8 - ret $REST_SKIP /* pop extended registers */ + pax_force_retaddr + ret CFI_ENDPROC -END(stub_\func) +ENDPROC(stub_\func) .endm .macro FIXED_FRAME label,func @@ -630,9 +1108,10 @@ ENTRY(\label) FIXUP_TOP_OF_STACK %r11, 8-ARGOFFSET call \func RESTORE_TOP_OF_STACK %r11, 8-ARGOFFSET + pax_force_retaddr ret CFI_ENDPROC -END(\label) +ENDPROC(\label) .endm FORK_LIKE clone @@ -640,19 +1119,6 @@ END(\label) FORK_LIKE vfork FIXED_FRAME stub_iopl, sys_iopl -ENTRY(ptregscall_common) - DEFAULT_FRAME 1 8 /* offset 8: return address */ - RESTORE_TOP_OF_STACK %r11, 8 - movq_cfi_restore R15+8, r15 - movq_cfi_restore R14+8, r14 - movq_cfi_restore R13+8, r13 - movq_cfi_restore R12+8, r12 - movq_cfi_restore RBP+8, rbp - movq_cfi_restore RBX+8, rbx - ret $REST_SKIP /* pop extended registers */ - CFI_ENDPROC -END(ptregscall_common) - ENTRY(stub_execve) CFI_STARTPROC addq $8, %rsp @@ -664,7 +1130,7 @@ 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. @@ -681,7 +1147,7 @@ ENTRY(stub_rt_sigreturn) RESTORE_REST jmp int_ret_from_sys_call CFI_ENDPROC -END(stub_rt_sigreturn) +ENDPROC(stub_rt_sigreturn) #ifdef CONFIG_X86_X32_ABI ENTRY(stub_x32_rt_sigreturn) @@ -695,7 +1161,7 @@ ENTRY(stub_x32_rt_sigreturn) RESTORE_REST jmp int_ret_from_sys_call CFI_ENDPROC -END(stub_x32_rt_sigreturn) +ENDPROC(stub_x32_rt_sigreturn) ENTRY(stub_x32_execve) CFI_STARTPROC @@ -709,7 +1175,7 @@ ENTRY(stub_x32_execve) RESTORE_REST jmp int_ret_from_sys_call CFI_ENDPROC -END(stub_x32_execve) +ENDPROC(stub_x32_execve) #endif @@ -746,7 +1212,7 @@ vector=vector+1 2: jmp common_interrupt .endr CFI_ENDPROC -END(irq_entries_start) +ENDPROC(irq_entries_start) .previous END(interrupt) @@ -763,8 +1229,8 @@ END(interrupt) /* 0(%rsp): ~(interrupt number) */ .macro interrupt func /* reserve pt_regs for scratch regs and rbp */ - subq $ORIG_RAX-RBP, %rsp - CFI_ADJUST_CFA_OFFSET ORIG_RAX-RBP + subq $ORIG_RAX, %rsp + CFI_ADJUST_CFA_OFFSET ORIG_RAX SAVE_ARGS_IRQ call \func .endm @@ -787,14 +1253,14 @@ ret_from_intr: /* Restore saved previous stack */ popq %rsi - CFI_DEF_CFA rsi,SS+8-RBP /* reg/off reset after def_cfa_expr */ - leaq ARGOFFSET-RBP(%rsi), %rsp + CFI_DEF_CFA rsi,SS+8 /* reg/off reset after def_cfa_expr */ + movq %rsi, %rsp CFI_DEF_CFA_REGISTER rsp - CFI_ADJUST_CFA_OFFSET RBP-ARGOFFSET + CFI_ADJUST_CFA_OFFSET -ARGOFFSET exit_intr: GET_THREAD_INFO(%rcx) - testl $3,CS-ARGOFFSET(%rsp) + testb $3,CS-ARGOFFSET(%rsp) je retint_kernel /* Interrupt came from user space */ @@ -816,12 +1282,35 @@ retint_swapgs: /* return to user-space */ * The iretq could re-enable interrupts: */ DISABLE_INTERRUPTS(CLBR_ANY) + pax_exit_kernel_user +retint_swapgs_pax: TRACE_IRQS_IRETQ SWAPGS jmp restore_args retint_restore_args: /* return to kernel space */ DISABLE_INTERRUPTS(CLBR_ANY) + pax_exit_kernel + +#if defined(CONFIG_EFI) && defined(CONFIG_PAX_KERNEXEC) + /* This is a quirk to allow IRQs/NMIs/MCEs during early EFI setup, + * namely calling EFI runtime services with a phys mapping. We're + * starting off with NOPs and patch in the real instrumentation + * (BTS/OR) before starting any userland process; even before starting + * up the APs. + */ + .pushsection .altinstr_replacement, "a" + 601: pax_force_retaddr (RIP-ARGOFFSET) + 602: + .popsection + 603: .fill 602b-601b, 1, 0x90 + .pushsection .altinstructions, "a" + altinstruction_entry 603b, 601b, X86_FEATURE_ALWAYS, 602b-601b, 602b-601b + .popsection +#else + pax_force_retaddr (RIP-ARGOFFSET) +#endif + /* * The iretq could re-enable interrupts: */ @@ -934,7 +1423,7 @@ ENTRY(retint_kernel) jmp exit_intr #endif CFI_ENDPROC -END(common_interrupt) +ENDPROC(common_interrupt) /* * If IRET takes a fault on the espfix stack, then we @@ -956,13 +1445,13 @@ __do_double_fault: cmpq $native_irq_return_iret,%rax jne do_double_fault /* This shouldn't happen... */ movq PER_CPU_VAR(kernel_stack),%rax - subq $(6*8-KERNEL_STACK_OFFSET),%rax /* Reset to original stack */ + subq $(6*8),%rax /* Reset to original stack */ movq %rax,RSP(%rdi) movq $0,(%rax) /* Missing (lost) #GP error code */ movq $general_protection,RIP(%rdi) retq CFI_ENDPROC -END(__do_double_fault) +ENDPROC(__do_double_fault) #else # define __do_double_fault do_double_fault #endif @@ -979,7 +1468,7 @@ ENTRY(\sym) interrupt \do_sym jmp ret_from_intr CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm #ifdef CONFIG_TRACING @@ -1052,7 +1541,7 @@ apicinterrupt IRQ_WORK_VECTOR \ /* * Exception entry points. */ -#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)(%r13) .macro idtentry sym do_sym has_error_code:req paranoid=0 shift_ist=-1 ENTRY(\sym) @@ -1103,6 +1592,12 @@ ENTRY(\sym) .endif .if \shift_ist != -1 +#ifdef CONFIG_SMP + imul $TSS_size, PER_CPU_VAR(cpu_number), %r13d + lea init_tss(%r13), %r13 +#else + lea init_tss(%rip), %r13 +#endif subq $EXCEPTION_STKSZ, INIT_TSS_IST(\shift_ist) .endif @@ -1119,7 +1614,7 @@ ENTRY(\sym) .endif CFI_ENDPROC -END(\sym) +ENDPROC(\sym) .endm #ifdef CONFIG_TRACING @@ -1160,9 +1655,10 @@ gs_change: 2: mfence /* workaround */ SWAPGS popfq_cfi + pax_force_retaddr ret CFI_ENDPROC -END(native_load_gs_index) +ENDPROC(native_load_gs_index) _ASM_EXTABLE(gs_change,bad_gs) .section .fixup,"ax" @@ -1190,9 +1686,10 @@ ENTRY(do_softirq_own_stack) CFI_DEF_CFA_REGISTER rsp CFI_ADJUST_CFA_OFFSET -8 decl PER_CPU_VAR(irq_count) + pax_force_retaddr ret CFI_ENDPROC -END(do_softirq_own_stack) +ENDPROC(do_softirq_own_stack) #ifdef CONFIG_XEN idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0 @@ -1230,7 +1727,7 @@ 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. @@ -1289,7 +1786,7 @@ ENTRY(xen_failsafe_callback) SAVE_ALL jmp error_exit CFI_ENDPROC -END(xen_failsafe_callback) +ENDPROC(xen_failsafe_callback) apicinterrupt3 HYPERVISOR_CALLBACK_VECTOR \ xen_hvm_callback_vector xen_evtchn_do_upcall @@ -1336,18 +1833,33 @@ ENTRY(paranoid_exit) DEFAULT_FRAME DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF_DEBUG - testl %ebx,%ebx /* swapgs needed? */ + testl $1,%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_DEBUG 0 RESTORE_ALL 8 + pax_force_retaddr_bts jmp irq_return paranoid_userspace: GET_THREAD_INFO(%rcx) @@ -1376,7 +1888,7 @@ 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. @@ -1403,12 +1915,23 @@ 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: +#ifdef CONFIG_PAX_MEMORY_UDEREF + testb $3, CS+8(%rsp) + jnz 1f + pax_enter_kernel + jmp 2f +1: pax_enter_kernel_user +2: +#else + pax_enter_kernel +#endif TRACE_IRQS_OFF + pax_force_retaddr ret /* @@ -1435,7 +1958,7 @@ 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) */ @@ -1446,7 +1969,7 @@ ENTRY(error_exit) DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF GET_THREAD_INFO(%rcx) - testl %eax,%eax + testl $1,%eax jne retint_kernel LOCKDEP_SYS_EXIT_IRQ movl TI_flags(%rcx),%edx @@ -1455,7 +1978,7 @@ 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. @@ -1513,9 +2036,11 @@ 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. @@ -1549,8 +2074,7 @@ nested_nmi: 1: /* Set up the interrupted NMIs stack to jump to repeat_nmi */ - leaq -1*8(%rsp), %rdx - movq %rdx, %rsp + subq $8, %rsp CFI_ADJUST_CFA_OFFSET 1*8 leaq -10*8(%rsp), %rdx pushq_cfi $__KERNEL_DS @@ -1568,6 +2092,7 @@ nested_nmi_out: CFI_RESTORE rdx /* No need to check faults here */ +# pax_force_retaddr_bts INTERRUPT_RETURN CFI_RESTORE_STATE @@ -1664,13 +2189,13 @@ end_repeat_nmi: subq $ORIG_RAX-R15, %rsp CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 /* - * Use save_paranoid to handle SWAPGS, but no need to use paranoid_exit + * Use save_paranoid_nmi to handle SWAPGS, but no need to use paranoid_exit * as we should not be calling schedule in NMI context. * Even with normal interrupts enabled. An NMI should not be * setting NEED_RESCHED or anything that normal interrupts and * exceptions might do. */ - call save_paranoid + call save_paranoid_nmi DEFAULT_FRAME 0 /* @@ -1680,9 +2205,9 @@ end_repeat_nmi: * NMI itself takes a page fault, the page fault that was preempted * will read the information from the NMI page fault and not the * origin fault. Save it off and restore it if it changes. - * Use the r12 callee-saved register. + * Use the r13 callee-saved register. */ - movq %cr2, %r12 + movq %cr2, %r13 /* paranoidentry do_nmi, 0; without TRACE_IRQS_OFF */ movq %rsp,%rdi @@ -1691,29 +2216,34 @@ end_repeat_nmi: /* Did the NMI take a page fault? Restore cr2 if it did */ movq %cr2, %rcx - cmpq %rcx, %r12 + cmpq %rcx, %r13 je 1f - movq %r12, %cr2 + movq %r13, %cr2 1: - testl %ebx,%ebx /* swapgs needed? */ + testl $1,%ebx /* swapgs needed? */ jnz nmi_restore nmi_swapgs: SWAPGS_UNSAFE_STACK nmi_restore: + pax_exit_kernel_nmi /* Pop the extra iret frame at once */ RESTORE_ALL 6*8 + testb $3, 8(%rsp) + jnz 1f + pax_force_retaddr_bts +1: /* Clear the NMI executing stack variable */ movq $0, 5*8(%rsp) jmp irq_return CFI_ENDPROC -END(nmi) +ENDPROC(nmi) ENTRY(ignore_sysret) CFI_STARTPROC mov $-ENOSYS,%eax sysret CFI_ENDPROC -END(ignore_sysret) +ENDPROC(ignore_sysret) diff --git a/arch/x86/kernel/espfix_64.c b/arch/x86/kernel/espfix_64.c index 94d857f..bf1f0bf 100644 --- a/arch/x86/kernel/espfix_64.c +++ b/arch/x86/kernel/espfix_64.c @@ -197,7 +197,7 @@ void init_espfix_ap(void) set_pte(&pte_p[n*PTE_STRIDE], pte); /* Job is done for this CPU and any CPU which shares this page */ - ACCESS_ONCE(espfix_pages[page]) = stack_page; + ACCESS_ONCE_RW(espfix_pages[page]) = stack_page; unlock_done: mutex_unlock(&espfix_init_mutex); diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index cbc4a91..b38ee45 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -88,7 +88,7 @@ static unsigned long text_ip_addr(unsigned long ip) * kernel identity mapping to modify code. */ if (within(ip, (unsigned long)_text, (unsigned long)_etext)) - ip = (unsigned long)__va(__pa_symbol(ip)); + ip = (unsigned long)__va(__pa_symbol(ktla_ktva(ip))); return ip; } @@ -104,6 +104,8 @@ ftrace_modify_code_direct(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 @@ -229,7 +231,7 @@ static int update_ftrace_func(unsigned long ip, void *new) unsigned char old[MCOUNT_INSN_SIZE]; int ret; - memcpy(old, (void *)ip, MCOUNT_INSN_SIZE); + memcpy(old, (void *)ktla_ktva(ip), MCOUNT_INSN_SIZE); ftrace_update_func = ip; /* Make sure the breakpoints see the ftrace_update_func update */ @@ -310,7 +312,7 @@ static int add_break(unsigned long ip, const char *old) unsigned char replaced[MCOUNT_INSN_SIZE]; unsigned char brk = BREAKPOINT_INSTRUCTION; - if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE)) + if (probe_kernel_read(replaced, (void *)ktla_ktva(ip), MCOUNT_INSN_SIZE)) return -EFAULT; /* Make sure it is what we expect it to be */ diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index eda1a86..8f6df48 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -67,12 +67,12 @@ again: pgd = *pgd_p; /* - * The use of __START_KERNEL_map rather than __PAGE_OFFSET here is - * critical -- __PAGE_OFFSET would point us back into the dynamic + * The use of __early_va rather than __va here is critical: + * __va would point us back into the dynamic * range and we might end up looping forever... */ if (pgd) - pud_p = (pudval_t *)((pgd & PTE_PFN_MASK) + __START_KERNEL_map - phys_base); + pud_p = (pudval_t *)(__early_va(pgd & PTE_PFN_MASK)); else { if (next_early_pgt >= EARLY_DYNAMIC_PAGE_TABLES) { reset_early_page_tables(); @@ -82,13 +82,13 @@ again: pud_p = (pudval_t *)early_dynamic_pgts[next_early_pgt++]; for (i = 0; i < PTRS_PER_PUD; i++) pud_p[i] = 0; - *pgd_p = (pgdval_t)pud_p - __START_KERNEL_map + phys_base + _KERNPG_TABLE; + *pgd_p = (pgdval_t)__pa(pud_p) + _KERNPG_TABLE; } pud_p += pud_index(address); pud = *pud_p; if (pud) - pmd_p = (pmdval_t *)((pud & PTE_PFN_MASK) + __START_KERNEL_map - phys_base); + pmd_p = (pmdval_t *)(__early_va(pud & PTE_PFN_MASK)); else { if (next_early_pgt >= EARLY_DYNAMIC_PAGE_TABLES) { reset_early_page_tables(); @@ -98,7 +98,7 @@ again: pmd_p = (pmdval_t *)early_dynamic_pgts[next_early_pgt++]; for (i = 0; i < PTRS_PER_PMD; i++) pmd_p[i] = 0; - *pud_p = (pudval_t)pmd_p - __START_KERNEL_map + phys_base + _KERNPG_TABLE; + *pud_p = (pudval_t)__pa(pmd_p) + _KERNPG_TABLE; } pmd = (physaddr & PMD_MASK) + early_pmd_flags; pmd_p[pmd_index(address)] = pmd; @@ -175,7 +175,6 @@ asmlinkage __visible void __init x86_64_start_kernel(char * real_mode_data) if (console_loglevel >= CONSOLE_LOGLEVEL_DEBUG) early_printk("Kernel alive\n"); - clear_page(init_level4_pgt); /* set init_level4_pgt kernel high mapping*/ init_level4_pgt[511] = early_level4_pgt[511]; diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index f36bd42..0ab4474 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -26,6 +26,12 @@ /* 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. */ @@ -55,11 +61,7 @@ * 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) @@ -78,6 +80,12 @@ 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 @@ -85,6 +93,13 @@ 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 @@ -106,6 +121,59 @@ 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,GDT_ENTRY_PERCPU * 8 + 2(%edi) + rorl $16,%eax + movb %al,GDT_ENTRY_PERCPU * 8 + 4(%edi) + movb %ah,GDT_ENTRY_PERCPU * 8 + 7(%edi) + movl $__per_cpu_end - 1,%eax + subl $__per_cpu_start,%eax + movw %ax,GDT_ENTRY_PERCPU * 8 + 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,GDT_ENTRY_BOOT_CS * 8 + 2(%edi) + rorl $16,%eax + movb %al,GDT_ENTRY_BOOT_CS * 8 + 4(%edi) + movb %ah,GDT_ENTRY_BOOT_CS * 8 + 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: + movb $0xc0,GDT_ENTRY_KERNEL_CS * 8 + 6(%edi) + movb $0xc0,GDT_ENTRY_KERNEXEC_KERNEL_CS * 8 + 6(%edi) + movw %ax,GDT_ENTRY_KERNEL_CS * 8 + 2(%edi) + movw %ax,GDT_ENTRY_KERNEXEC_KERNEL_CS * 8 + 2(%edi) + rorl $16,%eax + movb %al,GDT_ENTRY_KERNEL_CS * 8 + 4(%edi) + movb %al,GDT_ENTRY_KERNEXEC_KERNEL_CS * 8 + 4(%edi) + movb %ah,GDT_ENTRY_KERNEL_CS * 8 + 7(%edi) + movb %ah,GDT_ENTRY_KERNEXEC_KERNEL_CS * 8 + 7(%edi) + rorl $16,%eax + addl $PAGE_SIZE_asm,%edi + loop 1b +#endif + /* * Clear BSS first so that there are no surprises... */ @@ -201,8 +269,11 @@ 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); @@ -232,8 +303,11 @@ 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 @@ -247,9 +321,7 @@ 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) @@ -261,10 +333,10 @@ 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 @@ -354,6 +426,7 @@ default_entry: movl pa(mmu_cr4_features),%eax movl %eax,%cr4 +#ifdef CONFIG_X86_PAE testb $X86_CR4_PAE, %al # check if PAE is enabled jz enable_paging @@ -382,6 +455,9 @@ default_entry: /* Make changes effective */ wrmsr + btsl $_PAGE_BIT_NX-32,pa(__supported_pte_mask+4) +#endif + enable_paging: /* @@ -449,14 +525,20 @@ is486: 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 movl $(__KERNEL_PERCPU), %eax movl %eax,%fs # set this cpu's percpu +#ifdef CONFIG_CC_STACKPROTECTOR 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 @@ -512,8 +594,11 @@ setup_once: * relocation. Manually set base address in stack canary * segment descriptor. */ - 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) @@ -548,7 +633,7 @@ ENTRY(early_idt_handler) cmpl $2,(%esp) # X86_TRAP_NMI je is_nmi # Ignore NMI - cmpl $2,%ss:early_recursion_flag + cmpl $1,%ss:early_recursion_flag je hlt_loop incl %ss:early_recursion_flag @@ -586,8 +671,8 @@ ENTRY(early_idt_handler) pushl (20+6*4)(%esp) /* trapno */ pushl $fault_msg call printk -#endif call dump_stack +#endif hlt_loop: hlt jmp hlt_loop @@ -607,8 +692,11 @@ ENDPROC(early_idt_handler) /* 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 @@ -617,9 +705,6 @@ 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) @@ -653,29 +738,34 @@ ENTRY(setup_once_ref) /* * 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 /* * 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 @@ -694,12 +784,20 @@ 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 2*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 __INITRODATA int_msg: @@ -727,7 +825,7 @@ fault_msg: * segment size, and 32-bit linear address value: */ - .data +.section .rodata,"a",@progbits .globl boot_gdt_descr .globl idt_descr @@ -736,7 +834,7 @@ 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: @@ -747,7 +845,7 @@ 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 @@ -756,5 +854,65 @@ 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 0x00c093000000ffff /* 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 a468c0a..8b5a879 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -20,6 +20,8 @@ #include #include #include +#include +#include #ifdef CONFIG_PARAVIRT #include @@ -41,6 +43,12 @@ 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 @@ -89,11 +97,24 @@ startup_64: * Fixup the physical addresses in the page table */ addq %rbp, early_level4_pgt + (L4_START_KERNEL*8)(%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_kernel_pgt + (510*8)(%rip) - addq %rbp, level3_kernel_pgt + (511*8)(%rip) + addq %rbp, level3_ident_pgt + (0*8)(%rip) +#ifndef CONFIG_XEN + addq %rbp, level3_ident_pgt + (1*8)(%rip) +#endif + + 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+1)*8)(%rip) addq %rbp, level2_fixmap_pgt + (506*8)(%rip) + addq %rbp, level2_fixmap_pgt + (507*8)(%rip) /* * Set up the identity mapping for the switchover. These @@ -174,11 +195,12 @@ ENTRY(secondary_startup_64) * after the boot processor executes this code. */ + orq $-1, %rbp movq $(init_level4_pgt - __START_KERNEL_map), %rax 1: - /* Enable PAE mode and PGE */ - movl $(X86_CR4_PAE | X86_CR4_PGE), %ecx + /* Enable PAE mode and PSE/PGE */ + movl $(X86_CR4_PSE | X86_CR4_PAE | X86_CR4_PGE), %ecx movq %rcx, %cr4 /* Setup early boot stage 4 level pagetables. */ @@ -199,10 +221,19 @@ 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 + cmpq $-1, %rbp + je 1f btsq $_PAGE_BIT_NX,early_pmd_flags(%rip) + btsq $_PAGE_BIT_NX, init_level4_pgt + 8*L4_PAGE_OFFSET(%rip) + btsq $_PAGE_BIT_NX, init_level4_pgt + 8*L4_VMALLOC_START(%rip) + btsq $_PAGE_BIT_NX, init_level4_pgt + 8*L4_VMALLOC_END(%rip) + btsq $_PAGE_BIT_NX, init_level4_pgt + 8*L4_VMEMMAP_START(%rip) + btsq $_PAGE_BIT_NX, level2_fixmap_pgt + 8*506(%rip) + btsq $_PAGE_BIT_NX, level2_fixmap_pgt + 8*507(%rip) + btsq $_PAGE_BIT_NX, __supported_pte_mask(%rip) 1: wrmsr /* Make changes effective */ /* Setup cr0 */ @@ -282,6 +313,7 @@ ENTRY(secondary_startup_64) * REX.W + FF /5 JMP m16:64 Jump far, absolute indirect, * address given in m16:64. */ + pax_set_fptr_mask movq initial_code(%rip),%rax pushq $0 # fake return address to stop unwinder pushq $__KERNEL_CS # set correct cs @@ -313,7 +345,7 @@ ENDPROC(start_cpu0) .quad INIT_PER_CPU_VAR(irq_stack_union) GLOBAL(stack_start) - .quad init_thread_union+THREAD_SIZE-8 + .quad init_thread_union+THREAD_SIZE-16 .word 0 __FINITDATA @@ -391,7 +423,7 @@ ENTRY(early_idt_handler) call dump_stack #ifdef CONFIG_KALLSYMS leaq early_idt_ripmsg(%rip),%rdi - movq 40(%rsp),%rsi # %rip again + movq 88(%rsp),%rsi # %rip again call __print_symbol #endif #endif /* EARLY_PRINTK */ @@ -420,6 +452,7 @@ ENDPROC(early_idt_handler) early_recursion_flag: .long 0 + .section .rodata,"a",@progbits #ifdef CONFIG_EARLY_PRINTK early_idt_msg: .asciz "PANIC: early exception %02lx rip %lx:%lx error %lx cr2 %lx\n" @@ -447,29 +480,52 @@ NEXT_PAGE(early_level4_pgt) NEXT_PAGE(early_dynamic_pgts) .fill 512*EARLY_DYNAMIC_PAGE_TABLES,8,0 - .data + .section .rodata,"a",@progbits -#ifndef CONFIG_XEN NEXT_PAGE(init_level4_pgt) - .fill 512,8,0 -#else -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 2*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(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) -#endif + PMDS(0, __PAGE_KERNEL_IDENT_LARGE_EXEC, 2*PTRS_PER_PMD) NEXT_PAGE(level3_kernel_pgt) .fill L3_START_KERNEL,8,0 @@ -477,6 +533,9 @@ 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_kernel_pgt) /* * 512 MB kernel mapping. We spend a full page on this pagetable @@ -494,28 +553,64 @@ NEXT_PAGE(level2_kernel_pgt) 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 + .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) .fill 512,8,0 +NEXT_PAGE(level1_vsyscall_pgt) + .fill 512,8,0 + #undef PMDS - .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 */ + +#ifdef CONFIG_PAX_MEMORY_UDEREF + .quad 0x00cf93000000ffff /* __UDEREF_KERNEL_DS */ +#else + .quad 0x0 /* unused */ +#endif + + /* 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" - - __PAGE_ALIGNED_BSS + + .section .rodata,"a",@progbits NEXT_PAGE(empty_zero_page) .skip PAGE_SIZE diff --git a/arch/x86/kernel/i386_ksyms_32.c b/arch/x86/kernel/i386_ksyms_32.c index 05fd74f..c3548b1 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); @@ -44,3 +48,11 @@ EXPORT_SYMBOL(___preempt_schedule); EXPORT_SYMBOL(___preempt_schedule_context); #endif #endif + +#ifdef CONFIG_PAX_KERNEXEC +EXPORT_SYMBOL(__LOAD_PHYSICAL_ADDR); +#endif + +#ifdef CONFIG_PAX_PER_CPU_PGD +EXPORT_SYMBOL(cpu_pgd); +#endif diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c index d5dd808..b6432cf 100644 --- a/arch/x86/kernel/i387.c +++ b/arch/x86/kernel/i387.c @@ -51,7 +51,7 @@ 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/kernel/i8259.c b/arch/x86/kernel/i8259.c index 8af8171..f8c1169 100644 --- a/arch/x86/kernel/i8259.c +++ b/arch/x86/kernel/i8259.c @@ -110,7 +110,7 @@ static int i8259A_irq_pending(unsigned int irq) static void make_8259A_irq(unsigned int irq) { disable_irq_nosync(irq); - io_apic_irqs &= ~(1< #include #include +#include #include #include #include @@ -30,6 +31,12 @@ asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on) return -EINVAL; if (turn_on && !capable(CAP_SYS_RAWIO)) return -EPERM; +#ifdef CONFIG_GRKERNSEC_IO + if (turn_on && grsec_disable_privio) { + gr_handle_ioperm(); + return -ENODEV; + } +#endif /* * If it's the first ioperm() call in this thread's lifetime, set the @@ -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); @@ -105,6 +112,12 @@ SYSCALL_DEFINE1(iopl, unsigned int, level) if (level > old) { if (!capable(CAP_SYS_RAWIO)) return -EPERM; +#ifdef CONFIG_GRKERNSEC_IO + if (grsec_disable_privio) { + gr_handle_iopl(); + return -ENODEV; + } +#endif } regs->flags = (regs->flags & ~X86_EFLAGS_IOPL) | (level << 12); t->iopl = level << 12; diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c index 922d285..6d20692 100644 --- a/arch/x86/kernel/irq.c +++ b/arch/x86/kernel/irq.c @@ -22,7 +22,7 @@ #define CREATE_TRACE_POINTS #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; @@ -132,9 +132,9 @@ int arch_show_interrupts(struct seq_file *p, int prec) seq_printf(p, "%10u ", irq_stats(j)->irq_hv_callback_count); seq_printf(p, " Hypervisor callback interrupts\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; } @@ -174,7 +174,7 @@ 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); return sum; } diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index 63ce838..2ea3e06 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -29,6 +29,8 @@ EXPORT_PER_CPU_SYMBOL(irq_regs); #ifdef CONFIG_DEBUG_STACKOVERFLOW +extern void gr_handle_kernel_exploit(void); + int sysctl_panic_on_stackoverflow __read_mostly; /* Debugging check for stack overflow: is there less than 1KB free? */ @@ -39,13 +41,14 @@ 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) { printk(KERN_WARNING "low stack detected by irq handler\n"); dump_stack(); + gr_handle_kernel_exploit(); if (sysctl_panic_on_stackoverflow) panic("low stack detected by irq handler - check messages\n"); } @@ -84,10 +87,9 @@ static inline void *current_stack(void) static inline int execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) { - struct irq_stack *curstk, *irqstk; + struct irq_stack *irqstk; u32 *isp, *prev_esp, arg1, arg2; - curstk = (struct irq_stack *) current_stack(); irqstk = __this_cpu_read(hardirq_stack); /* @@ -96,15 +98,19 @@ 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(curstk == irqstk)) + if (unlikely((void *)current_stack_pointer - (void *)irqstk < THREAD_SIZE)) return 0; - isp = (u32 *) ((char *)irqstk + sizeof(*irqstk)); + isp = (u32 *) ((char *)irqstk + sizeof(*irqstk) - 8); /* Save the next esp at the bottom of the stack */ prev_esp = (u32 *)irqstk; *prev_esp = current_stack_pointer; +#ifdef CONFIG_PAX_MEMORY_UDEREF + __set_fs(MAKE_MM_SEG(0)); +#endif + if (unlikely(overflow)) call_on_stack(print_stack_overflow, isp); @@ -115,6 +121,11 @@ 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; } @@ -123,32 +134,18 @@ execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) */ void irq_ctx_init(int cpu) { - struct irq_stack *irqstk; - if (per_cpu(hardirq_stack, cpu)) return; - irqstk = page_address(alloc_pages_node(cpu_to_node(cpu), - THREADINFO_GFP, - THREAD_SIZE_ORDER)); - per_cpu(hardirq_stack, cpu) = irqstk; - - irqstk = page_address(alloc_pages_node(cpu_to_node(cpu), - THREADINFO_GFP, - THREAD_SIZE_ORDER)); - per_cpu(softirq_stack, cpu) = irqstk; - - printk(KERN_DEBUG "CPU %u irqstacks, hard=%p soft=%p\n", - cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); + per_cpu(hardirq_stack, cpu) = page_address(alloc_pages_node(cpu_to_node(cpu), THREADINFO_GFP, THREAD_SIZE_ORDER)); + per_cpu(softirq_stack, cpu) = page_address(alloc_pages_node(cpu_to_node(cpu), THREADINFO_GFP, THREAD_SIZE_ORDER)); } void do_softirq_own_stack(void) { - struct thread_info *curstk; struct irq_stack *irqstk; u32 *isp, *prev_esp; - curstk = current_stack(); irqstk = __this_cpu_read(softirq_stack); /* build the stack frame on the softirq stack */ @@ -158,7 +155,16 @@ void do_softirq_own_stack(void) prev_esp = (u32 *)irqstk; *prev_esp = current_stack_pointer; +#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 + } bool handle_irq(unsigned irq, struct pt_regs *regs) @@ -172,7 +178,7 @@ bool handle_irq(unsigned irq, struct pt_regs *regs) if (unlikely(!desc)) return false; - if (user_mode_vm(regs) || !execute_on_irq_stack(overflow, desc, irq)) { + if (user_mode(regs) || !execute_on_irq_stack(overflow, desc, irq)) { if (unlikely(overflow)) print_stack_overflow(); desc->handle_irq(irq, desc); diff --git a/arch/x86/kernel/irq_64.c b/arch/x86/kernel/irq_64.c index 4d1c746..55a22d6 100644 --- a/arch/x86/kernel/irq_64.c +++ b/arch/x86/kernel/irq_64.c @@ -26,6 +26,8 @@ EXPORT_PER_CPU_SYMBOL(irq_stat); DEFINE_PER_CPU(struct pt_regs *, irq_regs); EXPORT_PER_CPU_SYMBOL(irq_regs); +extern void gr_handle_kernel_exploit(void); + int sysctl_panic_on_stackoverflow; /* @@ -44,7 +46,7 @@ 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) + @@ -69,6 +71,8 @@ static inline void stack_overflow_check(struct pt_regs *regs) irq_stack_top, irq_stack_bottom, estack_top, estack_bottom); + gr_handle_kernel_exploit(); + if (sysctl_panic_on_stackoverflow) panic("low stack detected by irq handler - check messages\n"); #endif diff --git a/arch/x86/kernel/jump_label.c b/arch/x86/kernel/jump_label.c index 26d5a55..a01160a 100644 --- a/arch/x86/kernel/jump_label.c +++ b/arch/x86/kernel/jump_label.c @@ -51,7 +51,7 @@ static void __jump_label_transform(struct jump_entry *entry, * Jump label is enabled for the first time. * So we expect a default_nop... */ - if (unlikely(memcmp((void *)entry->code, default_nop, 5) + if (unlikely(memcmp((void *)ktla_ktva(entry->code), default_nop, 5) != 0)) bug_at((void *)entry->code, __LINE__); } else { @@ -59,7 +59,7 @@ static void __jump_label_transform(struct jump_entry *entry, * ...otherwise expect an ideal_nop. Otherwise * something went horribly wrong. */ - if (unlikely(memcmp((void *)entry->code, ideal_nop, 5) + if (unlikely(memcmp((void *)ktla_ktva(entry->code), ideal_nop, 5) != 0)) bug_at((void *)entry->code, __LINE__); } @@ -75,13 +75,13 @@ static void __jump_label_transform(struct jump_entry *entry, * are converting the default nop to the ideal nop. */ if (init) { - if (unlikely(memcmp((void *)entry->code, default_nop, 5) != 0)) + if (unlikely(memcmp((void *)ktla_ktva(entry->code), default_nop, 5) != 0)) bug_at((void *)entry->code, __LINE__); } else { code.jump = 0xe9; code.offset = entry->target - (entry->code + JUMP_LABEL_NOP_SIZE); - if (unlikely(memcmp((void *)entry->code, &code, 5) != 0)) + if (unlikely(memcmp((void *)ktla_ktva(entry->code), &code, 5) != 0)) bug_at((void *)entry->code, __LINE__); } memcpy(&code, ideal_nops[NOP_ATOMIC5], JUMP_LABEL_NOP_SIZE); diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c index 7ec1d5f..5a7d130 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: @@ -228,7 +228,10 @@ static void kgdb_correct_hw_break(void) bp->attr.bp_addr = breakinfo[breakno].addr; bp->attr.bp_len = breakinfo[breakno].len; bp->attr.bp_type = breakinfo[breakno].type; - info->address = breakinfo[breakno].addr; + if (breakinfo[breakno].type == X86_BREAKPOINT_EXECUTE) + info->address = ktla_ktva(breakinfo[breakno].addr); + else + info->address = breakinfo[breakno].addr; info->len = breakinfo[breakno].len; info->type = breakinfo[breakno].type; val = arch_install_hw_breakpoint(bp); @@ -475,12 +478,12 @@ 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 +548,7 @@ 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; @@ -750,11 +753,11 @@ int kgdb_arch_set_breakpoint(struct kgdb_bkpt *bpt) #endif /* CONFIG_DEBUG_RODATA */ bpt->type = BP_BREAKPOINT; - err = probe_kernel_read(bpt->saved_instr, (char *)bpt->bpt_addr, + err = probe_kernel_read(bpt->saved_instr, ktla_ktva((char *)bpt->bpt_addr), BREAK_INSTR_SIZE); if (err) return err; - err = probe_kernel_write((char *)bpt->bpt_addr, + err = probe_kernel_write(ktla_ktva((char *)bpt->bpt_addr), arch_kgdb_ops.gdb_bpt_instr, BREAK_INSTR_SIZE); #ifdef CONFIG_DEBUG_RODATA if (!err) @@ -767,7 +770,7 @@ int kgdb_arch_set_breakpoint(struct kgdb_bkpt *bpt) return -EBUSY; text_poke((void *)bpt->bpt_addr, arch_kgdb_ops.gdb_bpt_instr, BREAK_INSTR_SIZE); - err = probe_kernel_read(opc, (char *)bpt->bpt_addr, BREAK_INSTR_SIZE); + err = probe_kernel_read(opc, ktla_ktva((char *)bpt->bpt_addr), BREAK_INSTR_SIZE); if (err) return err; if (memcmp(opc, arch_kgdb_ops.gdb_bpt_instr, BREAK_INSTR_SIZE)) @@ -792,13 +795,13 @@ int kgdb_arch_remove_breakpoint(struct kgdb_bkpt *bpt) if (mutex_is_locked(&text_mutex)) goto knl_write; text_poke((void *)bpt->bpt_addr, bpt->saved_instr, BREAK_INSTR_SIZE); - err = probe_kernel_read(opc, (char *)bpt->bpt_addr, BREAK_INSTR_SIZE); + err = probe_kernel_read(opc, ktla_ktva((char *)bpt->bpt_addr), BREAK_INSTR_SIZE); if (err || memcmp(opc, bpt->saved_instr, BREAK_INSTR_SIZE)) goto knl_write; return err; knl_write: #endif /* CONFIG_DEBUG_RODATA */ - return probe_kernel_write((char *)bpt->bpt_addr, + return probe_kernel_write(ktla_ktva((char *)bpt->bpt_addr), (char *)bpt->saved_instr, BREAK_INSTR_SIZE); } diff --git a/arch/x86/kernel/kprobes/core.c b/arch/x86/kernel/kprobes/core.c index 67e6d19..731ed28 100644 --- a/arch/x86/kernel/kprobes/core.c +++ b/arch/x86/kernel/kprobes/core.c @@ -120,9 +120,12 @@ __synthesize_relative_insn(void *from, void *to, u8 op) s32 raddr; } __packed *insn; - insn = (struct __arch_relative_insn *)from; + insn = (struct __arch_relative_insn *)ktla_ktva(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'.*/ @@ -168,7 +171,7 @@ int 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: @@ -242,9 +245,9 @@ __recover_probed_insn(kprobe_opcode_t *buf, unsigned long addr) * for the first byte, we can recover the original instruction * from it and kp->opcode. */ - memcpy(buf, kp->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + memcpy(buf, ktla_ktva(kp->addr), MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); buf[0] = kp->opcode; - return (unsigned long)buf; + return ktva_ktla((unsigned long)buf); } /* @@ -336,7 +339,9 @@ int __copy_instruction(u8 *dest, u8 *src) /* Another subsystem puts a breakpoint, failed to recover */ if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) return 0; + pax_open_kernel(); memcpy(dest, insn.kaddr, insn.length); + pax_close_kernel(); #ifdef CONFIG_X86_64 if (insn_rip_relative(&insn)) { @@ -363,7 +368,9 @@ int __copy_instruction(u8 *dest, u8 *src) return 0; } disp = (u8 *) dest + insn_offset_displacement(&insn); + pax_open_kernel(); *(s32 *) disp = (s32) newdisp; + pax_close_kernel(); } #endif return insn.length; @@ -505,7 +512,7 @@ static void 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; } @@ -522,9 +529,9 @@ static void setup_singlestep(struct kprobe *p, struct pt_regs *regs, regs->flags &= ~X86_EFLAGS_IF; /* single step inline if the instruction is an int3 */ if (p->opcode == BREAKPOINT_INSTRUCTION) - regs->ip = (unsigned long)p->addr; + regs->ip = ktla_ktva((unsigned long)p->addr); else - regs->ip = (unsigned long)p->ainsn.insn; + regs->ip = ktva_ktla((unsigned long)p->ainsn.insn); } NOKPROBE_SYMBOL(setup_singlestep); @@ -574,7 +581,7 @@ int kprobe_int3_handler(struct pt_regs *regs) struct kprobe *p; struct kprobe_ctlblk *kcb; - if (user_mode_vm(regs)) + if (user_mode(regs)) return 0; addr = (kprobe_opcode_t *)(regs->ip - sizeof(kprobe_opcode_t)); @@ -609,7 +616,7 @@ int kprobe_int3_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 @@ -656,6 +663,9 @@ static void __used 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 @@ -796,7 +806,7 @@ static void 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; @@ -979,7 +989,7 @@ int kprobe_exceptions_notify(struct notifier_block *self, unsigned long val, 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; if (val == DIE_GPF) { diff --git a/arch/x86/kernel/kprobes/opt.c b/arch/x86/kernel/kprobes/opt.c index f304773..551e63c 100644 --- a/arch/x86/kernel/kprobes/opt.c +++ b/arch/x86/kernel/kprobes/opt.c @@ -79,6 +79,7 @@ found: /* Insert a move instruction which sets a pointer to eax/rdi (1st arg). */ static void synthesize_set_arg1(kprobe_opcode_t *addr, unsigned long val) { + pax_open_kernel(); #ifdef CONFIG_X86_64 *addr++ = 0x48; *addr++ = 0xbf; @@ -86,6 +87,7 @@ static void synthesize_set_arg1(kprobe_opcode_t *addr, unsigned long val) *addr++ = 0xb8; #endif *(unsigned long *)addr = val; + pax_close_kernel(); } asm ( @@ -337,7 +339,7 @@ int 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; @@ -352,16 +354,18 @@ int arch_prepare_optimized_kprobe(struct optimized_kprobe *op) op->optinsn.size = ret; /* Copy arch-dep-instance from template */ - memcpy(buf, &optprobe_template_entry, TMPL_END_IDX); + pax_open_kernel(); + memcpy(buf, ktla_ktva(&optprobe_template_entry), TMPL_END_IDX); + pax_close_kernel(); /* Set probe information */ synthesize_set_arg1(buf + TMPL_MOVE_IDX, (unsigned long)op); /* Set probe function call */ - synthesize_relcall(buf + TMPL_CALL_IDX, optimized_callback); + synthesize_relcall(ktva_ktla(buf) + TMPL_CALL_IDX, optimized_callback); /* Set returning jmp instruction at the tail of out-of-line buffer */ - synthesize_reljump(buf + TMPL_END_IDX + op->optinsn.size, + synthesize_reljump(ktva_ktla(buf) + TMPL_END_IDX + op->optinsn.size, (u8 *)op->kp.addr + op->optinsn.size); flush_icache_range((unsigned long) buf, @@ -386,7 +390,7 @@ void arch_optimize_kprobes(struct list_head *oplist) WARN_ON(kprobe_disabled(&op->kp)); /* 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; @@ -434,7 +438,7 @@ int setup_detour_execution(struct kprobe *p, struct pt_regs *regs, int reenter) /* This kprobe is really able to run optimized path. */ op = container_of(p, struct optimized_kprobe, kp); /* Detour through copied instructions */ - regs->ip = (unsigned long)op->optinsn.insn + TMPL_END_IDX; + regs->ip = ktva_ktla((unsigned long)op->optinsn.insn) + TMPL_END_IDX; if (!reenter) reset_current_kprobe(); preempt_enable_no_resched(); diff --git a/arch/x86/kernel/ksysfs.c b/arch/x86/kernel/ksysfs.c index c2bedae..25e7ab60 100644 --- a/arch/x86/kernel/ksysfs.c +++ b/arch/x86/kernel/ksysfs.c @@ -184,7 +184,7 @@ out: static struct kobj_attribute type_attr = __ATTR_RO(type); -static struct bin_attribute data_attr = { +static bin_attribute_no_const data_attr __read_only = { .attr = { .name = "data", .mode = S_IRUGO, diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c index c37886d..d851d32 100644 --- a/arch/x86/kernel/ldt.c +++ b/arch/x86/kernel/ldt.c @@ -66,13 +66,13 @@ 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) { @@ -94,7 +94,7 @@ 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; } @@ -115,6 +115,24 @@ 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; } @@ -229,6 +247,13 @@ 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 + if (!IS_ENABLED(CONFIG_X86_16BIT) && !ldt_info.seg_32bit) { error = -EINVAL; goto out_unlock; diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index 1667b1d..16492c5 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -25,7 +25,7 @@ #include #include -static void set_idt(void *newidt, __u16 limit) +static void set_idt(struct desc_struct *newidt, __u16 limit) { struct desc_ptr curidt; @@ -37,7 +37,7 @@ 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; @@ -215,7 +215,7 @@ 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/mcount_64.S b/arch/x86/kernel/mcount_64.S index c050a01..5774072 100644 --- a/arch/x86/kernel/mcount_64.S +++ b/arch/x86/kernel/mcount_64.S @@ -7,7 +7,7 @@ #include #include #include - +#include .code64 .section .entry.text, "ax" @@ -24,8 +24,9 @@ #ifdef CONFIG_DYNAMIC_FTRACE ENTRY(function_hook) + pax_force_retaddr retq -END(function_hook) +ENDPROC(function_hook) /* skip is set if stack has been adjusted */ .macro ftrace_caller_setup skip=0 @@ -66,8 +67,9 @@ GLOBAL(ftrace_graph_call) #endif GLOBAL(ftrace_stub) + pax_force_retaddr retq -END(ftrace_caller) +ENDPROC(ftrace_caller) ENTRY(ftrace_regs_caller) /* Save the current flags before compare (in SS location)*/ @@ -135,7 +137,7 @@ ftrace_restore_flags: popfq jmp ftrace_stub -END(ftrace_regs_caller) +ENDPROC(ftrace_regs_caller) #else /* ! CONFIG_DYNAMIC_FTRACE */ @@ -156,6 +158,7 @@ ENTRY(function_hook) #endif GLOBAL(ftrace_stub) + pax_force_retaddr retq trace: @@ -169,12 +172,13 @@ trace: #endif subq $MCOUNT_INSN_SIZE, %rdi + pax_force_fptr ftrace_trace_function call *ftrace_trace_function MCOUNT_RESTORE_FRAME jmp ftrace_stub -END(function_hook) +ENDPROC(function_hook) #endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_TRACER */ @@ -196,8 +200,9 @@ 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 @@ -213,5 +218,7 @@ GLOBAL(return_to_handler) movq 8(%rsp), %rdx movq (%rsp), %rax addq $24, %rsp + pax_force_fptr %rdi jmp *%rdi +ENDPROC(return_to_handler) #endif diff --git a/arch/x86/kernel/module.c b/arch/x86/kernel/module.c index e69f988..da078ea 100644 --- a/arch/x86/kernel/module.c +++ b/arch/x86/kernel/module.c @@ -81,17 +81,62 @@ static unsigned long int get_module_load_offset(void) } #endif -void *module_alloc(unsigned long size) +static inline void *__module_alloc(unsigned long size, pgprot_t prot) { - if (PAGE_ALIGN(size) > MODULES_LEN) + if (!size || PAGE_ALIGN(size) > MODULES_LEN) return NULL; return __vmalloc_node_range(size, 1, MODULES_VADDR + get_module_load_offset(), - MODULES_END, GFP_KERNEL | __GFP_HIGHMEM, - PAGE_KERNEL_EXEC, NUMA_NO_NODE, + MODULES_END, GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO, + prot, NUMA_NO_NODE, __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, @@ -102,14 +147,16 @@ 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 @@ -118,11 +165,15 @@ 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 position */ - *location += sym->st_value - (uint32_t)location; + pax_open_kernel(); + *plocation += sym->st_value - location; + pax_close_kernel(); break; default: pr_err("%s: Unknown relocation: %u\n", @@ -167,21 +218,30 @@ 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/msr.c b/arch/x86/kernel/msr.c index c9603ac..9f88728 100644 --- a/arch/x86/kernel/msr.c +++ b/arch/x86/kernel/msr.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include @@ -103,6 +104,11 @@ static ssize_t msr_write(struct file *file, const char __user *buf, int err = 0; ssize_t bytes = 0; +#ifdef CONFIG_GRKERNSEC_KMEM + gr_handle_msr_write(); + return -EPERM; +#endif + if (count % 8) return -EINVAL; /* Invalid chunk size */ @@ -150,6 +156,10 @@ static long msr_ioctl(struct file *file, unsigned int ioc, unsigned long arg) err = -EBADF; break; } +#ifdef CONFIG_GRKERNSEC_KMEM + gr_handle_msr_write(); + return -EPERM; +#endif if (copy_from_user(®s, uregs, sizeof regs)) { err = -EFAULT; break; @@ -233,7 +243,7 @@ static int msr_class_cpu_callback(struct notifier_block *nfb, return notifier_from_errno(err); } -static struct notifier_block __refdata msr_class_cpu_notifier = { +static struct notifier_block msr_class_cpu_notifier = { .notifier_call = msr_class_cpu_callback, }; diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c index c3e985d..110a36a 100644 --- a/arch/x86/kernel/nmi.c +++ b/arch/x86/kernel/nmi.c @@ -98,16 +98,16 @@ fs_initcall(nmi_warning_debugfs); static void nmi_max_handler(struct irq_work *w) { - struct nmiaction *a = container_of(w, struct nmiaction, irq_work); + struct nmiwork *n = container_of(w, struct nmiwork, irq_work); int remainder_ns, decimal_msecs; - u64 whole_msecs = ACCESS_ONCE(a->max_duration); + u64 whole_msecs = ACCESS_ONCE(n->max_duration); remainder_ns = do_div(whole_msecs, (1000 * 1000)); decimal_msecs = remainder_ns / 1000; printk_ratelimited(KERN_INFO "INFO: NMI handler (%ps) took too long to run: %lld.%03d msecs\n", - a->handler, whole_msecs, decimal_msecs); + n->action->handler, whole_msecs, decimal_msecs); } static int nmi_handle(unsigned int type, struct pt_regs *regs, bool b2b) @@ -134,11 +134,11 @@ static int nmi_handle(unsigned int type, struct pt_regs *regs, bool b2b) delta = sched_clock() - delta; trace_nmi_handler(a->handler, (int)delta, thishandled); - if (delta < nmi_longest_ns || delta < a->max_duration) + if (delta < nmi_longest_ns || delta < a->work->max_duration) continue; - a->max_duration = delta; - irq_work_queue(&a->irq_work); + a->work->max_duration = delta; + irq_work_queue(&a->work->irq_work); } rcu_read_unlock(); @@ -148,7 +148,7 @@ static int nmi_handle(unsigned int type, struct pt_regs *regs, bool b2b) } NOKPROBE_SYMBOL(nmi_handle); -int __register_nmi_handler(unsigned int type, struct nmiaction *action) +int __register_nmi_handler(unsigned int type, const struct nmiaction *action) { struct nmi_desc *desc = nmi_to_desc(type); unsigned long flags; @@ -156,7 +156,8 @@ int __register_nmi_handler(unsigned int type, struct nmiaction *action) if (!action->handler) return -EINVAL; - init_irq_work(&action->irq_work, nmi_max_handler); + action->work->action = action; + init_irq_work(&action->work->irq_work, nmi_max_handler); spin_lock_irqsave(&desc->lock, flags); @@ -174,9 +175,9 @@ int __register_nmi_handler(unsigned int type, struct nmiaction *action) * event confuses some handlers (kdump uses this flag) */ if (action->flags & NMI_FLAG_FIRST) - list_add_rcu(&action->list, &desc->head); + pax_list_add_rcu((struct list_head *)&action->list, &desc->head); else - list_add_tail_rcu(&action->list, &desc->head); + pax_list_add_tail_rcu((struct list_head *)&action->list, &desc->head); spin_unlock_irqrestore(&desc->lock, flags); return 0; @@ -199,7 +200,7 @@ void unregister_nmi_handler(unsigned int type, const char *name) if (!strcmp(n->name, name)) { WARN(in_nmi(), "Trying to free NMI (%s) from NMI context!\n", n->name); - list_del_rcu(&n->list); + pax_list_del_rcu((struct list_head *)&n->list); break; } } @@ -528,6 +529,17 @@ static inline void nmi_nesting_postprocess(void) dotraplinkage notrace 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/nmi_selftest.c b/arch/x86/kernel/nmi_selftest.c index 6d9582e..f746287 100644 --- a/arch/x86/kernel/nmi_selftest.c +++ b/arch/x86/kernel/nmi_selftest.c @@ -43,7 +43,7 @@ static void __init init_nmi_testsuite(void) { /* trap all the unknown NMIs we may generate */ register_nmi_handler(NMI_UNKNOWN, nmi_unk_cb, 0, "nmi_selftest_unk", - __initdata); + __initconst); } static void __init cleanup_nmi_testsuite(void) @@ -66,7 +66,7 @@ static void __init test_nmi_ipi(struct cpumask *mask) unsigned long timeout; if (register_nmi_handler(NMI_LOCAL, test_nmi_ipi_callback, - NMI_FLAG_FIRST, "nmi_selftest", __initdata)) { + NMI_FLAG_FIRST, "nmi_selftest", __initconst)) { nmi_fail = FAILURE; return; } diff --git a/arch/x86/kernel/paravirt-spinlocks.c b/arch/x86/kernel/paravirt-spinlocks.c index bbb6c73..24a58ef 100644 --- a/arch/x86/kernel/paravirt-spinlocks.c +++ b/arch/x86/kernel/paravirt-spinlocks.c @@ -8,7 +8,7 @@ #include -struct pv_lock_ops pv_lock_ops = { +struct pv_lock_ops pv_lock_ops __read_only = { #ifdef CONFIG_SMP .lock_spinning = __PV_IS_CALLEE_SAVE(paravirt_nop), .unlock_kick = paravirt_nop, diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 548d25f..f8fb99c 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -56,6 +56,9 @@ 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) { @@ -142,16 +145,20 @@ 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) + ret = paravirt_patch_insns(insnbuf, len, ktva_ktla(ud2a), ud2a+sizeof(ud2a)); + 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) || @@ -176,7 +183,7 @@ 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; } @@ -300,7 +307,7 @@ enum paravirt_lazy_mode paravirt_get_lazy_mode(void) return this_cpu_read(paravirt_lazy_mode); } -struct pv_info pv_info = { +struct pv_info pv_info __read_only = { .name = "bare hardware", .paravirt_enabled = 0, .kernel_rpl = 0, @@ -311,16 +318,16 @@ 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, }; -__visible struct pv_irq_ops pv_irq_ops = { +__visible 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), @@ -332,7 +339,7 @@ __visible struct pv_irq_ops pv_irq_ops = { #endif }; -__visible struct pv_cpu_ops pv_cpu_ops = { +__visible 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 @@ NOKPROBE_SYMBOL(native_get_debugreg); NOKPROBE_SYMBOL(native_set_debugreg); NOKPROBE_SYMBOL(native_load_idt); -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 */ @@ -479,6 +492,12 @@ 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-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index 0497f71..7186c0d 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -1347,7 +1347,7 @@ static void __init get_tce_space_from_tar(void) tce_space = be64_to_cpu(readq(target)); tce_space = tce_space & TAR_SW_BITS; - tce_space = tce_space & (~specified_table_size); + tce_space = tce_space & (~(unsigned long)specified_table_size); info->tce_space = (u64 *)__va(tce_space); } } 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/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c index 77dd0ad..9ec4723 100644 --- a/arch/x86/kernel/pci-swiotlb.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -33,7 +33,7 @@ void x86_swiotlb_free_coherent(struct device *dev, size_t size, struct dma_attrs *attrs) { if (is_swiotlb_buffer(dma_to_phys(dev, dma_addr))) - swiotlb_free_coherent(dev, size, vaddr, dma_addr); + swiotlb_free_coherent(dev, size, vaddr, dma_addr, attrs); else dma_generic_free_coherent(dev, size, vaddr, dma_addr, attrs); } diff --git a/arch/x86/kernel/preempt.S b/arch/x86/kernel/preempt.S index ca7f0d5..8996469 100644 --- a/arch/x86/kernel/preempt.S +++ b/arch/x86/kernel/preempt.S @@ -3,12 +3,14 @@ #include #include #include +#include ENTRY(___preempt_schedule) CFI_STARTPROC SAVE_ALL call preempt_schedule RESTORE_ALL + pax_force_retaddr ret CFI_ENDPROC @@ -19,6 +21,7 @@ ENTRY(___preempt_schedule_context) SAVE_ALL call preempt_schedule_context RESTORE_ALL + pax_force_retaddr ret CFI_ENDPROC diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 4505e2a..ae28b0d 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -36,7 +36,8 @@ * section. Since TSS's are completely CPU-local, we want them * on exact cacheline boundaries, to eliminate cacheline ping-pong. */ -__visible DEFINE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss) = INIT_TSS; +struct tss_struct init_tss[NR_CPUS] __visible ____cacheline_internodealigned_in_smp = { [0 ... NR_CPUS-1] = INIT_TSS }; +EXPORT_SYMBOL(init_tss); #ifdef CONFIG_X86_64 static DEFINE_PER_CPU(unsigned char, is_idle); @@ -92,7 +93,7 @@ void arch_task_cache_init(void) 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); } /* @@ -105,7 +106,7 @@ 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); @@ -125,6 +126,9 @@ 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)); drop_init_fpu(tsk); @@ -271,7 +275,7 @@ 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(); } @@ -324,7 +328,7 @@ bool xen_set_default_idle(void) return ret; } #endif -void stop_this_cpu(void *dummy) +__noreturn void stop_this_cpu(void *dummy) { local_irq_disable(); /* @@ -453,16 +457,37 @@ 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 + this_cpu_write(kernel_stack, thread->sp0); +#endif +} +#endif diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 7bc86bb..0ea06e8 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -64,6 +64,7 @@ asmlinkage void ret_from_kernel_thread(void) __asm__("ret_from_kernel_thread"); unsigned long thread_saved_pc(struct task_struct *tsk) { return ((unsigned long *)tsk->thread.sp)[3]; +//XXX return tsk->thread.eip; } void __show_regs(struct pt_regs *regs, int all) @@ -73,19 +74,18 @@ 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); printk(KERN_DEFAULT "EIP: %04x:[<%08lx>] EFLAGS: %08lx CPU: %d\n", (u16)regs->cs, regs->ip, regs->flags, - smp_processor_id()); + raw_smp_processor_id()); print_symbol("EIP is at %s\n", regs->ip); printk(KERN_DEFAULT "EAX: %08lx EBX: %08lx ECX: %08lx EDX: %08lx\n", @@ -132,20 +132,21 @@ void release_thread(struct task_struct *dead_task) int copy_thread(unsigned long clone_flags, unsigned long sp, unsigned long arg, struct task_struct *p) { - struct pt_regs *childregs = task_pt_regs(p); + struct pt_regs *childregs = task_stack_page(p) + THREAD_SIZE - sizeof(struct pt_regs) - 8; struct task_struct *tsk; int err; p->thread.sp = (unsigned long) childregs; p->thread.sp0 = (unsigned long) (childregs+1); + p->tinfo.lowest_stack = (unsigned long)task_stack_page(p); if (unlikely(p->flags & PF_KTHREAD)) { /* kernel thread */ memset(childregs, 0, sizeof(struct pt_regs)); p->thread.ip = (unsigned long) ret_from_kernel_thread; - task_user_gs(p) = __KERNEL_STACK_CANARY; - childregs->ds = __USER_DS; - childregs->es = __USER_DS; + savesegment(gs, childregs->gs); + childregs->ds = __KERNEL_DS; + childregs->es = __KERNEL_DS; childregs->fs = __KERNEL_PERCPU; childregs->bx = sp; /* function */ childregs->bp = arg; @@ -252,7 +253,7 @@ __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 */ @@ -276,6 +277,10 @@ __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. */ @@ -314,9 +319,9 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) */ arch_end_context_switch(next_p); - this_cpu_write(kernel_stack, - (unsigned long)task_stack_page(next_p) + - THREAD_SIZE - KERNEL_STACK_OFFSET); + this_cpu_write(current_task, next_p); + this_cpu_write(current_tinfo, &next_p->tinfo); + this_cpu_write(kernel_stack, next->sp0); /* * Restore %gs if needed (which is common) @@ -326,8 +331,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) switch_fpu_finish(next_p, fpu); - this_cpu_write(current_task, next_p); - return prev_p; } @@ -357,4 +360,3 @@ 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 ca5b02d..c0b2f6a 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -158,10 +158,11 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, struct pt_regs *childregs; struct task_struct *me = current; - p->thread.sp0 = (unsigned long)task_stack_page(p) + THREAD_SIZE; + p->thread.sp0 = (unsigned long)task_stack_page(p) + THREAD_SIZE - 16; childregs = task_pt_regs(p); p->thread.sp = (unsigned long) childregs; p->thread.usersp = me->thread.usersp; + p->tinfo.lowest_stack = (unsigned long)task_stack_page(p); set_tsk_thread_flag(p, TIF_FORK); p->thread.fpu_counter = 0; p->thread.io_bitmap_ptr = NULL; @@ -172,6 +173,8 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, p->thread.fs = p->thread.fsindex ? 0 : me->thread.fs; savesegment(es, p->thread.es); savesegment(ds, p->thread.ds); + savesegment(ss, p->thread.ss); + BUG_ON(p->thread.ss == __UDEREF_KERNEL_DS); memset(p->thread.ptrace_bps, 0, sizeof(p->thread.ptrace_bps)); if (unlikely(p->flags & PF_KTHREAD)) { @@ -280,7 +283,7 @@ __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; @@ -303,6 +306,9 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) if (unlikely(next->ds | prev->ds)) loadsegment(ds, next->ds); + savesegment(ss, prev->ss); + if (unlikely(next->ss != prev->ss)) + loadsegment(ss, next->ss); /* We must save %fs and %gs before load_TLS() because * %fs and %gs may be cleared by load_TLS(). @@ -362,6 +368,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) prev->usersp = this_cpu_read(old_rsp); this_cpu_write(old_rsp, next->usersp); this_cpu_write(current_task, next_p); + this_cpu_write(current_tinfo, &next_p->tinfo); /* * If it were not for PREEMPT_ACTIVE we could guarantee that the @@ -371,9 +378,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) task_thread_info(prev_p)->saved_preempt_count = this_cpu_read(__preempt_count); this_cpu_write(__preempt_count, task_thread_info(next_p)->saved_preempt_count); - this_cpu_write(kernel_stack, - (unsigned long)task_stack_page(next_p) + - THREAD_SIZE - KERNEL_STACK_OFFSET); + this_cpu_write(kernel_stack, next->sp0); /* * Now maybe reload the debug registers and handle I/O bitmaps @@ -443,12 +448,11 @@ 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 678c0ad..2fc2a7b 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -186,10 +186,10 @@ unsigned long kernel_stack_pointer(struct pt_regs *regs) unsigned long sp = (unsigned long)®s->sp; u32 *prev_esp; - if (context == (sp & ~(THREAD_SIZE - 1))) + if (context == ((sp + 8) & ~(THREAD_SIZE - 1))) return sp; - prev_esp = (u32 *)(context); + prev_esp = *(u32 **)(context); if (prev_esp) return (unsigned long)prev_esp; @@ -452,6 +452,20 @@ static int putreg(struct task_struct *child, if (child->thread.gs != value) return do_arch_prctl(child, ARCH_SET_GS, value); return 0; + + case offsetof(struct user_regs_struct,ip): + /* + * Protect against any attempt to set ip to an + * impossible address. There are dragons lurking if the + * address is noncanonical. (This explicitly allows + * setting ip to TASK_SIZE_MAX, because user code can do + * that all by itself by running off the end of its + * address space. + */ + if (value > TASK_SIZE_MAX) + return -EIO; + break; + #endif } @@ -588,7 +602,7 @@ static void ptrace_triggered(struct perf_event *bp, static unsigned long ptrace_get_dr7(struct perf_event *bp[]) { int i; - int dr7 = 0; + unsigned long dr7 = 0; struct arch_hw_breakpoint *info; for (i = 0; i < HBP_NUM; i++) { @@ -822,7 +836,7 @@ 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. */ @@ -907,14 +921,14 @@ 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 @@ -1292,7 +1306,7 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request, #ifdef CONFIG_X86_64 -static struct user_regset x86_64_regsets[] __read_mostly = { +static user_regset_no_const x86_64_regsets[] __read_only = { [REGSET_GENERAL] = { .core_note_type = NT_PRSTATUS, .n = sizeof(struct user_regs_struct) / sizeof(long), @@ -1333,7 +1347,7 @@ static const struct user_regset_view user_x86_64_view = { #endif /* CONFIG_X86_64 */ #if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION -static struct user_regset x86_32_regsets[] __read_mostly = { +static user_regset_no_const x86_32_regsets[] __read_only = { [REGSET_GENERAL] = { .core_note_type = NT_PRSTATUS, .n = sizeof(struct user_regs_struct32) / sizeof(u32), @@ -1386,7 +1400,7 @@ static const struct user_regset_view user_x86_32_view = { */ u64 xstate_fx_sw_bytes[USER_XSTATE_FX_SW_WORDS]; -void update_regset_xstate_info(unsigned int size, u64 xstate_mask) +void __init update_regset_xstate_info(unsigned int size, u64 xstate_mask) { #ifdef CONFIG_X86_64 x86_64_regsets[REGSET_XSTATE].n = size / sizeof(u64); @@ -1421,7 +1435,7 @@ 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, @@ -1450,6 +1464,10 @@ 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. @@ -1460,6 +1478,11 @@ long syscall_trace_enter(struct pt_regs *regs) user_exit(); +#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. @@ -1515,6 +1538,11 @@ void syscall_trace_leave(struct pt_regs *regs) */ user_exit(); +#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 2f355d2..e75ed0a 100644 --- a/arch/x86/kernel/pvclock.c +++ b/arch/x86/kernel/pvclock.c @@ -51,11 +51,11 @@ void pvclock_touch_watchdogs(void) reset_hung_task_detector(); } -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); } u8 pvclock_read_flags(struct pvclock_vcpu_time_info *src) @@ -105,11 +105,11 @@ 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 52b1157..c6e67c4 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -68,6 +68,11 @@ static int __init set_bios_reboot(const struct dmi_system_id *d) void __noreturn machine_real_restart(unsigned int type) { + +#if defined(CONFIG_X86_32) && (defined(CONFIG_PAX_KERNEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF)) + struct desc_struct *gdt; +#endif + local_irq_disable(); /* @@ -95,7 +100,29 @@ void __noreturn machine_real_restart(unsigned int type) /* Jump to the identity-mapped low memory code */ #ifdef CONFIG_X86_32 - asm volatile("jmpl *%0" : : + +#if 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; + loadsegment(ds, __KERNEL_DS); + loadsegment(es, __KERNEL_DS); + loadsegment(ss, __KERNEL_DS); +#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 + + asm volatile("ljmpl *%0" : : "rm" (real_mode_header->machine_real_restart_asm), "a" (type)); #else @@ -486,7 +513,7 @@ void __attribute__((weak)) mach_reboot_fixups(void) * This means that this function can never return, it can misbehave * by not rebooting properly and hanging. */ -static void native_machine_emergency_restart(void) +static void __noreturn native_machine_emergency_restart(void) { int i; int attempt = 0; @@ -610,13 +637,13 @@ void native_machine_shutdown(void) #endif } -static void __machine_emergency_restart(int emergency) +static void __noreturn __machine_emergency_restart(int emergency) { reboot_emergency = emergency; machine_ops.emergency_restart(); } -static void native_machine_restart(char *__unused) +static void __noreturn native_machine_restart(char *__unused) { pr_notice("machine restart\n"); @@ -625,7 +652,7 @@ static void native_machine_restart(char *__unused) __machine_emergency_restart(0); } -static void native_machine_halt(void) +static void __noreturn native_machine_halt(void) { /* Stop other cpus and apics */ machine_shutdown(); @@ -635,7 +662,7 @@ static void native_machine_halt(void) stop_this_cpu(NULL); } -static void native_machine_power_off(void) +static void __noreturn native_machine_power_off(void) { if (pm_power_off) { if (!reboot_force) @@ -644,9 +671,10 @@ 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 = { +struct machine_ops machine_ops __read_only = { .power_off = native_machine_power_off, .shutdown = native_machine_shutdown, .emergency_restart = native_machine_emergency_restart, diff --git a/arch/x86/kernel/reboot_fixups_32.c b/arch/x86/kernel/reboot_fixups_32.c index c8e41e9..64049ef 100644 --- a/arch/x86/kernel/reboot_fixups_32.c +++ b/arch/x86/kernel/reboot_fixups_32.c @@ -57,7 +57,7 @@ struct device_fixup { unsigned int vendor; unsigned int device; void (*reboot_fixup)(struct pci_dev *); -}; +} __do_const; /* * PCI ids solely used for fixups_table go here diff --git a/arch/x86/kernel/relocate_kernel_64.S b/arch/x86/kernel/relocate_kernel_64.S index 3fd2c69..a444264 100644 --- a/arch/x86/kernel/relocate_kernel_64.S +++ b/arch/x86/kernel/relocate_kernel_64.S @@ -96,8 +96,7 @@ relocate_kernel: /* jump to identity mapped page */ addq $(identity_mapped - relocate_kernel), %r8 - pushq %r8 - ret + jmp *%r8 identity_mapped: /* set return address to 0 if not preserving context */ diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 78a0e62..5c2e510 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -110,6 +110,7 @@ #include #include #include +#include /* * max_low_pfn_mapped: highest direct mapped pfn under 4GB @@ -205,12 +206,50 @@ EXPORT_SYMBOL(boot_cpu_data); #endif -#if !defined(CONFIG_X86_PAE) || defined(CONFIG_X86_64) -__visible unsigned long mmu_cr4_features; +#ifdef CONFIG_X86_64 +__visible unsigned long mmu_cr4_features __read_only = X86_CR4_PSE | X86_CR4_PAE | X86_CR4_PGE; +#elif defined(CONFIG_X86_PAE) +__visible unsigned long mmu_cr4_features __read_only = X86_CR4_PAE; #else -__visible unsigned long mmu_cr4_features = X86_CR4_PAE; +__visible unsigned long mmu_cr4_features __read_only; #endif +void set_in_cr4(unsigned long mask) +{ + unsigned long cr4 = read_cr4(); + + if ((cr4 & mask) == mask && cr4 == mmu_cr4_features) + return; + + pax_open_kernel(); + mmu_cr4_features |= mask; + pax_close_kernel(); + + if (trampoline_cr4_features) + *trampoline_cr4_features = mmu_cr4_features; + cr4 |= mask; + write_cr4(cr4); +} +EXPORT_SYMBOL(set_in_cr4); + +void clear_in_cr4(unsigned long mask) +{ + unsigned long cr4 = read_cr4(); + + if (!(cr4 & mask) && cr4 == mmu_cr4_features) + return; + + pax_open_kernel(); + mmu_cr4_features &= ~mask; + pax_close_kernel(); + + if (trampoline_cr4_features) + *trampoline_cr4_features = mmu_cr4_features; + cr4 &= ~mask; + write_cr4(cr4); +} +EXPORT_SYMBOL(clear_in_cr4); + /* Boot loader ID and version as integers, for the benefit of proc_dointvec */ int bootloader_type, bootloader_version; @@ -772,7 +811,7 @@ 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); } @@ -780,7 +819,7 @@ static void __init trim_bios_range(void) /* called before trim_bios_range() to spare extra sanitize */ static void __init e820_add_kernel_range(void) { - u64 start = __pa_symbol(_text); + u64 start = __pa_symbol(ktla_ktva(_text)); u64 size = __pa_symbol(_end) - start; /* @@ -856,8 +895,12 @@ dump_kernel_offset(struct notifier_block *self, unsigned long v, void *p) void __init setup_arch(char **cmdline_p) { +#ifdef CONFIG_X86_32 + memblock_reserve(LOAD_PHYSICAL_ADDR, __pa_symbol(__bss_stop) - LOAD_PHYSICAL_ADDR); +#else memblock_reserve(__pa_symbol(_text), (unsigned long)__bss_stop - (unsigned long)_text); +#endif early_reserve_initrd(); @@ -946,14 +989,14 @@ 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 = __pa_symbol(_text); - code_resource.end = __pa_symbol(_etext)-1; - data_resource.start = __pa_symbol(_etext); + code_resource.start = __pa_symbol(ktla_ktva(_text)); + code_resource.end = __pa_symbol(ktla_ktva(_etext))-1; + data_resource.start = __pa_symbol(_sdata); data_resource.end = __pa_symbol(_edata)-1; bss_resource.start = __pa_symbol(__bss_start); bss_resource.end = __pa_symbol(__bss_stop)-1; diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 5cdff03..80fa283 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -21,19 +21,17 @@ #include #include -DEFINE_PER_CPU_READ_MOSTLY(int, cpu_number); +#ifdef CONFIG_SMP +DEFINE_PER_CPU_READ_MOSTLY(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); @@ -66,7 +64,7 @@ static bool __init pcpu_need_numa(void) { #ifdef CONFIG_NEED_MULTIPLE_NODES pg_data_t *last = NULL; - unsigned int cpu; + int cpu; for_each_possible_cpu(cpu) { int node = early_cpu_to_node(cpu); @@ -155,10 +153,10 @@ 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 +217,11 @@ 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 +262,12 @@ 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 2851d63..83bf567 100644 --- a/arch/x86/kernel/signal.c +++ b/arch/x86/kernel/signal.c @@ -190,7 +190,7 @@ 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 @@ -298,10 +298,9 @@ __setup_frame(int sig, struct ksignal *ksig, sigset_t *set, } if (current->mm->context.vdso) - restorer = current->mm->context.vdso + - selected_vdso32->sym___kernel_sigreturn; + restorer = (void __force_user *)(current->mm->context.vdso + selected_vdso32->sym___kernel_sigreturn); else - restorer = &frame->retcode; + restorer = (void __user *)&frame->retcode; if (ksig->ka.sa.sa_flags & SA_RESTORER) restorer = ksig->ka.sa.sa_restorer; @@ -315,7 +314,7 @@ __setup_frame(int sig, struct ksignal *ksig, 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; @@ -362,8 +361,10 @@ static int __setup_rt_frame(int sig, struct ksignal *ksig, save_altstack_ex(&frame->uc.uc_stack, regs->sp); /* Set up to return from userspace. */ - restorer = current->mm->context.vdso + - selected_vdso32->sym___kernel_rt_sigreturn; + if (current->mm->context.vdso) + restorer = (void __force_user *)(current->mm->context.vdso + selected_vdso32->sym___kernel_rt_sigreturn); + else + restorer = (void __user *)&frame->retcode; if (ksig->ka.sa.sa_flags & SA_RESTORER) restorer = ksig->ka.sa.sa_restorer; put_user_ex(restorer, &frame->pretcode); @@ -375,7 +376,7 @@ static int __setup_rt_frame(int sig, struct ksignal *ksig, * 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); err |= copy_siginfo_to_user(&frame->info, &ksig->info); @@ -611,7 +612,12 @@ setup_rt_frame(struct ksignal *ksig, struct pt_regs *regs) { int usig = signr_convert(ksig->sig); sigset_t *set = sigmask_to_save(); - compat_sigset_t *cset = (compat_sigset_t *) set; + sigset_t sigcopy; + compat_sigset_t *cset; + + sigcopy = *set; + + cset = (compat_sigset_t *) &sigcopy; /* Set up the stack frame */ if (is_ia32_frame()) { @@ -622,7 +628,7 @@ setup_rt_frame(struct ksignal *ksig, struct pt_regs *regs) } else if (is_x32_frame()) { return x32_setup_rt_frame(ksig, cset, regs); } else { - return __setup_rt_frame(ksig->sig, ksig, set, regs); + return __setup_rt_frame(ksig->sig, ksig, &sigcopy, regs); } } diff --git a/arch/x86/kernel/smp.c b/arch/x86/kernel/smp.c index be8e1bd..a3d93fa 100644 --- a/arch/x86/kernel/smp.c +++ b/arch/x86/kernel/smp.c @@ -341,7 +341,7 @@ static int __init nonmi_ipi_setup(char *str) __setup("nonmi_ipi", nonmi_ipi_setup); -struct smp_ops smp_ops = { +struct smp_ops smp_ops __read_only = { .smp_prepare_boot_cpu = native_smp_prepare_boot_cpu, .smp_prepare_cpus = native_smp_prepare_cpus, .smp_cpus_done = native_smp_cpus_done, diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 215815b..9a814fd 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -230,14 +230,17 @@ static void notrace start_secondary(void *unused) enable_start_cpu0 = 0; -#ifdef CONFIG_X86_32 + /* otherwise gcc will move up smp_processor_id before the cpu_init */ + barrier(); + /* switch away from the initial page table */ +#ifdef CONFIG_PAX_PER_CPU_PGD + load_cr3(get_cpu_pgd(smp_processor_id(), kernel)); +#else load_cr3(swapper_pg_dir); +#endif __flush_tlb_all(); -#endif - /* otherwise gcc will move up smp_processor_id before the cpu_init */ - barrier(); /* * Check TSC synchronization with the BP: */ @@ -764,8 +767,9 @@ static int do_boot_cpu(int apicid, int cpu, struct task_struct *idle) alternatives_enable_smp(); idle->thread.sp = (unsigned long) (((struct pt_regs *) - (THREAD_SIZE + task_stack_page(idle))) - 1); + (THREAD_SIZE - 16 + task_stack_page(idle))) - 1); per_cpu(current_task, cpu) = idle; + per_cpu(current_tinfo, cpu) = &idle->tinfo; #ifdef CONFIG_X86_32 /* Stack for startup_32 can be just as for start_secondary onwards */ @@ -774,10 +778,10 @@ static int do_boot_cpu(int apicid, int cpu, struct task_struct *idle) clear_tsk_thread_flag(idle, TIF_FORK); initial_gs = per_cpu_offset(cpu); #endif - 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; + 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 = idle->thread.sp; @@ -923,6 +927,15 @@ int native_cpu_up(unsigned int cpu, struct task_struct *tidle) /* the FPU context is blank, nobody can own it */ __cpu_disable_lazy_restore(cpu); +#ifdef CONFIG_PAX_PER_CPU_PGD + clone_pgd_range(get_cpu_pgd(cpu, kernel) + KERNEL_PGD_BOUNDARY, + swapper_pg_dir + KERNEL_PGD_BOUNDARY, + KERNEL_PGD_PTRS); + clone_pgd_range(get_cpu_pgd(cpu, user) + KERNEL_PGD_BOUNDARY, + swapper_pg_dir + KERNEL_PGD_BOUNDARY, + KERNEL_PGD_PTRS); +#endif + err = do_boot_cpu(apicid, cpu, tidle); if (err) { pr_err("do_boot_cpu failed(%d) to wakeup CPU#%u\n", err, cpu); diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c index 9b4d51d..5d28b58 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 new file mode 100644 index 0000000..5877189 --- /dev/null +++ b/arch/x86/kernel/sys_i386_32.c @@ -0,0 +1,189 @@ +/* + * This file contains various random system calls that + * have a non-standard calling sequence on the Linux/i386 + * platform. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#include + +int i386_mmap_check(unsigned long addr, unsigned long len, unsigned long flags) +{ + 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 (flags & MAP_FIXED) + if (len > pax_task_size || addr > pax_task_size - len) + return -EINVAL; + + return 0; +} + +/* + * Align a virtual address to avoid aliasing in the I$ on AMD F15h. + */ +static unsigned long get_align_mask(void) +{ + if (va_align.flags < 0 || !(va_align.flags & ALIGN_VA_32)) + return 0; + + if (!(current->flags & PF_RANDOMIZE)) + return 0; + + return va_align.mask; +} + +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 pax_task_size = TASK_SIZE; + struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); + +#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, offset)) + return addr; + } + } + + info.flags = 0; + info.length = len; + info.align_mask = filp ? get_align_mask() : 0; + info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; + +#ifdef CONFIG_PAX_PAGEEXEC + if (!(__supported_pte_mask & _PAGE_NX) && (mm->pax_flags & MF_PAX_PAGEEXEC) && (flags & MAP_EXECUTABLE)) { + info.low_limit = 0x00110000UL; + info.high_limit = mm->start_code; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + info.low_limit += mm->delta_mmap & 0x03FFF000UL; +#endif + + if (info.low_limit < info.high_limit) { + addr = vm_unmapped_area(&info); + if (!IS_ERR_VALUE(addr)) + return addr; + } + } else +#endif + + info.low_limit = mm->mmap_base; + info.high_limit = pax_task_size; + + return vm_unmapped_area(&info); +} + +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 addr = addr0, pax_task_size = TASK_SIZE; + struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); + +#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, offset)) + return addr; + } + } + + info.flags = VM_UNMAPPED_AREA_TOPDOWN; + info.length = len; + info.low_limit = PAGE_SIZE; + info.high_limit = mm->mmap_base; + info.align_mask = filp ? get_align_mask() : 0; + info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; + + addr = vm_unmapped_area(&info); + if (!(addr & ~PAGE_MASK)) + return addr; + VM_BUG_ON(addr != -ENOMEM); + +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. + */ + return arch_get_unmapped_area(filp, addr0, len, pgoff, flags); +} diff --git a/arch/x86/kernel/sys_x86_64.c b/arch/x86/kernel/sys_x86_64.c index 30277e2..5664a29 100644 --- a/arch/x86/kernel/sys_x86_64.c +++ b/arch/x86/kernel/sys_x86_64.c @@ -81,8 +81,8 @@ 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_ADDR32) && (flags & MAP_32BIT)) { unsigned long new_begin; @@ -101,7 +101,7 @@ static void find_start_end(unsigned long flags, unsigned long *begin, *begin = new_begin; } } else { - *begin = current->mm->mmap_legacy_base; + *begin = mm->mmap_legacy_base; *end = TASK_SIZE; } } @@ -114,20 +114,24 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, struct vm_area_struct *vma; struct vm_unmapped_area_info info; unsigned long begin, end; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); 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, offset)) return addr; } @@ -137,6 +141,7 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, info.high_limit = end; info.align_mask = filp ? get_align_mask() : 0; info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; return vm_unmapped_area(&info); } @@ -149,6 +154,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, struct mm_struct *mm = current->mm; unsigned long addr = addr0; struct vm_unmapped_area_info info; + unsigned long offset = gr_rand_threadstack_offset(mm, filp, flags); /* requested length too big for entire address space */ if (len > TASK_SIZE) @@ -161,12 +167,15 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (!test_thread_flag(TIF_ADDR32) && (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)) + if (TASK_SIZE - len >= addr && check_heap_stack_gap(vma, addr, len, offset)) return addr; } @@ -176,6 +185,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, info.high_limit = mm->mmap_base; info.align_mask = filp ? get_align_mask() : 0; info.align_offset = pgoff << PAGE_SHIFT; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); if (!(addr & ~PAGE_MASK)) return addr; diff --git a/arch/x86/kernel/tboot.c b/arch/x86/kernel/tboot.c index 91a4496..bb87552 100644 --- a/arch/x86/kernel/tboot.c +++ b/arch/x86/kernel/tboot.c @@ -221,7 +221,7 @@ static int tboot_setup_sleep(void) void tboot_shutdown(u32 shutdown_type) { - void (*shutdown)(void); + void (* __noreturn shutdown)(void); if (!tboot_enabled()) return; @@ -243,7 +243,7 @@ void tboot_shutdown(u32 shutdown_type) switch_to_tboot_pt(); - shutdown = (void(*)(void))(unsigned long)tboot->shutdown_entry; + shutdown = (void *)(unsigned long)tboot->shutdown_entry; shutdown(); /* should not reach here */ @@ -310,7 +310,7 @@ static int tboot_extended_sleep(u8 sleep_state, u32 val_a, u32 val_b) return -ENODEV; } -static atomic_t ap_wfs_count; +static atomic_unchecked_t ap_wfs_count; static int tboot_wait_for_aps(int num_aps) { @@ -334,9 +334,9 @@ static int tboot_cpu_callback(struct notifier_block *nfb, unsigned long action, { 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; } @@ -422,7 +422,7 @@ 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); #ifdef CONFIG_DEBUG_FS diff --git a/arch/x86/kernel/time.c b/arch/x86/kernel/time.c index bf7ef5c..59d0ac9 100644 --- a/arch/x86/kernel/time.c +++ b/arch/x86/kernel/time.c @@ -30,9 +30,9 @@ 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); @@ -41,11 +41,17 @@ 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 f7fec09..9991981 100644 --- a/arch/x86/kernel/tls.c +++ b/arch/x86/kernel/tls.c @@ -84,6 +84,11 @@ 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; @@ -200,7 +205,7 @@ int regset_tls_set(struct task_struct *target, const struct user_regset *regset, if (kbuf) info = kbuf; - else if (__copy_from_user(infobuf, ubuf, count)) + else if (count > sizeof infobuf || __copy_from_user(infobuf, ubuf, count)) return -EFAULT; else info = infobuf; diff --git a/arch/x86/kernel/tracepoint.c b/arch/x86/kernel/tracepoint.c index 1c113db..287b42e 100644 --- a/arch/x86/kernel/tracepoint.c +++ b/arch/x86/kernel/tracepoint.c @@ -9,11 +9,11 @@ #include atomic_t trace_idt_ctr = ATOMIC_INIT(0); -struct desc_ptr trace_idt_descr = { NR_VECTORS * 16 - 1, +const struct desc_ptr trace_idt_descr = { NR_VECTORS * 16 - 1, (unsigned long) trace_idt_table }; /* No need to be aligned, but done to keep all IDTs defined the same way. */ -gate_desc trace_idt_table[NR_VECTORS] __page_aligned_bss; +gate_desc trace_idt_table[NR_VECTORS] __page_aligned_rodata; static int trace_irq_vector_refcount; static DEFINE_MUTEX(irq_vector_mutex); diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 0d0e922..0886373 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -67,7 +67,7 @@ #include /* No need to be aligned, but done to keep all IDTs defined the same way. */ -gate_desc debug_idt_table[NR_VECTORS] __page_aligned_bss; +gate_desc debug_idt_table[NR_VECTORS] __page_aligned_rodata; #else #include #include @@ -76,7 +76,7 @@ asmlinkage int system_call(void); #endif /* Must be page-aligned because the real IDT is used in a fixmap. */ -gate_desc idt_table[NR_VECTORS] __page_aligned_bss; +gate_desc idt_table[NR_VECTORS] __page_aligned_rodata; DECLARE_BITMAP(used_vectors, NR_VECTORS); EXPORT_SYMBOL_GPL(used_vectors); @@ -108,11 +108,11 @@ static inline void preempt_conditional_cli(struct pt_regs *regs) } static nokprobe_inline int -do_trap_no_signal(struct task_struct *tsk, int trapnr, char *str, +do_trap_no_signal(struct task_struct *tsk, int trapnr, const char *str, struct pt_regs *regs, long error_code) { #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. @@ -125,12 +125,24 @@ do_trap_no_signal(struct task_struct *tsk, int trapnr, char *str, return -1; } #endif - if (!user_mode(regs)) { + if (!user_mode_novm(regs)) { if (!fixup_exception(regs)) { tsk->thread.error_code = error_code; tsk->thread.trap_nr = trapnr; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + if (trapnr == X86_TRAP_SS && ((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 == X86_TRAP_OF) + pax_report_refcount_overflow(regs); +#endif + return 0; } @@ -169,7 +181,7 @@ static siginfo_t *fill_trap_info(struct pt_regs *regs, int signr, int trapnr, } static void -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; @@ -193,7 +205,7 @@ do_trap(int trapnr, int signr, char *str, struct pt_regs *regs, if (show_unhandled_signals && unhandled_signal(tsk, signr) && printk_ratelimit()) { pr_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); pr_cont("\n"); @@ -266,6 +278,11 @@ dotraplinkage void do_double_fault(struct pt_regs *regs, long error_code) tsk->thread.error_code = error_code; tsk->thread.trap_nr = X86_TRAP_DF; +#ifdef CONFIG_GRKERNSEC_KSTACKOVERFLOW + if ((unsigned long)tsk->stack - regs->sp <= PAGE_SIZE) + die("grsec: kernel stack overflow detected", regs, error_code); +#endif + #ifdef CONFIG_DOUBLEFAULT df_debug(regs, error_code); #endif @@ -288,7 +305,7 @@ 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)) { local_irq_enable(); handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code); goto exit; @@ -296,18 +313,42 @@ do_general_protection(struct pt_regs *regs, long error_code) #endif tsk = current; - if (!user_mode(regs)) { + if (!user_mode_novm(regs)) { if (fixup_exception(regs)) goto exit; tsk->thread.error_code = error_code; tsk->thread.trap_nr = X86_TRAP_GP; if (notify_die(DIE_GPF, "general protection fault", regs, error_code, - X86_TRAP_GP, SIGSEGV) != NOTIFY_STOP) + X86_TRAP_GP, SIGSEGV) != NOTIFY_STOP) { + +#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); + } goto exit; } +#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_nr = X86_TRAP_GP; @@ -481,7 +522,7 @@ dotraplinkage void 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, X86_TRAP_DB); preempt_conditional_cli(regs); @@ -496,7 +537,7 @@ dotraplinkage void 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; @@ -529,7 +570,7 @@ static 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/tsc.c b/arch/x86/kernel/tsc.c index ea03031..34a5cdda 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -150,7 +150,7 @@ static void cyc2ns_write_end(int cpu, struct cyc2ns_data *data) */ smp_wmb(); - ACCESS_ONCE(c2n->head) = data; + ACCESS_ONCE_RW(c2n->head) = data; } /* diff --git a/arch/x86/kernel/uprobes.c b/arch/x86/kernel/uprobes.c index 5d1cbfe..2a21feb 100644 --- a/arch/x86/kernel/uprobes.c +++ b/arch/x86/kernel/uprobes.c @@ -845,7 +845,7 @@ int arch_uprobe_exception_notify(struct notifier_block *self, unsigned long val, int ret = NOTIFY_DONE; /* We are only interested in userspace traps */ - if (regs && !user_mode_vm(regs)) + if (regs && !user_mode(regs)) return NOTIFY_DONE; switch (val) { @@ -919,7 +919,7 @@ arch_uretprobe_hijack_return_addr(unsigned long trampoline_vaddr, struct pt_regs if (nleft != rasize) { pr_err("uprobe: return address clobbered: pid=%d, %%sp=%#lx, " - "%%ip=%#lx\n", current->pid, regs->sp, regs->ip); + "%%ip=%#lx\n", task_pid_nr(current), regs->sp, regs->ip); force_sig_info(SIGSEGV, SEND_SIG_FORCED, current); } 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 e8edcf5..27f9344 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -44,6 +44,7 @@ #include #include #include +#include #include #include @@ -150,7 +151,7 @@ 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); @@ -214,6 +215,14 @@ SYSCALL_DEFINE1(vm86old, struct vm86_struct __user *, v86) if (tsk->thread.saved_sp0) return -EPERM; + +#ifdef CONFIG_GRKERNSEC_VM86 + if (!capable(CAP_SYS_RAWIO)) { + gr_handle_vm86(); + return -EPERM; + } +#endif + tmp = copy_vm86_regs_from_user(&info.regs, &v86->regs, offsetof(struct kernel_vm86_struct, vm86plus) - sizeof(info.regs)); @@ -238,6 +247,13 @@ SYSCALL_DEFINE2(vm86, unsigned long, cmd, unsigned long, arg) int tmp; struct vm86plus_struct __user *v86; +#ifdef CONFIG_GRKERNSEC_VM86 + if (!capable(CAP_SYS_RAWIO)) { + gr_handle_vm86(); + return -EPERM; + } +#endif + tsk = current; switch (cmd) { case VM86_REQUEST_IRQ: @@ -318,7 +334,7 @@ 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; @@ -525,7 +541,7 @@ 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 49edf2d..c0d1362 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 . = ALIGN(8); _stext = .; @@ -104,13 +124,47 @@ 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) { + +#ifdef CONFIG_PAX_KERNEXEC + MODULES_EXEC_VADDR = .; + BYTE(0) + . += (CONFIG_PAX_KERNEXEC_MODULE_TEXT * 1024 * 1024); + . = ALIGN(HPAGE_SIZE) - 1; + MODULES_EXEC_END = .; +#endif + + } :module +#endif + + .text.end : AT(ADDR(.text.end) - LOAD_OFFSET) { + /* End of text section */ + BYTE(0) + _etext = . - __KERNEL_TEXT_OFFSET; + } + +#ifdef CONFIG_X86_32 + . = ALIGN(PAGE_SIZE); + .rodata.page_aligned : AT(ADDR(.rodata.page_aligned) - LOAD_OFFSET) { + . = 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 */ @@ -122,16 +176,20 @@ 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) @@ -174,12 +232,19 @@ SECTIONS . = ALIGN(__vvar_page + PAGE_SIZE, PAGE_SIZE); /* 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 @@ -188,12 +253,27 @@ 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 .x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) { __x86_cpu_dev_start = .; @@ -264,19 +344,12 @@ 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 @@ -295,16 +368,10 @@ 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) { @@ -320,6 +387,7 @@ SECTIONS __brk_base = .; . += 64 * 1024; /* 64k alignment slop space */ *(.brk_reservation) /* areas brk users have reserved */ + . = ALIGN(HPAGE_SIZE); __brk_limit = .; } @@ -346,13 +414,12 @@ 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 e1e1e80..1400089 100644 --- a/arch/x86/kernel/vsyscall_64.c +++ b/arch/x86/kernel/vsyscall_64.c @@ -54,15 +54,13 @@ DEFINE_VVAR(int, vgetcpu_mode); -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 @@ -279,8 +277,7 @@ do_ret: return true; sigsegv: - force_sig(SIGSEGV, current); - return true; + do_group_exit(SIGKILL); } /* @@ -331,10 +328,7 @@ void __init map_vsyscall(void) extern char __vsyscall_page; unsigned long physaddr_vsyscall = __pa_symbol(&__vsyscall_page); - __set_fixmap(VSYSCALL_PAGE, physaddr_vsyscall, - vsyscall_mode == NATIVE - ? PAGE_KERNEL_VSYSCALL - : PAGE_KERNEL_VVAR); + __set_fixmap(VSYSCALL_PAGE, physaddr_vsyscall, PAGE_KERNEL_VVAR); BUILD_BUG_ON((unsigned long)__fix_to_virt(VSYSCALL_PAGE) != (unsigned long)VSYSCALL_ADDR); } diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c index 04068192..4d75aa6 100644 --- a/arch/x86/kernel/x8664_ksyms_64.c +++ b/arch/x86/kernel/x8664_ksyms_64.c @@ -34,8 +34,6 @@ EXPORT_SYMBOL(copy_user_generic_string); EXPORT_SYMBOL(copy_user_generic_unrolled); EXPORT_SYMBOL(copy_user_enhanced_fast_string); EXPORT_SYMBOL(__copy_user_nocache); -EXPORT_SYMBOL(_copy_from_user); -EXPORT_SYMBOL(_copy_to_user); EXPORT_SYMBOL(copy_page); EXPORT_SYMBOL(clear_page); @@ -73,3 +71,7 @@ EXPORT_SYMBOL(___preempt_schedule); EXPORT_SYMBOL(___preempt_schedule_context); #endif #endif + +#ifdef CONFIG_PAX_PER_CPU_PGD +EXPORT_SYMBOL(cpu_pgd); +#endif diff --git a/arch/x86/kernel/x86_init.c b/arch/x86/kernel/x86_init.c index e48b674..a451dd9 100644 --- a/arch/x86/kernel/x86_init.c +++ b/arch/x86/kernel/x86_init.c @@ -93,7 +93,7 @@ struct x86_cpuinit_ops x86_cpuinit = { static void default_nmi_init(void) { }; static int default_i8042_detect(void) { return 1; }; -struct x86_platform_ops x86_platform = { +struct x86_platform_ops x86_platform __read_only = { .calibrate_tsc = native_calibrate_tsc, .get_wallclock = mach_get_cmos_time, .set_wallclock = mach_set_rtc_mmss, @@ -109,7 +109,7 @@ struct x86_platform_ops x86_platform = { EXPORT_SYMBOL_GPL(x86_platform); #if defined(CONFIG_PCI_MSI) -struct x86_msi_ops x86_msi = { +struct x86_msi_ops x86_msi __read_only = { .setup_msi_irqs = native_setup_msi_irqs, .compose_msi_msg = native_compose_msi_msg, .teardown_msi_irq = native_teardown_msi_irq, @@ -150,7 +150,7 @@ u32 arch_msix_mask_irq(struct msi_desc *desc, u32 flag) } #endif -struct x86_io_apic_ops x86_io_apic_ops = { +struct x86_io_apic_ops x86_io_apic_ops __read_only = { .init = native_io_apic_init_mappings, .read = native_io_apic_read, .write = native_io_apic_write, diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index a4b451c..8dfe1ad 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -164,18 +164,18 @@ static inline int save_xstate_epilog(void __user *buf, int ia32_frame) /* Setup the bytes not touched by the [f]xsave and reserved for SW. */ sw_bytes = ia32_frame ? &fx_sw_reserved_ia32 : &fx_sw_reserved; - err = __copy_to_user(&x->i387.sw_reserved, sw_bytes, sizeof(*sw_bytes)); + err = __copy_to_user(x->i387.sw_reserved, sw_bytes, sizeof(*sw_bytes)); if (!use_xsave()) return err; - err |= __put_user(FP_XSTATE_MAGIC2, (__u32 *)(buf + xstate_size)); + err |= __put_user(FP_XSTATE_MAGIC2, (__u32 __user *)(buf + xstate_size)); /* * Read the xstate_bv which we copied (directly from the cpu or * from the state in task struct) to the user buffers. */ - err |= __get_user(xstate_bv, (__u32 *)&x->xsave_hdr.xstate_bv); + err |= __get_user(xstate_bv, (__u32 __user *)&x->xsave_hdr.xstate_bv); /* * For legacy compatible, we always set FP/SSE bits in the bit @@ -190,7 +190,7 @@ static inline int save_xstate_epilog(void __user *buf, int ia32_frame) */ xstate_bv |= XSTATE_FPSSE; - err |= __put_user(xstate_bv, (__u32 *)&x->xsave_hdr.xstate_bv); + err |= __put_user(xstate_bv, (__u32 __user *)&x->xsave_hdr.xstate_bv); return err; } @@ -199,6 +199,7 @@ static inline int save_user_xstate(struct xsave_struct __user *buf) { int err; + buf = (struct xsave_struct __user *)____m(buf); if (use_xsave()) err = xsave_user(buf); else if (use_fxsr()) @@ -311,6 +312,7 @@ sanitize_restored_xstate(struct task_struct *tsk, */ static inline int restore_user_xstate(void __user *buf, u64 xbv, int fx_only) { + buf = (void __user *)____m(buf); if (use_xsave()) { if ((unsigned long)buf % 64 || fx_only) { u64 init_bv = pcntxt_mask & ~XSTATE_FPSSE; diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c index 38a0afe..94421a9 100644 --- a/arch/x86/kvm/cpuid.c +++ b/arch/x86/kvm/cpuid.c @@ -166,15 +166,20 @@ 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); @@ -189,15 +194,19 @@ 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/lapic.c b/arch/x86/kvm/lapic.c index 453e5fb..214168f 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -55,7 +55,7 @@ #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 4107765..d9eb358 100644 --- a/arch/x86/kvm/paging_tmpl.h +++ b/arch/x86/kvm/paging_tmpl.h @@ -331,7 +331,7 @@ 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; walker->ptep_user[walker->level - 1] = ptep_user; diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index b5e994a..35b5866 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -3541,7 +3541,11 @@ 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(); } @@ -3942,6 +3946,10 @@ 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 801332e..eeff1cc 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -1339,12 +1339,12 @@ static void vmcs_write64(unsigned long field, u64 value) #endif } -static void vmcs_clear_bits(unsigned long field, u32 mask) +static void vmcs_clear_bits(unsigned long field, unsigned long mask) { vmcs_writel(field, vmcs_readl(field) & ~mask); } -static void vmcs_set_bits(unsigned long field, u32 mask) +static void vmcs_set_bits(unsigned long field, unsigned long mask) { vmcs_writel(field, vmcs_readl(field) | mask); } @@ -1604,7 +1604,11 @@ 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(); } @@ -1832,6 +1836,10 @@ static void vmx_vcpu_load(struct kvm_vcpu *vcpu, int cpu) vmcs_writel(HOST_TR_BASE, kvm_read_tr_base()); /* 22.2.4 */ vmcs_writel(HOST_GDTR_BASE, gdt->address); /* 22.2.4 */ +#ifdef CONFIG_PAX_PER_CPU_PGD + vmcs_writel(HOST_CR3, read_cr3()); /* 22.2.3 FIXME: shadow tables */ +#endif + rdmsrl(MSR_IA32_SYSENTER_ESP, sysenter_esp); vmcs_writel(HOST_IA32_SYSENTER_ESP, sysenter_esp); /* 22.2.3 */ vmx->loaded_vmcs->cpu = cpu; @@ -2121,7 +2129,7 @@ static void setup_msrs(struct vcpu_vmx *vmx) * reads and returns guest's timestamp counter "register" * guest_tsc = host_tsc + tsc_offset -- 21.3 */ -static u64 guest_read_tsc(void) +static u64 __intentional_overflow(-1) guest_read_tsc(void) { u64 host_tsc, tsc_offset; @@ -3093,8 +3101,11 @@ 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(); @@ -3105,13 +3116,15 @@ static __init int hardware_setup(void) if (!cpu_has_vmx_apicv()) enable_apicv = 0; + pax_open_kernel(); if (enable_apicv) - kvm_x86_ops->update_cr8_intercept = NULL; + *(void **)&kvm_x86_ops->update_cr8_intercept = NULL; else { - kvm_x86_ops->hwapic_irr_update = NULL; - kvm_x86_ops->deliver_posted_interrupt = NULL; - kvm_x86_ops->sync_pir_to_irr = vmx_sync_pir_to_irr_dummy; + *(void **)&kvm_x86_ops->hwapic_irr_update = NULL; + *(void **)&kvm_x86_ops->deliver_posted_interrupt = NULL; + *(void **)&kvm_x86_ops->sync_pir_to_irr = vmx_sync_pir_to_irr_dummy; } + pax_close_kernel(); if (nested) nested_vmx_setup_ctls_msrs(); @@ -4221,7 +4234,10 @@ static void vmx_set_constant_host_state(struct vcpu_vmx *vmx) vmcs_writel(HOST_CR0, read_cr0() & ~X86_CR0_TS); /* 22.2.3 */ vmcs_writel(HOST_CR4, read_cr4()); /* 22.2.3, 22.2.5 */ + +#ifndef CONFIG_PAX_PER_CPU_PGD vmcs_writel(HOST_CR3, read_cr3()); /* 22.2.3 FIXME: shadow tables */ +#endif vmcs_write16(HOST_CS_SELECTOR, __KERNEL_CS); /* 22.2.4 */ #ifdef CONFIG_X86_64 @@ -4243,7 +4259,7 @@ static void vmx_set_constant_host_state(struct vcpu_vmx *vmx) vmcs_writel(HOST_IDTR_BASE, dt.address); /* 22.2.4 */ vmx->host_idt_base = dt.address; - vmcs_writel(HOST_RIP, vmx_return); /* 22.2.5 */ + vmcs_writel(HOST_RIP, ktla_ktva(vmx_return)); /* 22.2.5 */ rdmsr(MSR_IA32_SYSENTER_CS, low32, high32); vmcs_write32(HOST_IA32_SYSENTER_CS, low32); @@ -7413,6 +7429,12 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) "jmp 2f \n\t" "1: " __ex(ASM_VMX_VMRESUME) "\n\t" "2: " + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + "ljmp %[cs],$3f\n\t" + "3: " +#endif + /* Save guest registers, load host registers, keep flags */ "mov %0, %c[wordsize](%%" _ASM_SP ") \n\t" "pop %0 \n\t" @@ -7465,6 +7487,11 @@ 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" #ifdef CONFIG_X86_64 , "rax", "rbx", "rdi", "rsi" @@ -7478,7 +7505,7 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) if (debugctlmsr) update_debugctlmsr(debugctlmsr); -#ifndef CONFIG_X86_64 +#ifdef CONFIG_X86_32 /* * The sysexit path does not restore ds/es, so we must set them to * a reasonable value ourselves. @@ -7487,8 +7514,18 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) * may be executed in interrupt context, which saves and restore segments * around it, nullifying its effect. */ - loadsegment(ds, __USER_DS); - loadsegment(es, __USER_DS); + loadsegment(ds, __KERNEL_DS); + loadsegment(es, __KERNEL_DS); + loadsegment(ss, __KERNEL_DS); + +#ifdef CONFIG_PAX_KERNEXEC + loadsegment(fs, __KERNEL_PERCPU); +#endif + +#ifdef CONFIG_PAX_MEMORY_UDEREF + __set_fs(current_thread_info()->addr_limit); +#endif + #endif vcpu->arch.regs_avail = ~((1 << VCPU_REGS_RIP) | (1 << VCPU_REGS_RSP) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index ef432f8..a630659 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1808,8 +1808,8 @@ 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; @@ -2729,6 +2729,8 @@ 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; @@ -5567,7 +5569,7 @@ static struct notifier_block pvclock_gtod_notifier = { }; #endif -int kvm_arch_init(void *opaque) +int kvm_arch_init(const void *opaque) { int r; struct kvm_x86_ops *ops = opaque; diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index aae9413..d11e829 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -1206,9 +1206,10 @@ 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_386_32.S b/arch/x86/lib/atomic64_386_32.S index 00933d5..3a64af9 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, %ecx 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 f5cc9eb..51fa319 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 @@ -60,12 +85,13 @@ 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 @@ -82,27 +108,44 @@ 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 @@ -112,21 +155,39 @@ 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 @@ -138,6 +199,13 @@ 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) @@ -147,6 +215,7 @@ 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) @@ -171,6 +240,13 @@ ENTRY(atomic64_add_unless_cx8) movl %edx, %ecx addl %ebp, %ebx adcl %edi, %ecx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 3f) +#endif + LOCK_PREFIX cmpxchg8b (%esi) jne 1b @@ -181,6 +257,7 @@ ENTRY(atomic64_add_unless_cx8) CFI_ADJUST_CFA_OFFSET -8 RESTORE ebx RESTORE ebp + pax_force_retaddr ret 4: cmpl %edx, 4(%esp) @@ -203,6 +280,13 @@ ENTRY(atomic64_inc_not_zero_cx8) xorl %ecx, %ecx addl $1, %ebx adcl %edx, %ecx + +#ifdef CONFIG_PAX_REFCOUNT + into +1234: + _ASM_EXTABLE(1234b, 3f) +#endif + LOCK_PREFIX cmpxchg8b (%esi) jne 1b @@ -210,6 +294,7 @@ ENTRY(atomic64_inc_not_zero_cx8) movl $1, %eax 3: RESTORE ebx + pax_force_retaddr ret CFI_ENDPROC ENDPROC(atomic64_inc_not_zero_cx8) diff --git a/arch/x86/lib/checksum_32.S b/arch/x86/lib/checksum_32.S index e78b8eee..7e173a8 100644 --- a/arch/x86/lib/checksum_32.S +++ b/arch/x86/lib/checksum_32.S @@ -29,7 +29,8 @@ #include #include #include - +#include + /* * computes a partial checksum, e.g. for TCP/UDP fragments */ @@ -293,9 +294,24 @@ 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 @@ -317,7 +333,7 @@ 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 @@ -329,30 +345,30 @@ 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 @@ -366,7 +382,7 @@ 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 @@ -378,12 +394,12 @@ 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: @@ -394,7 +410,7 @@ 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 @@ -407,11 +423,15 @@ 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 @@ -421,26 +441,43 @@ 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 @@ -461,7 +498,7 @@ 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 @@ -482,19 +519,19 @@ 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 @@ -502,10 +539,17 @@ 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 @@ -514,7 +558,7 @@ 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 176cca6..e0d658e 100644 --- a/arch/x86/lib/copy_page_64.S +++ b/arch/x86/lib/copy_page_64.S @@ -9,6 +9,7 @@ copy_page_rep: CFI_STARTPROC movl $4096/8, %ecx rep movsq + pax_force_retaddr ret CFI_ENDPROC ENDPROC(copy_page_rep) @@ -24,8 +25,8 @@ ENTRY(copy_page) CFI_ADJUST_CFA_OFFSET 2*8 movq %rbx, (%rsp) CFI_REL_OFFSET rbx, 0 - movq %r12, 1*8(%rsp) - CFI_REL_OFFSET r12, 1*8 + movq %r13, 1*8(%rsp) + CFI_REL_OFFSET r13, 1*8 movl $(4096/64)-5, %ecx .p2align 4 @@ -38,7 +39,7 @@ ENTRY(copy_page) movq 0x8*4(%rsi), %r9 movq 0x8*5(%rsi), %r10 movq 0x8*6(%rsi), %r11 - movq 0x8*7(%rsi), %r12 + movq 0x8*7(%rsi), %r13 prefetcht0 5*64(%rsi) @@ -49,7 +50,7 @@ ENTRY(copy_page) movq %r9, 0x8*4(%rdi) movq %r10, 0x8*5(%rdi) movq %r11, 0x8*6(%rdi) - movq %r12, 0x8*7(%rdi) + movq %r13, 0x8*7(%rdi) leaq 64 (%rsi), %rsi leaq 64 (%rdi), %rdi @@ -68,7 +69,7 @@ ENTRY(copy_page) movq 0x8*4(%rsi), %r9 movq 0x8*5(%rsi), %r10 movq 0x8*6(%rsi), %r11 - movq 0x8*7(%rsi), %r12 + movq 0x8*7(%rsi), %r13 movq %rax, 0x8*0(%rdi) movq %rbx, 0x8*1(%rdi) @@ -77,7 +78,7 @@ ENTRY(copy_page) movq %r9, 0x8*4(%rdi) movq %r10, 0x8*5(%rdi) movq %r11, 0x8*6(%rdi) - movq %r12, 0x8*7(%rdi) + movq %r13, 0x8*7(%rdi) leaq 64(%rdi), %rdi leaq 64(%rsi), %rsi @@ -85,10 +86,11 @@ ENTRY(copy_page) movq (%rsp), %rbx CFI_RESTORE rbx - movq 1*8(%rsp), %r12 - CFI_RESTORE r12 + movq 1*8(%rsp), %r13 + CFI_RESTORE r13 addq $2*8, %rsp CFI_ADJUST_CFA_OFFSET -2*8 + pax_force_retaddr ret .Lcopy_page_end: CFI_ENDPROC @@ -99,7 +101,7 @@ ENDPROC(copy_page) #include - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 1: .byte 0xeb /* jmp */ .byte (copy_page_rep - 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 dee945d..a84067b 100644 --- a/arch/x86/lib/copy_user_64.S +++ b/arch/x86/lib/copy_user_64.S @@ -18,31 +18,7 @@ #include #include #include - -/* - * By placing feature2 after feature1 in altinstructions section, we logically - * implement: - * If CPU has feature2, jmp to alt2 is used - * else if CPU has feature1, jmp to alt1 is used - * else jmp to orig is used. - */ - .macro ALTERNATIVE_JUMP feature1,feature2,orig,alt1,alt2 -0: - .byte 0xe9 /* 32bit jump */ - .long \orig-1f /* by default jump to orig */ -1: - .section .altinstr_replacement,"ax" -2: .byte 0xe9 /* near jump with 32bit immediate */ - .long \alt1-1b /* offset */ /* or alternatively to alt1 */ -3: .byte 0xe9 /* near jump with 32bit immediate */ - .long \alt2-1b /* offset */ /* or alternatively to alt2 */ - .previous - - .section .altinstructions,"a" - altinstruction_entry 0b,2b,\feature1,5,5 - altinstruction_entry 0b,3b,\feature2,5,5 - .previous - .endm +#include .macro ALIGN_DESTINATION #ifdef FIX_ALIGNMENT @@ -70,52 +46,6 @@ #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 - movl %edx,%ecx - xorl %eax,%eax - rep - stosb -bad_to_user: - movl %edx,%eax - ret - CFI_ENDPROC -ENDPROC(bad_from_user) - .previous - /* * copy_user_generic_unrolled - memory copy with exception handling. * This version is for CPUs like P4 that don't have efficient micro @@ -131,6 +61,7 @@ ENDPROC(bad_from_user) */ ENTRY(copy_user_generic_unrolled) CFI_STARTPROC + ASM_PAX_OPEN_USERLAND ASM_STAC cmpl $8,%edx jb 20f /* less then 8 bytes, go to byte copy loop */ @@ -180,6 +111,8 @@ ENTRY(copy_user_generic_unrolled) jnz 21b 23: xor %eax,%eax ASM_CLAC + ASM_PAX_CLOSE_USERLAND + pax_force_retaddr ret .section .fixup,"ax" @@ -235,6 +168,7 @@ ENDPROC(copy_user_generic_unrolled) */ ENTRY(copy_user_generic_string) CFI_STARTPROC + ASM_PAX_OPEN_USERLAND ASM_STAC cmpl $8,%edx jb 2f /* less than 8 bytes, go to byte copy loop */ @@ -249,6 +183,8 @@ ENTRY(copy_user_generic_string) movsb xorl %eax,%eax ASM_CLAC + ASM_PAX_CLOSE_USERLAND + pax_force_retaddr ret .section .fixup,"ax" @@ -276,12 +212,15 @@ ENDPROC(copy_user_generic_string) */ ENTRY(copy_user_enhanced_fast_string) CFI_STARTPROC + ASM_PAX_OPEN_USERLAND ASM_STAC movl %edx,%ecx 1: rep movsb xorl %eax,%eax ASM_CLAC + ASM_PAX_CLOSE_USERLAND + 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 6a4f43c..c70fb52 100644 --- a/arch/x86/lib/copy_user_nocache_64.S +++ b/arch/x86/lib/copy_user_nocache_64.S @@ -8,6 +8,7 @@ #include #include +#include #define FIX_ALIGNMENT 1 @@ -16,6 +17,7 @@ #include #include #include +#include .macro ALIGN_DESTINATION #ifdef FIX_ALIGNMENT @@ -49,6 +51,16 @@ */ 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 + + ASM_PAX_OPEN_USERLAND ASM_STAC cmpl $8,%edx jb 20f /* less then 8 bytes, go to byte copy loop */ @@ -98,7 +110,9 @@ ENTRY(__copy_user_nocache) jnz 21b 23: xorl %eax,%eax ASM_CLAC + ASM_PAX_CLOSE_USERLAND 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 2419d5f..fe52d0e 100644 --- a/arch/x86/lib/csum-copy_64.S +++ b/arch/x86/lib/csum-copy_64.S @@ -9,6 +9,7 @@ #include #include #include +#include /* * Checksum copy with exception handling. @@ -56,8 +57,8 @@ ENTRY(csum_partial_copy_generic) CFI_ADJUST_CFA_OFFSET 7*8 movq %rbx, 2*8(%rsp) CFI_REL_OFFSET rbx, 2*8 - movq %r12, 3*8(%rsp) - CFI_REL_OFFSET r12, 3*8 + movq %r15, 3*8(%rsp) + CFI_REL_OFFSET r15, 3*8 movq %r14, 4*8(%rsp) CFI_REL_OFFSET r14, 4*8 movq %r13, 5*8(%rsp) @@ -72,16 +73,16 @@ ENTRY(csum_partial_copy_generic) movl %edx, %ecx xorl %r9d, %r9d - movq %rcx, %r12 + movq %rcx, %r15 - shrq $6, %r12 + shrq $6, %r15 jz .Lhandle_tail /* < 64 */ clc /* main loop. clear in 64 byte blocks */ /* r9: zero, r8: temp2, rbx: temp1, rax: sum, rcx: saved length */ - /* r11: temp3, rdx: temp4, r12 loopcnt */ + /* r11: temp3, rdx: temp4, r15 loopcnt */ /* r10: temp5, rbp: temp6, r14 temp7, r13 temp8 */ .p2align 4 .Lloop: @@ -115,7 +116,7 @@ ENTRY(csum_partial_copy_generic) adcq %r14, %rax adcq %r13, %rax - decl %r12d + decl %r15d dest movq %rbx, (%rsi) @@ -210,8 +211,8 @@ ENTRY(csum_partial_copy_generic) .Lende: movq 2*8(%rsp), %rbx CFI_RESTORE rbx - movq 3*8(%rsp), %r12 - CFI_RESTORE r12 + movq 3*8(%rsp), %r15 + CFI_RESTORE r15 movq 4*8(%rsp), %r14 CFI_RESTORE r14 movq 5*8(%rsp), %r13 @@ -220,6 +221,7 @@ ENTRY(csum_partial_copy_generic) CFI_RESTORE rbp addq $7*8, %rsp CFI_ADJUST_CFA_OFFSET -7*8 + pax_force_retaddr ret CFI_RESTORE_STATE diff --git a/arch/x86/lib/csum-wrappers_64.c b/arch/x86/lib/csum-wrappers_64.c index 7609e0e..b449b98 100644 --- a/arch/x86/lib/csum-wrappers_64.c +++ b/arch/x86/lib/csum-wrappers_64.c @@ -53,10 +53,12 @@ csum_partial_copy_from_user(const void __user *src, void *dst, len -= 2; } } + pax_open_userland(); stac(); - isum = csum_partial_copy_generic((__force const void *)src, + isum = csum_partial_copy_generic((const void __force_kernel *)____m(src), dst, len, isum, errp, NULL); clac(); + pax_close_userland(); if (unlikely(*errp)) goto out_err; @@ -110,10 +112,12 @@ csum_partial_copy_to_user(const void *src, void __user *dst, } *errp = 0; + pax_open_userland(); stac(); - ret = csum_partial_copy_generic(src, (void __force *)dst, + ret = csum_partial_copy_generic(src, (void __force_kernel *)____m(dst), len, isum, NULL, errp); clac(); + pax_close_userland(); return ret; } EXPORT_SYMBOL(csum_partial_copy_to_user); diff --git a/arch/x86/lib/getuser.S b/arch/x86/lib/getuser.S index a451235..1daa956 100644 --- a/arch/x86/lib/getuser.S +++ b/arch/x86/lib/getuser.S @@ -33,17 +33,40 @@ #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 ASM_STAC -1: movzbl (%_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 movzbl (%_ASM_AX),%edx xor %eax,%eax ASM_CLAC + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__get_user_1) @@ -51,14 +74,28 @@ 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 ASM_STAC -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 ASM_CLAC + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__get_user_2) @@ -66,14 +103,28 @@ 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 ASM_STAC -3: movl -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 movl -3(%_ASM_AX),%edx xor %eax,%eax ASM_CLAC + pax_force_retaddr ret CFI_ENDPROC ENDPROC(__get_user_4) @@ -86,10 +137,20 @@ 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 + ASM_STAC 4: movq -7(%_ASM_AX),%rdx xor %eax,%eax ASM_CLAC + pax_force_retaddr ret #else add $7,%_ASM_AX @@ -98,10 +159,11 @@ ENTRY(__get_user_8) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user_8 ASM_STAC -4: movl -7(%_ASM_AX),%edx -5: movl -3(%_ASM_AX),%ecx +4: __copyuser_seg movl -7(%_ASM_AX),%edx +5: __copyuser_seg movl -3(%_ASM_AX),%ecx xor %eax,%eax ASM_CLAC + pax_force_retaddr ret #endif CFI_ENDPROC @@ -113,6 +175,7 @@ bad_get_user: xor %edx,%edx mov $(-EFAULT),%_ASM_AX ASM_CLAC + pax_force_retaddr ret CFI_ENDPROC END(bad_get_user) @@ -124,6 +187,7 @@ bad_get_user_8: xor %ecx,%ecx mov $(-EFAULT),%_ASM_AX ASM_CLAC + pax_force_retaddr ret CFI_ENDPROC END(bad_get_user_8) diff --git a/arch/x86/lib/insn.c b/arch/x86/lib/insn.c index 54fcffe..7be149e 100644 --- a/arch/x86/lib/insn.c +++ b/arch/x86/lib/insn.c @@ -20,8 +20,10 @@ #ifdef __KERNEL__ #include +#include #else #include +#define ktla_ktva(addr) addr #endif #include #include @@ -53,8 +55,8 @@ 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 56313a3..0db417e 100644 --- a/arch/x86/lib/memcpy_64.S +++ b/arch/x86/lib/memcpy_64.S @@ -24,7 +24,7 @@ * This gets patched over the unrolled variant (below) via the * alternative instructions framework: */ - .section .altinstr_replacement, "ax", @progbits + .section .altinstr_replacement, "a", @progbits .Lmemcpy_c: movq %rdi, %rax movq %rdx, %rcx @@ -33,6 +33,7 @@ rep movsq movl %edx, %ecx rep movsb + pax_force_retaddr ret .Lmemcpy_e: .previous @@ -44,11 +45,12 @@ * This gets patched over the unrolled variant (below) via the * alternative instructions framework: */ - .section .altinstr_replacement, "ax", @progbits + .section .altinstr_replacement, "a", @progbits .Lmemcpy_c_e: movq %rdi, %rax movq %rdx, %rcx rep movsb + pax_force_retaddr ret .Lmemcpy_e_e: .previous @@ -136,6 +138,7 @@ ENTRY(memcpy) movq %r9, 1*8(%rdi) movq %r10, -2*8(%rdi, %rdx) movq %r11, -1*8(%rdi, %rdx) + pax_force_retaddr retq .p2align 4 .Lless_16bytes: @@ -148,6 +151,7 @@ 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: @@ -161,6 +165,7 @@ ENTRY(memcpy) movl -4(%rsi, %rdx), %r8d movl %ecx, (%rdi) movl %r8d, -4(%rdi, %rdx) + pax_force_retaddr retq .p2align 4 .Lless_3bytes: @@ -179,6 +184,7 @@ ENTRY(memcpy) movb %cl, (%rdi) .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 65268a6..dd1de11 100644 --- a/arch/x86/lib/memmove_64.S +++ b/arch/x86/lib/memmove_64.S @@ -202,14 +202,16 @@ ENTRY(memmove) movb (%rsi), %r11b movb %r11b, (%rdi) 13: + pax_force_retaddr retq CFI_ENDPROC - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" .Lmemmove_begin_forward_efs: /* 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 2dcb380..2eb79fe 100644 --- a/arch/x86/lib/memset_64.S +++ b/arch/x86/lib/memset_64.S @@ -16,7 +16,7 @@ * * rax original destination */ - .section .altinstr_replacement, "ax", @progbits + .section .altinstr_replacement, "a", @progbits .Lmemset_c: movq %rdi,%r9 movq %rdx,%rcx @@ -30,6 +30,7 @@ movl %edx,%ecx rep stosb movq %r9,%rax + pax_force_retaddr ret .Lmemset_e: .previous @@ -45,13 +46,14 @@ * * rax original destination */ - .section .altinstr_replacement, "ax", @progbits + .section .altinstr_replacement, "a", @progbits .Lmemset_c_e: movq %rdi,%r9 movb %sil,%al movq %rdx,%rcx rep stosb movq %r9,%rax + pax_force_retaddr ret .Lmemset_e_e: .previous @@ -118,6 +120,7 @@ ENTRY(__memset) .Lende: movq %r10,%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 f6d13ee..d789440 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 /* @@ -37,6 +38,7 @@ ENTRY(\op\()_safe_regs) movl %edi, 28(%r10) 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 fc6ba17..d4d989d 100644 --- a/arch/x86/lib/putuser.S +++ b/arch/x86/lib/putuser.S @@ -16,7 +16,9 @@ #include #include #include - +#include +#include +#include /* * __put_user_X @@ -30,57 +32,125 @@ * as they get called from within inline assembly. */ -#define ENTER CFI_STARTPROC ; \ - GET_THREAD_INFO(%_ASM_BX) -#define EXIT ASM_CLAC ; \ - ret ; \ +#define ENTER CFI_STARTPROC +#define EXIT ASM_CLAC ; \ + 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 ASM_STAC -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 ASM_STAC -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 ASM_STAC -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 $7,%_ASM_BX cmp %_ASM_BX,%_ASM_CX jae bad_put_user ASM_STAC -4: mov %_ASM_AX,(%_ASM_CX) + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) + mov pax_user_shadow_base,%_ASM_BX + cmp %_ASM_BX,%_ASM_CX + jb 1234f + xor %ebx,%ebx +1234: +#endif + +#endif + +4: __copyuser_seg mov %_ASM_AX,(_DEST) #ifdef CONFIG_X86_32 -5: movl %edx,4(%_ASM_CX) +5: __copyuser_seg movl %edx,4(_DEST) #endif xor %eax,%eax EXIT diff --git a/arch/x86/lib/rwlock.S b/arch/x86/lib/rwlock.S index 1cad221..de671ee 100644 --- a/arch/x86/lib/rwlock.S +++ b/arch/x86/lib/rwlock.S @@ -16,13 +16,34 @@ ENTRY(__write_lock_failed) FRAME 0: LOCK_PREFIX WRITE_LOCK_ADD($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 92d9fea..b2762c8 100644 --- a/arch/x86/lib/thunk_64.S +++ b/arch/x86/lib/thunk_64.S @@ -9,6 +9,7 @@ #include #include #include +#include /* rdi: arg1 ... normal C conventions. rax is saved/restored. */ .macro THUNK name, func, put_ret_addr_in_rdi=0 @@ -16,11 +17,11 @@ \name: CFI_STARTPROC - /* this one pushes 9 elems, the next one would be %rIP */ - SAVE_ARGS + /* this one pushes 15+1 elems, the next one would be %rIP */ + SAVE_ARGS 8 .if \put_ret_addr_in_rdi - movq_cfi_restore 9*8, rdi + movq_cfi_restore RIP, rdi .endif call \func @@ -40,9 +41,10 @@ /* SAVE_ARGS below is used only for the .cfi directives it contains. */ CFI_STARTPROC - SAVE_ARGS + SAVE_ARGS 8 restore: - RESTORE_ARGS + RESTORE_ARGS 1,8 + pax_force_retaddr ret CFI_ENDPROC _ASM_NOKPROBE(restore) diff --git a/arch/x86/lib/usercopy_32.c b/arch/x86/lib/usercopy_32.c index e2f5e21..4b22130 100644 --- a/arch/x86/lib/usercopy_32.c +++ b/arch/x86/lib/usercopy_32.c @@ -42,11 +42,13 @@ do { \ int __d0; \ might_fault(); \ __asm__ __volatile__( \ + __COPYUSER_SET_ES \ ASM_STAC "\n" \ "0: rep; stosl\n" \ " movl %2,%0\n" \ "1: rep; stosb\n" \ "2: " ASM_CLAC "\n" \ + __COPYUSER_RESTORE_ES \ ".section .fixup,\"ax\"\n" \ "3: lea 0(%2,%0,4),%0\n" \ " jmp 2b\n" \ @@ -98,7 +100,7 @@ EXPORT_SYMBOL(__clear_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__( @@ -110,36 +112,36 @@ __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" @@ -149,10 +151,116 @@ __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" + ".previous\n" + _ASM_EXTABLE(1b,100b) + _ASM_EXTABLE(2b,100b) + _ASM_EXTABLE(3b,100b) + _ASM_EXTABLE(4b,100b) + _ASM_EXTABLE(5b,100b) + _ASM_EXTABLE(6b,100b) + _ASM_EXTABLE(7b,100b) + _ASM_EXTABLE(8b,100b) + _ASM_EXTABLE(9b,100b) + _ASM_EXTABLE(10b,100b) + _ASM_EXTABLE(11b,100b) + _ASM_EXTABLE(12b,100b) + _ASM_EXTABLE(13b,100b) + _ASM_EXTABLE(14b,100b) + _ASM_EXTABLE(15b,100b) + _ASM_EXTABLE(16b,100b) + _ASM_EXTABLE(17b,100b) + _ASM_EXTABLE(18b,100b) + _ASM_EXTABLE(19b,100b) + _ASM_EXTABLE(20b,100b) + _ASM_EXTABLE(21b,100b) + _ASM_EXTABLE(22b,100b) + _ASM_EXTABLE(23b,100b) + _ASM_EXTABLE(24b,100b) + _ASM_EXTABLE(25b,100b) + _ASM_EXTABLE(26b,100b) + _ASM_EXTABLE(27b,100b) + _ASM_EXTABLE(28b,100b) + _ASM_EXTABLE(29b,100b) + _ASM_EXTABLE(30b,100b) + _ASM_EXTABLE(31b,100b) + _ASM_EXTABLE(32b,100b) + _ASM_EXTABLE(33b,100b) + _ASM_EXTABLE(34b,100b) + _ASM_EXTABLE(35b,100b) + _ASM_EXTABLE(36b,100b) + _ASM_EXTABLE(37b,100b) + _ASM_EXTABLE(99b,101b) + : "=&c"(size), "=&D" (d0), "=&S" (d1) + : "1"(to), "2"(from), "0"(size) + : "eax", "edx", "memory"); + return 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" @@ -207,41 +315,41 @@ __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" @@ -253,9 +361,9 @@ __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" @@ -305,41 +413,41 @@ static unsigned long __copy_user_zeroing_intel_nocache(void *to, __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" @@ -352,9 +460,9 @@ 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" @@ -399,41 +507,41 @@ static unsigned long __copy_user_intel_nocache(void *to, __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" @@ -446,9 +554,9 @@ 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" @@ -488,32 +596,36 @@ 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" \ @@ -538,14 +650,14 @@ 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" \ @@ -572,9 +684,9 @@ unsigned long __copy_to_user_ll(void __user *to, const void *from, { stac(); 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); clac(); return n; } @@ -598,10 +710,9 @@ unsigned long __copy_from_user_ll_nozero(void *to, const void __user *from, { stac(); 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); clac(); return n; } @@ -632,58 +743,38 @@ 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 clac(); 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 n) +#ifdef CONFIG_PAX_MEMORY_UDEREF +void __set_fs(mm_segment_t x) { - if (access_ok(VERIFY_WRITE, to, n)) - n = __copy_to_user(to, from, n); - return n; + 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(); + } } -EXPORT_SYMBOL(_copy_to_user); +EXPORT_SYMBOL(__set_fs); -/** - * 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 n) +void set_fs(mm_segment_t x) { - if (access_ok(VERIFY_READ, from, n)) - n = __copy_from_user(to, from, n); - else - memset(to, 0, n); - return n; + current_thread_info()->addr_limit = x; + __set_fs(x); } -EXPORT_SYMBOL(_copy_from_user); +EXPORT_SYMBOL(set_fs); +#endif diff --git a/arch/x86/lib/usercopy_64.c b/arch/x86/lib/usercopy_64.c index c905e89..01ab928 100644 --- a/arch/x86/lib/usercopy_64.c +++ b/arch/x86/lib/usercopy_64.c @@ -18,6 +18,7 @@ unsigned long __clear_user(void __user *addr, unsigned long size) might_fault(); /* no memory constraint because it doesn't change any memory gcc knows about */ + pax_open_userland(); stac(); asm volatile( " testq %[size8],%[size8]\n" @@ -39,9 +40,10 @@ unsigned long __clear_user(void __user *addr, unsigned long size) _ASM_EXTABLE(0b,3b) _ASM_EXTABLE(1b,2b) : [size8] "=&c"(size), [dst] "=&D" (__d0) - : [size1] "r"(size & 7), "[size8]" (size / 8), "[dst]"(addr), + : [size1] "r"(size & 7), "[size8]" (size / 8), "[dst]"(____m(addr)), [zero] "r" (0UL), [eight] "r" (8UL)); clac(); + pax_close_userland(); return size; } EXPORT_SYMBOL(__clear_user); @@ -54,12 +56,11 @@ unsigned long clear_user(void __user *to, unsigned long n) } EXPORT_SYMBOL(clear_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)) + return copy_user_generic((void __force_kernel *)____m(to), (void __force_kernel *)____m(from), len); + return len; } EXPORT_SYMBOL(copy_in_user); @@ -69,11 +70,13 @@ EXPORT_SYMBOL(copy_in_user); * it is not necessary to optimize tail handling. */ __visible 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; + clac(); + pax_close_userland(); for (; len; --len, to++) { if (__get_user_nocheck(c, from++, sizeof(char))) break; @@ -84,6 +87,5 @@ copy_user_handle_tail(char *to, char *from, unsigned len, unsigned zerorest) for (c = 0, zero_len = len; zerorest && zero_len; --zero_len) if (__put_user_nocheck(c, to++, sizeof(char))) break; - clac(); return len; } diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 6a19ad9..1c48f9a 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -30,3 +30,7 @@ obj-$(CONFIG_ACPI_NUMA) += srat.o obj-$(CONFIG_NUMA_EMU) += numa_emulation.o obj-$(CONFIG_MEMTEST) += memtest.o + +quote:=" +obj-$(CONFIG_X86_64) += uderef_64.o +CFLAGS_uderef_64.o := $(subst $(quote),,$(CONFIG_ARCH_HWEIGHT_CFLAGS)) diff --git a/arch/x86/mm/extable.c b/arch/x86/mm/extable.c index 903ec1e..c4166b2 100644 --- a/arch/x86/mm/extable.c +++ b/arch/x86/mm/extable.c @@ -6,12 +6,24 @@ static inline unsigned long ex_insn_addr(const struct exception_table_entry *x) { - return (unsigned long)&x->insn + x->insn; + unsigned long reloc = 0; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + reloc = ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; +#endif + + return (unsigned long)&x->insn + x->insn + reloc; } static inline unsigned long ex_fixup_addr(const struct exception_table_entry *x) { - return (unsigned long)&x->fixup + x->fixup; + unsigned long reloc = 0; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + reloc = ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; +#endif + + return (unsigned long)&x->fixup + x->fixup + reloc; } int fixup_exception(struct pt_regs *regs) @@ -20,7 +32,7 @@ int fixup_exception(struct pt_regs *regs) unsigned long new_ip; #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; @@ -145,6 +157,13 @@ void sort_extable(struct exception_table_entry *start, i += 4; p->fixup -= i; i += 4; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + BUILD_BUG_ON(!IS_ENABLED(CONFIG_BUILDTIME_EXTABLE_SORT)); + p->insn -= ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; + p->fixup -= ____LOAD_PHYSICAL_ADDR - LOAD_PHYSICAL_ADDR; +#endif + } } diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index 3664279..c6a7830 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -14,12 +14,19 @@ #include /* hstate_index_to_shift */ #include /* prefetchw */ #include /* exception_enter(), ... */ +#include +#include #include /* dotraplinkage, ... */ #include /* pgd_*(), ... */ #include /* kmemcheck_*(), ... */ #include /* VSYSCALL_ADDR */ #include /* emulate_vsyscall */ +#include + +#if defined(CONFIG_X86_64) && defined(CONFIG_PAX_MEMORY_UDEREF) +#include +#endif #define CREATE_TRACE_POINTS #include @@ -60,7 +67,7 @@ static nokprobe_inline int kprobes_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; @@ -121,7 +128,10 @@ 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) && @@ -155,7 +165,10 @@ 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++; @@ -186,6 +199,34 @@ 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); @@ -236,10 +277,27 @@ 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, user); + pmd_t *ret; + + ret = vmalloc_sync_one(pgd, address); + if (!ret) + break; + pgd = get_cpu_pgd(cpu, kernel); +#else list_for_each_entry(page, &pgd_list, lru) { + pgd_t *pgd; spinlock_t *pgt_lock; pmd_t *ret; @@ -247,8 +305,14 @@ 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); + pgd = page_address(page); +#endif + + ret = vmalloc_sync_one(pgd, address); + +#ifndef CONFIG_PAX_PER_CPU_PGD spin_unlock(pgt_lock); +#endif if (!ret) break; @@ -282,6 +346,12 @@ static noinline 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(), kernel)) != (pgd_paddr & __PHYSICAL_MASK)); + vmalloc_sync_one(__va(pgd_paddr + PAGE_SIZE), address); +#endif + pmd_k = vmalloc_sync_one(__va(pgd_paddr), address); if (!pmd_k) return -1; @@ -378,11 +448,25 @@ static noinline int vmalloc_fault(unsigned long address) * happen within a race in page table update. In the later * case just flush: */ - pgd = pgd_offset(current->active_mm, address); + pgd_ref = pgd_offset_k(address); if (pgd_none(*pgd_ref)) return -1; +#ifdef CONFIG_PAX_PER_CPU_PGD + BUG_ON(__pa(get_cpu_pgd(smp_processor_id(), kernel)) != (read_cr3() & __PHYSICAL_MASK)); + pgd = pgd_offset_cpu(smp_processor_id(), user, address); + if (pgd_none(*pgd)) { + set_pgd(pgd, *pgd_ref); + arch_flush_lazy_mmu_mode(); + } else { + BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref)); + } + pgd = pgd_offset_cpu(smp_processor_id(), kernel, address); +#else + pgd = pgd_offset(current->active_mm, address); +#endif + if (pgd_none(*pgd)) { set_pgd(pgd, *pgd_ref); arch_flush_lazy_mmu_mode(); @@ -549,7 +633,7 @@ 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; @@ -576,7 +660,7 @@ 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, @@ -585,7 +669,7 @@ 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; pgd_t *pgd; pte_t *pte; @@ -596,9 +680,21 @@ show_fault_oops(struct pt_regs *regs, unsigned long error_code, pte = lookup_address_in_pgd(pgd, address, &level); if (pte && pte_present(*pte) && !pte_exec(*pte)) - printk(nx_warning, from_kuid(&init_user_ns, current_uid())); + printk(nx_warning, from_kuid_munged(&init_user_ns, 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), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, 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), + from_kuid_munged(&init_user_ns, current_uid()), from_kuid_munged(&init_user_ns, current_euid())); + } +#endif + printk(KERN_ALERT "BUG: unable to handle kernel "); if (address < PAGE_SIZE) printk(KERN_CONT "NULL pointer dereference"); @@ -779,6 +875,22 @@ __bad_area_nosemaphore(struct pt_regs *regs, unsigned long error_code, return; } #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 + /* Kernel addresses are always protection faults: */ if (address >= TASK_SIZE) error_code |= PF_PROT; @@ -864,7 +976,7 @@ 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 @@ -918,6 +1030,99 @@ 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. * @@ -985,6 +1190,9 @@ 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))) @@ -1019,7 +1227,7 @@ static inline bool smap_violation(int error_code, struct pt_regs *regs) if (error_code & PF_USER) return false; - if (!user_mode_vm(regs) && (regs->flags & X86_EFLAGS_AC)) + if (!user_mode(regs) && (regs->flags & X86_EFLAGS_AC)) return false; return true; @@ -1047,6 +1255,22 @@ __do_page_fault(struct pt_regs *regs, unsigned long error_code, tsk = current; mm = tsk->mm; +#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)) { + printk(KERN_ERR "PAX: please report this to pageexec@freemail.hu\n"); + 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 + /* * Detect and handle instructions that would cause a page fault for * both a tracked kernel page and a userspace page. @@ -1124,7 +1348,7 @@ __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; flags |= FAULT_FLAG_USER; @@ -1171,6 +1395,11 @@ 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); @@ -1182,18 +1411,24 @@ 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; @@ -1309,3 +1544,292 @@ trace_do_page_fault(struct pt_regs *regs, unsigned long error_code) } NOKPROBE_SYMBOL(trace_do_page_fault); #endif /* CONFIG_TRACING */ + +#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 207d9aef..69030980 100644 --- a/arch/x86/mm/gup.c +++ b/arch/x86/mm/gup.c @@ -268,7 +268,7 @@ 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_noprefault(write ? VERIFY_WRITE : VERIFY_READ, (void __user *)start, len))) return 0; @@ -344,6 +344,10 @@ int get_user_pages_fast(unsigned long start, int nr_pages, int write, goto slow_irqon; #endif + if (unlikely(!access_ok_noprefault(write ? VERIFY_WRITE : VERIFY_READ, + (void __user *)start, len))) + return 0; + /* * XXX: batch / limit 'nr', to avoid large irq off latency * needs some instrumenting to determine the common sizes used by diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c index 4500142..53a363c 100644 --- a/arch/x86/mm/highmem_32.c +++ b/arch/x86/mm/highmem_32.c @@ -45,7 +45,11 @@ 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 8b977eb..4732c33 100644 --- a/arch/x86/mm/hugetlbpage.c +++ b/arch/x86/mm/hugetlbpage.c @@ -80,23 +80,24 @@ int pud_huge(pud_t pud) #ifdef CONFIG_HUGETLB_PAGE static unsigned long hugetlb_get_unmapped_area_bottomup(struct file *file, unsigned long addr, unsigned long len, - unsigned long pgoff, unsigned long flags) + unsigned long pgoff, unsigned long flags, unsigned long offset) { struct hstate *h = hstate_file(file); struct vm_unmapped_area_info info; - + info.flags = 0; info.length = len; info.low_limit = current->mm->mmap_legacy_base; info.high_limit = TASK_SIZE; info.align_mask = PAGE_MASK & ~huge_page_mask(h); info.align_offset = 0; + info.threadstack_offset = offset; return vm_unmapped_area(&info); } static unsigned long hugetlb_get_unmapped_area_topdown(struct file *file, unsigned long addr0, unsigned long len, - unsigned long pgoff, unsigned long flags) + unsigned long pgoff, unsigned long flags, unsigned long offset) { struct hstate *h = hstate_file(file); struct vm_unmapped_area_info info; @@ -108,6 +109,7 @@ static unsigned long hugetlb_get_unmapped_area_topdown(struct file *file, info.high_limit = current->mm->mmap_base; info.align_mask = PAGE_MASK & ~huge_page_mask(h); info.align_offset = 0; + info.threadstack_offset = offset; addr = vm_unmapped_area(&info); /* @@ -120,6 +122,12 @@ static unsigned long hugetlb_get_unmapped_area_topdown(struct file *file, VM_BUG_ON(addr != -ENOMEM); info.flags = 0; info.low_limit = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (current->mm->pax_flags & MF_PAX_RANDMMAP) + info.low_limit += current->mm->delta_mmap; +#endif + info.high_limit = TASK_SIZE; addr = vm_unmapped_area(&info); } @@ -134,10 +142,20 @@ 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; + unsigned long offset = gr_rand_threadstack_offset(mm, file, flags); 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) { @@ -146,19 +164,22 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, return addr; } +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + 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, offset)) return addr; } if (mm->get_unmapped_area == arch_get_unmapped_area) return hugetlb_get_unmapped_area_bottomup(file, addr, len, - pgoff, flags); + pgoff, flags, offset); else return hugetlb_get_unmapped_area_topdown(file, addr, len, - pgoff, flags); + pgoff, flags, offset); } #endif /* CONFIG_HUGETLB_PAGE */ diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index f971306..e83e0f6 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -4,6 +4,7 @@ #include #include #include /* for max_low_pfn */ +#include #include #include @@ -17,6 +18,8 @@ #include #include /* for MAX_DMA_PFN */ #include +#include +#include #include "mm_internal.h" @@ -563,7 +566,18 @@ void __init init_mem_mapping(void) early_ioremap_page_table_range_init(); #endif +#ifdef CONFIG_PAX_PER_CPU_PGD + clone_pgd_range(get_cpu_pgd(0, kernel) + KERNEL_PGD_BOUNDARY, + swapper_pg_dir + KERNEL_PGD_BOUNDARY, + KERNEL_PGD_PTRS); + clone_pgd_range(get_cpu_pgd(0, user) + KERNEL_PGD_BOUNDARY, + swapper_pg_dir + KERNEL_PGD_BOUNDARY, + KERNEL_PGD_PTRS); + load_cr3(get_cpu_pgd(0, kernel)); +#else load_cr3(swapper_pg_dir); +#endif + __flush_tlb_all(); early_memtest(0, max_pfn_mapped << PAGE_SHIFT); @@ -579,10 +593,40 @@ void __init init_mem_mapping(void) * Access has to be given to non-kernel-ram areas as well, these contain the PCI * mmio resources as well as potential bios/acpi data regions. */ + +#ifdef CONFIG_GRKERNSEC_KMEM +static unsigned int ebda_start __read_only; +static unsigned int ebda_end __read_only; +#endif + int devmem_is_allowed(unsigned long pagenr) { - if (pagenr < 256) +#ifdef CONFIG_GRKERNSEC_KMEM + /* allow BDA */ + if (!pagenr) return 1; + /* allow EBDA */ + if (pagenr >= ebda_start && pagenr < ebda_end) + return 1; + /* if tboot is in use, allow access to its hardcoded serial log range */ + if (tboot_enabled() && ((0x60000 >> PAGE_SHIFT) <= pagenr) && (pagenr < (0x68000 >> PAGE_SHIFT))) + 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 0; +#endif if (iomem_is_exclusive(pagenr << PAGE_SHIFT)) return 0; if (!page_is_ram(pagenr)) @@ -628,8 +672,117 @@ void free_init_pages(char *what, unsigned long begin, unsigned long end) #endif } +#ifdef CONFIG_GRKERNSEC_KMEM +static inline void gr_init_ebda(void) +{ + unsigned int ebda_addr; + unsigned int ebda_size = 0; + + ebda_addr = get_bios_ebda(); + if (ebda_addr) { + ebda_size = *(unsigned char *)phys_to_virt(ebda_addr); + ebda_size <<= 10; + } + if (ebda_addr && ebda_size) { + ebda_start = ebda_addr >> PAGE_SHIFT; + ebda_end = min((unsigned int)PAGE_ALIGN(ebda_addr + ebda_size), (unsigned int)0xa0000) >> PAGE_SHIFT; + } else { + ebda_start = 0x9f000 >> PAGE_SHIFT; + ebda_end = 0xa0000 >> PAGE_SHIFT; + } +} +#else +static inline void gr_init_ebda(void) { } +#endif + 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; +#else + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + unsigned long addr, end; +#endif +#endif + + gr_init_ebda(); + +#ifdef CONFIG_PAX_KERNEXEC +#ifdef CONFIG_X86_32 + 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); + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_KERNEXEC_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 + /* 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", (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 e395048..cd38278 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -62,33 +62,6 @@ static noinline int do_test_wp_bit(void); bool __read_mostly __vmalloc_start_set = false; /* - * 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)) { - 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: */ @@ -98,13 +71,28 @@ static pte_t * __init one_page_table_init(pmd_t *pmd) pte_t *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); @@ -208,6 +196,7 @@ 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; unsigned long count = page_table_range_init_count(start, end); @@ -222,8 +211,13 @@ 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), @@ -235,11 +229,20 @@ 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; } /* @@ -256,9 +259,10 @@ 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; @@ -291,8 +295,13 @@ repeat: pfn = start_pfn; pgd_idx = pgd_index((pfn<> PAGE_SHIFT); +#endif if (pfn >= end_pfn) continue; @@ -304,14 +313,13 @@ 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 @@ -322,11 +330,7 @@ repeat: _PAGE_PSE); pfn &= PMD_MASK >> PAGE_SHIFT; - 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++; @@ -343,7 +347,7 @@ 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 +885,7 @@ 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 +900,7 @@ 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 +929,7 @@ 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 df1a992..94c272c 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -151,7 +151,7 @@ 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; @@ -184,12 +184,29 @@ 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, user, address); + + if (pgd_none(*pgd)) + set_pgd(pgd, *pgd_ref); + else + BUG_ON(pgd_page_vaddr(*pgd) + != pgd_page_vaddr(*pgd_ref)); + pgd = pgd_offset_cpu(cpu, kernel, address); +#else list_for_each_entry(page, &pgd_list, lru) { pgd_t *pgd; spinlock_t *pgt_lock; @@ -198,6 +215,7 @@ 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); @@ -205,7 +223,10 @@ 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); } @@ -238,7 +259,7 @@ 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)); @@ -250,7 +271,7 @@ 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)); @@ -279,7 +300,9 @@ 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. @@ -338,14 +361,12 @@ 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)); @@ -586,7 +607,7 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end, prot); spin_lock(&init_mm.page_table_lock); - pud_populate(&init_mm, pud, pmd); + pud_populate_kernel(&init_mm, pud, pmd); spin_unlock(&init_mm.page_table_lock); } __flush_tlb_all(); @@ -627,7 +648,7 @@ kernel_physical_mapping_init(unsigned long start, page_size_mask); spin_lock(&init_mm.page_table_lock); - pgd_populate(&init_mm, pgd, pud); + pgd_populate_kernel(&init_mm, pgd, pud); spin_unlock(&init_mm.page_table_lock); pgd_changed = true; } @@ -1195,8 +1216,8 @@ static struct vm_operations_struct gate_vma_ops = { static struct vm_area_struct gate_vma = { .vm_start = VSYSCALL_ADDR, .vm_end = VSYSCALL_ADDR + PAGE_SIZE, - .vm_page_prot = PAGE_READONLY_EXEC, - .vm_flags = VM_READ | VM_EXEC, + .vm_page_prot = PAGE_READONLY, + .vm_flags = VM_READ, .vm_ops = &gate_vma_ops, }; diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index 7b179b4..6bd17777 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 baff1da..2816ef4 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -56,8 +56,8 @@ static int __ioremap_check_ram(unsigned long start_pfn, unsigned long nr_pages, unsigned long i; for (i = 0; i < nr_pages; ++i) - if (pfn_valid(start_pfn + i) && - !PageReserved(pfn_to_page(start_pfn + i))) + if (pfn_valid(start_pfn + i) && (start_pfn + i >= 0x100 || + !PageReserved(pfn_to_page(start_pfn + i)))) return 1; WARN_ONCE(1, "ioremap on RAM pfn 0x%lx\n", start_pfn); @@ -268,7 +268,7 @@ EXPORT_SYMBOL(ioremap_prot); * * Caller must ensure there is only one unmapping for the same pointer. */ -void iounmap(volatile void __iomem *addr) +void iounmap(const volatile void __iomem *addr) { struct vm_struct *p, *o; @@ -322,6 +322,9 @@ 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); @@ -334,13 +337,16 @@ void *xlate_dev_mem_ptr(unsigned long phys) void unxlate_dev_mem_ptr(unsigned long phys, void *addr) { if (page_is_ram(phys >> PAGE_SHIFT)) +#ifdef CONFIG_HIGHMEM + if ((phys >> PAGE_SHIFT) < max_low_pfn) +#endif return; iounmap((void __iomem *)((unsigned long)addr & PAGE_MASK)); return; } -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) { @@ -376,8 +382,7 @@ void __init early_ioremap_init(void) early_ioremap_setup(); 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 dd89a13..d77bdcc 100644 --- a/arch/x86/mm/kmemcheck/kmemcheck.c +++ b/arch/x86/mm/kmemcheck/kmemcheck.c @@ -628,9 +628,9 @@ 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 25e7e13..1964579 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(); } @@ -112,8 +125,15 @@ static unsigned long mmap_legacy_base(void) */ void arch_pick_mmap_layout(struct mm_struct *mm) { - mm->mmap_legacy_base = mmap_legacy_base(); - mm->mmap_base = mmap_base(); + mm->mmap_legacy_base = mmap_legacy_base(mm); + mm->mmap_base = mmap_base(mm); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) { + mm->mmap_legacy_base += mm->delta_mmap; + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; + } +#endif if (mmap_is_legacy()) { mm->mmap_base = mm->mmap_legacy_base; diff --git a/arch/x86/mm/mmio-mod.c b/arch/x86/mm/mmio-mod.c index 0057a7a..95c7edd 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; @@ -290,7 +290,7 @@ void mmiotrace_ioremap(resource_size_t offset, unsigned long size, ioremap_trace_core(offset, size, addr); } -static void iounmap_trace_core(volatile void __iomem *addr) +static void iounmap_trace_core(const volatile void __iomem *addr) { struct mmiotrace_map map = { .phys = 0, @@ -328,7 +328,7 @@ not_enabled: } } -void mmiotrace_iounmap(volatile void __iomem *addr) +void mmiotrace_iounmap(const volatile void __iomem *addr) { might_sleep(); if (is_enabled()) /* recheck and proper locking in *_core() */ diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index a32b706..efb308b 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -478,7 +478,7 @@ static bool __init numa_meminfo_cover_memory(const struct numa_meminfo *mi) return true; } -static int __init numa_register_memblks(struct numa_meminfo *mi) +static int __init __intentional_overflow(-1) numa_register_memblks(struct numa_meminfo *mi) { unsigned long uninitialized_var(pfn_align); int i, nid; diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index ae242a7..1c7998f 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -262,7 +262,7 @@ 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 /* @@ -270,9 +270,10 @@ 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. @@ -280,6 +281,7 @@ static inline pgprot_t static_protections(pgprot_t prot, unsigned long address, if (within(pfn, __pa_symbol(__start_rodata) >> PAGE_SHIFT, __pa_symbol(__end_rodata) >> PAGE_SHIFT)) pgprot_val(forbidden) |= _PAGE_RW; +#endif #if defined(CONFIG_X86_64) && defined(CONFIG_DEBUG_RODATA) /* @@ -318,6 +320,13 @@ static inline pgprot_t static_protections(pgprot_t prot, unsigned long address, } #endif +#ifdef CONFIG_PAX_KERNEXEC + if (within(pfn, __pa(ktla_ktva((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; @@ -420,23 +429,37 @@ EXPORT_SYMBOL_GPL(slow_virt_to_phys); 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, kernel); +#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 6574388..87e9bef 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -376,7 +376,7 @@ int free_memtype(u64 start, u64 end) if (!entry) { printk(KERN_INFO "%s:%d freeing invalid memtype [mem %#010Lx-%#010Lx]\n", - current->comm, current->pid, start, end - 1); + current->comm, task_pid_nr(current), start, end - 1); return -EINVAL; } @@ -506,8 +506,8 @@ 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 [mem %#010Lx-%#010Lx]\n", - current->comm, from, to - 1); + printk(KERN_INFO "Program %s tried to access /dev/mem between [mem %#010Lx-%#010Lx] (%#010Lx)\n", + current->comm, from, to - 1, cursor); return 0; } cursor += PAGE_SIZE; @@ -577,7 +577,7 @@ int kernel_map_sync_memtype(u64 base, unsigned long size, unsigned long flags) if (ioremap_change_attr((unsigned long)__va(base), id_sz, flags) < 0) { printk(KERN_INFO "%s:%d ioremap_change_attr failed %s " "for [mem %#010Lx-%#010Lx]\n", - current->comm, current->pid, + current->comm, task_pid_nr(current), cattr_name(flags), base, (unsigned long long)(base + size-1)); return -EINVAL; @@ -612,7 +612,7 @@ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot, flags = lookup_memtype(paddr); if (want_flags != flags) { printk(KERN_WARNING "%s:%d map pfn RAM range req %s for [mem %#010Lx-%#010Lx], 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 - 1), @@ -634,7 +634,7 @@ 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 [mem %#010Lx-%#010Lx], 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 - 1), diff --git a/arch/x86/mm/pat_rbtree.c b/arch/x86/mm/pat_rbtree.c index 415f6c4..d319983 100644 --- a/arch/x86/mm/pat_rbtree.c +++ b/arch/x86/mm/pat_rbtree.c @@ -160,7 +160,7 @@ success: failure: printk(KERN_INFO "%s:%d conflicting memory types " - "%Lx-%Lx %s<->%s\n", current->comm, current->pid, start, + "%Lx-%Lx %s<->%s\n", current->comm, task_pid_nr(current), start, end, cattr_name(found_type), cattr_name(match->type)); return -EBUSY; } 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 6fb6927..4fc13c0 100644 --- a/arch/x86/mm/pgtable.c +++ b/arch/x86/mm/pgtable.c @@ -97,10 +97,71 @@ 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) +{ + unsigned int count = USER_PGD_PTRS; + if (!pax_user_shadow_base) + return; + + 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) +{ + unsigned int count = USER_PGD_PTRS; + + 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 pgtable_pxd_page_ctor(page) true +#define pgtable_pxd_page_dtor(page) +#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 pgtable_pxd_page_ctor(page) pgtable_pmd_page_ctor(page) +#define pgtable_pxd_page_dtor(page) pgtable_pmd_page_dtor(page) +#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)); @@ -141,6 +202,7 @@ 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 @@ -153,7 +215,7 @@ static void pgd_dtor(pgd_t *pgd) * -- nyc */ -#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 @@ -165,7 +227,7 @@ 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) { @@ -183,43 +245,45 @@ 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]) { - pgtable_pmd_page_dtor(virt_to_page(pmds[i])); - free_page((unsigned long)pmds[i]); + for(i = 0; i < PREALLOCATED_PXDS; i++) + if (pxds[i]) { + pgtable_pxd_page_dtor(virt_to_page(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) + for(i = 0; i < PREALLOCATED_PXDS; i++) { + pxd_t *pxd = (pxd_t *)__get_free_page(PGALLOC_GFP); + if (!pxd) failed = true; - if (pmd && !pgtable_pmd_page_ctor(virt_to_page(pmd))) { - free_page((unsigned long)pmd); - pmd = NULL; + if (pxd && !pgtable_pxd_page_ctor(virt_to_page(pxd))) { + free_page((unsigned long)pxd); + pxd = NULL; failed = true; } - pmds[i] = pmd; + pxds[i] = pxd; } if (failed) { - free_pmds(pmds); + free_pxds(pxds); return -ENOMEM; } @@ -232,49 +296,52 @@ 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; 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); - - for (i = 0; i < PREALLOCATED_PMDS; i++, pud++) { - pmd_t *pmd = pmds[i]; +#ifdef CONFIG_X86_64 + pyd = pyd_offset(mm, 0L); +#else + pyd = pyd_offset(pgd, 0L); +#endif + for (i = 0; i < PREALLOCATED_PXDS; i++, pyd++) { + 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); @@ -283,11 +350,11 @@ 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 @@ -297,14 +364,14 @@ 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: @@ -313,7 +380,7 @@ 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 4dd8cf6..f9d143e 100644 --- a/arch/x86/mm/pgtable_32.c +++ b/arch/x86/mm/pgtable_32.c @@ -47,10 +47,13 @@ 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/physaddr.c b/arch/x86/mm/physaddr.c index e666cbb..61788c45 100644 --- a/arch/x86/mm/physaddr.c +++ b/arch/x86/mm/physaddr.c @@ -10,7 +10,7 @@ #ifdef CONFIG_X86_64 #ifdef CONFIG_DEBUG_VIRTUAL -unsigned long __phys_addr(unsigned long x) +unsigned long __intentional_overflow(-1) __phys_addr(unsigned long x) { unsigned long y = x - __START_KERNEL_map; @@ -67,7 +67,7 @@ EXPORT_SYMBOL(__virt_addr_valid); #else #ifdef CONFIG_DEBUG_VIRTUAL -unsigned long __phys_addr(unsigned long x) +unsigned long __intentional_overflow(-1) __phys_addr(unsigned long x) { unsigned long phys_addr = x - PAGE_OFFSET; /* VMALLOC_* aren't constants */ diff --git a/arch/x86/mm/setup_nx.c b/arch/x86/mm/setup_nx.c index 90555bf..f5f1828 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; +#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 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 dd8dda1..9e9b0f6 100644 --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -48,7 +48,11 @@ void leave_mm(int cpu) BUG(); if (cpumask_test_cpu(cpu, mm_cpumask(active_mm))) { cpumask_clear_cpu(cpu, mm_cpumask(active_mm)); + +#ifndef CONFIG_PAX_PER_CPU_PGD load_cr3(swapper_pg_dir); +#endif + } } EXPORT_SYMBOL_GPL(leave_mm); diff --git a/arch/x86/mm/uderef_64.c b/arch/x86/mm/uderef_64.c new file mode 100644 index 0000000..dace51c --- /dev/null +++ b/arch/x86/mm/uderef_64.c @@ -0,0 +1,37 @@ +#include +#include +#include + +#ifdef CONFIG_PAX_MEMORY_UDEREF +/* PaX: due to the special call convention these functions must + * - remain leaf functions under all configurations, + * - never be called directly, only dereferenced from the wrappers. + */ +void __pax_open_userland(void) +{ + unsigned int cpu; + + if (unlikely(!segment_eq(get_fs(), USER_DS))) + return; + + cpu = raw_get_cpu(); + BUG_ON((read_cr3() & ~PAGE_MASK) != PCID_KERNEL); + write_cr3(__pa(get_cpu_pgd(cpu, user)) | PCID_USER | PCID_NOFLUSH); + raw_put_cpu_no_resched(); +} +EXPORT_SYMBOL(__pax_open_userland); + +void __pax_close_userland(void) +{ + unsigned int cpu; + + if (unlikely(!segment_eq(get_fs(), USER_DS))) + return; + + cpu = raw_get_cpu(); + BUG_ON((read_cr3() & ~PAGE_MASK) != PCID_USER); + write_cr3(__pa(get_cpu_pgd(cpu, kernel)) | PCID_KERNEL | PCID_NOFLUSH); + raw_put_cpu_no_resched(); +} +EXPORT_SYMBOL(__pax_close_userland); +#endif diff --git a/arch/x86/net/bpf_jit.S b/arch/x86/net/bpf_jit.S index 6440221..f746de8 100644 --- a/arch/x86/net/bpf_jit.S +++ b/arch/x86/net/bpf_jit.S @@ -9,19 +9,17 @@ */ #include #include +#include /* * Calling convention : - * rbx : skb pointer (callee saved) + * rdi : skb pointer * esi : offset of byte(s) to fetch in skb (can be scratched) - * r10 : copy of skb->data + * r8 : copy of skb->data * r9d : hlen = skb->len - skb->data_len */ -#define SKBDATA %r10 +#define SKBDATA %r8 #define SKF_MAX_NEG_OFF $(-0x200000) /* SKF_LL_OFF from filter.h */ -#define MAX_BPF_STACK (512 /* from filter.h */ + \ - 32 /* space for rbx,r13,r14,r15 */ + \ - 8 /* space for skb_copy_bits */) sk_load_word: .globl sk_load_word @@ -38,6 +36,7 @@ sk_load_word_positive_offset: jle bpf_slow_path_word mov (SKBDATA,%rsi),%eax bswap %eax /* ntohl() */ + pax_force_retaddr ret sk_load_half: @@ -55,6 +54,7 @@ sk_load_half_positive_offset: jle bpf_slow_path_half movzwl (SKBDATA,%rsi),%eax rol $8,%ax # ntohs() + pax_force_retaddr ret sk_load_byte: @@ -69,45 +69,83 @@ sk_load_byte_positive_offset: cmp %esi,%r9d /* if (offset >= hlen) goto bpf_slow_path_byte */ jle bpf_slow_path_byte movzbl (SKBDATA,%rsi),%eax + pax_force_retaddr + ret + +/** + * sk_load_byte_msh - BPF_S_LDX_B_MSH helper + * + * Implements BPF_S_LDX_B_MSH : ldxb 4*([offset]&0xf) + * Must preserve A accumulator (%eax) + * Inputs : %esi is the offset value + */ +sk_load_byte_msh: + .globl sk_load_byte_msh + test %esi,%esi + js bpf_slow_path_byte_msh_neg + +sk_load_byte_msh_positive_offset: + .globl sk_load_byte_msh_positive_offset + cmp %esi,%r9d /* if (offset >= hlen) goto bpf_slow_path_byte_msh */ + jle bpf_slow_path_byte_msh + movzbl (SKBDATA,%rsi),%ebx + and $15,%bl + shl $2,%bl + pax_force_retaddr ret /* rsi contains offset and can be scratched */ #define bpf_slow_path_common(LEN) \ - mov %rbx, %rdi; /* arg1 == skb */ \ + push %rdi; /* save skb */ \ push %r9; \ push SKBDATA; \ /* rsi already has offset */ \ mov $LEN,%ecx; /* len */ \ - lea - MAX_BPF_STACK + 32(%rbp),%rdx; \ + lea -12(%rbp),%rdx; \ call skb_copy_bits; \ test %eax,%eax; \ pop SKBDATA; \ - pop %r9; + pop %r9; \ + pop %rdi bpf_slow_path_word: bpf_slow_path_common(4) js bpf_error - mov - MAX_BPF_STACK + 32(%rbp),%eax + mov -12(%rbp),%eax bswap %eax + pax_force_retaddr ret bpf_slow_path_half: bpf_slow_path_common(2) js bpf_error - mov - MAX_BPF_STACK + 32(%rbp),%ax + 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 - MAX_BPF_STACK + 32(%rbp),%eax + movzbl -12(%rbp),%eax + pax_force_retaddr + ret + +bpf_slow_path_byte_msh: + xchg %eax,%ebx /* dont lose A , X is about to be scratched */ + bpf_slow_path_common(1) + js bpf_error + movzbl -12(%rbp),%eax + and $15,%al + shl $2,%al + xchg %eax,%ebx + pax_force_retaddr ret #define sk_negative_common(SIZE) \ - mov %rbx, %rdi; /* arg1 == skb */ \ + push %rdi; /* save skb */ \ push %r9; \ push SKBDATA; \ /* rsi already has offset */ \ @@ -116,8 +154,10 @@ bpf_slow_path_byte: test %rax,%rax; \ pop SKBDATA; \ pop %r9; \ + pop %rdi; \ jz bpf_error + bpf_slow_path_word_neg: cmp SKF_MAX_NEG_OFF, %esi /* test range */ jl bpf_error /* offset lower -> error */ @@ -126,6 +166,7 @@ sk_load_word_negative_offset: sk_negative_common(4) mov (%rax), %eax bswap %eax + pax_force_retaddr ret bpf_slow_path_half_neg: @@ -137,6 +178,7 @@ sk_load_half_negative_offset: mov (%rax),%ax rol $8,%ax movzwl %ax,%eax + pax_force_retaddr ret bpf_slow_path_byte_neg: @@ -146,14 +188,27 @@ sk_load_byte_negative_offset: .globl sk_load_byte_negative_offset sk_negative_common(1) movzbl (%rax), %eax + pax_force_retaddr + ret + +bpf_slow_path_byte_msh_neg: + cmp SKF_MAX_NEG_OFF, %esi + jl bpf_error +sk_load_byte_msh_negative_offset: + .globl sk_load_byte_msh_negative_offset + xchg %eax,%ebx /* dont lose A , X is about to be scratched */ + sk_negative_common(1) + movzbl (%rax),%eax + and $15,%al + shl $2,%al + xchg %eax,%ebx + pax_force_retaddr ret bpf_error: # force a return 0 from jit handler - xor %eax,%eax - mov - MAX_BPF_STACK(%rbp),%rbx - mov - MAX_BPF_STACK + 8(%rbp),%r13 - mov - MAX_BPF_STACK + 16(%rbp),%r14 - mov - MAX_BPF_STACK + 24(%rbp),%r15 + xor %eax,%eax + mov -8(%rbp),%rbx leaveq + pax_force_retaddr ret diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index 99bef86..bdfb5c6 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -1,7 +1,6 @@ /* bpf_jit_comp.c : BPF JIT compiler * * Copyright (C) 2011-2013 Eric Dumazet (eric.dumazet@gmail.com) - * Internal BPF Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -15,16 +14,28 @@ #include #include +/* + * Conventions : + * EAX : BPF A accumulator + * EBX : BPF X accumulator + * RDI : pointer to skb (first argument given to JIT function) + * RBP : frame pointer (even if CONFIG_FRAME_POINTER=n) + * ECX,EDX,ESI : scratch registers + * r9d : skb->len - skb->data_len (headlen) + * r8 : skb->data + * -8(RBP) : saved RBX value + * -16(RBP)..-80(RBP) : BPF_MEMWORDS values + */ int bpf_jit_enable __read_mostly; /* * assembly code in arch/x86/net/bpf_jit.S */ -extern u8 sk_load_word[], sk_load_half[], sk_load_byte[]; +extern u8 sk_load_word[], sk_load_half[], sk_load_byte[], sk_load_byte_msh[]; extern u8 sk_load_word_positive_offset[], sk_load_half_positive_offset[]; -extern u8 sk_load_byte_positive_offset[]; +extern u8 sk_load_byte_positive_offset[], sk_load_byte_msh_positive_offset[]; extern u8 sk_load_word_negative_offset[], sk_load_half_negative_offset[]; -extern u8 sk_load_byte_negative_offset[]; +extern u8 sk_load_byte_negative_offset[], sk_load_byte_msh_negative_offset[]; static inline u8 *emit_code(u8 *ptr, u32 bytes, unsigned int len) { @@ -39,50 +50,113 @@ static inline u8 *emit_code(u8 *ptr, u32 bytes, unsigned int len) return ptr + len; } +#ifdef CONFIG_GRKERNSEC_JIT_HARDEN +#define MAX_INSTR_CODE_SIZE 96 +#else +#define MAX_INSTR_CODE_SIZE 64 +#endif + #define EMIT(bytes, len) do { prog = emit_code(prog, bytes, len); } while (0) #define EMIT1(b1) EMIT(b1, 1) #define EMIT2(b1, b2) EMIT((b1) + ((b2) << 8), 2) #define EMIT3(b1, b2, b3) EMIT((b1) + ((b2) << 8) + ((b3) << 16), 3) #define EMIT4(b1, b2, b3, b4) EMIT((b1) + ((b2) << 8) + ((b3) << 16) + ((b4) << 24), 4) -#define EMIT1_off32(b1, off) \ - do {EMIT1(b1); EMIT(off, 4); } while (0) -#define EMIT2_off32(b1, b2, off) \ - do {EMIT2(b1, b2); EMIT(off, 4); } while (0) -#define EMIT3_off32(b1, b2, b3, off) \ - do {EMIT3(b1, b2, b3); EMIT(off, 4); } while (0) -#define EMIT4_off32(b1, b2, b3, b4, off) \ - do {EMIT4(b1, b2, b3, b4); EMIT(off, 4); } while (0) + +#ifdef CONFIG_GRKERNSEC_JIT_HARDEN +/* original constant will appear in ecx */ +#define DILUTE_CONST_SEQUENCE(_off, _key) \ +do { \ + /* mov ecx, randkey */ \ + EMIT1(0xb9); \ + EMIT(_key, 4); \ + /* xor ecx, randkey ^ off */ \ + EMIT2(0x81, 0xf1); \ + EMIT((_key) ^ (_off), 4); \ +} while (0) + +#define EMIT1_off32(b1, _off) \ +do { \ + switch (b1) { \ + case 0x05: /* add eax, imm32 */ \ + case 0x2d: /* sub eax, imm32 */ \ + case 0x25: /* and eax, imm32 */ \ + case 0x0d: /* or eax, imm32 */ \ + case 0xb8: /* mov eax, imm32 */ \ + case 0x35: /* xor eax, imm32 */ \ + case 0x3d: /* cmp eax, imm32 */ \ + case 0xa9: /* test eax, imm32 */ \ + DILUTE_CONST_SEQUENCE(_off, randkey); \ + EMIT2((b1) - 4, 0xc8); /* convert imm instruction to eax, ecx */\ + break; \ + case 0xbb: /* mov ebx, imm32 */ \ + DILUTE_CONST_SEQUENCE(_off, randkey); \ + /* mov ebx, ecx */ \ + EMIT2(0x89, 0xcb); \ + break; \ + case 0xbe: /* mov esi, imm32 */ \ + DILUTE_CONST_SEQUENCE(_off, randkey); \ + /* mov esi, ecx */ \ + EMIT2(0x89, 0xce); \ + break; \ + case 0xe8: /* call rel imm32, always to known funcs */ \ + EMIT1(b1); \ + EMIT(_off, 4); \ + break; \ + case 0xe9: /* jmp rel imm32 */ \ + EMIT1(b1); \ + EMIT(_off, 4); \ + /* prevent fall-through, we're not called if off = 0 */ \ + EMIT(0xcccccccc, 4); \ + EMIT(0xcccccccc, 4); \ + break; \ + default: \ + BUILD_BUG(); \ + } \ +} while (0) + +#define EMIT2_off32(b1, b2, _off) \ +do { \ + if ((b1) == 0x8d && (b2) == 0xb3) { /* lea esi, [rbx+imm32] */ \ + EMIT2(0x8d, 0xb3); /* lea esi, [rbx+randkey] */ \ + EMIT(randkey, 4); \ + EMIT2(0x8d, 0xb6); /* lea esi, [esi+off-randkey] */ \ + EMIT((_off) - randkey, 4); \ + } else if ((b1) == 0x69 && (b2) == 0xc0) { /* imul eax, imm32 */\ + DILUTE_CONST_SEQUENCE(_off, randkey); \ + /* imul eax, ecx */ \ + EMIT3(0x0f, 0xaf, 0xc1); \ + } else { \ + BUILD_BUG(); \ + } \ +} while (0) +#else +#define EMIT1_off32(b1, off) do { EMIT1(b1); EMIT(off, 4);} while (0) +#define EMIT2_off32(b1, b2, off) do { EMIT2(b1, b2); EMIT(off, 4);} while (0) +#endif + +#define CLEAR_A() EMIT2(0x31, 0xc0) /* xor %eax,%eax */ +#define CLEAR_X() EMIT2(0x31, 0xdb) /* xor %ebx,%ebx */ static inline bool is_imm8(int value) { return value <= 127 && value >= -128; } -static inline bool is_simm32(s64 value) +static inline bool is_near(int offset) { - return value == (s64) (s32) value; + return offset <= 127 && offset >= -128; } -/* mov dst, src */ -#define EMIT_mov(DST, SRC) \ - do {if (DST != SRC) \ - EMIT3(add_2mod(0x48, DST, SRC), 0x89, add_2reg(0xC0, DST, SRC)); \ - } while (0) - -static int bpf_size_to_x86_bytes(int bpf_size) -{ - if (bpf_size == BPF_W) - return 4; - else if (bpf_size == BPF_H) - return 2; - else if (bpf_size == BPF_B) - return 1; - else if (bpf_size == BPF_DW) - return 4; /* imm32 */ - else - return 0; -} +#define EMIT_JMP(offset) \ +do { \ + if (offset) { \ + if (is_near(offset)) \ + EMIT2(0xeb, offset); /* jmp .+off8 */ \ + else \ + EMIT1_off32(0xe9, offset); /* jmp .+off32 */ \ + } \ +} while (0) /* list of x86 cond jumps opcodes (. + s8) * Add 0x10 (and an extra 0x0f) to generate far jumps (. + s32) @@ -93,8 +167,46 @@ static int bpf_size_to_x86_bytes(int bpf_size) #define X86_JNE 0x75 #define X86_JBE 0x76 #define X86_JA 0x77 -#define X86_JGE 0x7D -#define X86_JG 0x7F + +#ifdef CONFIG_GRKERNSEC_JIT_HARDEN +#define APPEND_FLOW_VERIFY() \ +do { \ + /* mov ecx, randkey */ \ + EMIT1(0xb9); \ + EMIT(randkey, 4); \ + /* cmp ecx, randkey */ \ + EMIT2(0x81, 0xf9); \ + EMIT(randkey, 4); \ + /* jz after 8 int 3s */ \ + EMIT2(0x74, 0x08); \ + EMIT(0xcccccccc, 4); \ + EMIT(0xcccccccc, 4); \ +} while (0) +#else +#define APPEND_FLOW_VERIFY() do { } while (0) +#endif + +#define EMIT_COND_JMP(op, offset) \ +do { \ + if (is_near(offset)) \ + EMIT2(op, offset); /* jxx .+off8 */ \ + else { \ + EMIT2(0x0f, op + 0x10); \ + EMIT(offset, 4); /* jxx .+off32 */ \ + APPEND_FLOW_VERIFY(); \ + } \ +} while (0) + +#define COND_SEL(CODE, TOP, FOP) \ + case CODE: \ + t_op = TOP; \ + f_op = FOP; \ + goto cond_branch + + +#define SEEN_DATAREF 1 /* might call external helpers */ +#define SEEN_XREG 2 /* ebx is used */ +#define SEEN_MEM 4 /* use mem[] for temporary storage */ static inline void bpf_flush_icache(void *start, void *end) { @@ -109,804 +221,646 @@ static inline void bpf_flush_icache(void *start, void *end) #define CHOOSE_LOAD_FUNC(K, func) \ ((int)K < 0 ? ((int)K >= SKF_LL_OFF ? func##_negative_offset : func) : func##_positive_offset) -struct bpf_binary_header { - unsigned int pages; - /* Note : for security reasons, bpf code will follow a randomly - * sized amount of int3 instructions - */ - u8 image[]; -}; +/* Helper to find the offset of pkt_type in sk_buff + * We want to make sure its still a 3bit field starting at a byte boundary. + */ +#define PKT_TYPE_MAX 7 +static int pkt_type_offset(void) +{ + struct sk_buff skb_probe = { + .pkt_type = ~0, + }; + char *ct = (char *)&skb_probe; + unsigned int off; -static struct bpf_binary_header *bpf_alloc_binary(unsigned int proglen, + for (off = 0; off < sizeof(struct sk_buff); off++) { + if (ct[off] == PKT_TYPE_MAX) + return off; + } + pr_err_once("Please fix pkt_type_offset(), as pkt_type couldn't be found\n"); + return -1; +} + +/* Note : for security reasons, bpf code will follow a randomly + * sized amount of int3 instructions + */ +static u8 *bpf_alloc_binary(unsigned int proglen, u8 **image_ptr) { unsigned int sz, hole; - struct bpf_binary_header *header; + u8 *header; /* Most of BPF filters are really small, * but if some of them fill a page, allow at least * 128 extra bytes to insert a random section of int3 */ - sz = round_up(proglen + sizeof(*header) + 128, PAGE_SIZE); - header = module_alloc(sz); + sz = round_up(proglen + 128, PAGE_SIZE); + header = module_alloc_exec(sz); if (!header) return NULL; + pax_open_kernel(); memset(header, 0xcc, sz); /* fill whole space with int3 instructions */ + pax_close_kernel(); - header->pages = sz / PAGE_SIZE; - hole = min(sz - (proglen + sizeof(*header)), PAGE_SIZE - sizeof(*header)); + hole = PAGE_SIZE - (proglen & ~PAGE_MASK); /* insert a random number of int3 instructions before BPF code */ - *image_ptr = &header->image[prandom_u32() % hole]; + *image_ptr = &header[prandom_u32() % hole]; return header; } -/* pick a register outside of BPF range for JIT internal work */ -#define AUX_REG (MAX_BPF_REG + 1) - -/* the following table maps BPF registers to x64 registers. - * x64 register r12 is unused, since if used as base address register - * in load/store instructions, it always needs an extra byte of encoding - */ -static const int reg2hex[] = { - [BPF_REG_0] = 0, /* rax */ - [BPF_REG_1] = 7, /* rdi */ - [BPF_REG_2] = 6, /* rsi */ - [BPF_REG_3] = 2, /* rdx */ - [BPF_REG_4] = 1, /* rcx */ - [BPF_REG_5] = 0, /* r8 */ - [BPF_REG_6] = 3, /* rbx callee saved */ - [BPF_REG_7] = 5, /* r13 callee saved */ - [BPF_REG_8] = 6, /* r14 callee saved */ - [BPF_REG_9] = 7, /* r15 callee saved */ - [BPF_REG_FP] = 5, /* rbp readonly */ - [AUX_REG] = 3, /* r11 temp register */ -}; - -/* is_ereg() == true if BPF register 'reg' maps to x64 r8..r15 - * which need extra byte of encoding. - * rax,rcx,...,rbp have simpler encoding - */ -static inline bool is_ereg(u32 reg) -{ - if (reg == BPF_REG_5 || reg == AUX_REG || - (reg >= BPF_REG_7 && reg <= BPF_REG_9)) - return true; - else - return false; -} - -/* add modifiers if 'reg' maps to x64 registers r8..r15 */ -static inline u8 add_1mod(u8 byte, u32 reg) -{ - if (is_ereg(reg)) - byte |= 1; - return byte; -} - -static inline u8 add_2mod(u8 byte, u32 r1, u32 r2) -{ - if (is_ereg(r1)) - byte |= 1; - if (is_ereg(r2)) - byte |= 4; - return byte; -} - -/* encode 'dst_reg' register into x64 opcode 'byte' */ -static inline u8 add_1reg(u8 byte, u32 dst_reg) -{ - return byte + reg2hex[dst_reg]; -} - -/* encode 'dst_reg' and 'src_reg' registers into x64 opcode 'byte' */ -static inline u8 add_2reg(u8 byte, u32 dst_reg, u32 src_reg) -{ - return byte + reg2hex[dst_reg] + (reg2hex[src_reg] << 3); -} - -struct jit_context { - unsigned int cleanup_addr; /* epilogue code offset */ - bool seen_ld_abs; -}; - -static int do_jit(struct sk_filter *bpf_prog, int *addrs, u8 *image, - int oldproglen, struct jit_context *ctx) -{ - struct sock_filter_int *insn = bpf_prog->insnsi; - int insn_cnt = bpf_prog->len; - u8 temp[64]; - int i; - int proglen = 0; - u8 *prog = temp; - int stacksize = MAX_BPF_STACK + - 32 /* space for rbx, r13, r14, r15 */ + - 8 /* space for skb_copy_bits() buffer */; - - EMIT1(0x55); /* push rbp */ - EMIT3(0x48, 0x89, 0xE5); /* mov rbp,rsp */ - - /* sub rsp, stacksize */ - EMIT3_off32(0x48, 0x81, 0xEC, stacksize); - - /* all classic BPF filters use R6(rbx) save it */ - - /* mov qword ptr [rbp-X],rbx */ - EMIT3_off32(0x48, 0x89, 0x9D, -stacksize); - - /* sk_convert_filter() maps classic BPF register X to R7 and uses R8 - * as temporary, so all tcpdump filters need to spill/fill R7(r13) and - * R8(r14). R9(r15) spill could be made conditional, but there is only - * one 'bpf_error' return path out of helper functions inside bpf_jit.S - * The overhead of extra spill is negligible for any filter other - * than synthetic ones. Therefore not worth adding complexity. - */ - - /* mov qword ptr [rbp-X],r13 */ - EMIT3_off32(0x4C, 0x89, 0xAD, -stacksize + 8); - /* mov qword ptr [rbp-X],r14 */ - EMIT3_off32(0x4C, 0x89, 0xB5, -stacksize + 16); - /* mov qword ptr [rbp-X],r15 */ - EMIT3_off32(0x4C, 0x89, 0xBD, -stacksize + 24); - - /* clear A and X registers */ - EMIT2(0x31, 0xc0); /* xor eax, eax */ - EMIT3(0x4D, 0x31, 0xED); /* xor r13, r13 */ - - if (ctx->seen_ld_abs) { - /* r9d : skb->len - skb->data_len (headlen) - * r10 : skb->data - */ - if (is_imm8(offsetof(struct sk_buff, len))) - /* mov %r9d, off8(%rdi) */ - EMIT4(0x44, 0x8b, 0x4f, - offsetof(struct sk_buff, len)); - else - /* mov %r9d, off32(%rdi) */ - EMIT3_off32(0x44, 0x8b, 0x8f, - offsetof(struct sk_buff, len)); - - if (is_imm8(offsetof(struct sk_buff, data_len))) - /* sub %r9d, off8(%rdi) */ - EMIT4(0x44, 0x2b, 0x4f, - offsetof(struct sk_buff, data_len)); - else - EMIT3_off32(0x44, 0x2b, 0x8f, - offsetof(struct sk_buff, data_len)); - - if (is_imm8(offsetof(struct sk_buff, data))) - /* mov %r10, off8(%rdi) */ - EMIT4(0x4c, 0x8b, 0x57, - offsetof(struct sk_buff, data)); - else - /* mov %r10, off32(%rdi) */ - EMIT3_off32(0x4c, 0x8b, 0x97, - offsetof(struct sk_buff, data)); - } - - for (i = 0; i < insn_cnt; i++, insn++) { - const s32 imm32 = insn->imm; - u32 dst_reg = insn->dst_reg; - u32 src_reg = insn->src_reg; - u8 b1 = 0, b2 = 0, b3 = 0; - s64 jmp_offset; - u8 jmp_cond; - int ilen; - u8 *func; - - switch (insn->code) { - /* ALU */ - case BPF_ALU | BPF_ADD | BPF_X: - case BPF_ALU | BPF_SUB | BPF_X: - case BPF_ALU | BPF_AND | BPF_X: - case BPF_ALU | BPF_OR | BPF_X: - case BPF_ALU | BPF_XOR | BPF_X: - case BPF_ALU64 | BPF_ADD | BPF_X: - case BPF_ALU64 | BPF_SUB | BPF_X: - case BPF_ALU64 | BPF_AND | BPF_X: - case BPF_ALU64 | BPF_OR | BPF_X: - case BPF_ALU64 | BPF_XOR | BPF_X: - switch (BPF_OP(insn->code)) { - case BPF_ADD: b2 = 0x01; break; - case BPF_SUB: b2 = 0x29; break; - case BPF_AND: b2 = 0x21; break; - case BPF_OR: b2 = 0x09; break; - case BPF_XOR: b2 = 0x31; break; - } - if (BPF_CLASS(insn->code) == BPF_ALU64) - EMIT1(add_2mod(0x48, dst_reg, src_reg)); - else if (is_ereg(dst_reg) || is_ereg(src_reg)) - EMIT1(add_2mod(0x40, dst_reg, src_reg)); - EMIT2(b2, add_2reg(0xC0, dst_reg, src_reg)); - break; - - /* mov dst, src */ - case BPF_ALU64 | BPF_MOV | BPF_X: - EMIT_mov(dst_reg, src_reg); - break; - - /* mov32 dst, src */ - case BPF_ALU | BPF_MOV | BPF_X: - if (is_ereg(dst_reg) || is_ereg(src_reg)) - EMIT1(add_2mod(0x40, dst_reg, src_reg)); - EMIT2(0x89, add_2reg(0xC0, dst_reg, src_reg)); - break; - - /* neg dst */ - case BPF_ALU | BPF_NEG: - case BPF_ALU64 | BPF_NEG: - if (BPF_CLASS(insn->code) == BPF_ALU64) - EMIT1(add_1mod(0x48, dst_reg)); - else if (is_ereg(dst_reg)) - EMIT1(add_1mod(0x40, dst_reg)); - EMIT2(0xF7, add_1reg(0xD8, dst_reg)); - break; - - case BPF_ALU | BPF_ADD | BPF_K: - case BPF_ALU | BPF_SUB | BPF_K: - case BPF_ALU | BPF_AND | BPF_K: - case BPF_ALU | BPF_OR | BPF_K: - case BPF_ALU | BPF_XOR | BPF_K: - case BPF_ALU64 | BPF_ADD | BPF_K: - case BPF_ALU64 | BPF_SUB | BPF_K: - case BPF_ALU64 | BPF_AND | BPF_K: - case BPF_ALU64 | BPF_OR | BPF_K: - case BPF_ALU64 | BPF_XOR | BPF_K: - if (