diff -urNp linux-2.6.28.8/arch/alpha/include/asm/elf.h linux-2.6.28.8/arch/alpha/include/asm/elf.h --- linux-2.6.28.8/arch/alpha/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/alpha/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -91,6 +91,13 @@ typedef elf_fpreg_t elf_fpregset_t[ELF_N #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 -urNp linux-2.6.28.8/arch/alpha/include/asm/kmap_types.h linux-2.6.28.8/arch/alpha/include/asm/kmap_types.h --- linux-2.6.28.8/arch/alpha/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/alpha/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -24,7 +24,8 @@ D(9) KM_IRQ0, D(10) KM_IRQ1, D(11) KM_SOFTIRQ0, D(12) KM_SOFTIRQ1, -D(13) KM_TYPE_NR +D(13) KM_CLEARPAGE, +D(14) KM_TYPE_NR }; #undef D diff -urNp linux-2.6.28.8/arch/alpha/include/asm/pgtable.h linux-2.6.28.8/arch/alpha/include/asm/pgtable.h --- linux-2.6.28.8/arch/alpha/include/asm/pgtable.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/alpha/include/asm/pgtable.h 2009-02-21 09:37:48.000000000 -0500 @@ -101,6 +101,17 @@ struct vm_area_struct; #define PAGE_SHARED __pgprot(_PAGE_VALID | __ACCESS_BITS) #define PAGE_COPY __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW) #define PAGE_READONLY __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW) + +#ifdef CONFIG_PAX_PAGEEXEC +# define PAGE_SHARED_NOEXEC __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOE) +# define PAGE_COPY_NOEXEC __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW | _PAGE_FOE) +# define PAGE_READONLY_NOEXEC __pgprot(_PAGE_VALID | __ACCESS_BITS | _PAGE_FOW | _PAGE_FOE) +#else +# define PAGE_SHARED_NOEXEC PAGE_SHARED +# define PAGE_COPY_NOEXEC PAGE_COPY +# define PAGE_READONLY_NOEXEC PAGE_READONLY +#endif + #define PAGE_KERNEL __pgprot(_PAGE_VALID | _PAGE_ASM | _PAGE_KRE | _PAGE_KWE) #define _PAGE_NORMAL(x) __pgprot(_PAGE_VALID | __ACCESS_BITS | (x)) diff -urNp linux-2.6.28.8/arch/alpha/kernel/module.c linux-2.6.28.8/arch/alpha/kernel/module.c --- linux-2.6.28.8/arch/alpha/kernel/module.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/alpha/kernel/module.c 2009-02-21 09:37:48.000000000 -0500 @@ -182,7 +182,7 @@ apply_relocate_add(Elf64_Shdr *sechdrs, /* 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 -urNp linux-2.6.28.8/arch/alpha/kernel/osf_sys.c linux-2.6.28.8/arch/alpha/kernel/osf_sys.c --- linux-2.6.28.8/arch/alpha/kernel/osf_sys.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/alpha/kernel/osf_sys.c 2009-02-21 09:37:48.000000000 -0500 @@ -1230,6 +1230,10 @@ arch_get_unmapped_area(struct file *filp merely specific addresses, but regions of memory -- perhaps this feature should be incorporated into all ports? */ +#ifdef CONFIG_PAX_RANDMMAP + if (!(current->mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { addr = arch_get_unmapped_area_1 (PAGE_ALIGN(addr), len, limit); if (addr != (unsigned long) -ENOMEM) @@ -1237,8 +1241,8 @@ arch_get_unmapped_area(struct file *filp } /* Next, try allocating at TASK_UNMAPPED_BASE. */ - addr = arch_get_unmapped_area_1 (PAGE_ALIGN(TASK_UNMAPPED_BASE), - len, limit); + addr = arch_get_unmapped_area_1 (PAGE_ALIGN(current->mm->mmap_base), len, limit); + if (addr != (unsigned long) -ENOMEM) return addr; diff -urNp linux-2.6.28.8/arch/alpha/kernel/ptrace.c linux-2.6.28.8/arch/alpha/kernel/ptrace.c --- linux-2.6.28.8/arch/alpha/kernel/ptrace.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/alpha/kernel/ptrace.c 2009-02-21 09:37:48.000000000 -0500 @@ -266,6 +266,9 @@ long arch_ptrace(struct task_struct *chi size_t copied; long ret; + if (gr_handle_ptrace(child, request)) + return -EPERM; + switch (request) { /* When I and D space are separate, these will need to be fixed. */ case PTRACE_PEEKTEXT: /* read word at location addr. */ diff -urNp linux-2.6.28.8/arch/alpha/mm/fault.c linux-2.6.28.8/arch/alpha/mm/fault.c --- linux-2.6.28.8/arch/alpha/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/alpha/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -54,6 +54,124 @@ __load_new_mm_context(struct mm_struct * __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(void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 5; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif /* * This routine handles page faults. It determines the address, @@ -131,8 +249,29 @@ do_page_fault(unsigned long address, uns 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 -urNp linux-2.6.28.8/arch/arm/include/asm/elf.h linux-2.6.28.8/arch/arm/include/asm/elf.h --- linux-2.6.28.8/arch/arm/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/arm/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -99,7 +99,14 @@ extern int arm_elf_read_implies_exec(con 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 diff -urNp linux-2.6.28.8/arch/arm/include/asm/kmap_types.h linux-2.6.28.8/arch/arm/include/asm/kmap_types.h --- linux-2.6.28.8/arch/arm/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/arm/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -18,6 +18,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/arm/mm/mmap.c linux-2.6.28.8/arch/arm/mm/mmap.c --- linux-2.6.28.8/arch/arm/mm/mmap.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/arm/mm/mmap.c 2009-02-21 09:37:48.000000000 -0500 @@ -62,6 +62,10 @@ arch_get_unmapped_area(struct file *filp 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); @@ -74,10 +78,10 @@ arch_get_unmapped_area(struct file *filp return addr; } if (len > mm->cached_hole_size) { - start_addr = addr = mm->free_area_cache; + start_addr = addr = mm->free_area_cache; } else { - start_addr = addr = TASK_UNMAPPED_BASE; - mm->cached_hole_size = 0; + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; } full_search: @@ -93,8 +97,8 @@ full_search: * Start a new search - just in case we missed * some holes. */ - if (start_addr != TASK_UNMAPPED_BASE) { - start_addr = addr = TASK_UNMAPPED_BASE; + if (start_addr != mm->mmap_base) { + start_addr = addr = mm->mmap_base; mm->cached_hole_size = 0; goto full_search; } diff -urNp linux-2.6.28.8/arch/avr32/include/asm/elf.h linux-2.6.28.8/arch/avr32/include/asm/elf.h --- linux-2.6.28.8/arch/avr32/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/avr32/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -85,8 +85,14 @@ typedef struct user_fpu_struct elf_fpreg 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 -urNp linux-2.6.28.8/arch/avr32/include/asm/kmap_types.h linux-2.6.28.8/arch/avr32/include/asm/kmap_types.h --- linux-2.6.28.8/arch/avr32/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/avr32/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -22,7 +22,8 @@ D(10) KM_IRQ0, D(11) KM_IRQ1, D(12) KM_SOFTIRQ0, D(13) KM_SOFTIRQ1, -D(14) KM_TYPE_NR +D(14) KM_CLEARPAGE, +D(15) KM_TYPE_NR }; #undef D diff -urNp linux-2.6.28.8/arch/avr32/mm/fault.c linux-2.6.28.8/arch/avr32/mm/fault.c --- linux-2.6.28.8/arch/avr32/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/avr32/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -41,6 +41,23 @@ static inline int notify_page_fault(stru int exception_trace = 1; +#ifdef CONFIG_PAX_PAGEEXEC +void pax_report_insns(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. @@ -157,6 +174,16 @@ 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 -urNp linux-2.6.28.8/arch/blackfin/include/asm/kmap_types.h linux-2.6.28.8/arch/blackfin/include/asm/kmap_types.h --- linux-2.6.28.8/arch/blackfin/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/blackfin/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -15,6 +15,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/cris/include/asm/kmap_types.h linux-2.6.28.8/arch/cris/include/asm/kmap_types.h --- linux-2.6.28.8/arch/cris/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/cris/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -19,6 +19,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/h8300/include/asm/kmap_types.h linux-2.6.28.8/arch/h8300/include/asm/kmap_types.h --- linux-2.6.28.8/arch/h8300/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/h8300/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -15,6 +15,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/ia64/ia32/binfmt_elf32.c linux-2.6.28.8/arch/ia64/ia32/binfmt_elf32.c --- linux-2.6.28.8/arch/ia64/ia32/binfmt_elf32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/ia32/binfmt_elf32.c 2009-02-21 09:37:48.000000000 -0500 @@ -45,6 +45,13 @@ randomize_stack_top(unsigned long stack_ #define elf_read_implies_exec(ex, have_pt_gnu_stack) (!(have_pt_gnu_stack)) +#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 + /* Ugly but avoids duplication */ #include "../../../fs/binfmt_elf.c" diff -urNp linux-2.6.28.8/arch/ia64/ia32/ia32priv.h linux-2.6.28.8/arch/ia64/ia32/ia32priv.h --- linux-2.6.28.8/arch/ia64/ia32/ia32priv.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/ia32/ia32priv.h 2009-02-21 09:37:48.000000000 -0500 @@ -296,7 +296,14 @@ typedef struct compat_siginfo { #define ELF_DATA ELFDATA2LSB #define ELF_ARCH EM_386 -#define IA32_STACK_TOP IA32_PAGE_OFFSET +#ifdef CONFIG_PAX_RANDUSTACK +#define __IA32_DELTA_STACK (current->mm->delta_stack) +#else +#define __IA32_DELTA_STACK 0UL +#endif + +#define IA32_STACK_TOP (IA32_PAGE_OFFSET - __IA32_DELTA_STACK) + #define IA32_GATE_OFFSET IA32_PAGE_OFFSET #define IA32_GATE_END IA32_PAGE_OFFSET + PAGE_SIZE diff -urNp linux-2.6.28.8/arch/ia64/include/asm/elf.h linux-2.6.28.8/arch/ia64/include/asm/elf.h --- linux-2.6.28.8/arch/ia64/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -43,6 +43,13 @@ */ #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 -urNp linux-2.6.28.8/arch/ia64/include/asm/kmap_types.h linux-2.6.28.8/arch/ia64/include/asm/kmap_types.h --- linux-2.6.28.8/arch/ia64/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -22,7 +22,8 @@ D(9) KM_IRQ0, D(10) KM_IRQ1, D(11) KM_SOFTIRQ0, D(12) KM_SOFTIRQ1, -D(13) KM_TYPE_NR +D(13) KM_CLEARPAGE, +D(14) KM_TYPE_NR }; #undef D diff -urNp linux-2.6.28.8/arch/ia64/include/asm/pgtable.h linux-2.6.28.8/arch/ia64/include/asm/pgtable.h --- linux-2.6.28.8/arch/ia64/include/asm/pgtable.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/include/asm/pgtable.h 2009-02-21 09:37:48.000000000 -0500 @@ -143,6 +143,17 @@ #define PAGE_READONLY __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) #define PAGE_COPY __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) #define PAGE_COPY_EXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RX) + +#ifdef CONFIG_PAX_PAGEEXEC +# define PAGE_SHARED_NOEXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RW) +# define PAGE_READONLY_NOEXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) +# define PAGE_COPY_NOEXEC __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_R) +#else +# define PAGE_SHARED_NOEXEC PAGE_SHARED +# define PAGE_READONLY_NOEXEC PAGE_READONLY +# define PAGE_COPY_NOEXEC PAGE_COPY +#endif + #define PAGE_GATE __pgprot(__ACCESS_BITS | _PAGE_PL_0 | _PAGE_AR_X_RX) #define PAGE_KERNEL __pgprot(__DIRTY_BITS | _PAGE_PL_0 | _PAGE_AR_RWX) #define PAGE_KERNELRX __pgprot(__ACCESS_BITS | _PAGE_PL_0 | _PAGE_AR_RX) diff -urNp linux-2.6.28.8/arch/ia64/kernel/module.c linux-2.6.28.8/arch/ia64/kernel/module.c --- linux-2.6.28.8/arch/ia64/kernel/module.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/kernel/module.c 2009-02-21 09:37:48.000000000 -0500 @@ -312,8 +312,7 @@ module_alloc (unsigned long size) 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; } @@ -491,15 +490,39 @@ module_frob_arch_sections (Elf_Ehdr *ehd } 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 @@ -683,7 +706,14 @@ do_reloc (struct module *mod, uint8_t r_ 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: @@ -817,15 +847,15 @@ apply_relocate_add (Elf64_Shdr *sechdrs, * 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 -urNp linux-2.6.28.8/arch/ia64/kernel/sys_ia64.c linux-2.6.28.8/arch/ia64/kernel/sys_ia64.c --- linux-2.6.28.8/arch/ia64/kernel/sys_ia64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/kernel/sys_ia64.c 2009-02-21 09:37:48.000000000 -0500 @@ -43,6 +43,13 @@ arch_get_unmapped_area (struct file *fil if (REGION_NUMBER(addr) == RGN_HPAGE) addr = 0; #endif + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + addr = mm->free_area_cache; + else +#endif + if (!addr) addr = mm->free_area_cache; @@ -61,9 +68,9 @@ arch_get_unmapped_area (struct file *fil for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { /* At this point: (!vma || addr < vma->vm_end). */ if (TASK_SIZE - len < addr || RGN_MAP_LIMIT - len < REGION_OFFSET(addr)) { - if (start_addr != TASK_UNMAPPED_BASE) { + if (start_addr != mm->mmap_base) { /* Start a new search --- just in case we missed some holes. */ - addr = TASK_UNMAPPED_BASE; + addr = mm->mmap_base; goto full_search; } return -ENOMEM; diff -urNp linux-2.6.28.8/arch/ia64/mm/fault.c linux-2.6.28.8/arch/ia64/mm/fault.c --- linux-2.6.28.8/arch/ia64/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -72,6 +72,23 @@ mapped_kernel_page_is_present (unsigned return pte_present(pte); } +#ifdef CONFIG_PAX_PAGEEXEC +void pax_report_insns(void *pc, void *sp) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 8; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + void __kprobes ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *regs) { @@ -145,9 +162,23 @@ ia64_do_page_fault (unsigned long addres mask = ( (((isr >> IA64_ISR_X_BIT) & 1UL) << VM_EXEC_BIT) | (((isr >> IA64_ISR_W_BIT) & 1UL) << VM_WRITE_BIT)); - if ((vma->vm_flags & mask) != mask) + if ((vma->vm_flags & mask) != mask) { + +#ifdef CONFIG_PAX_PAGEEXEC + if (!(vma->vm_flags & VM_EXEC) && (mask & VM_EXEC)) { + if (!(mm->pax_flags & MF_PAX_PAGEEXEC) || address != regs->cr_iip) + goto bad_area; + + up_read(&mm->mmap_sem); + pax_report_fault(regs, (void *)regs->cr_iip, (void *)regs->r12); + do_group_exit(SIGKILL); + } +#endif + goto bad_area; + } + survive: /* * If for any reason at all we couldn't handle the fault, make diff -urNp linux-2.6.28.8/arch/ia64/mm/init.c linux-2.6.28.8/arch/ia64/mm/init.c --- linux-2.6.28.8/arch/ia64/mm/init.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/ia64/mm/init.c 2009-02-21 09:37:48.000000000 -0500 @@ -121,6 +121,19 @@ ia64_init_addr_space (void) vma->vm_start = current->thread.rbs_bot & PAGE_MASK; vma->vm_end = vma->vm_start + PAGE_SIZE; vma->vm_flags = VM_DATA_DEFAULT_FLAGS|VM_GROWSUP|VM_ACCOUNT; + +#ifdef CONFIG_PAX_PAGEEXEC + if (current->mm->pax_flags & MF_PAX_PAGEEXEC) { + vma->vm_flags &= ~VM_EXEC; + +#ifdef CONFIG_PAX_MPROTECT + if (current->mm->pax_flags & MF_PAX_MPROTECT) + vma->vm_flags &= ~VM_MAYEXEC; +#endif + + } +#endif + vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); down_write(¤t->mm->mmap_sem); if (insert_vm_struct(current->mm, vma)) { diff -urNp linux-2.6.28.8/arch/m68knommu/include/asm/kmap_types.h linux-2.6.28.8/arch/m68knommu/include/asm/kmap_types.h --- linux-2.6.28.8/arch/m68knommu/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/m68knommu/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -15,6 +15,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/mips/include/asm/elf.h linux-2.6.28.8/arch/mips/include/asm/elf.h --- linux-2.6.28.8/arch/mips/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -364,4 +364,11 @@ extern int dump_task_fpu(struct task_str #define ELF_ET_DYN_BASE (TASK_SIZE / 3 * 2) #endif +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (test_thread_flag(TIF_32BIT_ADDR) ? 0x00400000UL : 0x00400000UL) + +#define PAX_DELTA_MMAP_LEN (test_thread_flag(TIF_32BIT_ADDR) ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#define PAX_DELTA_STACK_LEN (test_thread_flag(TIF_32BIT_ADDR) ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#endif + #endif /* _ASM_ELF_H */ diff -urNp linux-2.6.28.8/arch/mips/include/asm/kmap_types.h linux-2.6.28.8/arch/mips/include/asm/kmap_types.h --- linux-2.6.28.8/arch/mips/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -22,7 +22,8 @@ D(9) KM_IRQ0, D(10) KM_IRQ1, D(11) KM_SOFTIRQ0, D(12) KM_SOFTIRQ1, -D(13) KM_TYPE_NR +D(13) KM_CLEARPAGE, +D(14) KM_TYPE_NR }; #undef D diff -urNp linux-2.6.28.8/arch/mips/include/asm/page.h linux-2.6.28.8/arch/mips/include/asm/page.h --- linux-2.6.28.8/arch/mips/include/asm/page.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/include/asm/page.h 2009-02-21 09:37:48.000000000 -0500 @@ -82,7 +82,7 @@ extern void copy_user_highpage(struct pa #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 -urNp linux-2.6.28.8/arch/mips/include/asm/system.h linux-2.6.28.8/arch/mips/include/asm/system.h --- linux-2.6.28.8/arch/mips/include/asm/system.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/include/asm/system.h 2009-02-21 09:37:48.000000000 -0500 @@ -217,6 +217,6 @@ extern void per_cpu_trap_init(void); */ #define __ARCH_WANT_UNLOCKED_CTXSW -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ALMASK) #endif /* _ASM_SYSTEM_H */ diff -urNp linux-2.6.28.8/arch/mips/kernel/binfmt_elfn32.c linux-2.6.28.8/arch/mips/kernel/binfmt_elfn32.c --- linux-2.6.28.8/arch/mips/kernel/binfmt_elfn32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/kernel/binfmt_elfn32.c 2009-02-21 09:37:48.000000000 -0500 @@ -50,6 +50,13 @@ typedef elf_fpreg_t elf_fpregset_t[ELF_N #undef ELF_ET_DYN_BASE #define ELF_ET_DYN_BASE (TASK32_SIZE / 3 * 2) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (test_thread_flag(TIF_32BIT_ADDR) ? 0x00400000UL : 0x00400000UL) + +#define PAX_DELTA_MMAP_LEN (test_thread_flag(TIF_32BIT_ADDR) ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#define PAX_DELTA_STACK_LEN (test_thread_flag(TIF_32BIT_ADDR) ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#endif + #include #include #include diff -urNp linux-2.6.28.8/arch/mips/kernel/binfmt_elfo32.c linux-2.6.28.8/arch/mips/kernel/binfmt_elfo32.c --- linux-2.6.28.8/arch/mips/kernel/binfmt_elfo32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/kernel/binfmt_elfo32.c 2009-02-21 09:37:48.000000000 -0500 @@ -52,6 +52,13 @@ typedef elf_fpreg_t elf_fpregset_t[ELF_N #undef ELF_ET_DYN_BASE #define ELF_ET_DYN_BASE (TASK32_SIZE / 3 * 2) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (test_thread_flag(TIF_32BIT_ADDR) ? 0x00400000UL : 0x00400000UL) + +#define PAX_DELTA_MMAP_LEN (test_thread_flag(TIF_32BIT_ADDR) ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#define PAX_DELTA_STACK_LEN (test_thread_flag(TIF_32BIT_ADDR) ? 27-PAGE_SHIFT : 36-PAGE_SHIFT) +#endif + #include #include #include diff -urNp linux-2.6.28.8/arch/mips/kernel/process.c linux-2.6.28.8/arch/mips/kernel/process.c --- linux-2.6.28.8/arch/mips/kernel/process.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/kernel/process.c 2009-02-21 09:37:48.000000000 -0500 @@ -457,15 +457,3 @@ unsigned long get_wchan(struct task_stru 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 -urNp linux-2.6.28.8/arch/mips/kernel/syscall.c linux-2.6.28.8/arch/mips/kernel/syscall.c --- linux-2.6.28.8/arch/mips/kernel/syscall.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/kernel/syscall.c 2009-02-21 09:37:48.000000000 -0500 @@ -99,6 +99,11 @@ unsigned long arch_get_unmapped_area(str do_color_align = 0; if (filp || (flags & MAP_SHARED)) do_color_align = 1; + +#ifdef CONFIG_PAX_RANDMMAP + if (!(current->mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_color_align) addr = COLOUR_ALIGN(addr, pgoff); @@ -109,7 +114,7 @@ unsigned long arch_get_unmapped_area(str (!vmm || addr + len <= vmm->vm_start)) return addr; } - addr = TASK_UNMAPPED_BASE; + addr = current->mm->mmap_base; if (do_color_align) addr = COLOUR_ALIGN(addr, pgoff); else diff -urNp linux-2.6.28.8/arch/mips/mm/fault.c linux-2.6.28.8/arch/mips/mm/fault.c --- linux-2.6.28.8/arch/mips/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/mips/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -26,6 +26,23 @@ #include #include /* For VMALLOC_END */ +#ifdef CONFIG_PAX_PAGEEXEC +void pax_report_insns(void *pc) +{ + unsigned long i; + + printk(KERN_ERR "PAX: bytes at PC: "); + for (i = 0; i < 5; i++) { + unsigned int c; + if (get_user(c, (unsigned int *)pc+i)) + printk(KERN_CONT "???????? "); + else + printk(KERN_CONT "%08x ", c); + } + printk("\n"); +} +#endif + /* * This routine handles page faults. It determines the address, * and the problem, and then passes it off to one of the appropriate diff -urNp linux-2.6.28.8/arch/parisc/include/asm/elf.h linux-2.6.28.8/arch/parisc/include/asm/elf.h --- linux-2.6.28.8/arch/parisc/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/parisc/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -333,6 +333,13 @@ 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 -urNp linux-2.6.28.8/arch/parisc/include/asm/kmap_types.h linux-2.6.28.8/arch/parisc/include/asm/kmap_types.h --- linux-2.6.28.8/arch/parisc/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/parisc/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -22,7 +22,8 @@ D(9) KM_IRQ0, D(10) KM_IRQ1, D(11) KM_SOFTIRQ0, D(12) KM_SOFTIRQ1, -D(13) KM_TYPE_NR +D(13) KM_CLEARPAGE, +D(14) KM_TYPE_NR }; #undef D diff -urNp linux-2.6.28.8/arch/parisc/include/asm/pgtable.h linux-2.6.28.8/arch/parisc/include/asm/pgtable.h --- linux-2.6.28.8/arch/parisc/include/asm/pgtable.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/parisc/include/asm/pgtable.h 2009-02-21 09:37:48.000000000 -0500 @@ -202,6 +202,17 @@ #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_RO __pgprot(_PAGE_KERNEL & ~_PAGE_WRITE) #define PAGE_KERNEL_UNC __pgprot(_PAGE_KERNEL | _PAGE_NO_CACHE) diff -urNp linux-2.6.28.8/arch/parisc/kernel/module.c linux-2.6.28.8/arch/parisc/kernel/module.c --- linux-2.6.28.8/arch/parisc/kernel/module.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/parisc/kernel/module.c 2009-02-21 09:37:48.000000000 -0500 @@ -75,16 +75,38 @@ /* 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) @@ -298,21 +320,21 @@ int module_frob_arch_sections(CONST Elf_ } /* 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 = ALIGN(me->core_size, 16); - me->arch.fdesc_offset = me->core_size; - me->core_size += fdescs * sizeof(Elf_Fdesc); - - me->core_size = ALIGN(me->core_size, 16); - me->arch.stub_offset = me->core_size; - me->core_size += stubs * sizeof(struct stub_entry); - - me->init_size = ALIGN(me->init_size, 16); - me->arch.init_stub_offset = me->init_size; - me->init_size += init_stubs * sizeof(struct stub_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_rw = ALIGN(me->core_size_rw, 16); + me->arch.fdesc_offset = me->core_size_rw; + me->core_size_rw += fdescs * sizeof(Elf_Fdesc); + + me->core_size_rx = ALIGN(me->core_size_rx, 16); + me->arch.stub_offset = me->core_size_rx; + me->core_size_rx += stubs * sizeof(struct stub_entry); + + me->init_size_rx = ALIGN(me->init_size_rx, 16); + me->arch.init_stub_offset = me->init_size_rx; + me->init_size_rx += init_stubs * sizeof(struct stub_entry); me->arch.got_max = gots; me->arch.fdesc_max = fdescs; @@ -332,7 +354,7 @@ static Elf64_Word get_got(struct module 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; @@ -350,7 +372,7 @@ static Elf64_Word get_got(struct module #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); @@ -368,7 +390,7 @@ static Elf_Addr get_fdesc(struct module /* 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 */ @@ -388,12 +410,12 @@ static Elf_Addr get_stub(struct module * if(init_section) { i = me->arch.init_stub_count++; BUG_ON(me->arch.init_stub_count > me->arch.init_stub_max); - stub = me->module_init + me->arch.init_stub_offset + + stub = me->module_init_rx + me->arch.init_stub_offset + i * sizeof(struct stub_entry); } else { i = me->arch.stub_count++; BUG_ON(me->arch.stub_count > me->arch.stub_max); - stub = me->module_core + me->arch.stub_offset + + stub = me->module_core_rx + me->arch.stub_offset + i * sizeof(struct stub_entry); } @@ -761,7 +783,7 @@ 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 -urNp linux-2.6.28.8/arch/parisc/kernel/sys_parisc.c linux-2.6.28.8/arch/parisc/kernel/sys_parisc.c --- linux-2.6.28.8/arch/parisc/kernel/sys_parisc.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/parisc/kernel/sys_parisc.c 2009-02-21 09:37:48.000000000 -0500 @@ -98,7 +98,7 @@ unsigned long arch_get_unmapped_area(str if (flags & MAP_FIXED) return addr; if (!addr) - addr = TASK_UNMAPPED_BASE; + addr = current->mm->mmap_base; if (filp) { addr = get_shared_area(filp->f_mapping, addr, len, pgoff); diff -urNp linux-2.6.28.8/arch/parisc/kernel/traps.c linux-2.6.28.8/arch/parisc/kernel/traps.c --- linux-2.6.28.8/arch/parisc/kernel/traps.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/parisc/kernel/traps.c 2009-02-21 09:37:48.000000000 -0500 @@ -731,9 +731,7 @@ void handle_interruption(int code, struc 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 -urNp linux-2.6.28.8/arch/parisc/mm/fault.c linux-2.6.28.8/arch/parisc/mm/fault.c --- linux-2.6.28.8/arch/parisc/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/parisc/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -16,6 +16,7 @@ #include #include #include +#include #include #include @@ -53,7 +54,7 @@ DEFINE_PER_CPU(struct exception_data, ex 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) { @@ -139,6 +140,116 @@ parisc_acctyp(unsigned long code, unsign } #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(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 + void do_page_fault(struct pt_regs *regs, unsigned long code, unsigned long address) { @@ -165,8 +276,33 @@ good_area: acc_type = parisc_acctyp(code,regs->iir); - if ((vma->vm_flags & acc_type) != acc_type) + if ((vma->vm_flags & acc_type) != acc_type) { + +#ifdef CONFIG_PAX_PAGEEXEC + if ((mm->pax_flags & MF_PAX_PAGEEXEC) && (acc_type & VM_EXEC) && + (address & ~3UL) == instruction_pointer(regs)) + { + up_read(&mm->mmap_sem); + switch (pax_handle_fetch_fault(regs)) { + +#ifdef CONFIG_PAX_EMUPLT + case 3: + return; +#endif + +#ifdef CONFIG_PAX_EMUTRAMP + case 2: + return; +#endif + + } + pax_report_fault(regs, (void *)instruction_pointer(regs), (void *)regs->gr[30]); + do_group_exit(SIGKILL); + } +#endif + goto bad_area; + } /* * If for any reason at all we couldn't handle the fault, make diff -urNp linux-2.6.28.8/arch/powerpc/include/asm/elf.h linux-2.6.28.8/arch/powerpc/include/asm/elf.h --- linux-2.6.28.8/arch/powerpc/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -180,6 +180,18 @@ typedef elf_fpreg_t elf_vsrreghalf_t32[E #define ELF_ET_DYN_BASE (0x20000000) +#ifdef CONFIG_PAX_ASLR +#define PAX_ELF_ET_DYN_BASE (0x10000000UL) + +#ifdef __powerpc64__ +#define PAX_DELTA_MMAP_LEN (test_thread_flag(TIF_32BIT) ? 16 : 28) +#define PAX_DELTA_STACK_LEN (test_thread_flag(TIF_32BIT) ? 16 : 28) +#else +#define PAX_DELTA_MMAP_LEN 15 +#define PAX_DELTA_STACK_LEN 15 +#endif +#endif + /* * Our registers are always unsigned longs, whether we're a 32 bit * process or 64 bit, on either a 64 bit or 32 bit kernel. diff -urNp linux-2.6.28.8/arch/powerpc/include/asm/kmap_types.h linux-2.6.28.8/arch/powerpc/include/asm/kmap_types.h --- linux-2.6.28.8/arch/powerpc/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -26,6 +26,7 @@ enum km_type { KM_SOFTIRQ1, KM_PPC_SYNC_PAGE, KM_PPC_SYNC_ICACHE, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/powerpc/include/asm/page_64.h linux-2.6.28.8/arch/powerpc/include/asm/page_64.h --- linux-2.6.28.8/arch/powerpc/include/asm/page_64.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/include/asm/page_64.h 2009-02-21 09:37:48.000000000 -0500 @@ -170,15 +170,18 @@ do { \ * stack by default, so in the absense 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 \ (test_thread_flag(TIF_32BIT) ? \ VM_STACK_DEFAULT_FLAGS32 : VM_STACK_DEFAULT_FLAGS64) +#endif #include diff -urNp linux-2.6.28.8/arch/powerpc/include/asm/page.h linux-2.6.28.8/arch/powerpc/include/asm/page.h --- linux-2.6.28.8/arch/powerpc/include/asm/page.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/include/asm/page.h 2009-02-21 09:37:48.000000000 -0500 @@ -111,8 +111,9 @@ extern phys_addr_t kernstart_addr; * 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) diff -urNp linux-2.6.28.8/arch/powerpc/kernel/module_32.c linux-2.6.28.8/arch/powerpc/kernel/module_32.c --- linux-2.6.28.8/arch/powerpc/kernel/module_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/kernel/module_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -158,7 +158,7 @@ int module_frob_arch_sections(Elf32_Ehdr 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; } @@ -199,11 +199,16 @@ static uint32_t do_plt_call(void *locati DEBUGP("Doing plt for call to 0x%x at 0x%x\n", val, (unsigned int)location); /* Init, or core PLT? */ - if (location >= mod->module_core - && location < mod->module_core + mod->core_size) + if ((location >= mod->module_core_rx && location < mod->module_core_rx + mod->core_size_rx) || + (location >= mod->module_core_rw && location < mod->module_core_rw + mod->core_size_rw)) entry = (void *)sechdrs[mod->arch.core_plt_section].sh_addr; - else + else if ((location >= mod->module_init_rx && location < mod->module_init_rx + mod->init_size_rx) || + (location >= mod->module_init_rw && location < mod->module_init_rw + mod->init_size_rw)) entry = (void *)sechdrs[mod->arch.init_plt_section].sh_addr; + else { + printk(KERN_ERR "%s: invalid R_PPC_REL24 entry found\n", mod->name); + return ~0UL; + } /* Find this entry, or if that fails, the next avail. entry */ while (entry->jump[0]) { diff -urNp linux-2.6.28.8/arch/powerpc/kernel/signal_32.c linux-2.6.28.8/arch/powerpc/kernel/signal_32.c --- linux-2.6.28.8/arch/powerpc/kernel/signal_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/kernel/signal_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -857,7 +857,7 @@ int handle_rt_signal32(unsigned long sig /* Save user registers on the stack */ frame = &rt_sf->uc.uc_mcontext; addr = frame; - if (vdso32_rt_sigtramp && current->mm->context.vdso_base) { + if (vdso32_rt_sigtramp && current->mm->context.vdso_base != ~0UL) { if (save_user_regs(regs, frame, 0, 1)) goto badframe; regs->link = current->mm->context.vdso_base + vdso32_rt_sigtramp; diff -urNp linux-2.6.28.8/arch/powerpc/kernel/signal_64.c linux-2.6.28.8/arch/powerpc/kernel/signal_64.c --- linux-2.6.28.8/arch/powerpc/kernel/signal_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/kernel/signal_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -429,7 +429,7 @@ int handle_rt_signal64(int signr, struct current->thread.fpscr.val = 0; /* Set up to return from userspace. */ - if (vdso64_rt_sigtramp && current->mm->context.vdso_base) { + if (vdso64_rt_sigtramp && current->mm->context.vdso_base != ~0UL) { regs->link = current->mm->context.vdso_base + vdso64_rt_sigtramp; } else { err |= setup_trampoline(__NR_rt_sigreturn, &frame->tramp[0]); diff -urNp linux-2.6.28.8/arch/powerpc/kernel/vdso.c linux-2.6.28.8/arch/powerpc/kernel/vdso.c --- linux-2.6.28.8/arch/powerpc/kernel/vdso.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/kernel/vdso.c 2009-02-21 09:37:48.000000000 -0500 @@ -212,7 +212,7 @@ int arch_setup_additional_pages(struct l 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 @@ -229,7 +229,7 @@ int arch_setup_additional_pages(struct l */ down_write(&mm->mmap_sem); vdso_base = get_unmapped_area(NULL, vdso_base, - vdso_pages << PAGE_SHIFT, 0, 0); + vdso_pages << PAGE_SHIFT, 0, MAP_PRIVATE | MAP_EXECUTABLE); if (IS_ERR_VALUE(vdso_base)) { rc = vdso_base; goto fail_mmapsem; diff -urNp linux-2.6.28.8/arch/powerpc/mm/fault.c linux-2.6.28.8/arch/powerpc/mm/fault.c --- linux-2.6.28.8/arch/powerpc/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -29,6 +29,10 @@ #include #include #include +#include +#include +#include +#include #include #include @@ -62,6 +66,363 @@ static inline int notify_page_fault(stru } #endif +#ifdef CONFIG_PAX_EMUSIGRT +void pax_syscall_close(struct vm_area_struct *vma) +{ + vma->vm_mm->call_syscall = 0UL; +} + +static int pax_syscall_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] = 0x44000002U; /* sc */ + __flush_dcache_icache(kaddr); + kunmap(vmf->page); + return VM_FAULT_MAJOR; +} + +static struct vm_operations_struct pax_vm_ops = { + .close = pax_syscall_close, + .fault = pax_syscall_fault +}; + +static int pax_insert_vma(struct vm_area_struct *vma, unsigned long addr) +{ + int ret; + + 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 + +#ifdef CONFIG_PAX_PAGEEXEC +/* + * PaX: decide what to do with offenders (regs->nip = fault address) + * + * returns 1 when task should be killed + * 2 when patched GOT trampoline was detected + * 3 when patched PLT trampoline was detected + * 4 when unpatched PLT trampoline was detected + * 5 when sigreturn trampoline was detected + * 6 when rt_sigreturn trampoline was detected + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + +#if defined(CONFIG_PAX_EMUPLT) || defined(CONFIG_PAX_EMUSIGRT) + int err; +#endif + +#ifdef CONFIG_PAX_EMUPLT + do { /* PaX: patched GOT emulation */ + unsigned int blrl; + + err = get_user(blrl, (unsigned int *)regs->nip); + + if (!err && blrl == 0x4E800021U) { + unsigned long temp = regs->nip; + + regs->nip = regs->link & 0xFFFFFFFCUL; + regs->link = temp + 4UL; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #1 */ + unsigned int b; + + err = get_user(b, (unsigned int *)regs->nip); + + if (!err && (b & 0xFC000003U) == 0x48000000U) { + regs->nip += (((b | 0xFC000000UL) ^ 0x02000000UL) + 0x02000000UL); + return 3; + } + } while (0); + + do { /* PaX: unpatched PLT emulation #1 */ + unsigned int li, b; + + err = get_user(li, (unsigned int *)regs->nip); + err |= get_user(b, (unsigned int *)(regs->nip+4)); + + if (!err && (li & 0xFFFF0000U) == 0x39600000U && (b & 0xFC000003U) == 0x48000000U) { + unsigned int rlwinm, add, li2, addis2, mtctr, li3, addis3, bctr; + unsigned long addr = b | 0xFC000000UL; + + addr = regs->nip + 4 + ((addr ^ 0x02000000UL) + 0x02000000UL); + err = get_user(rlwinm, (unsigned int *)addr); + err |= get_user(add, (unsigned int *)(addr+4)); + err |= get_user(li2, (unsigned int *)(addr+8)); + err |= get_user(addis2, (unsigned int *)(addr+12)); + err |= get_user(mtctr, (unsigned int *)(addr+16)); + err |= get_user(li3, (unsigned int *)(addr+20)); + err |= get_user(addis3, (unsigned int *)(addr+24)); + err |= get_user(bctr, (unsigned int *)(addr+28)); + + if (err) + break; + + if (rlwinm == 0x556C083CU && + add == 0x7D6C5A14U && + (li2 & 0xFFFF0000U) == 0x39800000U && + (addis2 & 0xFFFF0000U) == 0x3D8C0000U && + mtctr == 0x7D8903A6U && + (li3 & 0xFFFF0000U) == 0x39800000U && + (addis3 & 0xFFFF0000U) == 0x3D8C0000U && + bctr == 0x4E800420U) + { + regs->gpr[PT_R11] = 3 * (((li | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + regs->gpr[PT_R12] = (((li3 | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + regs->gpr[PT_R12] += (addis3 & 0xFFFFU) << 16; + regs->ctr = (((li2 | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + regs->ctr += (addis2 & 0xFFFFU) << 16; + regs->nip = regs->ctr; + return 4; + } + } + } while (0); + +#if 0 + do { /* PaX: unpatched PLT emulation #2 */ + unsigned int lis, lwzu, b, bctr; + + err = get_user(lis, (unsigned int *)regs->nip); + err |= get_user(lwzu, (unsigned int *)(regs->nip+4)); + err |= get_user(b, (unsigned int *)(regs->nip+8)); + err |= get_user(bctr, (unsigned int *)(regs->nip+12)); + + if (err) + break; + + if ((lis & 0xFFFF0000U) == 0x39600000U && + (lwzu & 0xU) == 0xU && + (b & 0xFC000003U) == 0x48000000U && + bctr == 0x4E800420U) + { + unsigned int addis, addi, rlwinm, add, li2, addis2, mtctr, li3, addis3, bctr; + unsigned long addr = b | 0xFC000000UL; + + addr = regs->nip + 12 + ((addr ^ 0x02000000UL) + 0x02000000UL); + err = get_user(addis, (unsigned int *)addr); + err |= get_user(addi, (unsigned int *)(addr+4)); + err |= get_user(rlwinm, (unsigned int *)(addr+8)); + err |= get_user(add, (unsigned int *)(addr+12)); + err |= get_user(li2, (unsigned int *)(addr+16)); + err |= get_user(addis2, (unsigned int *)(addr+20)); + err |= get_user(mtctr, (unsigned int *)(addr+24)); + err |= get_user(li3, (unsigned int *)(addr+28)); + err |= get_user(addis3, (unsigned int *)(addr+32)); + err |= get_user(bctr, (unsigned int *)(addr+36)); + + if (err) + break; + + if ((addis & 0xFFFF0000U) == 0x3D6B0000U && + (addi & 0xFFFF0000U) == 0x396B0000U && + rlwinm == 0x556C083CU && + add == 0x7D6C5A14U && + (li2 & 0xFFFF0000U) == 0x39800000U && + (addis2 & 0xFFFF0000U) == 0x3D8C0000U && + mtctr == 0x7D8903A6U && + (li3 & 0xFFFF0000U) == 0x39800000U && + (addis3 & 0xFFFF0000U) == 0x3D8C0000U && + bctr == 0x4E800420U) + { + regs->gpr[PT_R11] = 3 * (((li | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + regs->gpr[PT_R12] = (((li3 | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + regs->gpr[PT_R12] += (addis3 & 0xFFFFU) << 16; + regs->ctr = (((li2 | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + regs->ctr += (addis2 & 0xFFFFU) << 16; + regs->nip = regs->ctr; + return 4; + } + } + } while (0); +#endif + + do { /* PaX: unpatched PLT emulation #3 */ + unsigned int li, b; + + err = get_user(li, (unsigned int *)regs->nip); + err |= get_user(b, (unsigned int *)(regs->nip+4)); + + if (!err && (li & 0xFFFF0000U) == 0x39600000U && (b & 0xFC000003U) == 0x48000000U) { + unsigned int addis, lwz, mtctr, bctr; + unsigned long addr = b | 0xFC000000UL; + + addr = regs->nip + 4 + ((addr ^ 0x02000000UL) + 0x02000000UL); + err = get_user(addis, (unsigned int *)addr); + err |= get_user(lwz, (unsigned int *)(addr+4)); + err |= get_user(mtctr, (unsigned int *)(addr+8)); + err |= get_user(bctr, (unsigned int *)(addr+12)); + + if (err) + break; + + if ((addis & 0xFFFF0000U) == 0x3D6B0000U && + (lwz & 0xFFFF0000U) == 0x816B0000U && + mtctr == 0x7D6903A6U && + bctr == 0x4E800420U) + { + unsigned int r11; + + addr = (addis << 16) + (((li | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + addr += (((lwz | 0xFFFF0000UL) ^ 0x00008000UL) + 0x00008000UL); + + err = get_user(r11, (unsigned int *)addr); + if (err) + break; + + regs->gpr[PT_R11] = r11; + regs->ctr = r11; + regs->nip = r11; + return 4; + } + } + } while (0); +#endif + +#ifdef CONFIG_PAX_EMUSIGRT + do { /* PaX: sigreturn emulation */ + unsigned int li, sc; + + err = get_user(li, (unsigned int *)regs->nip); + err |= get_user(sc, (unsigned int *)(regs->nip+4)); + + if (!err && li == 0x38000000U + __NR_sigreturn && sc == 0x44000002U) { + struct vm_area_struct *vma; + unsigned long call_syscall; + + down_read(¤t->mm->mmap_sem); + call_syscall = current->mm->call_syscall; + up_read(¤t->mm->mmap_sem); + if (likely(call_syscall)) + goto emulate; + + vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); + + down_write(¤t->mm->mmap_sem); + if (current->mm->call_syscall) { + call_syscall = current->mm->call_syscall; + up_write(¤t->mm->mmap_sem); + if (vma) + kmem_cache_free(vm_area_cachep, vma); + goto emulate; + } + + call_syscall = get_unmapped_area(NULL, 0UL, PAGE_SIZE, 0UL, MAP_PRIVATE); + if (!vma || (call_syscall & ~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_syscall)) { + up_write(¤t->mm->mmap_sem); + kmem_cache_free(vm_area_cachep, vma); + return 1; + } + + current->mm->call_syscall = call_syscall; + up_write(¤t->mm->mmap_sem); + +emulate: + regs->gpr[PT_R0] = __NR_sigreturn; + regs->nip = call_syscall; + return 5; + } + } while (0); + + do { /* PaX: rt_sigreturn emulation */ + unsigned int li, sc; + + err = get_user(li, (unsigned int *)regs->nip); + err |= get_user(sc, (unsigned int *)(regs->nip+4)); + + if (!err && li == 0x38000000U + __NR_rt_sigreturn && sc == 0x44000002U) { + struct vm_area_struct *vma; + unsigned int call_syscall; + + down_read(¤t->mm->mmap_sem); + call_syscall = current->mm->call_syscall; + up_read(¤t->mm->mmap_sem); + if (likely(call_syscall)) + goto rt_emulate; + + vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); + + down_write(¤t->mm->mmap_sem); + if (current->mm->call_syscall) { + call_syscall = current->mm->call_syscall; + up_write(¤t->mm->mmap_sem); + if (vma) + kmem_cache_free(vm_area_cachep, vma); + goto rt_emulate; + } + + call_syscall = get_unmapped_area(NULL, 0UL, PAGE_SIZE, 0UL, MAP_PRIVATE); + if (!vma || (call_syscall & ~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_syscall)) { + up_write(¤t->mm->mmap_sem); + kmem_cache_free(vm_area_cachep, vma); + return 1; + } + + current->mm->call_syscall = call_syscall; + up_write(¤t->mm->mmap_sem); + +rt_emulate: + regs->gpr[PT_R0] = __NR_rt_sigreturn; + regs->nip = call_syscall; + return 6; + } + } while (0); +#endif + + return 1; +} + +void pax_report_insns(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 + /* * Check whether the instruction at regs->nip is a store using * an update addressing form which will update r1. @@ -132,7 +493,7 @@ int __kprobes do_page_fault(struct pt_re * 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 @@ -331,6 +692,37 @@ 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_PPC64 + if (is_exec && (error_code & DSISR_PROTFAULT)) { +#else + if (is_exec && regs->nip == address) { +#endif + switch (pax_handle_fetch_fault(regs)) { + +#ifdef CONFIG_PAX_EMUPLT + case 2: + case 3: + case 4: + return 0; +#endif + +#ifdef CONFIG_PAX_EMUSIGRT + case 5: + case 6: + return 0; +#endif + + } + + pax_report_fault(regs, (void *)regs->nip, (void *)regs->gpr[PT_R1]); + do_group_exit(SIGKILL); + } + } +#endif + _exception(SIGSEGV, regs, code, address); return 0; } diff -urNp linux-2.6.28.8/arch/powerpc/mm/mmap.c linux-2.6.28.8/arch/powerpc/mm/mmap.c --- linux-2.6.28.8/arch/powerpc/mm/mmap.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/powerpc/mm/mmap.c 2009-02-21 09:37:48.000000000 -0500 @@ -75,10 +75,22 @@ void arch_pick_mmap_layout(struct mm_str */ if (mmap_is_legacy()) { mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { mm->mmap_base = mmap_base(); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff -urNp linux-2.6.28.8/arch/s390/include/asm/kmap_types.h linux-2.6.28.8/arch/s390/include/asm/kmap_types.h --- linux-2.6.28.8/arch/s390/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/s390/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -16,6 +16,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/s390/kernel/module.c linux-2.6.28.8/arch/s390/kernel/module.c --- linux-2.6.28.8/arch/s390/kernel/module.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/s390/kernel/module.c 2009-02-21 09:37:48.000000000 -0500 @@ -166,11 +166,11 @@ module_frob_arch_sections(Elf_Ehdr *hdr, /* 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; } @@ -256,7 +256,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base 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; @@ -280,7 +280,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base else if (r_type == R_390_GOTENT || r_type == R_390_GOTPLTENT) *(unsigned int *) loc = - (val + (Elf_Addr) me->module_core - loc) >> 1; + (val + (Elf_Addr) me->module_core_rw - loc) >> 1; else if (r_type == R_390_GOT64 || r_type == R_390_GOTPLT64) *(unsigned long *) loc = val; @@ -294,7 +294,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base 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 */ @@ -316,7 +316,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base val = me->arch.plt_offset - me->arch.got_offset + info->plt_offset + rela->r_addend; else - val = (Elf_Addr) me->module_core + + val = (Elf_Addr) me->module_core_rx + me->arch.plt_offset + info->plt_offset + rela->r_addend - loc; if (r_type == R_390_PLT16DBL) @@ -336,7 +336,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base case R_390_GOTOFF32: /* 32 bit offset to GOT. */ case R_390_GOTOFF64: /* 64 bit offset to GOT. */ val = val + rela->r_addend - - ((Elf_Addr) me->module_core + me->arch.got_offset); + ((Elf_Addr) me->module_core_rw + me->arch.got_offset); if (r_type == R_390_GOTOFF16) *(unsigned short *) loc = val; else if (r_type == R_390_GOTOFF32) @@ -346,7 +346,7 @@ apply_rela(Elf_Rela *rela, Elf_Addr base break; case R_390_GOTPC: /* 32 bit PC relative offset to GOT. */ case R_390_GOTPCDBL: /* 32 bit PC rel. off. to GOT shifted by 1. */ - val = (Elf_Addr) me->module_core + me->arch.got_offset + + val = (Elf_Addr) me->module_core_rw + me->arch.got_offset + rela->r_addend - loc; if (r_type == R_390_GOTPC) *(unsigned int *) loc = val; diff -urNp linux-2.6.28.8/arch/sh/include/asm/kmap_types.h linux-2.6.28.8/arch/sh/include/asm/kmap_types.h --- linux-2.6.28.8/arch/sh/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sh/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -24,7 +24,8 @@ D(9) KM_IRQ0, D(10) KM_IRQ1, D(11) KM_SOFTIRQ0, D(12) KM_SOFTIRQ1, -D(13) KM_TYPE_NR +D(13) KM_CLEARPAGE, +D(14) KM_TYPE_NR }; #undef D diff -urNp linux-2.6.28.8/arch/sparc/include/asm/elf_32.h linux-2.6.28.8/arch/sparc/include/asm/elf_32.h --- linux-2.6.28.8/arch/sparc/include/asm/elf_32.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/include/asm/elf_32.h 2009-02-21 09:37:48.000000000 -0500 @@ -116,6 +116,13 @@ 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 -urNp linux-2.6.28.8/arch/sparc/include/asm/elf_64.h linux-2.6.28.8/arch/sparc/include/asm/elf_64.h --- linux-2.6.28.8/arch/sparc/include/asm/elf_64.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/include/asm/elf_64.h 2009-02-21 09:37:48.000000000 -0500 @@ -163,6 +163,12 @@ 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 /* This yields a mask that user programs can use to figure out what instruction set this cpu supports. */ diff -urNp linux-2.6.28.8/arch/sparc/include/asm/kmap_types.h linux-2.6.28.8/arch/sparc/include/asm/kmap_types.h --- linux-2.6.28.8/arch/sparc/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -19,6 +19,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/sparc/include/asm/pgtable_32.h linux-2.6.28.8/arch/sparc/include/asm/pgtable_32.h --- linux-2.6.28.8/arch/sparc/include/asm/pgtable_32.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/include/asm/pgtable_32.h 2009-02-21 09:37:48.000000000 -0500 @@ -43,6 +43,13 @@ BTFIXUPDEF_SIMM13(user_ptrs_per_pgd) BTFIXUPDEF_INT(page_none) BTFIXUPDEF_INT(page_copy) BTFIXUPDEF_INT(page_readonly) + +#ifdef CONFIG_PAX_PAGEEXEC +BTFIXUPDEF_INT(page_shared_noexec) +BTFIXUPDEF_INT(page_copy_noexec) +BTFIXUPDEF_INT(page_readonly_noexec) +#endif + BTFIXUPDEF_INT(page_kernel) #define PMD_SHIFT SUN4C_PMD_SHIFT @@ -64,6 +71,16 @@ extern pgprot_t PAGE_SHARED; #define PAGE_COPY __pgprot(BTFIXUP_INT(page_copy)) #define PAGE_READONLY __pgprot(BTFIXUP_INT(page_readonly)) +#ifdef CONFIG_PAX_PAGEEXEC +extern pgprot_t PAGE_SHARED_NOEXEC; +# define PAGE_COPY_NOEXEC __pgprot(BTFIXUP_INT(page_copy_noexec)) +# define PAGE_READONLY_NOEXEC __pgprot(BTFIXUP_INT(page_readonly_noexec)) +#else +# define PAGE_SHARED_NOEXEC PAGE_SHARED +# define PAGE_COPY_NOEXEC PAGE_COPY +# define PAGE_READONLY_NOEXEC PAGE_READONLY +#endif + extern unsigned long page_kernel; #ifdef MODULE diff -urNp linux-2.6.28.8/arch/sparc/include/asm/pgtsrmmu.h linux-2.6.28.8/arch/sparc/include/asm/pgtsrmmu.h --- linux-2.6.28.8/arch/sparc/include/asm/pgtsrmmu.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/include/asm/pgtsrmmu.h 2009-02-21 09:37:48.000000000 -0500 @@ -115,6 +115,13 @@ SRMMU_EXEC | SRMMU_REF) #define SRMMU_PAGE_RDONLY __pgprot(SRMMU_VALID | SRMMU_CACHE | \ SRMMU_EXEC | SRMMU_REF) + +#ifdef CONFIG_PAX_PAGEEXEC +#define SRMMU_PAGE_SHARED_NOEXEC __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_WRITE | SRMMU_REF) +#define SRMMU_PAGE_COPY_NOEXEC __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_REF) +#define SRMMU_PAGE_RDONLY_NOEXEC __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_REF) +#endif + #define SRMMU_PAGE_KERNEL __pgprot(SRMMU_VALID | SRMMU_CACHE | SRMMU_PRIV | \ SRMMU_DIRTY | SRMMU_REF) diff -urNp linux-2.6.28.8/arch/sparc/kernel/sys_sparc.c linux-2.6.28.8/arch/sparc/kernel/sys_sparc.c --- linux-2.6.28.8/arch/sparc/kernel/sys_sparc.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/kernel/sys_sparc.c 2009-02-21 09:37:48.000000000 -0500 @@ -56,7 +56,7 @@ unsigned long arch_get_unmapped_area(str if (ARCH_SUN4C && len > 0x20000000) return -ENOMEM; if (!addr) - addr = TASK_UNMAPPED_BASE; + addr = current->mm->mmap_base; if (flags & MAP_SHARED) addr = COLOUR_ALIGN(addr); diff -urNp linux-2.6.28.8/arch/sparc/Makefile linux-2.6.28.8/arch/sparc/Makefile --- linux-2.6.28.8/arch/sparc/Makefile 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/Makefile 2009-02-21 09:37:48.000000000 -0500 @@ -37,7 +37,7 @@ drivers-$(CONFIG_OPROFILE) += arch/sparc # Renaming is done to avoid confusing pattern matching rules in 2.5.45 (multy-) INIT_Y := $(patsubst %/, %/built-in.o, $(init-y)) CORE_Y := $(core-y) -CORE_Y += kernel/ mm/ fs/ ipc/ security/ crypto/ block/ +CORE_Y += kernel/ mm/ fs/ ipc/ security/ crypto/ block/ grsecurity/ CORE_Y := $(patsubst %/, %/built-in.o, $(CORE_Y)) DRIVERS_Y := $(patsubst %/, %/built-in.o, $(drivers-y)) NET_Y := $(patsubst %/, %/built-in.o, $(net-y)) diff -urNp linux-2.6.28.8/arch/sparc/mm/fault.c linux-2.6.28.8/arch/sparc/mm/fault.c --- linux-2.6.28.8/arch/sparc/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -21,6 +21,9 @@ #include #include #include +#include +#include +#include #include #include @@ -167,6 +170,249 @@ static unsigned long compute_si_addr(str return safe_compute_effective_address(regs, insn); } +#ifdef CONFIG_PAX_PAGEEXEC +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 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; + + 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; +} + +/* + * PaX: decide what to do with offenders (regs->pc = fault address) + * + * returns 1 when task should be killed + * 2 when patched PLT trampoline was detected + * 3 when unpatched PLT trampoline was detected + */ +static int pax_handle_fetch_fault(struct pt_regs *regs) +{ + +#ifdef CONFIG_PAX_EMUPLT + int err; + + do { /* PaX: patched PLT emulation #1 */ + unsigned int sethi1, sethi2, jmpl; + + err = get_user(sethi1, (unsigned int *)regs->pc); + err |= get_user(sethi2, (unsigned int *)(regs->pc+4)); + err |= get_user(jmpl, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((sethi1 & 0xFFC00000U) == 0x03000000U && + (sethi2 & 0xFFC00000U) == 0x03000000U && + (jmpl & 0xFFFFE000U) == 0x81C06000U) + { + unsigned int addr; + + regs->u_regs[UREG_G1] = (sethi2 & 0x003FFFFFU) << 10; + addr = regs->u_regs[UREG_G1]; + addr += (((jmpl | 0xFFFFE000U) ^ 0x00001000U) + 0x00001000U); + regs->pc = addr; + regs->npc = addr+4; + return 2; + } + } while (0); + + { /* PaX: patched PLT emulation #2 */ + unsigned int ba; + + err = get_user(ba, (unsigned int *)regs->pc); + + if (!err && (ba & 0xFFC00000U) == 0x30800000U) { + unsigned int addr; + + addr = regs->pc + ((((ba | 0xFFC00000U) ^ 0x00200000U) + 0x00200000U) << 2); + regs->pc = addr; + regs->npc = addr+4; + return 2; + } + } + + do { /* PaX: patched PLT emulation #3 */ + unsigned int sethi, jmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->pc); + err |= get_user(jmpl, (unsigned int *)(regs->pc+4)); + err |= get_user(nop, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + (jmpl & 0xFFFFE000U) == 0x81C06000U && + nop == 0x01000000U) + { + unsigned int addr; + + addr = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] = addr; + addr += (((jmpl | 0xFFFFE000U) ^ 0x00001000U) + 0x00001000U); + regs->pc = addr; + regs->npc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: unpatched PLT emulation step 1 */ + unsigned int sethi, ba, nop; + + err = get_user(sethi, (unsigned int *)regs->pc); + err |= get_user(ba, (unsigned int *)(regs->pc+4)); + err |= get_user(nop, (unsigned int *)(regs->pc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + ((ba & 0xFFC00000U) == 0x30800000U || (ba & 0xFFF80000U) == 0x30680000U) && + nop == 0x01000000U) + { + unsigned int addr, save, call; + + if ((ba & 0xFFC00000U) == 0x30800000U) + addr = regs->pc + 4 + ((((ba | 0xFFC00000U) ^ 0x00200000U) + 0x00200000U) << 2); + else + addr = regs->pc + 4 + ((((ba | 0xFFF80000U) ^ 0x00040000U) + 0x00040000U) << 2); + + err = get_user(save, (unsigned int *)addr); + err |= get_user(call, (unsigned int *)(addr+4)); + err |= get_user(nop, (unsigned int *)(addr+8)); + if (err) + break; + + 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; + } + } + } 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(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 + asmlinkage void do_sparc_fault(struct pt_regs *regs, int text_fault, int write, unsigned long address) { @@ -231,6 +477,24 @@ 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 -urNp linux-2.6.28.8/arch/sparc/mm/init.c linux-2.6.28.8/arch/sparc/mm/init.c --- linux-2.6.28.8/arch/sparc/mm/init.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/mm/init.c 2009-02-21 09:37:48.000000000 -0500 @@ -313,6 +313,9 @@ extern void device_scan(void); pgprot_t PAGE_SHARED __read_mostly; EXPORT_SYMBOL(PAGE_SHARED); +pgprot_t PAGE_SHARED_NOEXEC __read_mostly; +EXPORT_SYMBOL(PAGE_SHARED_NOEXEC); + void __init paging_init(void) { switch(sparc_cpu_model) { @@ -338,17 +341,17 @@ void __init paging_init(void) /* Initialize the protection map with non-constant, MMU dependent values. */ protection_map[0] = PAGE_NONE; - protection_map[1] = PAGE_READONLY; - protection_map[2] = PAGE_COPY; - protection_map[3] = PAGE_COPY; + protection_map[1] = PAGE_READONLY_NOEXEC; + protection_map[2] = PAGE_COPY_NOEXEC; + protection_map[3] = PAGE_COPY_NOEXEC; protection_map[4] = PAGE_READONLY; protection_map[5] = PAGE_READONLY; protection_map[6] = PAGE_COPY; protection_map[7] = PAGE_COPY; protection_map[8] = PAGE_NONE; - protection_map[9] = PAGE_READONLY; - protection_map[10] = PAGE_SHARED; - protection_map[11] = PAGE_SHARED; + protection_map[9] = PAGE_READONLY_NOEXEC; + protection_map[10] = PAGE_SHARED_NOEXEC; + protection_map[11] = PAGE_SHARED_NOEXEC; protection_map[12] = PAGE_READONLY; protection_map[13] = PAGE_READONLY; protection_map[14] = PAGE_SHARED; diff -urNp linux-2.6.28.8/arch/sparc/mm/srmmu.c linux-2.6.28.8/arch/sparc/mm/srmmu.c --- linux-2.6.28.8/arch/sparc/mm/srmmu.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc/mm/srmmu.c 2009-02-21 09:37:48.000000000 -0500 @@ -2162,6 +2162,13 @@ void __init ld_mmu_srmmu(void) PAGE_SHARED = pgprot_val(SRMMU_PAGE_SHARED); BTFIXUPSET_INT(page_copy, pgprot_val(SRMMU_PAGE_COPY)); BTFIXUPSET_INT(page_readonly, pgprot_val(SRMMU_PAGE_RDONLY)); + +#ifdef CONFIG_PAX_PAGEEXEC + PAGE_SHARED_NOEXEC = pgprot_val(SRMMU_PAGE_SHARED_NOEXEC); + BTFIXUPSET_INT(page_copy_noexec, pgprot_val(SRMMU_PAGE_COPY_NOEXEC)); + BTFIXUPSET_INT(page_readonly_noexec, pgprot_val(SRMMU_PAGE_RDONLY_NOEXEC)); +#endif + BTFIXUPSET_INT(page_kernel, pgprot_val(SRMMU_PAGE_KERNEL)); page_kernel = pgprot_val(SRMMU_PAGE_KERNEL); diff -urNp linux-2.6.28.8/arch/sparc64/kernel/Makefile linux-2.6.28.8/arch/sparc64/kernel/Makefile --- linux-2.6.28.8/arch/sparc64/kernel/Makefile 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc64/kernel/Makefile 2009-02-21 09:37:48.000000000 -0500 @@ -3,7 +3,7 @@ # EXTRA_AFLAGS := -ansi -EXTRA_CFLAGS := -Werror +#EXTRA_CFLAGS := -Werror CFLAGS_REMOVE_ftrace.o = -pg diff -urNp linux-2.6.28.8/arch/sparc64/kernel/sys_sparc.c linux-2.6.28.8/arch/sparc64/kernel/sys_sparc.c --- linux-2.6.28.8/arch/sparc64/kernel/sys_sparc.c 2009-02-08 00:54:27.000000000 -0500 +++ linux-2.6.28.8/arch/sparc64/kernel/sys_sparc.c 2009-02-21 09:37:48.000000000 -0500 @@ -124,7 +124,7 @@ unsigned long arch_get_unmapped_area(str /* We do not accept a shared mapping if it would violate * cache aliasing constraints. */ - if ((flags & MAP_SHARED) && + if ((filp || (flags & MAP_SHARED)) && ((addr - (pgoff << PAGE_SHIFT)) & (SHMLBA - 1))) return -EINVAL; return addr; @@ -139,6 +139,10 @@ unsigned long arch_get_unmapped_area(str if (filp || (flags & MAP_SHARED)) do_color_align = 1; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + if (addr) { if (do_color_align) addr = COLOUR_ALIGN(addr, pgoff); @@ -152,9 +156,9 @@ unsigned long arch_get_unmapped_area(str } if (len > mm->cached_hole_size) { - start_addr = addr = mm->free_area_cache; + start_addr = addr = mm->free_area_cache; } else { - start_addr = addr = TASK_UNMAPPED_BASE; + start_addr = addr = mm->mmap_base; mm->cached_hole_size = 0; } @@ -174,8 +178,8 @@ full_search: vma = find_vma(mm, VA_EXCLUDE_END); } if (unlikely(task_size < addr)) { - if (start_addr != TASK_UNMAPPED_BASE) { - start_addr = addr = TASK_UNMAPPED_BASE; + if (start_addr != mm->mmap_base) { + start_addr = addr = mm->mmap_base; mm->cached_hole_size = 0; goto full_search; } @@ -215,7 +219,7 @@ arch_get_unmapped_area_topdown(struct fi /* 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; @@ -378,6 +382,12 @@ void arch_pick_mmap_layout(struct mm_str current->signal->rlim[RLIMIT_STACK].rlim_cur == RLIM_INFINITY || sysctl_legacy_va_layout) { mm->mmap_base = TASK_UNMAPPED_BASE + random_factor; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { @@ -392,6 +402,12 @@ void arch_pick_mmap_layout(struct mm_str gap = (task_size / 6 * 5); mm->mmap_base = PAGE_ALIGN(task_size - gap - random_factor); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff -urNp linux-2.6.28.8/arch/sparc64/mm/fault.c linux-2.6.28.8/arch/sparc64/mm/fault.c --- linux-2.6.28.8/arch/sparc64/mm/fault.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc64/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -19,6 +19,9 @@ #include #include #include +#include +#include +#include #include #include @@ -224,6 +227,367 @@ cannot_handle: unhandled_fault (address, current, regs); } +#ifdef CONFIG_PAX_PAGEEXEC +#ifdef CONFIG_PAX_EMUPLT +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 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; + + 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); + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + { /* PaX: patched PLT emulation #2 */ + unsigned int ba; + + err = get_user(ba, (unsigned int *)regs->tpc); + + if (!err && (ba & 0xFFC00000U) == 0x30800000U) { + unsigned long addr; + + addr = regs->tpc + ((((ba | 0xFFFFFFFFFFC00000UL) ^ 0x00200000UL) + 0x00200000UL) << 2); + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } + + do { /* PaX: patched PLT emulation #3 */ + unsigned int sethi, jmpl, nop; + + err = get_user(sethi, (unsigned int *)regs->tpc); + err |= get_user(jmpl, (unsigned int *)(regs->tpc+4)); + err |= get_user(nop, (unsigned int *)(regs->tpc+8)); + + if (err) + break; + + if ((sethi & 0xFFC00000U) == 0x03000000U && + (jmpl & 0xFFFFE000U) == 0x81C06000U && + nop == 0x01000000U) + { + unsigned long addr; + + addr = (sethi & 0x003FFFFFU) << 10; + regs->u_regs[UREG_G1] = addr; + addr += (((jmpl | 0xFFFFFFFFFFFFE000UL) ^ 0x00001000UL) + 0x00001000UL); + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #4 */ + unsigned int mov1, call, mov2; + + err = get_user(mov1, (unsigned int *)regs->tpc); + err |= get_user(call, (unsigned int *)(regs->tpc+4)); + err |= get_user(mov2, (unsigned int *)(regs->tpc+8)); + + if (err) + break; + + if (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); + regs->tpc = addr; + regs->tnpc = addr+4; + return 2; + } + } while (0); + + do { /* PaX: patched PLT emulation #5 */ + unsigned int sethi1, sethi2, or1, or2, sllx, jmpl, nop; + + err = get_user(sethi1, (unsigned int *)regs->tpc); + err |= get_user(sethi2, (unsigned int *)(regs->tpc+4)); + err |= get_user(or1, (unsigned int *)(regs->tpc+8)); + err |= get_user(or2, (unsigned int *)(regs->tpc+12)); + err |= get_user(sllx, (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 ((sethi1 & 0xFFC00000U) == 0x03000000U && + (sethi2 & 0xFFC00000U) == 0x0B000000U && + (or1 & 0xFFFFE000U) == 0x82106000U && + (or2 & 0xFFFFE000U) == 0x8A116000U && + sllx == 0x83287020 && + 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 sethi1, sethi2, sllx, or, jmpl, nop; + + err = get_user(sethi1, (unsigned int *)regs->tpc); + err |= get_user(sethi2, (unsigned int *)(regs->tpc+4)); + err |= get_user(sllx, (unsigned int *)(regs->tpc+8)); + err |= get_user(or, (unsigned int *)(regs->tpc+12)); + err |= get_user(jmpl, (unsigned int *)(regs->tpc+16)); + err |= get_user(nop, (unsigned int *)(regs->tpc+20)); + + if (err) + break; + + if ((sethi1 & 0xFFC00000U) == 0x03000000U && + (sethi2 & 0xFFC00000U) == 0x0B000000U && + sllx == 0x83287020 && + (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: patched PLT emulation #7 */ + 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); + 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; + + if ((ba & 0xFFC00000U) == 0x30800000U) + addr = regs->tpc + 4 + ((((ba | 0xFFFFFFFFFFC00000UL) ^ 0x00200000UL) + 0x00200000UL) << 2); + else + addr = regs->tpc + 4 + ((((ba | 0xFFFFFFFFFFF80000UL) ^ 0x00040000UL) + 0x00040000UL) << 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; + + 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; + } + } + } while (0); + + 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); + + regs->u_regs[UREG_RETPC] = regs->tpc; + regs->tpc = dl_resolve; + regs->tnpc = dl_resolve+4; + return 3; + } + } while (0); +#endif + + return 1; +} + +void pax_report_insns(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 + asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs) { struct mm_struct *mm = current->mm; @@ -265,8 +629,10 @@ asmlinkage void __kprobes do_sparc64_fau goto intr_or_no_mm; if (test_thread_flag(TIF_32BIT)) { - if (!(regs->tstate & TSTATE_PRIV)) + if (!(regs->tstate & TSTATE_PRIV)) { regs->tpc &= 0xffffffff; + regs->tnpc &= 0xffffffff; + } address &= 0xffffffff; } @@ -283,6 +649,29 @@ asmlinkage void __kprobes do_sparc64_fau 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 -urNp linux-2.6.28.8/arch/sparc64/mm/Makefile linux-2.6.28.8/arch/sparc64/mm/Makefile --- linux-2.6.28.8/arch/sparc64/mm/Makefile 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/sparc64/mm/Makefile 2009-02-21 09:37:48.000000000 -0500 @@ -2,7 +2,7 @@ # EXTRA_AFLAGS := -ansi -EXTRA_CFLAGS := -Werror +#EXTRA_CFLAGS := -Werror obj-y := ultra.o tlb.o tsb.o fault.o init.o generic.o diff -urNp linux-2.6.28.8/arch/um/include/asm/kmap_types.h linux-2.6.28.8/arch/um/include/asm/kmap_types.h --- linux-2.6.28.8/arch/um/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/um/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -23,6 +23,7 @@ enum km_type { KM_IRQ1, KM_SOFTIRQ0, KM_SOFTIRQ1, + KM_CLEARPAGE, KM_TYPE_NR }; diff -urNp linux-2.6.28.8/arch/um/include/asm/page.h linux-2.6.28.8/arch/um/include/asm/page.h --- linux-2.6.28.8/arch/um/include/asm/page.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/um/include/asm/page.h 2009-02-21 09:37:48.000000000 -0500 @@ -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 -urNp linux-2.6.28.8/arch/um/sys-i386/syscalls.c linux-2.6.28.8/arch/um/sys-i386/syscalls.c --- linux-2.6.28.8/arch/um/sys-i386/syscalls.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/um/sys-i386/syscalls.c 2009-02-21 09:37:48.000000000 -0500 @@ -11,6 +11,21 @@ #include "asm/uaccess.h" #include "asm/unistd.h" +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 (len > pax_task_size || addr > pax_task_size - len) + return -EINVAL; + + return 0; +} + /* * Perform the select(nd, in, out, ex, tv) and mmap() system * calls. Linux/i386 didn't use to be able to handle more than diff -urNp linux-2.6.28.8/arch/x86/boot/bitops.h linux-2.6.28.8/arch/x86/boot/bitops.h --- linux-2.6.28.8/arch/x86/boot/bitops.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/bitops.h 2009-02-21 09:37:48.000000000 -0500 @@ -26,7 +26,7 @@ static inline int variable_test_bit(int 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 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 -urNp linux-2.6.28.8/arch/x86/boot/boot.h linux-2.6.28.8/arch/x86/boot/boot.h --- linux-2.6.28.8/arch/x86/boot/boot.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/boot.h 2009-02-21 09:37:48.000000000 -0500 @@ -80,7 +80,7 @@ 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; } @@ -176,7 +176,7 @@ static inline void wrgs32(u32 v, addr_t static inline int memcmp(const void *s1, const void *s2, size_t len) { u8 diff; - asm("repe; cmpsb; setnz %0" + asm volatile("repe; cmpsb; setnz %0" : "=qm" (diff), "+D" (s1), "+S" (s2), "+c" (len)); return diff; } diff -urNp linux-2.6.28.8/arch/x86/boot/compressed/head_32.S linux-2.6.28.8/arch/x86/boot/compressed/head_32.S --- linux-2.6.28.8/arch/x86/boot/compressed/head_32.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/compressed/head_32.S 2009-03-07 10:28:24.000000000 -0500 @@ -70,7 +70,7 @@ startup_32: addl $(CONFIG_PHYSICAL_ALIGN - 1), %ebx andl $(~(CONFIG_PHYSICAL_ALIGN - 1)), %ebx #else - movl $LOAD_PHYSICAL_ADDR, %ebx + movl $____LOAD_PHYSICAL_ADDR, %ebx #endif /* Replace the compressed data size with the uncompressed size */ @@ -80,8 +80,8 @@ startup_32: /* Add 8 bytes for every 32K input block */ shrl $12, %eax addl %eax, %ebx - /* Add 32K + 18 bytes of extra slack */ - addl $(32768 + 18), %ebx + /* Add 64K of extra slack */ + addl $65536, %ebx /* Align on a 4K boundary */ addl $4095, %ebx andl $~4095, %ebx @@ -105,7 +105,7 @@ startup_32: addl $(CONFIG_PHYSICAL_ALIGN - 1), %ebp andl $(~(CONFIG_PHYSICAL_ALIGN - 1)), %ebp #else - movl $LOAD_PHYSICAL_ADDR, %ebp + movl $____LOAD_PHYSICAL_ADDR, %ebp #endif /* @@ -160,16 +160,15 @@ relocated: * and where it was actually loaded. */ movl %ebp, %ebx - subl $LOAD_PHYSICAL_ADDR, %ebx + subl $____LOAD_PHYSICAL_ADDR, %ebx jz 2f /* Nothing to be done if loaded at compiled addr. */ /* * Process relocations. */ 1: subl $4, %edi - movl 0(%edi), %ecx - testl %ecx, %ecx - jz 2f + movl (%edi), %ecx + jecxz 2f addl %ebx, -__PAGE_OFFSET(%ebx, %ecx) jmp 1b 2: diff -urNp linux-2.6.28.8/arch/x86/boot/compressed/misc.c linux-2.6.28.8/arch/x86/boot/compressed/misc.c --- linux-2.6.28.8/arch/x86/boot/compressed/misc.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/compressed/misc.c 2009-02-21 09:37:48.000000000 -0500 @@ -373,7 +373,7 @@ static void parse_elf(void *output) case PT_LOAD: #ifdef CONFIG_RELOCATABLE dest = output; - dest += (phdr->p_paddr - LOAD_PHYSICAL_ADDR); + dest += (phdr->p_paddr - ____LOAD_PHYSICAL_ADDR); #else dest = (void *)(phdr->p_paddr); #endif @@ -425,7 +425,7 @@ asmlinkage void decompress_kernel(void * if (heap > ((-__PAGE_OFFSET-(512<<20)-1) & 0x7fffffff)) error("Destination address too large"); #ifndef CONFIG_RELOCATABLE - if ((u32)output != LOAD_PHYSICAL_ADDR) + if ((u32)output != ____LOAD_PHYSICAL_ADDR) error("Wrong destination address"); #endif #endif diff -urNp linux-2.6.28.8/arch/x86/boot/compressed/relocs.c linux-2.6.28.8/arch/x86/boot/compressed/relocs.c --- linux-2.6.28.8/arch/x86/boot/compressed/relocs.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/compressed/relocs.c 2009-02-21 09:37:48.000000000 -0500 @@ -10,8 +10,11 @@ #define USE_BSD #include +#include "../../../../include/linux/autoconf.h" + #define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0])) static Elf32_Ehdr ehdr; +static Elf32_Phdr *phdr; static unsigned long reloc_count, reloc_idx; static unsigned long *relocs; @@ -245,6 +248,36 @@ static void read_ehdr(FILE *fp) } } +static void read_phdrs(FILE *fp) +{ + int i; + + phdr = calloc(ehdr.e_phnum, sizeof(Elf32_Phdr)); + if (!phdr) { + die("Unable to allocate %d program headers\n", + ehdr.e_phnum); + } + if (fseek(fp, ehdr.e_phoff, SEEK_SET) < 0) { + die("Seek to %d failed: %s\n", + ehdr.e_phoff, strerror(errno)); + } + if (fread(phdr, sizeof(*phdr), ehdr.e_phnum, fp) != ehdr.e_phnum) { + die("Cannot read ELF program headers: %s\n", + strerror(errno)); + } + for(i = 0; i < ehdr.e_phnum; i++) { + phdr[i].p_type = elf32_to_cpu(phdr[i].p_type); + phdr[i].p_offset = elf32_to_cpu(phdr[i].p_offset); + phdr[i].p_vaddr = elf32_to_cpu(phdr[i].p_vaddr); + phdr[i].p_paddr = elf32_to_cpu(phdr[i].p_paddr); + phdr[i].p_filesz = elf32_to_cpu(phdr[i].p_filesz); + phdr[i].p_memsz = elf32_to_cpu(phdr[i].p_memsz); + phdr[i].p_flags = elf32_to_cpu(phdr[i].p_flags); + phdr[i].p_align = elf32_to_cpu(phdr[i].p_align); + } + +} + static void read_shdrs(FILE *fp) { int i; @@ -341,6 +374,8 @@ static void read_symtabs(FILE *fp) static void read_relocs(FILE *fp) { int i,j; + uint32_t base; + for (i = 0; i < ehdr.e_shnum; i++) { struct section *sec = &secs[i]; if (sec->shdr.sh_type != SHT_REL) { @@ -360,9 +395,18 @@ static void read_relocs(FILE *fp) die("Cannot read symbol table: %s\n", strerror(errno)); } + base = 0; + for (j = 0; j < ehdr.e_phnum; j++) { + if (phdr[j].p_type != PT_LOAD ) + continue; + if (secs[sec->shdr.sh_info].shdr.sh_offset < phdr[j].p_offset || secs[sec->shdr.sh_info].shdr.sh_offset >= phdr[j].p_offset + phdr[j].p_filesz) + continue; + base = CONFIG_PAGE_OFFSET + phdr[j].p_paddr - phdr[j].p_vaddr; + break; + } for (j = 0; j < sec->shdr.sh_size/sizeof(Elf32_Rel); j++) { Elf32_Rel *rel = &sec->reltab[j]; - rel->r_offset = elf32_to_cpu(rel->r_offset); + rel->r_offset = elf32_to_cpu(rel->r_offset) + base; rel->r_info = elf32_to_cpu(rel->r_info); } } @@ -504,6 +548,23 @@ static void walk_relocs(void (*visit)(El if (sym->st_shndx == SHN_ABS) { continue; } + /* Don't relocate actual per-cpu variables, they are absolute indices, not addresses */ + if (!strcmp(sec_name(sym->st_shndx), ".data.percpu") && strncmp(sym_name(sym_strtab, sym), "__per_cpu_", 10)) + continue; +#if defined(CONFIG_PAX_KERNEXEC) && defined(CONFIG_X86_32) + /* Don't relocate actual code, they are relocated implicitly by the base address of KERNEL_CS */ + if (!strcmp(sec_name(sym->st_shndx), ".init.text")) + continue; + if (!strcmp(sec_name(sym->st_shndx), ".exit.text")) + continue; + if (!strcmp(sec_name(sym->st_shndx), ".text.head")) { + if (strcmp(sym_name(sym_strtab, sym), "__init_end") && + strcmp(sym_name(sym_strtab, sym), "KERNEL_TEXT_OFFSET")) + continue; + } + if (!strcmp(sec_name(sym->st_shndx), ".text")) + continue; +#endif if (r_type == R_386_PC32) { /* PC relative relocations don't need to be adjusted */ } @@ -631,6 +692,7 @@ int main(int argc, char **argv) fname, strerror(errno)); } read_ehdr(fp); + read_phdrs(fp); read_shdrs(fp); read_strtabs(fp); read_symtabs(fp); diff -urNp linux-2.6.28.8/arch/x86/boot/cpucheck.c linux-2.6.28.8/arch/x86/boot/cpucheck.c --- linux-2.6.28.8/arch/x86/boot/cpucheck.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/cpucheck.c 2009-02-21 09:37:48.000000000 -0500 @@ -74,7 +74,7 @@ static int has_fpu(void) u16 fcw = -1, fsw = -1; u32 cr0; - asm("movl %%cr0,%0" : "=r" (cr0)); + asm volatile("movl %%cr0,%0" : "=r" (cr0)); if (cr0 & (X86_CR0_EM|X86_CR0_TS)) { cr0 &= ~(X86_CR0_EM|X86_CR0_TS); asm volatile("movl %0,%%cr0" : : "r" (cr0)); @@ -90,7 +90,7 @@ static int has_eflag(u32 mask) { u32 f0, f1; - asm("pushfl ; " + asm volatile("pushfl ; " "pushfl ; " "popl %0 ; " "movl %0,%1 ; " @@ -115,7 +115,7 @@ static void get_flags(void) set_bit(X86_FEATURE_FPU, cpu.flags); if (has_eflag(X86_EFLAGS_ID)) { - asm("cpuid" + asm volatile("cpuid" : "=a" (max_intel_level), "=b" (cpu_vendor[0]), "=d" (cpu_vendor[1]), @@ -124,7 +124,7 @@ static void get_flags(void) if (max_intel_level >= 0x00000001 && max_intel_level <= 0x0000ffff) { - asm("cpuid" + asm volatile("cpuid" : "=a" (tfms), "=c" (cpu.flags[4]), "=d" (cpu.flags[0]) @@ -136,7 +136,7 @@ static void get_flags(void) cpu.model += ((tfms >> 16) & 0xf) << 4; } - asm("cpuid" + asm volatile("cpuid" : "=a" (max_amd_level) : "a" (0x80000000) : "ebx", "ecx", "edx"); @@ -144,7 +144,7 @@ static void get_flags(void) if (max_amd_level >= 0x80000001 && max_amd_level <= 0x8000ffff) { u32 eax = 0x80000001; - asm("cpuid" + asm volatile("cpuid" : "+a" (eax), "=c" (cpu.flags[6]), "=d" (cpu.flags[1]) @@ -203,9 +203,9 @@ int check_cpu(int *cpu_level_ptr, int *r u32 ecx = MSR_K7_HWCR; u32 eax, edx; - asm("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); + asm volatile("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); eax &= ~(1 << 15); - asm("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); + asm volatile("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); get_flags(); /* Make sure it really did something */ err = check_flags(); @@ -218,9 +218,9 @@ int check_cpu(int *cpu_level_ptr, int *r u32 ecx = MSR_VIA_FCR; u32 eax, edx; - asm("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); + asm volatile("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); eax |= (1<<1)|(1<<7); - asm("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); + asm volatile("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); set_bit(X86_FEATURE_CX8, cpu.flags); err = check_flags(); @@ -231,12 +231,12 @@ int check_cpu(int *cpu_level_ptr, int *r u32 eax, edx; u32 level = 1; - asm("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); - asm("wrmsr" : : "a" (~0), "d" (edx), "c" (ecx)); - asm("cpuid" + asm volatile("rdmsr" : "=a" (eax), "=d" (edx) : "c" (ecx)); + asm volatile("wrmsr" : : "a" (~0), "d" (edx), "c" (ecx)); + asm volatile("cpuid" : "+a" (level), "=d" (cpu.flags[0]) : : "ecx", "ebx"); - asm("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); + asm volatile("wrmsr" : : "a" (eax), "d" (edx), "c" (ecx)); err = check_flags(); } diff -urNp linux-2.6.28.8/arch/x86/boot/edd.c linux-2.6.28.8/arch/x86/boot/edd.c --- linux-2.6.28.8/arch/x86/boot/edd.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/edd.c 2009-02-21 09:37:48.000000000 -0500 @@ -81,7 +81,7 @@ static int get_edd_info(u8 devno, struct ax = 0x4100; bx = EDDMAGIC1; dx = devno; - asm("pushfl; stc; int $0x13; setc %%al; popfl" + asm volatile("pushfl; stc; int $0x13; setc %%al; popfl" : "+a" (ax), "+b" (bx), "=c" (cx), "+d" (dx) : : "esi", "edi"); @@ -100,7 +100,7 @@ static int get_edd_info(u8 devno, struct ei->params.length = sizeof(ei->params); ax = 0x4800; dx = devno; - asm("pushfl; int $0x13; popfl" + asm volatile("pushfl; int $0x13; popfl" : "+a" (ax), "+d" (dx), "=m" (ei->params) : "S" (&ei->params) : "ebx", "ecx", "edi"); @@ -111,7 +111,7 @@ static int get_edd_info(u8 devno, struct ax = 0x0800; dx = devno; di = 0; - asm("pushw %%es; " + asm volatile("pushw %%es; " "movw %%di,%%es; " "pushfl; stc; int $0x13; setc %%al; popfl; " "popw %%es" diff -urNp linux-2.6.28.8/arch/x86/boot/main.c linux-2.6.28.8/arch/x86/boot/main.c --- linux-2.6.28.8/arch/x86/boot/main.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/main.c 2009-02-21 09:37:48.000000000 -0500 @@ -78,7 +78,7 @@ static void query_ist(void) if (cpu.level < 6) return; - asm("int $0x15" + asm volatile("int $0x15" : "=a" (boot_params.ist_info.signature), "=b" (boot_params.ist_info.command), "=c" (boot_params.ist_info.event), diff -urNp linux-2.6.28.8/arch/x86/boot/mca.c linux-2.6.28.8/arch/x86/boot/mca.c --- linux-2.6.28.8/arch/x86/boot/mca.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/mca.c 2009-02-21 09:37:48.000000000 -0500 @@ -19,7 +19,7 @@ int query_mca(void) u8 err; u16 es, bx, len; - asm("pushw %%es ; " + asm volatile("pushw %%es ; " "int $0x15 ; " "setc %0 ; " "movw %%es, %1 ; " diff -urNp linux-2.6.28.8/arch/x86/boot/memory.c linux-2.6.28.8/arch/x86/boot/memory.c --- linux-2.6.28.8/arch/x86/boot/memory.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/memory.c 2009-02-21 09:37:48.000000000 -0500 @@ -30,7 +30,7 @@ static int detect_memory_e820(void) /* Important: %edx is clobbered by some BIOSes, so it must be either used for the error output or explicitly marked clobbered. */ - asm("int $0x15; setc %0" + asm volatile("int $0x15; setc %0" : "=d" (err), "+b" (next), "=a" (id), "+c" (size), "=m" (*desc) : "D" (desc), "d" (SMAP), "a" (0xe820)); @@ -65,7 +65,7 @@ static int detect_memory_e801(void) bx = cx = dx = 0; ax = 0xe801; - asm("stc; int $0x15; setc %0" + asm volatile("stc; int $0x15; setc %0" : "=m" (err), "+a" (ax), "+b" (bx), "+c" (cx), "+d" (dx)); if (err) @@ -95,7 +95,7 @@ static int detect_memory_88(void) u8 err; ax = 0x8800; - asm("stc; int $0x15; setc %0" : "=bcdm" (err), "+a" (ax)); + asm volatile("stc; int $0x15; setc %0" : "=bcdm" (err), "+a" (ax)); boot_params.screen_info.ext_mem_k = ax; diff -urNp linux-2.6.28.8/arch/x86/boot/video.c linux-2.6.28.8/arch/x86/boot/video.c --- linux-2.6.28.8/arch/x86/boot/video.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/video.c 2009-02-21 09:37:48.000000000 -0500 @@ -23,7 +23,7 @@ static void store_cursor_position(void) ax = 0x0300; bx = 0; - asm(INT10 + asm volatile(INT10 : "=d" (curpos), "+a" (ax), "+b" (bx) : : "ecx", "esi", "edi"); @@ -38,7 +38,7 @@ static void store_video_mode(void) /* N.B.: the saving of the video page here is a bit silly, since we pretty much assume page 0 everywhere. */ ax = 0x0f00; - asm(INT10 + asm volatile(INT10 : "+a" (ax), "=b" (page) : : "ecx", "edx", "esi", "edi"); diff -urNp linux-2.6.28.8/arch/x86/boot/video-vesa.c linux-2.6.28.8/arch/x86/boot/video-vesa.c --- linux-2.6.28.8/arch/x86/boot/video-vesa.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/video-vesa.c 2009-02-21 09:37:48.000000000 -0500 @@ -41,7 +41,7 @@ static int vesa_probe(void) ax = 0x4f00; di = (size_t)&vginfo; - asm(INT10 + asm volatile(INT10 : "+a" (ax), "+D" (di), "=m" (vginfo) : : "ebx", "ecx", "edx", "esi"); @@ -68,7 +68,7 @@ static int vesa_probe(void) ax = 0x4f01; cx = mode; di = (size_t)&vminfo; - asm(INT10 + asm volatile(INT10 : "+a" (ax), "+c" (cx), "+D" (di), "=m" (vminfo) : : "ebx", "edx", "esi"); @@ -120,7 +120,7 @@ static int vesa_set_mode(struct mode_inf ax = 0x4f01; cx = vesa_mode; di = (size_t)&vminfo; - asm(INT10 + asm volatile(INT10 : "+a" (ax), "+c" (cx), "+D" (di), "=m" (vminfo) : : "ebx", "edx", "esi"); @@ -202,19 +202,20 @@ static void vesa_dac_set_8bits(void) /* Save the VESA protected mode info */ static void vesa_store_pm_info(void) { - u16 ax, bx, di, es; + u16 ax, bx, cx, di, es; ax = 0x4f0a; - bx = di = 0; - asm("pushw %%es; "INT10"; movw %%es,%0; popw %%es" - : "=d" (es), "+a" (ax), "+b" (bx), "+D" (di) - : : "ecx", "esi"); + bx = cx = di = 0; + asm volatile("pushw %%es; "INT10"; movw %%es,%0; popw %%es" + : "=d" (es), "+a" (ax), "+b" (bx), "+c" (cx), "+D" (di) + : : "esi"); if (ax != 0x004f) return; boot_params.screen_info.vesapm_seg = es; boot_params.screen_info.vesapm_off = di; + boot_params.screen_info.vesapm_size = cx; } /* @@ -268,7 +269,7 @@ void vesa_store_edid(void) /* Note: The VBE DDC spec is different from the main VESA spec; we genuinely have to assume all registers are destroyed here. */ - asm("pushw %%es; movw %2,%%es; "INT10"; popw %%es" + asm volatile("pushw %%es; movw %2,%%es; "INT10"; popw %%es" : "+a" (ax), "+b" (bx) : "c" (cx), "D" (di) : "esi"); @@ -284,7 +285,7 @@ void vesa_store_edid(void) cx = 0; /* Controller 0 */ dx = 0; /* EDID block number */ di =(size_t) &boot_params.edid_info; /* (ES:)Pointer to block */ - asm(INT10 + asm volatile(INT10 : "+a" (ax), "+b" (bx), "+d" (dx), "=m" (boot_params.edid_info) : "c" (cx), "D" (di) : "esi"); diff -urNp linux-2.6.28.8/arch/x86/boot/video-vga.c linux-2.6.28.8/arch/x86/boot/video-vga.c --- linux-2.6.28.8/arch/x86/boot/video-vga.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/video-vga.c 2009-02-21 09:37:48.000000000 -0500 @@ -225,7 +225,7 @@ static int vga_probe(void) }; u8 vga_flag; - asm(INT10 + asm volatile(INT10 : "=b" (ega_bx) : "a" (0x1200), "b" (0x10) /* Check EGA/VGA */ : "ecx", "edx", "esi", "edi"); @@ -237,7 +237,7 @@ static int vga_probe(void) /* If we have MDA/CGA/HGC then BL will be unchanged at 0x10 */ if ((u8)ega_bx != 0x10) { /* EGA/VGA */ - asm(INT10 + asm volatile(INT10 : "=a" (vga_flag) : "a" (0x1a00) : "ebx", "ecx", "edx", "esi", "edi"); diff -urNp linux-2.6.28.8/arch/x86/boot/voyager.c linux-2.6.28.8/arch/x86/boot/voyager.c --- linux-2.6.28.8/arch/x86/boot/voyager.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/boot/voyager.c 2009-02-21 09:37:48.000000000 -0500 @@ -23,7 +23,7 @@ int query_voyager(void) data_ptr[0] = 0xff; /* Flag on config not found(?) */ - asm("pushw %%es ; " + asm volatile("pushw %%es ; " "int $0x15 ; " "setc %0 ; " "movw %%es, %1 ; " diff -urNp linux-2.6.28.8/arch/x86/ia32/ia32_signal.c linux-2.6.28.8/arch/x86/ia32/ia32_signal.c --- linux-2.6.28.8/arch/x86/ia32/ia32_signal.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/ia32/ia32_signal.c 2009-02-21 09:37:48.000000000 -0500 @@ -518,6 +518,7 @@ int ia32_setup_rt_frame(int sig, struct __NR_ia32_rt_sigreturn, 0x80cd, 0, + 0 }; frame = get_sigframe(ka, regs, sizeof(*frame), &fpstate); diff -urNp linux-2.6.28.8/arch/x86/include/asm/alternative.h linux-2.6.28.8/arch/x86/include/asm/alternative.h --- linux-2.6.28.8/arch/x86/include/asm/alternative.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/alternative.h 2009-02-21 09:37:48.000000000 -0500 @@ -96,7 +96,7 @@ const unsigned char *const *find_nop_tab " .byte 662b-661b\n" /* sourcelen */ \ " .byte 664f-663f\n" /* replacementlen */ \ ".previous\n" \ - ".section .altinstr_replacement,\"ax\"\n" \ + ".section .altinstr_replacement,\"a\"\n" \ "663:\n\t" newinstr "\n664:\n" /* replacement */ \ ".previous" :: "i" (feature) : "memory") @@ -120,7 +120,7 @@ const unsigned char *const *find_nop_tab " .byte 662b-661b\n" /* sourcelen */ \ " .byte 664f-663f\n" /* replacementlen */ \ ".previous\n" \ - ".section .altinstr_replacement,\"ax\"\n" \ + ".section .altinstr_replacement,\"a\"\n" \ "663:\n\t" newinstr "\n664:\n" /* replacement */ \ ".previous" :: "i" (feature), ##input) @@ -135,7 +135,7 @@ const unsigned char *const *find_nop_tab " .byte 662b-661b\n" /* sourcelen */ \ " .byte 664f-663f\n" /* replacementlen */ \ ".previous\n" \ - ".section .altinstr_replacement,\"ax\"\n" \ + ".section .altinstr_replacement,\"a\"\n" \ "663:\n\t" newinstr "\n664:\n" /* replacement */ \ ".previous" : output : [feat] "i" (feature), ##input) diff -urNp linux-2.6.28.8/arch/x86/include/asm/atomic_32.h linux-2.6.28.8/arch/x86/include/asm/atomic_32.h --- linux-2.6.28.8/arch/x86/include/asm/atomic_32.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/atomic_32.h 2009-02-21 09:37:48.000000000 -0500 @@ -47,7 +47,29 @@ typedef struct { */ 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" + "into\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_t + * + * Atomically adds @i to @v. + */ +static inline void atomic_add_unchecked(int i, atomic_t *v) +{ + asm volatile(LOCK_PREFIX "addl %1,%0\n" : "+m" (v->counter) : "ir" (i)); } @@ -61,7 +83,15 @@ static inline void atomic_add(int i, ato */ 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" + "into\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+m" (v->counter) : "ir" (i)); } @@ -79,7 +109,16 @@ static inline int atomic_sub_and_test(in { unsigned char c; - asm volatile(LOCK_PREFIX "subl %2,%0; sete %1" + asm volatile(LOCK_PREFIX "subl %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "addl %2,%0\n" + "into\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "+m" (v->counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -93,7 +132,18 @@ static inline int atomic_sub_and_test(in */ 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 + "into\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "decl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (v->counter)); } @@ -105,7 +155,18 @@ static inline void atomic_inc(atomic_t * */ 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 + "into\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "incl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (v->counter)); } @@ -121,7 +182,19 @@ static inline int atomic_dec_and_test(at { unsigned char c; - asm volatile(LOCK_PREFIX "decl %0; sete %1" + asm volatile(LOCK_PREFIX "decl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "into\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "incl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "+m" (v->counter), "=qm" (c) : : "memory"); return c != 0; @@ -139,7 +212,19 @@ static inline int atomic_inc_and_test(at { unsigned char c; - asm volatile(LOCK_PREFIX "incl %0; sete %1" + asm volatile(LOCK_PREFIX "incl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "into\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "decl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "+m" (v->counter), "=qm" (c) : : "memory"); return c != 0; @@ -158,7 +243,16 @@ static inline int atomic_add_negative(in { unsigned char c; - asm volatile(LOCK_PREFIX "addl %2,%0; sets %1" + asm volatile(LOCK_PREFIX "addl %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "subl %2,%0\n" + "into\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sets %1\n" : "+m" (v->counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -181,7 +275,15 @@ static inline int atomic_add_return(int #endif /* Modern 486+ processor */ __i = i; - asm volatile(LOCK_PREFIX "xaddl %0, %1" + asm volatile(LOCK_PREFIX "xaddl %0, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "movl %0, %1\n" + "into\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+r" (i), "+m" (v->counter) : : "memory"); return i + __i; diff -urNp linux-2.6.28.8/arch/x86/include/asm/atomic_64.h linux-2.6.28.8/arch/x86/include/asm/atomic_64.h --- linux-2.6.28.8/arch/x86/include/asm/atomic_64.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/atomic_64.h 2009-02-21 09:37:48.000000000 -0500 @@ -48,7 +48,29 @@ typedef struct { */ 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), "m" (v->counter)); +} + +/** + * atomic_add_unchecked - add integer to atomic variable + * @i: integer value to add + * @v: pointer of type atomic_t + * + * Atomically adds @i to @v. + */ +static inline void atomic_add_unchecked(int i, atomic_t *v) +{ + asm volatile(LOCK_PREFIX "addl %1,%0\n" : "=m" (v->counter) : "ir" (i), "m" (v->counter)); } @@ -62,7 +84,15 @@ static inline void atomic_add(int i, ato */ 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), "m" (v->counter)); } @@ -80,7 +110,16 @@ static inline int atomic_sub_and_test(in { unsigned char c; - asm volatile(LOCK_PREFIX "subl %2,%0; sete %1" + asm volatile(LOCK_PREFIX "subl %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "addl %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "ir" (i), "m" (v->counter) : "memory"); return c; @@ -94,7 +133,19 @@ static inline int atomic_sub_and_test(in */ 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" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "decl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "=m" (v->counter) : "m" (v->counter)); } @@ -107,7 +158,19 @@ static inline void atomic_inc(atomic_t * */ 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" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "incl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "=m" (v->counter) : "m" (v->counter)); } @@ -124,7 +187,20 @@ static inline int atomic_dec_and_test(at { unsigned char c; - asm volatile(LOCK_PREFIX "decl %0; sete %1" + asm volatile(LOCK_PREFIX "decl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "incl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "m" (v->counter) : "memory"); return c != 0; @@ -142,7 +218,20 @@ static inline int atomic_inc_and_test(at { unsigned char c; - asm volatile(LOCK_PREFIX "incl %0; sete %1" + asm volatile(LOCK_PREFIX "incl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "decl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "m" (v->counter) : "memory"); return c != 0; @@ -161,7 +250,16 @@ static inline int atomic_add_negative(in { unsigned char c; - asm volatile(LOCK_PREFIX "addl %2,%0; sets %1" + asm volatile(LOCK_PREFIX "addl %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "subl %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sets %1\n" : "=m" (v->counter), "=qm" (c) : "ir" (i), "m" (v->counter) : "memory"); return c; @@ -177,7 +275,15 @@ static inline int atomic_add_negative(in static inline int atomic_add_return(int i, atomic_t *v) { int __i = i; - asm volatile(LOCK_PREFIX "xaddl %0, %1" + asm volatile(LOCK_PREFIX "xaddl %0, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "movl %0, %1\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+r" (i), "+m" (v->counter) : : "memory"); return i + __i; @@ -226,7 +332,15 @@ typedef struct { */ static inline void atomic64_add(long i, atomic64_t *v) { - asm volatile(LOCK_PREFIX "addq %1,%0" + 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)); } @@ -240,7 +354,15 @@ static inline void atomic64_add(long i, */ 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)); } @@ -258,7 +380,16 @@ static inline int atomic64_sub_and_test( { unsigned char c; - asm volatile(LOCK_PREFIX "subq %2,%0; sete %1" + asm volatile(LOCK_PREFIX "subq %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "addq %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "er" (i), "m" (v->counter) : "memory"); return c; @@ -272,7 +403,19 @@ static inline int atomic64_sub_and_test( */ static inline void atomic64_inc(atomic64_t *v) { - asm volatile(LOCK_PREFIX "incq %0" + asm volatile(LOCK_PREFIX "incq %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "decq %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "=m" (v->counter) : "m" (v->counter)); } @@ -285,7 +428,19 @@ static inline void atomic64_inc(atomic64 */ 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" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "incq %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "=m" (v->counter) : "m" (v->counter)); } @@ -302,7 +457,20 @@ static inline int atomic64_dec_and_test( { unsigned char c; - asm volatile(LOCK_PREFIX "decq %0; sete %1" + asm volatile(LOCK_PREFIX "decq %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "incq %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "m" (v->counter) : "memory"); return c != 0; @@ -320,7 +488,20 @@ static inline int atomic64_inc_and_test( { unsigned char c; - asm volatile(LOCK_PREFIX "incq %0; sete %1" + asm volatile(LOCK_PREFIX "incq %0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "int $4\n0:\n" + ".pushsection .fixup,\"ax\"\n" + "1: \n" + LOCK_PREFIX "decq %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "=m" (v->counter), "=qm" (c) : "m" (v->counter) : "memory"); return c != 0; @@ -339,7 +520,16 @@ static inline int atomic64_add_negative( { unsigned char c; - asm volatile(LOCK_PREFIX "addq %2,%0; sets %1" + asm volatile(LOCK_PREFIX "addq %2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + LOCK_PREFIX "subq %2,%0\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + + "sets %1\n" : "=m" (v->counter), "=qm" (c) : "er" (i), "m" (v->counter) : "memory"); return c; @@ -355,7 +545,15 @@ static inline int atomic64_add_negative( static inline long atomic64_add_return(long i, atomic64_t *v) { long __i = i; - asm volatile(LOCK_PREFIX "xaddq %0, %1;" + asm volatile(LOCK_PREFIX "xaddq %0, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT + "jno 0f\n" + "movq %0, %1\n" + "int $4\n0:\n" + _ASM_EXTABLE(0b, 0b) +#endif + : "+r" (i), "+m" (v->counter) : : "memory"); return i + __i; diff -urNp linux-2.6.28.8/arch/x86/include/asm/boot.h linux-2.6.28.8/arch/x86/include/asm/boot.h --- linux-2.6.28.8/arch/x86/include/asm/boot.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/boot.h 2009-02-21 09:37:48.000000000 -0500 @@ -11,10 +11,15 @@ #define ASK_VGA 0xfffd /* ask for it at bootup */ /* 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 + #ifdef CONFIG_X86_64 #define BOOT_HEAP_SIZE 0x7000 #define BOOT_STACK_SIZE 0x4000 diff -urNp linux-2.6.28.8/arch/x86/include/asm/cache.h linux-2.6.28.8/arch/x86/include/asm/cache.h --- linux-2.6.28.8/arch/x86/include/asm/cache.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/cache.h 2009-02-21 09:37:48.000000000 -0500 @@ -6,6 +6,7 @@ #define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) #define __read_mostly __attribute__((__section__(".data.read_mostly"))) +#define __read_only __attribute__((__section__(".data.read_only"))) #ifdef CONFIG_X86_VSMP /* vSMP Internode cacheline shift */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/checksum_32.h linux-2.6.28.8/arch/x86/include/asm/checksum_32.h --- linux-2.6.28.8/arch/x86/include/asm/checksum_32.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/checksum_32.h 2009-02-21 09:37:48.000000000 -0500 @@ -31,6 +31,14 @@ asmlinkage __wsum csum_partial_copy_gene int len, __wsum sum, int *src_err_ptr, int *dst_err_ptr); +asmlinkage __wsum csum_partial_copy_generic_to_user(const void *src, void *dst, + int len, __wsum sum, + int *src_err_ptr, int *dst_err_ptr); + +asmlinkage __wsum csum_partial_copy_generic_from_user(const void *src, void *dst, + int len, __wsum sum, + int *src_err_ptr, int *dst_err_ptr); + /* * Note: when you get a NULL pointer exception here this means someone * passed in an incorrect kernel address to one of these functions. @@ -50,7 +58,7 @@ static inline __wsum csum_partial_copy_f int *err_ptr) { might_sleep(); - return csum_partial_copy_generic((__force void *)src, dst, + return csum_partial_copy_generic_from_user((__force void *)src, dst, len, sum, err_ptr, NULL); } @@ -177,7 +185,7 @@ static inline __wsum csum_and_copy_to_us { might_sleep(); if (access_ok(VERIFY_WRITE, dst, len)) - return csum_partial_copy_generic(src, (__force void *)dst, + return csum_partial_copy_generic_to_user(src, (__force void *)dst, len, sum, NULL, err_ptr); if (len) diff -urNp linux-2.6.28.8/arch/x86/include/asm/cpufeature.h linux-2.6.28.8/arch/x86/include/asm/cpufeature.h --- linux-2.6.28.8/arch/x86/include/asm/cpufeature.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/cpufeature.h 2009-02-21 09:37:48.000000000 -0500 @@ -80,7 +80,6 @@ #define X86_FEATURE_UP (3*32+ 9) /* smp kernel running on up */ #define X86_FEATURE_FXSAVE_LEAK (3*32+10) /* "" FXSAVE leaks FOP/FIP/FOP */ #define X86_FEATURE_ARCH_PERFMON (3*32+11) /* Intel Architectural PerfMon */ -#define X86_FEATURE_NOPL (3*32+20) /* The NOPL (0F 1F) instructions */ #define X86_FEATURE_PEBS (3*32+12) /* Precise-Event Based Sampling */ #define X86_FEATURE_BTS (3*32+13) /* Branch Trace Store */ #define X86_FEATURE_SYSCALL32 (3*32+14) /* "" syscall in ia32 userspace */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/desc.h linux-2.6.28.8/arch/x86/include/asm/desc.h --- linux-2.6.28.8/arch/x86/include/asm/desc.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/desc.h 2009-02-21 09:37:48.000000000 -0500 @@ -16,6 +16,7 @@ static inline void fill_ldt(struct desc_ desc->base1 = (info->base_addr & 0x00ff0000) >> 16; 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; desc->p = info->seg_not_present ^ 1; @@ -32,16 +33,12 @@ static inline void fill_ldt(struct desc_ } extern struct desc_ptr idt_descr; -extern gate_desc idt_table[]; - -struct gdt_page { - struct desc_struct gdt[GDT_ENTRIES]; -} __attribute__((aligned(PAGE_SIZE))); -DECLARE_PER_CPU(struct gdt_page, gdt_page); +extern gate_desc idt_table[256]; +extern struct desc_struct cpu_gdt_table[NR_CPUS][PAGE_SIZE / sizeof(struct desc_struct)]; static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu) { - return per_cpu(gdt_page, cpu).gdt; + return cpu_gdt_table[cpu]; } #ifdef CONFIG_X86_64 @@ -115,19 +112,48 @@ static inline void paravirt_free_ldt(str static inline void native_write_idt_entry(gate_desc *idt, int entry, const gate_desc *gate) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + memcpy(&idt[entry], gate, sizeof(*gate)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } static inline void native_write_ldt_entry(struct desc_struct *ldt, int entry, const void *desc) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + memcpy(&ldt[entry], desc, 8); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } static inline void native_write_gdt_entry(struct desc_struct *gdt, int entry, const void *desc, int type) { unsigned int size; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + switch (type) { case DESC_TSS: size = sizeof(tss_desc); @@ -139,7 +165,17 @@ static inline void native_write_gdt_entr size = sizeof(struct desc_struct); break; } + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + memcpy(&gdt[entry], desc, size); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } static inline void pack_descriptor(struct desc_struct *desc, unsigned long base, @@ -211,7 +247,19 @@ static inline void native_set_ldt(const static inline void native_load_tr_desc(void) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + asm volatile("ltr %w0"::"q" (GDT_ENTRY_TSS*8)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } static inline void native_load_gdt(const struct desc_ptr *dtr) @@ -246,8 +294,19 @@ static inline void native_load_tls(struc unsigned int i; struct desc_struct *gdt = get_cpu_gdt_table(cpu); +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + for (i = 0; i < GDT_ENTRY_TLS_ENTRIES; i++) gdt[GDT_ENTRY_TLS_MIN + i] = t->tls_array[i]; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } #define _LDT_empty(info) \ @@ -381,6 +440,18 @@ static inline void set_system_intr_gate_ _set_gate(n, GATE_INTERRUPT, addr, 0x3, ist, __KERNEL_CS); } +#ifdef CONFIG_X86_32 +static inline void set_user_cs(unsigned long base, unsigned long limit, int cpu) +{ + struct desc_struct d; + + if (likely(limit)) + limit = (limit - 1UL) >> PAGE_SHIFT; + pack_descriptor(&d, base, limit, 0xFB, 0xC); + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_DEFAULT_USER_CS, &d, DESCTYPE_S); +} +#endif + #else /* * GET_DESC_BASE reads the descriptor base of the specified segment. diff -urNp linux-2.6.28.8/arch/x86/include/asm/e820.h linux-2.6.28.8/arch/x86/include/asm/e820.h --- linux-2.6.28.8/arch/x86/include/asm/e820.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/e820.h 2009-02-21 09:37:48.000000000 -0500 @@ -134,7 +134,7 @@ extern char *memory_setup(void); #define ISA_END_ADDRESS 0x100000 #define is_ISA_range(s, e) ((s) >= ISA_START_ADDRESS && (e) < ISA_END_ADDRESS) -#define BIOS_BEGIN 0x000a0000 +#define BIOS_BEGIN 0x000c0000 #define BIOS_END 0x00100000 #ifdef __KERNEL__ diff -urNp linux-2.6.28.8/arch/x86/include/asm/elf.h linux-2.6.28.8/arch/x86/include/asm/elf.h --- linux-2.6.28.8/arch/x86/include/asm/elf.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/elf.h 2009-02-21 09:37:48.000000000 -0500 @@ -252,7 +252,25 @@ extern int force_personality32; the loader. We need to make sure that it is out of the way of the program that it will "exec", and that there is sufficient room for the brk. */ +#ifdef CONFIG_PAX_SEGMEXEC +#define ELF_ET_DYN_BASE ((current->mm->pax_flags & MF_PAX_SEGMEXEC) ? SEGMEXEC_TASK_SIZE/3*2 : TASK_SIZE/3*2) +#else #define ELF_ET_DYN_BASE (TASK_SIZE / 3 * 2) +#endif + +#ifdef CONFIG_PAX_ASLR +#ifdef CONFIG_X86_32 +#define PAX_ELF_ET_DYN_BASE 0x10000000UL + +#define PAX_DELTA_MMAP_LEN (current->mm->pax_flags & MF_PAX_SEGMEXEC ? 15 : 16) +#define PAX_DELTA_STACK_LEN (current->mm->pax_flags & MF_PAX_SEGMEXEC ? 15 : 16) +#else +#define PAX_ELF_ET_DYN_BASE 0x400000UL + +#define PAX_DELTA_MMAP_LEN ((test_thread_flag(TIF_IA32)) ? 16 : 32) +#define PAX_DELTA_STACK_LEN ((test_thread_flag(TIF_IA32)) ? 16 : 32) +#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, @@ -304,8 +322,7 @@ do { \ #define ARCH_DLINFO \ do { \ if (vdso_enabled) \ - NEW_AUX_ENT(AT_SYSINFO_EHDR, \ - (unsigned long)current->mm->context.vdso); \ + NEW_AUX_ENT(AT_SYSINFO_EHDR, current->mm->context.vdso);\ } while (0) #define AT_SYSINFO 32 @@ -316,7 +333,7 @@ do { \ #endif /* !CONFIG_X86_32 */ -#define VDSO_CURRENT_BASE ((unsigned long)current->mm->context.vdso) +#define VDSO_CURRENT_BASE (current->mm->context.vdso) #define VDSO_ENTRY \ ((unsigned long)VDSO32_SYMBOL(VDSO_CURRENT_BASE, vsyscall)) @@ -330,7 +347,4 @@ extern int arch_setup_additional_pages(s extern int syscall32_setup_pages(struct linux_binprm *, int exstack); #define compat_arch_setup_additional_pages syscall32_setup_pages -extern unsigned long arch_randomize_brk(struct mm_struct *mm); -#define arch_randomize_brk arch_randomize_brk - #endif /* _ASM_X86_ELF_H */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/futex.h linux-2.6.28.8/arch/x86/include/asm/futex.h --- linux-2.6.28.8/arch/x86/include/asm/futex.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/futex.h 2009-02-21 09:37:48.000000000 -0500 @@ -11,6 +11,40 @@ #include #include +#ifdef CONFIG_X86_32 +#define __futex_atomic_op1(insn, ret, oldval, uaddr, oparg) \ + asm volatile( \ + "movw\t%w6, %%ds\n" \ + "1:\t" insn "\n" \ + "2:\tpushl\t%%ss\n" \ + "\tpopl\t%%ds\n" \ + "\t.section .fixup,\"ax\"\n" \ + "3:\tmov\t%3, %1\n" \ + "\tjmp\t2b\n" \ + "\t.previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r" (oldval), "=r" (ret), "+m" (*uaddr) \ + : "i" (-EFAULT), "0" (oparg), "1" (0), "r" (__USER_DS)) + +#define __futex_atomic_op2(insn, ret, oldval, uaddr, oparg) \ + asm volatile("movw\t%w7, %%es\n" \ + "1:\tmovl\t%%es:%2, %0\n" \ + "\tmovl\t%0, %3\n" \ + "\t" insn "\n" \ + "2:\t" LOCK_PREFIX "cmpxchgl %3, %%es:%2\n"\ + "\tjnz\t1b\n" \ + "3:\tpushl\t%%ss\n" \ + "\tpopl\t%%es\n" \ + "\t.section .fixup,\"ax\"\n" \ + "4:\tmov\t%5, %1\n" \ + "\tjmp\t3b\n" \ + "\t.previous\n" \ + _ASM_EXTABLE(1b, 4b) \ + _ASM_EXTABLE(2b, 4b) \ + : "=&a" (oldval), "=&r" (ret), \ + "+m" (*uaddr), "=&r" (tem) \ + : "r" (oparg), "i" (-EFAULT), "1" (0), "r" (__USER_DS)) +#else #define __futex_atomic_op1(insn, ret, oldval, uaddr, oparg) \ asm volatile("1:\t" insn "\n" \ "2:\t.section .fixup,\"ax\"\n" \ @@ -36,8 +70,9 @@ : "=&a" (oldval), "=&r" (ret), \ "+m" (*uaddr), "=&r" (tem) \ : "r" (oparg), "i" (-EFAULT), "1" (0)) +#endif -static inline int futex_atomic_op_inuser(int encoded_op, int __user *uaddr) +static inline int futex_atomic_op_inuser(int encoded_op, u32 __user *uaddr) { int op = (encoded_op >> 28) & 7; int cmp = (encoded_op >> 24) & 15; @@ -61,11 +96,20 @@ static inline int futex_atomic_op_inuser switch (op) { case FUTEX_OP_SET: +#ifdef CONFIG_X86_32 + __futex_atomic_op1("xchgl %0, %%ds:%2", ret, oldval, uaddr, oparg); +#else __futex_atomic_op1("xchgl %0, %2", ret, oldval, uaddr, oparg); +#endif break; case FUTEX_OP_ADD: +#ifdef CONFIG_X86_32 + __futex_atomic_op1(LOCK_PREFIX "xaddl %0, %%ds:%2", ret, oldval, + uaddr, oparg); +#else __futex_atomic_op1(LOCK_PREFIX "xaddl %0, %2", ret, oldval, uaddr, oparg); +#endif break; case FUTEX_OP_OR: __futex_atomic_op2("orl %4, %3", ret, oldval, uaddr, oparg); @@ -109,7 +153,7 @@ static inline int futex_atomic_op_inuser return ret; } -static inline int futex_atomic_cmpxchg_inatomic(int __user *uaddr, int oldval, +static inline int futex_atomic_cmpxchg_inatomic(u32 __user *uaddr, int oldval, int newval) { @@ -122,14 +166,27 @@ static inline int futex_atomic_cmpxchg_i if (!access_ok(VERIFY_WRITE, uaddr, sizeof(int))) return -EFAULT; - asm volatile("1:\t" LOCK_PREFIX "cmpxchgl %3, %1\n" + asm volatile( +#ifdef CONFIG_X86_32 + "\tmovw %w5, %%ds\n" + "1:\t" LOCK_PREFIX "cmpxchgl %3, %1\n" + "2:\tpushl %%ss\n" + "\tpopl %%ds\n" + "\t.section .fixup, \"ax\"\n" +#else + "1:\t" LOCK_PREFIX "cmpxchgl %3, %1\n" "2:\t.section .fixup, \"ax\"\n" +#endif "3:\tmov %2, %0\n" "\tjmp 2b\n" "\t.previous\n" _ASM_EXTABLE(1b, 3b) : "=a" (oldval), "+m" (*uaddr) +#ifdef CONFIG_X86_32 + : "i" (-EFAULT), "r" (newval), "0" (oldval), "r" (__USER_DS) +#else : "i" (-EFAULT), "r" (newval), "0" (oldval) +#endif : "memory" ); diff -urNp linux-2.6.28.8/arch/x86/include/asm/i387.h linux-2.6.28.8/arch/x86/include/asm/i387.h --- linux-2.6.28.8/arch/x86/include/asm/i387.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/i387.h 2009-02-21 09:37:48.000000000 -0500 @@ -197,13 +197,8 @@ static inline void restore_fpu(struct ta } /* We need a safe address that is cheap to find and that is already - in L1 during context switch. The best choices are unfortunately - different for UP and SMP */ -#ifdef CONFIG_SMP -#define safe_address (__per_cpu_offset[0]) -#else -#define safe_address (kstat_cpu(0).cpustat.user) -#endif + in L1 during context switch. */ +#define safe_address (init_tss[smp_processor_id()].x86_tss.sp0) /* * These must be called with preempt disabled diff -urNp linux-2.6.28.8/arch/x86/include/asm/io_64.h linux-2.6.28.8/arch/x86/include/asm/io_64.h --- linux-2.6.28.8/arch/x86/include/asm/io_64.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/io_64.h 2009-02-21 09:37:48.000000000 -0500 @@ -158,6 +158,17 @@ static inline void *phys_to_virt(unsigne } #endif +#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) < (1 << (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)) < (1 << (boot_cpu_data.x86_phys_bits - PAGE_SHIFT)) ? 1 : 0; +} + /* * Change "struct page" to physical address. */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/irqflags.h linux-2.6.28.8/arch/x86/include/asm/irqflags.h --- linux-2.6.28.8/arch/x86/include/asm/irqflags.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/irqflags.h 2009-02-21 09:37:48.000000000 -0500 @@ -141,6 +141,8 @@ static inline unsigned long __raw_local_ #define INTERRUPT_RETURN iret #define ENABLE_INTERRUPTS_SYSEXIT sti; sysexit #define GET_CR0_INTO_EAX movl %cr0, %eax +#define GET_CR0_INTO_EDX movl %cr0, %edx +#define SET_CR0_FROM_EDX movl %edx, %cr0 #endif diff -urNp linux-2.6.28.8/arch/x86/include/asm/kmap_types.h linux-2.6.28.8/arch/x86/include/asm/kmap_types.h --- linux-2.6.28.8/arch/x86/include/asm/kmap_types.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/kmap_types.h 2009-02-21 09:37:48.000000000 -0500 @@ -21,7 +21,8 @@ D(9) KM_IRQ0, D(10) KM_IRQ1, D(11) KM_SOFTIRQ0, D(12) KM_SOFTIRQ1, -D(13) KM_TYPE_NR +D(13) KM_CLEARPAGE, +D(14) KM_TYPE_NR }; #undef D diff -urNp linux-2.6.28.8/arch/x86/include/asm/kvm_host.h linux-2.6.28.8/arch/x86/include/asm/kvm_host.h --- linux-2.6.28.8/arch/x86/include/asm/kvm_host.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/kvm_host.h 2009-02-21 09:37:48.000000000 -0500 @@ -479,7 +479,7 @@ struct kvm_x86_ops { int (*get_tdp_level)(void); }; -extern struct kvm_x86_ops *kvm_x86_ops; +extern const struct kvm_x86_ops *kvm_x86_ops; int kvm_mmu_module_init(void); void kvm_mmu_module_exit(void); diff -urNp linux-2.6.28.8/arch/x86/include/asm/linkage.h linux-2.6.28.8/arch/x86/include/asm/linkage.h --- linux-2.6.28.8/arch/x86/include/asm/linkage.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/linkage.h 2009-02-21 09:37:48.000000000 -0500 @@ -7,6 +7,11 @@ #ifdef CONFIG_X86_64 #define __ALIGN .p2align 4,,15 #define __ALIGN_STR ".p2align 4,,15" +#else +#ifdef CONFIG_X86_ALIGNMENT_16 +#define __ALIGN .align 16,0x90 +#define __ALIGN_STR ".align 16,0x90" +#endif #endif #ifdef CONFIG_X86_32 @@ -52,10 +57,5 @@ #endif -#ifdef CONFIG_X86_ALIGNMENT_16 -#define __ALIGN .align 16,0x90 -#define __ALIGN_STR ".align 16,0x90" -#endif - #endif /* _ASM_X86_LINKAGE_H */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/local.h linux-2.6.28.8/arch/x86/include/asm/local.h --- linux-2.6.28.8/arch/x86/include/asm/local.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/local.h 2009-02-21 09:37:48.000000000 -0500 @@ -18,26 +18,90 @@ typedef struct { static inline void local_inc(local_t *l) { - asm volatile(_ASM_INC "%0" + asm volatile(_ASM_INC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_DEC "%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (l->a.counter)); } static inline void local_dec(local_t *l) { - asm volatile(_ASM_DEC "%0" + asm volatile(_ASM_DEC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_INC "%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (l->a.counter)); } static inline void local_add(long i, local_t *l) { - asm volatile(_ASM_ADD "%1,%0" + asm volatile(_ASM_ADD "%1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_SUB "%1,%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (l->a.counter) : "ir" (i)); } static inline void local_sub(long i, local_t *l) { - asm volatile(_ASM_SUB "%1,%0" + asm volatile(_ASM_SUB "%1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_ADD "%1,%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (l->a.counter) : "ir" (i)); } @@ -55,7 +119,24 @@ static inline int local_sub_and_test(lon { unsigned char c; - asm volatile(_ASM_SUB "%2,%0; sete %1" + asm volatile(_ASM_SUB "%2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_ADD "%2,%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "+m" (l->a.counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -73,7 +154,24 @@ static inline int local_dec_and_test(loc { unsigned char c; - asm volatile(_ASM_DEC "%0; sete %1" + asm volatile(_ASM_DEC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_INC "%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "+m" (l->a.counter), "=qm" (c) : : "memory"); return c != 0; @@ -91,7 +189,24 @@ static inline int local_inc_and_test(loc { unsigned char c; - asm volatile(_ASM_INC "%0; sete %1" + asm volatile(_ASM_INC "%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_DEC "%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sete %1\n" : "+m" (l->a.counter), "=qm" (c) : : "memory"); return c != 0; @@ -110,7 +225,24 @@ static inline int local_add_negative(lon { unsigned char c; - asm volatile(_ASM_ADD "%2,%0; sets %1" + asm volatile(_ASM_ADD "%2,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_SUB "%2,%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "sets %1\n" : "+m" (l->a.counter), "=qm" (c) : "ir" (i) : "memory"); return c; @@ -133,7 +265,23 @@ static inline long local_add_return(long #endif /* Modern 486+ processor */ __i = i; - asm volatile(_ASM_XADD "%0, %1;" + asm volatile(_ASM_XADD "%0, %1\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + _ASM_MOV "%0,%1\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+r" (i), "+m" (l->a.counter) : : "memory"); return i + __i; diff -urNp linux-2.6.28.8/arch/x86/include/asm/mach-default/apm.h linux-2.6.28.8/arch/x86/include/asm/mach-default/apm.h --- linux-2.6.28.8/arch/x86/include/asm/mach-default/apm.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/mach-default/apm.h 2009-02-21 09:37:48.000000000 -0500 @@ -34,7 +34,7 @@ static inline void apm_bios_call_asm(u32 __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_as __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 -urNp linux-2.6.28.8/arch/x86/include/asm/mman.h linux-2.6.28.8/arch/x86/include/asm/mman.h --- linux-2.6.28.8/arch/x86/include/asm/mman.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/mman.h 2009-02-21 09:37:48.000000000 -0500 @@ -17,4 +17,14 @@ #define MCL_CURRENT 1 /* lock all current mappings */ #define MCL_FUTURE 2 /* lock all future mappings */ +#ifdef __KERNEL__ +#ifndef __ASSEMBLY__ +#ifdef CONFIG_X86_32 +#define arch_mmap_check i386_mmap_check +int i386_mmap_check(unsigned long addr, unsigned long len, + unsigned long flags); +#endif +#endif +#endif + #endif /* _ASM_X86_MMAN_H */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/mmu_context_32.h linux-2.6.28.8/arch/x86/include/asm/mmu_context_32.h --- linux-2.6.28.8/arch/x86/include/asm/mmu_context_32.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/mmu_context_32.h 2009-02-21 09:37:48.000000000 -0500 @@ -33,6 +33,22 @@ static inline void switch_mm(struct mm_s */ if (unlikely(prev->context.ldt != next->context.ldt)) load_LDT_nolock(&next->context); + +#if defined(CONFIG_PAX_PAGEEXEC) && defined(CONFIG_SMP) + if (!nx_enabled) { + smp_mb__before_clear_bit(); + cpu_clear(cpu, prev->context.cpu_user_cs_mask); + smp_mb__after_clear_bit(); + cpu_set(cpu, next->context.cpu_user_cs_mask); + } +#endif + +#if defined(CONFIG_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); +#endif + } #ifdef CONFIG_SMP else { @@ -45,6 +61,19 @@ static inline void switch_mm(struct mm_s */ load_cr3(next->pgd); load_LDT_nolock(&next->context); + +#ifdef CONFIG_PAX_PAGEEXEC + if (!nx_enabled) + cpu_set(cpu, next->context.cpu_user_cs_mask); +#endif + +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) +#ifdef CONFIG_PAX_PAGEEXEC + if (!((next->pax_flags & MF_PAX_PAGEEXEC) && nx_enabled)) +#endif + set_user_cs(next->context.user_cs_base, next->context.user_cs_limit, cpu); +#endif + } } #endif diff -urNp linux-2.6.28.8/arch/x86/include/asm/mmu.h linux-2.6.28.8/arch/x86/include/asm/mmu.h --- linux-2.6.28.8/arch/x86/include/asm/mmu.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/mmu.h 2009-02-21 09:37:48.000000000 -0500 @@ -9,10 +9,23 @@ * we put the segment information here. */ typedef struct { - void *ldt; + struct desc_struct *ldt; int size; struct mutex lock; - void *vdso; + unsigned long vdso; + +#ifdef CONFIG_X86_32 +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) + unsigned long user_cs_base; + unsigned long user_cs_limit; + +#if defined(CONFIG_PAX_PAGEEXEC) && defined(CONFIG_SMP) + cpumask_t cpu_user_cs_mask; +#endif + +#endif +#endif + } mm_context_t; #ifdef CONFIG_SMP diff -urNp linux-2.6.28.8/arch/x86/include/asm/module.h linux-2.6.28.8/arch/x86/include/asm/module.h --- linux-2.6.28.8/arch/x86/include/asm/module.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/module.h 2009-02-21 09:37:48.000000000 -0500 @@ -74,7 +74,12 @@ struct mod_arch_specific {}; # else # define MODULE_STACKSIZE "" # endif -# define MODULE_ARCH_VERMAGIC MODULE_PROC_FAMILY MODULE_STACKSIZE +# ifdef CONFIG_GRKERNSEC +# define MODULE_GRSEC "GRSECURITY " +# else +# define MODULE_GRSEC "" +# endif +# define MODULE_ARCH_VERMAGIC MODULE_PROC_FAMILY MODULE_STACKSIZE MODULE_GRSEC #endif #endif /* _ASM_X86_MODULE_H */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/page_32.h linux-2.6.28.8/arch/x86/include/asm/page_32.h --- linux-2.6.28.8/arch/x86/include/asm/page_32.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/page_32.h 2009-02-21 09:37:48.000000000 -0500 @@ -13,6 +13,23 @@ */ #define __PAGE_OFFSET _AC(CONFIG_PAGE_OFFSET, UL) +#ifdef CONFIG_PAX_KERNEXEC +#ifndef __ASSEMBLY__ +extern unsigned char MODULES_VADDR[]; +extern unsigned char MODULES_END[]; +extern unsigned char KERNEL_TEXT_OFFSET[]; +#define ktla_ktva(addr) (addr + (unsigned long)KERNEL_TEXT_OFFSET) +#define ktva_ktla(addr) (addr - (unsigned long)KERNEL_TEXT_OFFSET) +#endif +#else +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) +#endif + +#ifdef CONFIG_PAX_PAGEEXEC +#define CONFIG_ARCH_TRACK_EXEC_LIMIT 1 +#endif + #ifdef CONFIG_4KSTACKS #define THREAD_ORDER 0 #else diff -urNp linux-2.6.28.8/arch/x86/include/asm/page_64.h linux-2.6.28.8/arch/x86/include/asm/page_64.h --- linux-2.6.28.8/arch/x86/include/asm/page_64.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/page_64.h 2009-02-21 09:37:48.000000000 -0500 @@ -49,6 +49,9 @@ #define __START_KERNEL (__START_KERNEL_map + __PHYSICAL_START) #define __START_KERNEL_map _AC(0xffffffff80000000, UL) +#define ktla_ktva(addr) (addr) +#define ktva_ktla(addr) (addr) + /* See Documentation/x86_64/mm.txt for a description of the memory map. */ #define __PHYSICAL_MASK_SHIFT 46 #define __VIRTUAL_MASK_SHIFT 48 @@ -101,5 +104,6 @@ extern void init_extra_mapping_wb(unsign #define pfn_valid(pfn) ((pfn) < max_pfn) #endif +#define nx_enabled (1) #endif /* _ASM_X86_PAGE_64_H */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/paravirt.h linux-2.6.28.8/arch/x86/include/asm/paravirt.h --- linux-2.6.28.8/arch/x86/include/asm/paravirt.h 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/paravirt.h 2009-03-07 10:29:51.000000000 -0500 @@ -1557,7 +1557,7 @@ static inline unsigned long __raw_local_ #define PV_RESTORE_REGS popl %edx; popl %ecx; popl %edi; popl %eax #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 \ diff -urNp linux-2.6.28.8/arch/x86/include/asm/pda.h linux-2.6.28.8/arch/x86/include/asm/pda.h --- linux-2.6.28.8/arch/x86/include/asm/pda.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/pda.h 2009-02-21 09:37:48.000000000 -0500 @@ -16,11 +16,9 @@ struct x8664_pda { unsigned long oldrsp; /* 24 user rsp for system call */ int irqcount; /* 32 Irq nesting counter. Starts -1 */ unsigned int cpunumber; /* 36 Logical CPU number */ -#ifdef CONFIG_CC_STACKPROTECTOR unsigned long stack_canary; /* 40 stack canary value */ /* gcc-ABI: this canary MUST be at offset 40!!! */ -#endif char *irqstackptr; short nodenumber; /* number of current node (32k max) */ short in_bootmem; /* pda lives in bootmem */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/percpu.h linux-2.6.28.8/arch/x86/include/asm/percpu.h --- linux-2.6.28.8/arch/x86/include/asm/percpu.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/percpu.h 2009-02-21 09:37:48.000000000 -0500 @@ -93,6 +93,12 @@ DECLARE_PER_CPU(struct x8664_pda, pda); #define __my_cpu_offset x86_read_percpu(this_cpu_off) +#include +#include +#define __per_cpu_offset __per_cpu_offset +extern unsigned long __per_cpu_offset[NR_CPUS]; +#define per_cpu_offset(x) (__per_cpu_offset[x] + (unsigned long)__per_cpu_start) + /* fs segment starts at (positive) offset == __per_cpu_offset[cpu] */ #define __percpu_seg "%%fs:" diff -urNp linux-2.6.28.8/arch/x86/include/asm/pgalloc.h linux-2.6.28.8/arch/x86/include/asm/pgalloc.h --- linux-2.6.28.8/arch/x86/include/asm/pgalloc.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/pgalloc.h 2009-02-21 09:37:48.000000000 -0500 @@ -52,7 +52,7 @@ static inline void pmd_populate_kernel(s pmd_t *pmd, pte_t *pte) { paravirt_alloc_pte(mm, __pa(pte) >> PAGE_SHIFT); - set_pmd(pmd, __pmd(__pa(pte) | _PAGE_TABLE)); + set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE)); } static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, diff -urNp linux-2.6.28.8/arch/x86/include/asm/pgtable-2level.h linux-2.6.28.8/arch/x86/include/asm/pgtable-2level.h --- linux-2.6.28.8/arch/x86/include/asm/pgtable-2level.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/pgtable-2level.h 2009-02-21 09:37:48.000000000 -0500 @@ -18,7 +18,19 @@ static inline void native_set_pte(pte_t static inline void native_set_pmd(pmd_t *pmdp, pmd_t pmd) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + *pmdp = pmd; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } static inline void native_set_pte_atomic(pte_t *ptep, pte_t pte) diff -urNp linux-2.6.28.8/arch/x86/include/asm/pgtable_32.h linux-2.6.28.8/arch/x86/include/asm/pgtable_32.h --- linux-2.6.28.8/arch/x86/include/asm/pgtable_32.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/pgtable_32.h 2009-02-21 09:37:48.000000000 -0500 @@ -25,8 +25,6 @@ struct mm_struct; struct vm_area_struct; -extern pgd_t swapper_pg_dir[1024]; - static inline void pgtable_cache_init(void) { } static inline void check_pgt_cache(void) { } void paging_init(void); @@ -46,6 +44,11 @@ extern void set_pmd_pfn(unsigned long, u # include #endif +extern pgd_t swapper_pg_dir[PTRS_PER_PGD]; +#ifdef CONFIG_X86_PAE +extern pmd_t swapper_pm_dir[PTRS_PER_PGD][PTRS_PER_PMD]; +#endif + #define PGDIR_SIZE (1UL << PGDIR_SHIFT) #define PGDIR_MASK (~(PGDIR_SIZE - 1)) @@ -83,7 +86,7 @@ extern void set_pmd_pfn(unsigned long, u #undef TEST_ACCESS_OK /* The boot page tables (all created as a single array) */ -extern unsigned long pg0[]; +extern pte_t pg0[]; #define pte_present(x) ((x).pte_low & (_PAGE_PRESENT | _PAGE_PROTNONE)) @@ -175,6 +178,9 @@ do { \ #endif /* !__ASSEMBLY__ */ +#define HAVE_ARCH_UNMAPPED_AREA +#define HAVE_ARCH_UNMAPPED_AREA_TOPDOWN + /* * kern_addr_valid() is (1) for FLATMEM and (0) for * SPARSEMEM and DISCONTIGMEM diff -urNp linux-2.6.28.8/arch/x86/include/asm/pgtable-3level.h linux-2.6.28.8/arch/x86/include/asm/pgtable-3level.h --- linux-2.6.28.8/arch/x86/include/asm/pgtable-3level.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/pgtable-3level.h 2009-02-21 09:37:48.000000000 -0500 @@ -70,12 +70,36 @@ static inline void native_set_pte_atomic static inline void native_set_pmd(pmd_t *pmdp, pmd_t pmd) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + set_64bit((unsigned long long *)(pmdp), native_pmd_val(pmd)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } static inline void native_set_pud(pud_t *pudp, pud_t pud) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + set_64bit((unsigned long long *)(pudp), native_pud_val(pud)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } /* diff -urNp linux-2.6.28.8/arch/x86/include/asm/pgtable_64.h linux-2.6.28.8/arch/x86/include/asm/pgtable_64.h --- linux-2.6.28.8/arch/x86/include/asm/pgtable_64.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/pgtable_64.h 2009-02-21 09:37:48.000000000 -0500 @@ -15,9 +15,12 @@ extern pud_t level3_kernel_pgt[512]; extern pud_t level3_ident_pgt[512]; +extern pud_t level3_vmalloc_pgt[512]; +extern pud_t level3_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*4]; +extern pte_t level1_fixmap_pgt[512]; extern pgd_t init_level4_pgt[]; #define swapper_pg_dir init_level4_pgt @@ -106,7 +109,19 @@ static inline pte_t native_ptep_get_and_ static inline void native_set_pmd(pmd_t *pmdp, pmd_t pmd) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + *pmdp = pmd; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } static inline void native_pmd_clear(pmd_t *pmd) @@ -158,17 +173,17 @@ static inline void native_pgd_clear(pgd_ static inline int pgd_bad(pgd_t pgd) { - return (pgd_val(pgd) & ~(PTE_PFN_MASK | _PAGE_USER)) != _KERNPG_TABLE; + return (pgd_val(pgd) & ~(PTE_PFN_MASK | _PAGE_USER | _PAGE_NX)) != _KERNPG_TABLE; } static inline int pud_bad(pud_t pud) { - return (pud_val(pud) & ~(PTE_PFN_MASK | _PAGE_USER)) != _KERNPG_TABLE; + return (pud_val(pud) & ~(PTE_PFN_MASK | _PAGE_USER | _PAGE_NX)) != _KERNPG_TABLE; } static inline int pmd_bad(pmd_t pmd) { - return (pmd_val(pmd) & ~(PTE_PFN_MASK | _PAGE_USER)) != _KERNPG_TABLE; + return (pmd_val(pmd) & ~(PTE_PFN_MASK | _PAGE_USER | _PAGE_NX)) != _KERNPG_TABLE; } #define pte_none(x) (!pte_val((x))) diff -urNp linux-2.6.28.8/arch/x86/include/asm/pgtable.h linux-2.6.28.8/arch/x86/include/asm/pgtable.h --- linux-2.6.28.8/arch/x86/include/asm/pgtable.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/pgtable.h 2009-02-21 09:37:48.000000000 -0500 @@ -14,12 +14,11 @@ #define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page */ #define _PAGE_BIT_PAT 7 /* on 4KB pages */ #define _PAGE_BIT_GLOBAL 8 /* Global TLB entry PPro+ */ -#define _PAGE_BIT_UNUSED1 9 /* available for programmer */ +#define _PAGE_BIT_SPECIAL 9 /* special mappings, no associated struct page */ #define _PAGE_BIT_IOMAP 10 /* flag used to indicate IO mapping */ #define _PAGE_BIT_UNUSED3 11 #define _PAGE_BIT_PAT_LARGE 12 /* On 2MB or 1GB pages */ -#define _PAGE_BIT_SPECIAL _PAGE_BIT_UNUSED1 -#define _PAGE_BIT_CPA_TEST _PAGE_BIT_UNUSED1 +#define _PAGE_BIT_CPA_TEST _PAGE_BIT_SPECIAL #define _PAGE_BIT_NX 63 /* No execute: only valid after cpuid check */ #define _PAGE_PRESENT (_AT(pteval_t, 1) << _PAGE_BIT_PRESENT) @@ -31,7 +30,6 @@ #define _PAGE_DIRTY (_AT(pteval_t, 1) << _PAGE_BIT_DIRTY) #define _PAGE_PSE (_AT(pteval_t, 1) << _PAGE_BIT_PSE) #define _PAGE_GLOBAL (_AT(pteval_t, 1) << _PAGE_BIT_GLOBAL) -#define _PAGE_UNUSED1 (_AT(pteval_t, 1) << _PAGE_BIT_UNUSED1) #define _PAGE_IOMAP (_AT(pteval_t, 1) << _PAGE_BIT_IOMAP) #define _PAGE_UNUSED3 (_AT(pteval_t, 1) << _PAGE_BIT_UNUSED3) #define _PAGE_PAT (_AT(pteval_t, 1) << _PAGE_BIT_PAT) @@ -43,7 +41,7 @@ #if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) #define _PAGE_NX (_AT(pteval_t, 1) << _PAGE_BIT_NX) #else -#define _PAGE_NX (_AT(pteval_t, 0)) +#define _PAGE_NX (_AT(pteval_t, 1) << _PAGE_BIT_UNUSED3) #endif /* If _PAGE_PRESENT is clear, we use these: */ @@ -83,6 +81,9 @@ #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) @@ -94,7 +95,7 @@ #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_NOCACHE (__PAGE_KERNEL_VSYSCALL | _PAGE_PCD | _PAGE_PWT) +#define __PAGE_KERNEL_VSYSCALL_NOCACHE (__PAGE_KERNEL_RO | _PAGE_PCD | _PAGE_PWT | _PAGE_USER) #define __PAGE_KERNEL_LARGE (__PAGE_KERNEL | _PAGE_PSE) #define __PAGE_KERNEL_LARGE_NOCACHE (__PAGE_KERNEL | _PAGE_CACHE_UC | _PAGE_PSE) #define __PAGE_KERNEL_LARGE_EXEC (__PAGE_KERNEL_EXEC | _PAGE_PSE) @@ -153,7 +154,7 @@ * bits are combined, this will alow user to access the high address mapped * VDSO in the presence of CONFIG_COMPAT_VDSO */ -#define PTE_IDENT_ATTR 0x003 /* PRESENT+RW */ +#define PTE_IDENT_ATTR 0x063 /* PRESENT+RW+DIRTY+ACCESSED */ #define PDE_IDENT_ATTR 0x067 /* PRESENT+RW+USER+DIRTY+ACCESSED */ #define PGD_IDENT_ATTR 0x001 /* PRESENT (no other attributes) */ #endif @@ -170,10 +171,17 @@ extern unsigned long empty_zero_page[PAG extern spinlock_t pgd_lock; extern struct list_head pgd_list; +extern pteval_t __supported_pte_mask; + /* * 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; @@ -242,9 +250,29 @@ static inline pte_t pte_wrprotect(pte_t return __pte(pte_val(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(pte_val(pte) & ~_PAGE_NX); +#ifdef CONFIG_X86_PAE + if (__supported_pte_mask & _PAGE_NX) + return __pte(pte_val(pte) & ~(pteval_t)_PAGE_NX); + else +#endif + return __pte(pte_val(pte) | _PAGE_USER); +} + +static inline pte_t pte_exprotect(pte_t pte) +{ +#ifdef CONFIG_X86_PAE + if (__supported_pte_mask & _PAGE_NX) + return __pte(pte_val(pte) | _PAGE_NX); + else +#endif + return __pte(pte_val(pte) & ~_PAGE_USER); } static inline pte_t pte_mkdirty(pte_t pte) @@ -287,8 +315,6 @@ static inline pte_t pte_mkspecial(pte_t return __pte(pte_val(pte) | _PAGE_SPECIAL); } -extern pteval_t __supported_pte_mask; - static inline pte_t pfn_pte(unsigned long page_nr, pgprot_t pgprot) { return __pte((((phys_addr_t)page_nr << PAGE_SHIFT) | @@ -552,7 +578,19 @@ static inline void ptep_set_wrprotect(st */ static inline void clone_pgd_range(pgd_t *dst, pgd_t *src, int count) { - memcpy(dst, src, count * sizeof(pgd_t)); + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + + memcpy(dst, src, count * sizeof(pgd_t)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } diff -urNp linux-2.6.28.8/arch/x86/include/asm/processor.h linux-2.6.28.8/arch/x86/include/asm/processor.h --- linux-2.6.28.8/arch/x86/include/asm/processor.h 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/processor.h 2009-03-07 10:29:51.000000000 -0500 @@ -271,7 +271,7 @@ struct tss_struct { } ____cacheline_aligned; -DECLARE_PER_CPU(struct tss_struct, init_tss); +extern struct tss_struct init_tss[NR_CPUS]; /* * Save the original ist values for checking stack pointers during debugging @@ -822,11 +822,20 @@ static inline void spin_lock_prefetch(co * User space process size: 3GB (default). */ #define TASK_SIZE PAGE_OFFSET + +#ifdef CONFIG_PAX_SEGMEXEC +#define SEGMEXEC_TASK_SIZE (TASK_SIZE / 2) +#endif + +#ifdef CONFIG_PAX_SEGMEXEC +#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, \ @@ -841,7 +850,7 @@ static inline void spin_lock_prefetch(co */ #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, \ @@ -852,11 +861,7 @@ static inline void spin_lock_prefetch(co 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) ((info)->task.thread.sp0) /* * The below -8 is to reserve 8 bytes on top of the ring0 stack. @@ -871,7 +876,7 @@ extern unsigned long thread_saved_pc(str #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; \ }) @@ -887,7 +892,7 @@ extern unsigned long thread_saved_pc(str * space during mmap's. */ #define IA32_PAGE_OFFSET ((current->personality & ADDR_LIMIT_3GB) ? \ - 0xc0000000 : 0xFFFFe000) + 0xc0000000 : 0xFFFFf000) #define TASK_SIZE (test_thread_flag(TIF_IA32) ? \ IA32_PAGE_OFFSET : TASK_SIZE64) @@ -924,6 +929,10 @@ extern void start_thread(struct pt_regs */ #define TASK_UNMAPPED_BASE (PAGE_ALIGN(TASK_SIZE / 3)) +#ifdef CONFIG_PAX_SEGMEXEC +#define SEGMEXEC_TASK_UNMAPPED_BASE (PAGE_ALIGN(SEGMEXEC_TASK_SIZE / 3)) +#endif + #define KSTK_EIP(task) (task_pt_regs(task)->ip) /* Get/set a process' ability to use the timestamp counter instruction */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/ptrace.h linux-2.6.28.8/arch/x86/include/asm/ptrace.h --- linux-2.6.28.8/arch/x86/include/asm/ptrace.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/ptrace.h 2009-02-21 09:37:48.000000000 -0500 @@ -187,28 +187,29 @@ static inline unsigned long regs_return_ } /* - * 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 } diff -urNp linux-2.6.28.8/arch/x86/include/asm/reboot.h linux-2.6.28.8/arch/x86/include/asm/reboot.h --- linux-2.6.28.8/arch/x86/include/asm/reboot.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/reboot.h 2009-02-21 09:37:48.000000000 -0500 @@ -16,6 +16,6 @@ extern struct machine_ops machine_ops; void native_machine_crash_shutdown(struct pt_regs *regs); void native_machine_shutdown(void); -void machine_real_restart(const unsigned char *code, int length); +void machine_real_restart(const unsigned char *code, unsigned int length); #endif /* _ASM_X86_REBOOT_H */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/rwsem.h linux-2.6.28.8/arch/x86/include/asm/rwsem.h --- linux-2.6.28.8/arch/x86/include/asm/rwsem.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/rwsem.h 2009-02-21 09:37:48.000000000 -0500 @@ -106,10 +106,26 @@ static inline void __down_read(struct rw { asm volatile("# beginning down_read\n\t" LOCK_PREFIX " incl (%%eax)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "decl (%%eax)\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + /* adds 0x00000001, returns the old value */ - " jns 1f\n" + " jns 2f\n" " call call_rwsem_down_read_failed\n" - "1:\n\t" + "2:\n\t" "# ending down_read\n\t" : "+m" (sem->count) : "a" (sem) @@ -124,13 +140,29 @@ static inline int __down_read_trylock(st __s32 result, tmp; asm volatile("# beginning __down_read_trylock\n\t" " movl %0,%1\n\t" - "1:\n\t" + "2:\n\t" " movl %1,%2\n\t" " addl %3,%2\n\t" - " jle 2f\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + "subl %3,%2\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + " jle 3f\n\t" LOCK_PREFIX " cmpxchgl %2,%0\n\t" - " jnz 1b\n\t" - "2:\n\t" + " jnz 2b\n\t" + "3:\n\t" "# ending __down_read_trylock\n\t" : "+m" (sem->count), "=&a" (result), "=&r" (tmp) : "i" (RWSEM_ACTIVE_READ_BIAS) @@ -148,12 +180,28 @@ static inline void __down_write_nested(s tmp = RWSEM_ACTIVE_WRITE_BIAS; asm volatile("# beginning down_write\n\t" LOCK_PREFIX " xadd %%edx,(%%eax)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + "movl %%edx,(%%eax)\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + /* subtract 0x0000ffff, returns the old value */ " testl %%edx,%%edx\n\t" /* was the count 0 before? */ - " jz 1f\n" + " jz 2f\n" " call call_rwsem_down_write_failed\n" - "1:\n" + "2:\n" "# ending down_write" : "+m" (sem->count), "=d" (tmp) : "a" (sem), "1" (tmp) @@ -186,10 +234,26 @@ static inline void __up_read(struct rw_s __s32 tmp = -RWSEM_ACTIVE_READ_BIAS; asm volatile("# beginning __up_read\n\t" LOCK_PREFIX " xadd %%edx,(%%eax)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + "movl %%edx,(%%eax)\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + /* subtracts 1, returns the old value */ - " jns 1f\n\t" + " jns 2f\n\t" " call call_rwsem_wake\n" - "1:\n" + "2:\n" "# ending __up_read\n" : "+m" (sem->count), "=d" (tmp) : "a" (sem), "1" (tmp) @@ -204,11 +268,27 @@ static inline void __up_write(struct rw_ asm volatile("# beginning __up_write\n\t" " movl %2,%%edx\n\t" LOCK_PREFIX " xaddl %%edx,(%%eax)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + "movl %%edx,(%%eax)\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + /* tries to transition 0xffff0001 -> 0x00000000 */ - " jz 1f\n" + " jz 2f\n" " call call_rwsem_wake\n" - "1:\n\t" + "2:\n\t" "# ending __up_write\n" : "+m" (sem->count) : "a" (sem), "i" (-RWSEM_ACTIVE_WRITE_BIAS) @@ -222,10 +302,26 @@ static inline void __downgrade_write(str { asm volatile("# beginning __downgrade_write\n\t" LOCK_PREFIX " addl %2,(%%eax)\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "subl %2,(%%eax)\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + /* transitions 0xZZZZ0001 -> 0xYYYY0001 */ - " jns 1f\n\t" + " jns 2f\n\t" " call call_rwsem_downgrade_wake\n" - "1:\n\t" + "2:\n\t" "# ending __downgrade_write\n" : "+m" (sem->count) : "a" (sem), "i" (-RWSEM_WAITING_BIAS) @@ -237,7 +333,23 @@ static inline void __downgrade_write(str */ static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem) { - asm volatile(LOCK_PREFIX "addl %1,%0" + asm volatile(LOCK_PREFIX "addl %1,%0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "subl %1,%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (sem->count) : "ir" (delta)); } @@ -249,7 +361,23 @@ static inline int rwsem_atomic_update(in { int tmp = delta; - asm volatile(LOCK_PREFIX "xadd %0,%1" + asm volatile(LOCK_PREFIX "xadd %0,%1\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + "movl %0,%1\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+r" (tmp), "+m" (sem->count) : : "memory"); diff -urNp linux-2.6.28.8/arch/x86/include/asm/segment.h linux-2.6.28.8/arch/x86/include/asm/segment.h --- linux-2.6.28.8/arch/x86/include/asm/segment.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/segment.h 2009-02-21 09:37:48.000000000 -0500 @@ -88,13 +88,19 @@ #define GDT_ENTRY_ESPFIX_SS (GDT_ENTRY_KERNEL_BASE + 14) #define __ESPFIX_SS (GDT_ENTRY_ESPFIX_SS * 8) -#define GDT_ENTRY_PERCPU (GDT_ENTRY_KERNEL_BASE + 15) +#define GDT_ENTRY_PERCPU (GDT_ENTRY_KERNEL_BASE + 15) #ifdef CONFIG_SMP #define __KERNEL_PERCPU (GDT_ENTRY_PERCPU * 8) #else #define __KERNEL_PERCPU 0 #endif +#define GDT_ENTRY_PCIBIOS_CS (GDT_ENTRY_KERNEL_BASE + 16) +#define __PCIBIOS_CS (GDT_ENTRY_PCIBIOS_CS * 8) + +#define GDT_ENTRY_PCIBIOS_DS (GDT_ENTRY_KERNEL_BASE + 17) +#define __PCIBIOS_DS (GDT_ENTRY_PCIBIOS_DS * 8) + #define GDT_ENTRY_DOUBLEFAULT_TSS 31 /* @@ -132,7 +138,7 @@ */ /* 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 diff -urNp linux-2.6.28.8/arch/x86/include/asm/spinlock.h linux-2.6.28.8/arch/x86/include/asm/spinlock.h --- linux-2.6.28.8/arch/x86/include/asm/spinlock.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/spinlock.h 2009-02-21 09:37:48.000000000 -0500 @@ -310,18 +310,50 @@ static inline int __raw_write_can_lock(r static inline void __raw_read_lock(raw_rwlock_t *rw) { asm volatile(LOCK_PREFIX " subl $1,(%0)\n\t" - "jns 1f\n" - "call __read_lock_failed\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" "1:\n" + LOCK_PREFIX " addl $1,(%0)\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "jns 2f\n" + "call __read_lock_failed\n\t" + "2:\n" ::LOCK_PTR_REG (rw) : "memory"); } static inline void __raw_write_lock(raw_rwlock_t *rw) { asm volatile(LOCK_PREFIX " subl %1,(%0)\n\t" - "jz 1f\n" - "call __write_lock_failed\n\t" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" "1:\n" + LOCK_PREFIX " addl %1,(%0)\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + "jz 2f\n" + "call __write_lock_failed\n\t" + "2:\n" ::LOCK_PTR_REG (rw), "i" (RW_LOCK_BIAS) : "memory"); } @@ -348,12 +380,45 @@ static inline int __raw_write_trylock(ra static inline void __raw_read_unlock(raw_rwlock_t *rw) { - asm volatile(LOCK_PREFIX "incl %0" :"+m" (rw->lock) : : "memory"); + asm volatile(LOCK_PREFIX "incl %0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "decl %0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + + :"+m" (rw->lock) : : "memory"); } static inline void __raw_write_unlock(raw_rwlock_t *rw) { - asm volatile(LOCK_PREFIX "addl %1, %0" + asm volatile(LOCK_PREFIX "addl %1, %0\n" + +#ifdef CONFIG_PAX_REFCOUNT +#ifdef CONFIG_X86_32 + "into\n0:\n" +#else + "jno 0f\n" + "int $4\n0:\n" +#endif + ".pushsection .fixup,\"ax\"\n" + "1:\n" + LOCK_PREFIX "subl %1,%0\n" + "jmp 0b\n" + ".popsection\n" + _ASM_EXTABLE(0b, 1b) +#endif + : "+m" (rw->lock) : "i" (RW_LOCK_BIAS) : "memory"); } diff -urNp linux-2.6.28.8/arch/x86/include/asm/system.h linux-2.6.28.8/arch/x86/include/asm/system.h --- linux-2.6.28.8/arch/x86/include/asm/system.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/system.h 2009-02-21 09:37:48.000000000 -0500 @@ -95,6 +95,8 @@ do { \ ".globl thread_return\n" \ "thread_return:\n\t" \ "movq %%gs:%P[pda_pcurrent],%%rsi\n\t" \ + "movq %P[task_canary](%%rsi),%%r8\n\t" \ + "movq %%r8,%%gs:%P[pda_canary]\n\t" \ "movq %P[thread_info](%%rsi),%%r8\n\t" \ LOCK_PREFIX "btr %[tif_fork],%P[ti_flags](%%r8)\n\t" \ "movq %%rax,%%rdi\n\t" \ @@ -106,7 +108,9 @@ do { \ [ti_flags] "i" (offsetof(struct thread_info, flags)), \ [tif_fork] "i" (TIF_FORK), \ [thread_info] "i" (offsetof(struct task_struct, stack)), \ - [pda_pcurrent] "i" (offsetof(struct x8664_pda, pcurrent)) \ + [task_canary] "i" (offsetof(struct task_struct, stack_canary)), \ + [pda_pcurrent] "i" (offsetof(struct x8664_pda, pcurrent)), \ + [pda_canary] "i" (offsetof(struct x8664_pda, stack_canary))\ : "memory", "cc" __EXTRA_CLOBBER) #endif @@ -169,7 +173,7 @@ static inline unsigned long get_limit(un { unsigned long __limit; asm("lsll %1,%0" : "=r" (__limit) : "r" (segment)); - return __limit + 1; + return __limit; } static inline void native_clts(void) @@ -295,6 +299,21 @@ static inline void native_wbinvd(void) #define stts() write_cr0(read_cr0() | X86_CR0_TS) +#define pax_open_kernel(cr0) \ +do { \ + typecheck(unsigned long, cr0); \ + preempt_disable(); \ + cr0 = read_cr0(); \ + write_cr0(cr0 & ~X86_CR0_WP); \ +} while (0) + +#define pax_close_kernel(cr0) \ +do { \ + typecheck(unsigned long, cr0); \ + write_cr0(cr0); \ + preempt_enable_no_resched(); \ +} while (0) + #endif /* __KERNEL__ */ static inline void clflush(volatile void *__p) @@ -309,7 +328,7 @@ void enable_hlt(void); void cpu_idle_wait(void); -extern unsigned long arch_align_stack(unsigned long sp); +#define arch_align_stack(x) ((x) & ~0xfUL) extern void free_init_pages(char *what, unsigned long begin, unsigned long end); void default_idle(void); diff -urNp linux-2.6.28.8/arch/x86/include/asm/uaccess_64.h linux-2.6.28.8/arch/x86/include/asm/uaccess_64.h --- linux-2.6.28.8/arch/x86/include/asm/uaccess_64.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/uaccess_64.h 2009-02-21 09:37:48.000000000 -0500 @@ -10,6 +10,8 @@ #include #include +#define set_fs(x) (current_thread_info()->addr_limit = (x)) + /* * Copy To/From Userspace */ diff -urNp linux-2.6.28.8/arch/x86/include/asm/uaccess.h linux-2.6.28.8/arch/x86/include/asm/uaccess.h --- linux-2.6.28.8/arch/x86/include/asm/uaccess.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/include/asm/uaccess.h 2009-02-21 09:37:48.000000000 -0500 @@ -10,6 +10,7 @@ #include #include #include +#include #define VERIFY_READ 0 #define VERIFY_WRITE 1 @@ -29,7 +30,12 @@ #define get_ds() (KERNEL_DS) #define get_fs() (current_thread_info()->addr_limit) +#ifdef CONFIG_X86_32 +void __set_fs(mm_segment_t x, int cpu); +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) @@ -186,9 +192,12 @@ extern int __get_user_bad(void); #ifdef CONFIG_X86_32 #define __put_user_u64(x, addr, err) \ - asm volatile("1: movl %%eax,0(%2)\n" \ - "2: movl %%edx,4(%2)\n" \ + asm volatile(" movw %w5,%%ds\n" \ + "1: movl %%eax,%%ds:0(%2)\n" \ + "2: movl %%edx,%%ds:4(%2)\n" \ "3:\n" \ + " pushl %%ss\n" \ + " popl %%ds\n" \ ".section .fixup,\"ax\"\n" \ "4: movl %3,%0\n" \ " jmp 3b\n" \ @@ -196,7 +205,8 @@ extern int __get_user_bad(void); _ASM_EXTABLE(1b, 4b) \ _ASM_EXTABLE(2b, 4b) \ : "=r" (err) \ - : "A" (x), "r" (addr), "i" (-EFAULT), "0" (err)) + : "A" (x), "r" (addr), "i" (-EFAULT), "0" (err), \ + "r"(__USER_DS)) #define __put_user_x8(x, ptr, __ret_pu) \ asm volatile("call __put_user_8" : "=a" (__ret_pu) \ @@ -336,6 +346,22 @@ do { \ } \ } while (0) +#ifdef CONFIG_X86_32 +#define __get_user_asm(x, addr, err, itype, rtype, ltype, errret) \ + asm volatile(" movw %w5,%%ds\n" \ + "1: mov"itype" %%ds:%2,%"rtype"1\n" \ + "2:\n" \ + " pushl %%ss\n" \ + " popl %%ds\n" \ + ".section .fixup,\"ax\"\n" \ + "3: movl %3,%0\n" \ + " xor"itype" %"rtype"1,%"rtype"1\n" \ + " jmp 2b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r" (err), ltype (x) \ + : "m" (__m(addr)), "i" (errret), "0" (err), "r"(__USER_DS)) +#else #define __get_user_asm(x, addr, err, itype, rtype, ltype, errret) \ asm volatile("1: mov"itype" %2,%"rtype"1\n" \ "2:\n" \ @@ -347,6 +373,7 @@ do { \ _ASM_EXTABLE(1b, 3b) \ : "=r" (err), ltype(x) \ : "m" (__m(addr)), "i" (errret), "0" (err)) +#endif #define __put_user_nocheck(x, ptr, size) \ ({ \ @@ -373,6 +400,22 @@ struct __large_struct { unsigned long bu * we do not write to any memory gcc knows about, so there are no * aliasing issues. */ +#ifdef CONFIG_X86_32 +#define __put_user_asm(x, addr, err, itype, rtype, ltype, errret) \ + asm volatile(" movw %w5,%%ds\n" \ + "1: mov"itype" %"rtype"1,%%ds:%2\n" \ + "2:\n" \ + " pushl %%ss\n" \ + " popl %%ds\n" \ + ".section .fixup,\"ax\"\n" \ + "3: movl %3,%0\n" \ + " jmp 2b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r"(err) \ + : ltype (x), "m" (__m(addr)), "i" (errret), "0" (err),\ + "r"(__USER_DS)) +#else #define __put_user_asm(x, addr, err, itype, rtype, ltype, errret) \ asm volatile("1: mov"itype" %"rtype"1,%2\n" \ "2:\n" \ @@ -383,6 +426,7 @@ struct __large_struct { unsigned long bu _ASM_EXTABLE(1b, 3b) \ : "=r"(err) \ : ltype(x), "m" (__m(addr)), "i" (errret), "0" (err)) +#endif /** * __get_user: - Get a simple variable from user space, with less checking. * @x: Variable to store result. @@ -443,6 +487,7 @@ extern struct movsl_mask { #define ARCH_HAS_NOCACHE_UACCESS 1 +#define ARCH_HAS_SORT_EXTABLE #ifdef CONFIG_X86_32 # include "uaccess_32.h" #else diff -urNp linux-2.6.28.8/arch/x86/Kconfig linux-2.6.28.8/arch/x86/Kconfig --- linux-2.6.28.8/arch/x86/Kconfig 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/Kconfig 2009-02-21 09:37:48.000000000 -0500 @@ -935,7 +935,7 @@ 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 @@ -1337,8 +1337,7 @@ config KEXEC_JUMP config PHYSICAL_START hex "Physical address where the kernel is loaded" if (EMBEDDED || CRASH_DUMP) default "0x1000000" if X86_NUMAQ - default "0x200000" if X86_64 - default "0x100000" + default "0x200000" help This gives the physical address where the kernel is loaded. @@ -1430,9 +1429,9 @@ config HOTPLUG_CPU Say N if you want to disable CPU hotplug. config COMPAT_VDSO - def_bool y + def_bool n prompt "Compat VDSO support" - depends on X86_32 || IA32_EMULATION + depends on (X86_32 || IA32_EMULATION) && !PAX_NOEXEC help Map the 32-bit VDSO to the predictable old-style address too. ---help--- diff -urNp linux-2.6.28.8/arch/x86/Kconfig.cpu linux-2.6.28.8/arch/x86/Kconfig.cpu --- linux-2.6.28.8/arch/x86/Kconfig.cpu 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/Kconfig.cpu 2009-02-21 09:37:48.000000000 -0500 @@ -331,7 +331,7 @@ config X86_PPRO_FENCE config X86_F00F_BUG def_bool y - depends on M586MMX || M586TSC || M586 || M486 || M386 + depends on (M586MMX || M586TSC || M586 || M486 || M386) && !PAX_KERNEXEC config X86_WP_WORKS_OK def_bool y @@ -351,7 +351,7 @@ config X86_POPAD_OK config X86_ALIGNMENT_16 def_bool y - depends on MWINCHIP3D || MWINCHIPC6 || MCYRIXIII || X86_ELAN || 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 @@ -397,7 +397,7 @@ 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) + depends on (MK8 || MK7 || MCORE2 || MPSC || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || MVIAC3_2 || MVIAC7 || MCRUSOE || MEFFICEON || X86_64) config X86_MINIMUM_CPU_FAMILY int diff -urNp linux-2.6.28.8/arch/x86/Kconfig.debug linux-2.6.28.8/arch/x86/Kconfig.debug --- linux-2.6.28.8/arch/x86/Kconfig.debug 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/Kconfig.debug 2009-02-21 09:37:48.000000000 -0500 @@ -107,7 +107,7 @@ config X86_PTDUMP config DEBUG_RODATA bool "Write protect kernel read-only data structures" default y - depends on DEBUG_KERNEL + depends on DEBUG_KERNEL && BROKEN help Mark the kernel read-only data as write-protected in the pagetables, in order to catch accidental (and incorrect) writes to such const diff -urNp linux-2.6.28.8/arch/x86/kernel/acpi/boot.c linux-2.6.28.8/arch/x86/kernel/acpi/boot.c --- linux-2.6.28.8/arch/x86/kernel/acpi/boot.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/acpi/boot.c 2009-02-21 09:37:48.000000000 -0500 @@ -1645,7 +1645,7 @@ static struct dmi_system_id __initdata a DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq 6715b"), }, }, - {} + { NULL, NULL, {{0, {0}}}, NULL} }; /* diff -urNp linux-2.6.28.8/arch/x86/kernel/acpi/realmode/wakeup.S linux-2.6.28.8/arch/x86/kernel/acpi/realmode/wakeup.S --- linux-2.6.28.8/arch/x86/kernel/acpi/realmode/wakeup.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/acpi/realmode/wakeup.S 2009-02-21 09:37:48.000000000 -0500 @@ -104,7 +104,7 @@ _start: movl %eax, %ecx orl %edx, %ecx jz 1f - movl $0xc0000080, %ecx + mov $MSR_EFER, %ecx wrmsr 1: diff -urNp linux-2.6.28.8/arch/x86/kernel/acpi/sleep.c linux-2.6.28.8/arch/x86/kernel/acpi/sleep.c --- linux-2.6.28.8/arch/x86/kernel/acpi/sleep.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/acpi/sleep.c 2009-02-21 09:37:48.000000000 -0500 @@ -37,6 +37,10 @@ int acpi_save_state_mem(void) { struct wakeup_header *header; +#if defined(CONFIG_64BIT) && defined(CONFIG_SMP) && defined(CONFIG_PAX_KERNEXEC) + unsigned long cr0; +#endif + if (!acpi_realmode) { printk(KERN_ERR "Could not allocate memory during boot, " "S3 disabled\n"); @@ -99,8 +103,18 @@ int acpi_save_state_mem(void) header->trampoline_segment = setup_trampoline() >> 4; #ifdef CONFIG_SMP stack_start.sp = temp_stack + sizeof(temp_stack); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(smp_processor_id()); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + #endif initial_code = (unsigned long)wakeup_long64; saved_magic = 0x123456789abcdef0; diff -urNp linux-2.6.28.8/arch/x86/kernel/acpi/wakeup_32.S linux-2.6.28.8/arch/x86/kernel/acpi/wakeup_32.S --- linux-2.6.28.8/arch/x86/kernel/acpi/wakeup_32.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/acpi/wakeup_32.S 2009-02-21 09:37:48.000000000 -0500 @@ -30,13 +30,11 @@ wakeup_pmode_return: # and restore the stack ... but you need gdt for this to work movl saved_context_esp, %esp - movl %cs:saved_magic, %eax - cmpl $0x12345678, %eax + cmpl $0x12345678, saved_magic jne bogus_magic # jump to place where we left off - movl saved_eip, %eax - jmp *%eax + jmp *(saved_eip) bogus_magic: jmp bogus_magic diff -urNp linux-2.6.28.8/arch/x86/kernel/alternative.c linux-2.6.28.8/arch/x86/kernel/alternative.c --- linux-2.6.28.8/arch/x86/kernel/alternative.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/alternative.c 2009-02-21 09:37:48.000000000 -0500 @@ -393,7 +393,7 @@ void apply_paravirt(struct paravirt_patc 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); @@ -473,11 +473,26 @@ void __init alternative_instructions(voi * instructions. And on the local CPU you need to be protected again NMI or MCE * handlers seeing an inconsistent instruction while you patch. */ -void *text_poke_early(void *addr, const void *opcode, size_t len) +void *__kprobes text_poke_early(void *addr, const void *opcode, size_t len) { unsigned long flags; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + local_irq_save(flags); - memcpy(addr, opcode, len); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + + memcpy(ktla_ktva(addr), opcode, len); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + local_irq_restore(flags); sync_core(); /* Could also do a CLFLUSH here to speed up CPU recovery; but @@ -498,33 +513,27 @@ void *text_poke_early(void *addr, const */ void *__kprobes text_poke(void *addr, const void *opcode, size_t len) { - unsigned long flags; - char *vaddr; - int nr_pages = 2; + unsigned char *vaddr = ktla_ktva(addr); struct page *pages[2]; - int i; + size_t i; + + if (!core_kernel_text((unsigned long)addr) - if (!core_kernel_text((unsigned long)addr)) { - pages[0] = vmalloc_to_page(addr); - pages[1] = vmalloc_to_page(addr + PAGE_SIZE); +#if defined(CONFIG_X86_32) && defined(CONFIG_MODULES) && defined(CONFIG_PAX_KERNEXEC) + && (vaddr < MODULES_VADDR || MODULES_END < vaddr) +#endif + + ) { + 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]); - if (!pages[1]) - nr_pages = 1; - vaddr = vmap(pages, nr_pages, VM_MAP, PAGE_KERNEL); - BUG_ON(!vaddr); - local_irq_save(flags); - memcpy(&vaddr[(unsigned long)addr & ~PAGE_MASK], opcode, len); - local_irq_restore(flags); - vunmap(vaddr); - 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]); + BUG_ON((vaddr)[i] != ((unsigned char *)opcode)[i]); return addr; } diff -urNp linux-2.6.28.8/arch/x86/kernel/apm_32.c linux-2.6.28.8/arch/x86/kernel/apm_32.c --- linux-2.6.28.8/arch/x86/kernel/apm_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/apm_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -407,7 +407,7 @@ static DECLARE_WAIT_QUEUE_HEAD(apm_waitq static DECLARE_WAIT_QUEUE_HEAD(apm_suspend_waitqueue); static struct apm_user *user_list; static DEFINE_SPINLOCK(user_list_lock); -static const struct desc_struct bad_bios_desc = { { { 0, 0x00409200 } } }; +static const struct desc_struct bad_bios_desc = { { { 0, 0x00409300 } } }; static const char driver_version[] = "1.16ac"; /* no spaces */ @@ -602,19 +602,42 @@ static u8 apm_bios_call(u32 func, u32 eb struct desc_struct save_desc_40; struct desc_struct *gdt; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + cpus = apm_save_cpus(); cpu = get_cpu(); gdt = get_cpu_gdt_table(cpu); save_desc_40 = gdt[0x40 / 8]; + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + gdt[0x40 / 8] = bad_bios_desc; +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + apm_irq_save(flags); APM_DO_SAVE_SEGS; apm_bios_call_asm(func, ebx_in, ecx_in, eax, ebx, ecx, edx, esi); APM_DO_RESTORE_SEGS; apm_irq_restore(flags); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + gdt[0x40 / 8] = save_desc_40; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + put_cpu(); apm_restore_cpus(cpus); @@ -645,19 +668,42 @@ static u8 apm_bios_call_simple(u32 func, struct desc_struct save_desc_40; struct desc_struct *gdt; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + cpus = apm_save_cpus(); cpu = get_cpu(); gdt = get_cpu_gdt_table(cpu); save_desc_40 = gdt[0x40 / 8]; + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + gdt[0x40 / 8] = bad_bios_desc; +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + apm_irq_save(flags); APM_DO_SAVE_SEGS; error = apm_bios_call_simple_asm(func, ebx_in, ecx_in, eax); APM_DO_RESTORE_SEGS; apm_irq_restore(flags); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + gdt[0x40 / 8] = save_desc_40; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + put_cpu(); apm_restore_cpus(cpus); return error; @@ -929,7 +975,7 @@ recalc: static void apm_power_off(void) { - unsigned char po_bios_call[] = { + const unsigned char po_bios_call[] = { 0xb8, 0x00, 0x10, /* movw $0x1000,ax */ 0x8e, 0xd0, /* movw ax,ss */ 0xbc, 0x00, 0xf0, /* movw $0xf000,sp */ @@ -1876,7 +1922,10 @@ static const struct file_operations apm_ static struct miscdevice apm_device = { APM_MINOR_DEV, "apm_bios", - &apm_bios_fops + &apm_bios_fops, + {NULL, NULL}, + NULL, + NULL }; @@ -2197,7 +2246,7 @@ static struct dmi_system_id __initdata a { DMI_MATCH(DMI_SYS_VENDOR, "IBM"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL} }; /* @@ -2215,6 +2264,10 @@ static int __init apm_init(void) struct desc_struct *gdt; int err; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + dmi_check_system(apm_dmi_table); if (apm_info.bios.version == 0 || paravirt_enabled() || machine_is_olpc()) { @@ -2288,9 +2341,18 @@ static int __init apm_init(void) * This is for buggy BIOS's that refer to (real mode) segment 0x40 * even though they are called in protected mode. */ + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + set_base(bad_bios_desc, __va((unsigned long)0x40 << 4)); _set_limit((char *)&bad_bios_desc, 4095 - (0x40 << 4)); +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + /* * Set up the long jump entry point to the APM BIOS, which is called * from inline assembly. @@ -2309,6 +2371,11 @@ static int __init apm_init(void) * code to that CPU. */ gdt = get_cpu_gdt_table(0); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + set_base(gdt[APM_CS >> 3], __va((unsigned long)apm_info.bios.cseg << 4)); set_base(gdt[APM_CS_16 >> 3], @@ -2316,6 +2383,10 @@ static int __init apm_init(void) set_base(gdt[APM_DS >> 3], __va((unsigned long)apm_info.bios.dseg << 4)); +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + proc_create("apm", 0, NULL, &apm_file_ops); kapmd_task = kthread_create(apm, NULL, "kapmd"); diff -urNp linux-2.6.28.8/arch/x86/kernel/asm-offsets_32.c linux-2.6.28.8/arch/x86/kernel/asm-offsets_32.c --- linux-2.6.28.8/arch/x86/kernel/asm-offsets_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/asm-offsets_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -100,6 +100,7 @@ void foo(void) DEFINE(PTRS_PER_PTE, PTRS_PER_PTE); DEFINE(PTRS_PER_PMD, PTRS_PER_PMD); DEFINE(PTRS_PER_PGD, PTRS_PER_PGD); + DEFINE(PERCPU_MODULE_RESERVE, PERCPU_MODULE_RESERVE); OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx); @@ -113,6 +114,7 @@ void foo(void) OFFSET(PV_CPU_iret, pv_cpu_ops, iret); OFFSET(PV_CPU_irq_enable_sysexit, pv_cpu_ops, irq_enable_sysexit); OFFSET(PV_CPU_read_cr0, pv_cpu_ops, read_cr0); + OFFSET(PV_CPU_write_cr0, pv_cpu_ops, write_cr0); #endif #ifdef CONFIG_XEN diff -urNp linux-2.6.28.8/arch/x86/kernel/asm-offsets_64.c linux-2.6.28.8/arch/x86/kernel/asm-offsets_64.c --- linux-2.6.28.8/arch/x86/kernel/asm-offsets_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/asm-offsets_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -122,6 +122,7 @@ int main(void) ENTRY(cr8); BLANK(); #undef ENTRY + DEFINE(TSS_size, sizeof(struct tss_struct)); DEFINE(TSS_ist, offsetof(struct tss_struct, x86_tss.ist)); BLANK(); DEFINE(crypto_tfm_ctx_offset, offsetof(struct crypto_tfm, __crt_ctx)); diff -urNp linux-2.6.28.8/arch/x86/kernel/cpu/common.c linux-2.6.28.8/arch/x86/kernel/cpu/common.c --- linux-2.6.28.8/arch/x86/kernel/cpu/common.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/cpu/common.c 2009-02-21 09:37:48.000000000 -0500 @@ -2,7 +2,6 @@ #include #include #include -#include #include #include #include @@ -41,59 +40,6 @@ static struct cpu_dev *this_cpu __cpuinitdata; -#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 - */ -/* The TLS descriptors are currently at a different place compared to i386. - Hopefully nobody expects them at a fixed place (Wine?) */ -DEFINE_PER_CPU(struct gdt_page, gdt_page) = { .gdt = { - [GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } }, - [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } }, - [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } }, - [GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } }, - [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } }, - [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } }, -} }; -#else -DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { - [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00cf9a00 } } }, - [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9200 } } }, - [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00cffa00 } } }, - [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff200 } } }, - /* - * 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] = { { { 0x0000ffff, 0x00409a00 } } }, - /* 16-bit code */ - [GDT_ENTRY_PNPBIOS_CS16] = { { { 0x0000ffff, 0x00009a00 } } }, - /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_DS] = { { { 0x0000ffff, 0x00009200 } } }, - /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_TS1] = { { { 0x00000000, 0x00009200 } } }, - /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_TS2] = { { { 0x00000000, 0x00009200 } } }, - /* - * 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] = { { { 0x0000ffff, 0x00409a00 } } }, - /* 16-bit code */ - [GDT_ENTRY_APMBIOS_BASE+1] = { { { 0x0000ffff, 0x00009a00 } } }, - /* data */ - [GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } }, - - [GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } }, - [GDT_ENTRY_PERCPU] = { { { 0x00000000, 0x00000000 } } }, -} }; -#endif -EXPORT_PER_CPU_SYMBOL_GPL(gdt_page); - #ifdef CONFIG_X86_32 static int cachesize_override __cpuinitdata = -1; static int disable_x86_serial_nr __cpuinitdata = 1; @@ -227,7 +173,7 @@ void switch_to_new_gdt(void) { struct desc_ptr gdt_descr; - gdt_descr.address = (long)get_cpu_gdt_table(smp_processor_id()); + gdt_descr.address = (unsigned long)get_cpu_gdt_table(smp_processor_id()); gdt_descr.size = GDT_SIZE - 1; load_gdt(&gdt_descr); #ifdef CONFIG_X86_32 @@ -687,6 +633,10 @@ static void __cpuinit identify_cpu(struc * we do "generic changes." */ +#if 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]) { char *p; @@ -854,13 +804,13 @@ static __init int setup_disablecpuid(cha } __setup("clearcpuid=", setup_disablecpuid); -cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE; +cpumask_t cpu_initialized = CPU_MASK_NONE; #ifdef CONFIG_X86_64 struct x8664_pda **_cpu_pda __read_mostly; EXPORT_SYMBOL(_cpu_pda); -struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table }; +struct desc_ptr idt_descr __read_only = { 256 * 16 - 1, (unsigned long) idt_table }; char boot_cpu_stack[IRQSTACKSIZE] __page_aligned_bss; @@ -959,7 +909,7 @@ struct pt_regs * __cpuinit idle_regs(str void __cpuinit cpu_init(void) { int cpu = stack_smp_processor_id(); - struct tss_struct *t = &per_cpu(init_tss, cpu); + struct tss_struct *t = init_tss + cpu; struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu); unsigned long v; char *estacks = NULL; @@ -1080,7 +1030,7 @@ void __cpuinit cpu_init(void) { int cpu = smp_processor_id(); struct task_struct *curr = current; - struct tss_struct *t = &per_cpu(init_tss, cpu); + struct tss_struct *t = init_tss + cpu; struct thread_struct *thread = &curr->thread; if (cpu_test_and_set(cpu, cpu_initialized)) { diff -urNp linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c --- linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c 2009-02-21 09:37:48.000000000 -0500 @@ -561,7 +561,7 @@ static const struct dmi_system_id sw_any DMI_MATCH(DMI_PRODUCT_NAME, "X6DLP"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; #endif diff -urNp linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c --- linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c 2009-02-21 09:37:48.000000000 -0500 @@ -225,7 +225,7 @@ static struct cpu_model models[] = { &cpu_ids[CPU_MP4HT_D0], NULL, 0, NULL }, { &cpu_ids[CPU_MP4HT_E0], NULL, 0, NULL }, - { NULL, } + { NULL, NULL, 0, NULL} }; #undef _BANIAS #undef BANIAS diff -urNp linux-2.6.28.8/arch/x86/kernel/cpu/intel.c linux-2.6.28.8/arch/x86/kernel/cpu/intel.c --- linux-2.6.28.8/arch/x86/kernel/cpu/intel.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/cpu/intel.c 2009-02-21 09:37:48.000000000 -0500 @@ -72,7 +72,7 @@ static void __cpuinit trap_init_f00f_bug * Update the IDT descriptor and reload the IDT so that * it uses the read-only mapped virtual address. */ - idt_descr.address = fix_to_virt(FIX_F00F_IDT); + idt_descr.address = (struct desc_struct *)fix_to_virt(FIX_F00F_IDT); load_idt(&idt_descr); } #endif diff -urNp linux-2.6.28.8/arch/x86/kernel/cpu/mcheck/mce_64.c linux-2.6.28.8/arch/x86/kernel/cpu/mcheck/mce_64.c --- linux-2.6.28.8/arch/x86/kernel/cpu/mcheck/mce_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/cpu/mcheck/mce_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -678,6 +678,7 @@ static struct miscdevice mce_log_device MISC_MCELOG_MINOR, "mcelog", &mce_chrdev_ops, + {NULL, NULL}, NULL, NULL }; static unsigned long old_cr4 __initdata; diff -urNp linux-2.6.28.8/arch/x86/kernel/cpu/mtrr/generic.c linux-2.6.28.8/arch/x86/kernel/cpu/mtrr/generic.c --- linux-2.6.28.8/arch/x86/kernel/cpu/mtrr/generic.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/cpu/mtrr/generic.c 2009-02-21 09:37:48.000000000 -0500 @@ -31,11 +31,11 @@ static struct fixed_range_block fixed_ra { MTRRfix64K_00000_MSR, 1 }, /* one 64k MTRR */ { MTRRfix16K_80000_MSR, 2 }, /* two 16k MTRRs */ { MTRRfix4K_C0000_MSR, 8 }, /* eight 4k MTRRs */ - {} + { 0, 0 } }; static unsigned long smp_changes_mask; -static struct mtrr_state mtrr_state = {}; +static struct mtrr_state mtrr_state; static int mtrr_state_set; u64 mtrr_tom2; diff -urNp linux-2.6.28.8/arch/x86/kernel/crash.c linux-2.6.28.8/arch/x86/kernel/crash.c --- linux-2.6.28.8/arch/x86/kernel/crash.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/crash.c 2009-02-21 09:37:48.000000000 -0500 @@ -59,7 +59,7 @@ static int crash_nmi_callback(struct not local_irq_disable(); #ifdef CONFIG_X86_32 - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { crash_fixup_ss_esp(&fixed_regs, regs); regs = &fixed_regs; } diff -urNp linux-2.6.28.8/arch/x86/kernel/doublefault_32.c linux-2.6.28.8/arch/x86/kernel/doublefault_32.c --- linux-2.6.28.8/arch/x86/kernel/doublefault_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/doublefault_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -11,7 +11,7 @@ #define DOUBLEFAULT_STACKSIZE (1024) static unsigned long doublefault_stack[DOUBLEFAULT_STACKSIZE]; -#define STACK_START (unsigned long)(doublefault_stack+DOUBLEFAULT_STACKSIZE) +#define STACK_START (unsigned long)(doublefault_stack+DOUBLEFAULT_STACKSIZE-2) #define ptr_ok(x) ((x) > PAGE_OFFSET && (x) < PAGE_OFFSET + MAXMEM) @@ -21,7 +21,7 @@ static void doublefault_fn(void) unsigned long gdt, tss; store_gdt(&gdt_desc); - gdt = gdt_desc.address; + gdt = (unsigned long)gdt_desc.address; printk(KERN_EMERG "PANIC: double fault, gdt at %08lx [%d bytes]\n", gdt, gdt_desc.size); @@ -60,10 +60,10 @@ struct tss_struct doublefault_tss __cach /* 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 -urNp linux-2.6.28.8/arch/x86/kernel/dumpstack_32.c linux-2.6.28.8/arch/x86/kernel/dumpstack_32.c --- linux-2.6.28.8/arch/x86/kernel/dumpstack_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/dumpstack_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -238,11 +238,12 @@ void show_registers(struct pt_regs *regs * When in-kernel, we also print out the stack and code at the * time of the fault.. */ - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { unsigned int code_prologue = code_bytes * 43 / 64; unsigned int code_len = code_bytes; unsigned char c; u8 *ip; + unsigned long cs_base = get_desc_base(&get_cpu_gdt_table(smp_processor_id())[(0xffff & regs->cs) >> 3]); printk(KERN_EMERG "Stack:\n"); show_stack_log_lvl(NULL, regs, ®s->sp, @@ -250,10 +251,10 @@ void show_registers(struct pt_regs *regs printk(KERN_EMERG "Code: "); - ip = (u8 *)regs->ip - code_prologue; + ip = (u8 *)regs->ip - code_prologue + cs_base; if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { /* try starting at IP */ - ip = (u8 *)regs->ip; + ip = (u8 *)regs->ip + cs_base; code_len = code_len - code_prologue + 1; } for (i = 0; i < code_len; i++, ip++) { @@ -262,7 +263,7 @@ void show_registers(struct pt_regs *regs printk(" Bad EIP value."); break; } - if (ip == (u8 *)regs->ip) + if (ip == (u8 *)regs->ip + cs_base) printk("<%02x> ", c); else printk("%02x ", c); @@ -275,6 +276,7 @@ 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)) @@ -410,7 +412,7 @@ die_nmi(char *str, struct pt_regs *regs, * If we are in kernel we are probably nested up pretty bad * and might aswell get out now while we still can: */ - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { current->thread.trap_no = 2; crash_kexec(regs); } diff -urNp linux-2.6.28.8/arch/x86/kernel/efi_32.c linux-2.6.28.8/arch/x86/kernel/efi_32.c --- linux-2.6.28.8/arch/x86/kernel/efi_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/efi_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -38,70 +38,38 @@ */ static unsigned long efi_rt_eflags; -static pgd_t efi_bak_pg_dir_pointer[2]; +static pgd_t __initdata efi_bak_pg_dir_pointer[KERNEL_PGD_PTRS]; -void efi_call_phys_prelog(void) +void __init efi_call_phys_prelog(void) { - unsigned long cr4; - unsigned long temp; struct desc_ptr gdt_descr; local_irq_save(efi_rt_eflags); - /* - * If I don't have PAE, I should just duplicate two entries in page - * directory. If I have PAE, I just need to duplicate one entry in - * page directory. - */ - cr4 = read_cr4_safe(); - if (cr4 & X86_CR4_PAE) { - efi_bak_pg_dir_pointer[0].pgd = - swapper_pg_dir[pgd_index(0)].pgd; - swapper_pg_dir[0].pgd = - swapper_pg_dir[pgd_index(PAGE_OFFSET)].pgd; - } else { - efi_bak_pg_dir_pointer[0].pgd = - swapper_pg_dir[pgd_index(0)].pgd; - efi_bak_pg_dir_pointer[1].pgd = - swapper_pg_dir[pgd_index(0x400000)].pgd; - swapper_pg_dir[pgd_index(0)].pgd = - swapper_pg_dir[pgd_index(PAGE_OFFSET)].pgd; - temp = PAGE_OFFSET + 0x400000; - swapper_pg_dir[pgd_index(0x400000)].pgd = - swapper_pg_dir[pgd_index(temp)].pgd; - } + clone_pgd_range(efi_bak_pg_dir_pointer, swapper_pg_dir, KERNEL_PGD_PTRS); + clone_pgd_range(swapper_pg_dir, swapper_pg_dir + KERNEL_PGD_BOUNDARY, + min_t(unsigned long, KERNEL_PGD_PTRS, KERNEL_PGD_BOUNDARY)); /* * After the lock is released, the original page table is restored. */ __flush_tlb_all(); - gdt_descr.address = __pa(get_cpu_gdt_table(0)); + gdt_descr.address = (struct desc_struct *)__pa(get_cpu_gdt_table(0)); gdt_descr.size = GDT_SIZE - 1; load_gdt(&gdt_descr); } -void efi_call_phys_epilog(void) +void __init efi_call_phys_epilog(void) { - unsigned long cr4; struct desc_ptr gdt_descr; - gdt_descr.address = (unsigned long)get_cpu_gdt_table(0); + gdt_descr.address = get_cpu_gdt_table(0); gdt_descr.size = GDT_SIZE - 1; load_gdt(&gdt_descr); - cr4 = read_cr4_safe(); - - if (cr4 & X86_CR4_PAE) { - swapper_pg_dir[pgd_index(0)].pgd = - efi_bak_pg_dir_pointer[0].pgd; - } else { - swapper_pg_dir[pgd_index(0)].pgd = - efi_bak_pg_dir_pointer[0].pgd; - swapper_pg_dir[pgd_index(0x400000)].pgd = - efi_bak_pg_dir_pointer[1].pgd; - } + clone_pgd_range(swapper_pg_dir, efi_bak_pg_dir_pointer, KERNEL_PGD_PTRS); /* * After the lock is released, the original page table is restored. diff -urNp linux-2.6.28.8/arch/x86/kernel/efi_stub_32.S linux-2.6.28.8/arch/x86/kernel/efi_stub_32.S --- linux-2.6.28.8/arch/x86/kernel/efi_stub_32.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/efi_stub_32.S 2009-02-21 09:37:48.000000000 -0500 @@ -6,6 +6,7 @@ */ #include +#include #include /* @@ -20,7 +21,7 @@ * service functions will comply with gcc calling convention, too. */ -.text +__INIT ENTRY(efi_call_phys) /* * 0. The function can only be called in Linux kernel. So CS has been @@ -36,9 +37,7 @@ ENTRY(efi_call_phys) * The mapping of lower virtual memory has been created in prelog and * epilog. */ - movl $1f, %edx - subl $__PAGE_OFFSET, %edx - jmp *%edx + jmp 1f-__PAGE_OFFSET 1: /* @@ -47,14 +46,8 @@ ENTRY(efi_call_phys) * parameter 2, ..., param n. To make things easy, we save the return * address of efi_call_phys in a global variable. */ - popl %edx - movl %edx, saved_return_addr - /* get the function pointer into ECX*/ - popl %ecx - movl %ecx, efi_rt_function_ptr - movl $2f, %edx - subl $__PAGE_OFFSET, %edx - pushl %edx + popl (saved_return_addr) + popl (efi_rt_function_ptr) /* * 3. Clear PG bit in %CR0. @@ -73,9 +66,8 @@ ENTRY(efi_call_phys) /* * 5. Call the physical function. */ - jmp *%ecx + call *(efi_rt_function_ptr-__PAGE_OFFSET) -2: /* * 6. After EFI runtime service returns, control will return to * following instruction. We'd better readjust stack pointer first. @@ -88,34 +80,27 @@ ENTRY(efi_call_phys) movl %cr0, %edx orl $0x80000000, %edx movl %edx, %cr0 - jmp 1f -1: + /* * 8. Now restore the virtual mode from flat mode by * adding EIP with PAGE_OFFSET. */ - movl $1f, %edx - jmp *%edx + jmp 1f+__PAGE_OFFSET 1: /* * 9. Balance the stack. And because EAX contain the return value, * we'd better not clobber it. */ - leal efi_rt_function_ptr, %edx - movl (%edx), %ecx - pushl %ecx + pushl (efi_rt_function_ptr) /* - * 10. Push the saved return address onto the stack and return. + * 10. Return to the saved return address. */ - leal saved_return_addr, %edx - movl (%edx), %ecx - pushl %ecx - ret + jmpl *(saved_return_addr) .previous -.data +__INITDATA saved_return_addr: .long 0 efi_rt_function_ptr: diff -urNp linux-2.6.28.8/arch/x86/kernel/entry_32.S linux-2.6.28.8/arch/x86/kernel/entry_32.S --- linux-2.6.28.8/arch/x86/kernel/entry_32.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/entry_32.S 2009-02-21 09:37:48.000000000 -0500 @@ -101,7 +101,7 @@ #define resume_userspace_sig resume_userspace #endif -#define SAVE_ALL \ +#define __SAVE_ALL(_DS) \ cld; \ pushl %fs; \ CFI_ADJUST_CFA_OFFSET 4;\ @@ -133,12 +133,26 @@ pushl %ebx; \ CFI_ADJUST_CFA_OFFSET 4;\ CFI_REL_OFFSET ebx, 0;\ - movl $(__USER_DS), %edx; \ + movl $(_DS), %edx; \ movl %edx, %ds; \ movl %edx, %es; \ movl $(__KERNEL_PERCPU), %edx; \ movl %edx, %fs +#ifdef CONFIG_PAX_KERNEXEC +#define SAVE_ALL \ + __SAVE_ALL(__KERNEL_DS); \ + GET_CR0_INTO_EDX; \ + movl %edx, %esi; \ + orl $X86_CR0_WP, %edx; \ + xorl %edx, %esi; \ + SET_CR0_FROM_EDX +#elif defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) || defined(CONFIG_PAX_MEMORY_UDEREF) +#define SAVE_ALL __SAVE_ALL(__KERNEL_DS) +#else +#define SAVE_ALL __SAVE_ALL(__USER_DS) +#endif + #define RESTORE_INT_REGS \ popl %ebx; \ CFI_ADJUST_CFA_OFFSET -4;\ @@ -229,6 +243,11 @@ ENTRY(ret_from_fork) CFI_ADJUST_CFA_OFFSET 4 popfl CFI_ADJUST_CFA_OFFSET -4 + +#ifdef CONFIG_PAX_KERNEXEC + xorl %esi, %esi +#endif + jmp syscall_exit CFI_ENDPROC END(ret_from_fork) @@ -252,7 +271,17 @@ check_userspace: movb PT_CS(%esp), %al andl $(X86_EFLAGS_VM | SEGMENT_RPL_MASK), %eax cmpl $USER_RPL, %eax + +#ifdef CONFIG_PAX_KERNEXEC + jae resume_userspace + + GET_CR0_INTO_EDX + xorl %esi, %edx + SET_CR0_FROM_EDX + jmp resume_kernel +#else jb resume_kernel # not returning to v8086 or userspace +#endif ENTRY(resume_userspace) LOCKDEP_SYS_EXIT @@ -314,10 +343,9 @@ 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 (TI_sysenter_return-THREAD_SIZE+8+4*4)(%esp) + GET_THREAD_INFO(%ebp) + pushl TI_sysenter_return(%ebp) CFI_ADJUST_CFA_OFFSET 4 CFI_REL_OFFSET eip, 0 @@ -330,9 +358,19 @@ sysenter_past_esp: * Load the potential sixth argument from user stack. * Careful about security. */ + movl PT_OLDESP(%esp),%ebp + +#ifdef CONFIG_PAX_MEMORY_UDEREF + mov PT_OLDSS(%esp),%ds +1: movl %ds:(%ebp),%ebp + push %ss + pop %ds +#else cmpl $__PAGE_OFFSET-3,%ebp jae syscall_fault 1: movl (%ebp),%ebp +#endif + movl %ebp,PT_EBP(%esp) .section __ex_table,"a" .align 4 @@ -356,12 +394,23 @@ sysenter_do_call: testw $_TIF_ALLWORK_MASK, %cx jne sysexit_audit sysenter_exit: + +#ifdef CONFIG_PAX_RANDKSTACK + pushl %eax + CFI_ADJUST_CFA_OFFSET 4 + call pax_randomize_kstack + popl %eax + CFI_ADJUST_CFA_OFFSET -4 +#endif + /* 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 ENABLE_INTERRUPTS_SYSEXIT #ifdef CONFIG_AUDITSYSCALL @@ -404,11 +453,17 @@ sysexit_audit: CFI_ENDPROC .pushsection .fixup,"ax" -2: movl $0,PT_FS(%esp) +4: movl $0,PT_FS(%esp) + jmp 1b +5: movl $0,PT_DS(%esp) + jmp 1b +6: movl $0,PT_ES(%esp) jmp 1b .section __ex_table,"a" .align 4 - .long 1b,2b + .long 1b,4b + .long 2b,5b + .long 3b,6b .popsection ENDPROC(ia32_sysenter_target) @@ -438,6 +493,10 @@ syscall_exit: testw $_TIF_ALLWORK_MASK, %cx # current->work jne syscall_exit_work +#ifdef CONFIG_PAX_RANDKSTACK + call pax_randomize_kstack +#endif + restore_all: movl PT_EFLAGS(%esp), %eax # mix EFLAGS, SS and CS # Warning: PT_OLDSS(%esp) contains the wrong/random values if we @@ -531,25 +590,19 @@ work_resched: work_notifysig: # deal with pending signals and # notify-resume requests + movl %esp, %eax #ifdef CONFIG_VM86 testl $X86_EFLAGS_VM, PT_EFLAGS(%esp) - movl %esp, %eax - jne work_notifysig_v86 # returning to kernel-space or + jz 1f # returning to kernel-space or # vm86-space - xorl %edx, %edx - call do_notify_resume - jmp resume_userspace_sig - ALIGN -work_notifysig_v86: pushl %ecx # save ti_flags for do_notify_resume CFI_ADJUST_CFA_OFFSET 4 call save_v86_state # %eax contains pt_regs pointer popl %ecx CFI_ADJUST_CFA_OFFSET -4 movl %eax, %esp -#else - movl %esp, %eax +1: #endif xorl %edx, %edx call do_notify_resume @@ -584,6 +637,10 @@ END(syscall_exit_work) RING0_INT_FRAME # can't unwind into user space anyway syscall_fault: +#ifdef CONFIG_PAX_MEMORY_UDEREF + push %ss + pop %ds +#endif GET_THREAD_INFO(%ebp) movl $-EFAULT,PT_EAX(%esp) jmp resume_userspace @@ -595,17 +652,24 @@ syscall_badsys: END(syscall_badsys) CFI_ENDPROC -#define FIXUP_ESPFIX_STACK \ - /* since we are on a wrong stack, we cant make it a C code :( */ \ - PER_CPU(gdt_page, %ebx); \ - GET_DESC_BASE(GDT_ENTRY_ESPFIX_SS, %ebx, %eax, %ax, %al, %ah); \ - addl %esp, %eax; \ - pushl $__KERNEL_DS; \ - CFI_ADJUST_CFA_OFFSET 4; \ - pushl %eax; \ - CFI_ADJUST_CFA_OFFSET 4; \ - lss (%esp), %esp; \ +.macro FIXUP_ESPFIX_STACK + /* since we are on a wrong stack, we cant make it a C code :( */ +#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 + GET_DESC_BASE(GDT_ENTRY_ESPFIX_SS, %ebx, %eax, %ax, %al, %ah); + addl %esp, %eax; + pushl $__KERNEL_DS; + CFI_ADJUST_CFA_OFFSET 4; + pushl %eax; + CFI_ADJUST_CFA_OFFSET 4; + lss (%esp), %esp; CFI_ADJUST_CFA_OFFSET -8; +.endm #define UNWIND_ESPFIX_STACK \ movl %ss, %eax; \ /* see if on espfix stack */ \ @@ -622,7 +686,7 @@ END(syscall_badsys) * Build the entry stubs and pointer table with * some assembler magic. */ -.section .rodata,"a" +.section .rodata,"a",@progbits ENTRY(interrupt) .text @@ -722,12 +786,21 @@ error_code: popl %ecx CFI_ADJUST_CFA_OFFSET -4 /*CFI_REGISTER es, ecx*/ + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_EDX + movl %edx, %esi + orl $X86_CR0_WP, %edx + xorl %edx, %esi + SET_CR0_FROM_EDX +#endif + movl PT_FS(%esp), %edi # get the function address movl PT_ORIG_EAX(%esp), %edx # get the error code movl $-1, PT_ORIG_EAX(%esp) # no syscall to restart mov %ecx, PT_FS(%esp) /*CFI_REL_OFFSET fs, ES*/ - movl $(__USER_DS), %ecx + movl $(__KERNEL_DS), %ecx movl %ecx, %ds movl %ecx, %es TRACE_IRQS_OFF @@ -853,6 +926,13 @@ nmi_stack_correct: xorl %edx,%edx # zero error code movl %esp,%eax # pt_regs pointer call do_nmi + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_EDX + xorl %esi, %edx + SET_CR0_FROM_EDX +#endif + jmp restore_nocheck_notrace CFI_ENDPROC @@ -894,6 +974,13 @@ nmi_espfix_stack: FIXUP_ESPFIX_STACK # %eax == %esp xorl %edx,%edx # zero error code call do_nmi + +#ifdef CONFIG_PAX_KERNEXEC + GET_CR0_INTO_EDX + xorl %esi, %edx + SET_CR0_FROM_EDX +#endif + RESTORE_REGS lss 12+4(%esp), %esp # back to espfix stack CFI_ADJUST_CFA_OFFSET -24 @@ -1206,7 +1293,6 @@ END(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_TRACER */ -.section .rodata,"a" #include "syscall_table_32.S" syscall_table_size=(.-sys_call_table) diff -urNp linux-2.6.28.8/arch/x86/kernel/entry_64.S linux-2.6.28.8/arch/x86/kernel/entry_64.S --- linux-2.6.28.8/arch/x86/kernel/entry_64.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/entry_64.S 2009-02-21 09:37:48.000000000 -0500 @@ -911,7 +911,8 @@ END(spurious_interrupt) xorl %ebx,%ebx 1: .if \ist - movq %gs:pda_data_offset, %rbp + imul $TSS_size, %gs:pda_cpunumber, %ebp + lea init_tss(%rbp), %rbp .endif .if \irqtrace TRACE_IRQS_OFF @@ -920,11 +921,11 @@ END(spurious_interrupt) movq ORIG_RAX(%rsp),%rsi movq $-1,ORIG_RAX(%rsp) .if \ist - subq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp) + subq $EXCEPTION_STKSZ, TSS_ist + (\ist - 1) * 8(%rbp) .endif call \sym .if \ist - addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp) + addq $EXCEPTION_STKSZ, TSS_ist + (\ist - 1) * 8(%rbp) .endif DISABLE_INTERRUPTS(CLBR_NONE) .if \irqtrace diff -urNp linux-2.6.28.8/arch/x86/kernel/ftrace.c linux-2.6.28.8/arch/x86/kernel/ftrace.c --- linux-2.6.28.8/arch/x86/kernel/ftrace.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/ftrace.c 2009-02-21 09:37:48.000000000 -0500 @@ -95,9 +95,9 @@ int ftrace_update_ftrace_func(ftrace_fun unsigned char old[MCOUNT_INSN_SIZE], *new; int ret; - memcpy(old, &ftrace_call, MCOUNT_INSN_SIZE); + memcpy(old, (void *)ktla_ktva((unsigned long)ftrace_call), MCOUNT_INSN_SIZE); new = ftrace_call_replace(ip, (unsigned long)func); - ret = ftrace_modify_code(ip, old, new); + ret = ftrace_modify_code(ktla_ktva(ip), old, new); return ret; } diff -urNp linux-2.6.28.8/arch/x86/kernel/head32.c linux-2.6.28.8/arch/x86/kernel/head32.c --- linux-2.6.28.8/arch/x86/kernel/head32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/head32.c 2009-02-21 09:37:48.000000000 -0500 @@ -12,10 +12,11 @@ #include #include #include +#include void __init i386_start_kernel(void) { - reserve_early(__pa_symbol(&_text), __pa_symbol(&_end), "TEXT DATA BSS"); + reserve_early(LOAD_PHYSICAL_ADDR, __pa_symbol(&_end), "TEXT DATA BSS"); #ifdef CONFIG_BLK_DEV_INITRD /* Reserve INITRD */ diff -urNp linux-2.6.28.8/arch/x86/kernel/head_32.S linux-2.6.28.8/arch/x86/kernel/head_32.S --- linux-2.6.28.8/arch/x86/kernel/head_32.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/head_32.S 2009-02-21 09:37:48.000000000 -0500 @@ -19,6 +19,7 @@ #include #include #include +#include /* Physical address */ #define pa(X) ((X) - __PAGE_OFFSET) @@ -64,17 +65,22 @@ LOW_PAGES = 1<<(32-PAGE_SHIFT_asm) LOW_PAGES = LOW_PAGES + 0x1000000 #endif -#if PTRS_PER_PMD > 1 -PAGE_TABLE_SIZE = (LOW_PAGES / PTRS_PER_PMD) + PTRS_PER_PGD -#else -PAGE_TABLE_SIZE = (LOW_PAGES / PTRS_PER_PGD) -#endif +PAGE_TABLE_SIZE = (LOW_PAGES / PTRS_PER_PTE) BOOTBITMAP_SIZE = LOW_PAGES / 8 ALLOCATOR_SLOP = 4 INIT_MAP_BEYOND_END = BOOTBITMAP_SIZE + (PAGE_TABLE_SIZE + ALLOCATOR_SLOP)*PAGE_SIZE_asm /* + * Real beginning of normal "text" segment + */ +ENTRY(stext) +ENTRY(_stext) + +.section .text.startup,"ax",@progbits + ljmp $(__BOOT_CS),$phys_startup_32 + +/* * 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 @@ -82,6 +88,12 @@ INIT_MAP_BEYOND_END = BOOTBITMAP_SIZE + * can. */ .section .text.head,"ax",@progbits + +#ifdef CONFIG_PAX_KERNEXEC +/* PaX: fill first page in .text with int3 to catch NULL derefs in kernel mode */ +.fill 4096,1,0xcc +#endif + ENTRY(startup_32) /* test KEEP_SEGMENTS flag to see if the bootloader is asking us to not reload segments */ @@ -99,6 +111,56 @@ ENTRY(startup_32) movl %eax,%gs 2: + movl $pa(cpu_gdt_table),%edi + movl $__per_cpu_start,%eax + movw %ax,__KERNEL_PERCPU + 2(%edi) + rorl $16,%eax + movb %al,__KERNEL_PERCPU + 4(%edi) + movb %ah,__KERNEL_PERCPU + 7(%edi) + movl $__per_cpu_end + PERCPU_MODULE_RESERVE - 1,%eax + subl $__per_cpu_start,%eax + movw %ax,__KERNEL_PERCPU + 0(%edi) + +#ifdef CONFIG_PAX_MEMORY_UDEREF + /* check for VMware */ + movl $0x564d5868,%eax + xorl %ebx,%ebx + movl $0xa,%ecx + movl $0x5658,%edx + in (%dx),%eax + cmpl $0x564d5868,%ebx + jz 2f + + 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) + addl $PAGE_SIZE_asm,%edi + loop 1b +2: +#endif + +#ifdef CONFIG_PAX_KERNEXEC + movl $pa(boot_gdt),%edi + movl $KERNEL_TEXT_OFFSET,%eax + movw %ax,__BOOT_CS + 2(%edi) + rorl $16,%eax + movb %al,__BOOT_CS + 4(%edi) + movb %ah,__BOOT_CS + 7(%edi) + rorl $16,%eax + + movl $NR_CPUS,%ecx + movl $pa(cpu_gdt_table),%edi +1: + movw %ax,__KERNEL_CS + 2(%edi) + rorl $16,%eax + movb %al,__KERNEL_CS + 4(%edi) + movb %ah,__KERNEL_CS + 7(%edi) + rorl $16,%eax + addl $PAGE_SIZE_asm,%edi + loop 1b +#endif + /* * Clear BSS first so that there are no surprises... */ @@ -142,9 +204,7 @@ ENTRY(startup_32) 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) @@ -156,9 +216,9 @@ WEAK(xen_entry) __INITDATA subarch_entries: - .long default_entry /* normal x86/PC */ - .long lguest_entry /* lguest hypervisor */ - .long xen_entry /* Xen hypervisor */ + .long pa(default_entry) /* normal x86/PC */ + .long pa(lguest_entry) /* lguest hypervisor */ + .long pa(xen_entry) /* Xen hypervisor */ num_subarch_entries = (. - subarch_entries) / 4 .previous #endif /* CONFIG_PARAVIRT */ @@ -220,8 +280,7 @@ default_entry: movl %eax, pa(max_pfn_mapped) /* Do early initialization of the fixmap area */ - movl $pa(swapper_pg_fixmap)+PDE_IDENT_ATTR,%eax - movl %eax,pa(swapper_pg_pmd+0x1000*KPMDS-8) + movl $pa(swapper_pg_fixmap)+PDE_IDENT_ATTR,pa(swapper_pg_pmd+0x1000*KPMDS-8) #else /* Not PAE */ page_pde_offset = (__PAGE_OFFSET >> 20); @@ -253,8 +312,7 @@ page_pde_offset = (__PAGE_OFFSET >> 20); movl %eax, pa(max_pfn_mapped) /* Do early initialization of the fixmap area */ - movl $pa(swapper_pg_fixmap)+PDE_IDENT_ATTR,%eax - movl %eax,pa(swapper_pg_dir+0xffc) + movl $pa(swapper_pg_fixmap)+PDE_IDENT_ATTR,pa(swapper_pg_dir+0xffc) #endif jmp 3f /* @@ -318,13 +376,16 @@ ENTRY(startup_32_smp) jnc 6f /* Setup EFER (Extended Feature Enable Register) */ - movl $0xc0000080, %ecx + movl $MSR_EFER, %ecx rdmsr btsl $11, %eax /* Make changes effective */ wrmsr + btsl $_PAGE_BIT_NX-32,pa(__supported_pte_mask+4) + movl $1,pa(nx_enabled) + 6: /* @@ -350,9 +411,7 @@ ENTRY(startup_32_smp) #ifdef CONFIG_SMP cmpb $0, ready - jz 1f /* Initial CPU cleans BSS */ - jmp checkCPUtype -1: + jnz checkCPUtype /* Initial CPU cleans BSS */ #endif /* CONFIG_SMP */ /* @@ -429,12 +488,12 @@ is386: movl $2,%ecx # set MP ljmp $(__KERNEL_CS),$1f 1: movl $(__KERNEL_DS),%eax # reload all the segment registers movl %eax,%ss # after changing gdt. - movl %eax,%fs # gets reset once there's real percpu - - movl $(__USER_DS),%eax # DS/ES contains default USER segment movl %eax,%ds movl %eax,%es + movl $(__KERNEL_PERCPU), %eax + movl %eax,%fs # set this cpu's percpu + xorl %eax,%eax # Clear GS and LDT movl %eax,%gs lldt %ax @@ -444,12 +503,6 @@ is386: movl $2,%ecx # set MP #ifdef CONFIG_SMP movb ready, %cl movb $1, ready - cmpb $0,%cl # the first CPU calls start_kernel - je 1f - movl $(__KERNEL_PERCPU), %eax - movl %eax,%fs # set this cpu's percpu - movl (stack_start), %esp -1: #endif /* CONFIG_SMP */ jmp *(initial_code) @@ -535,15 +588,15 @@ early_page_fault: jmp early_fault early_fault: - cld #ifdef CONFIG_PRINTK + cmpl $2,%ss:early_recursion_flag + je hlt_loop + incl %ss:early_recursion_flag + cld pusha movl $(__KERNEL_DS),%eax movl %eax,%ds movl %eax,%es - cmpl $2,early_recursion_flag - je hlt_loop - incl early_recursion_flag movl %cr2,%eax pushl %eax pushl %edx /* trapno */ @@ -553,8 +606,8 @@ early_fault: #else call printk #endif -#endif call dump_stack +#endif hlt_loop: hlt jmp hlt_loop @@ -562,8 +615,11 @@ hlt_loop: /* This is the default interrupt "handler" :-) */ ALIGN ignore_int: - cld #ifdef CONFIG_PRINTK + cmpl $2,%ss:early_recursion_flag + je hlt_loop + incl %ss:early_recursion_flag + cld pushl %eax pushl %ecx pushl %edx @@ -572,9 +628,6 @@ 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) @@ -599,36 +652,41 @@ ignore_int: ENTRY(initial_code) .long i386_start_kernel -.section .text -/* - * Real beginning of normal "text" segment - */ -ENTRY(stext) -ENTRY(_stext) - /* * BSS section */ -.section ".bss.page_aligned","wa" - .align PAGE_SIZE_asm #ifdef CONFIG_X86_PAE +.section .swapper_pg_pmd,"a",@progbits swapper_pg_pmd: .fill 1024*KPMDS,4,0 #else +.section .swapper_pg_dir,"a",@progbits ENTRY(swapper_pg_dir) .fill 1024,4,0 #endif swapper_pg_fixmap: .fill 1024,4,0 + +.section .empty_zero_page,"a",@progbits ENTRY(empty_zero_page) .fill 4096,1,0 + +/* + * The IDT has to be page-aligned to simplify the Pentium + * F0 0F bug workaround.. We have a special link segment + * for this. + */ +.section .idt,"a",@progbits +ENTRY(idt_table) + .fill 256,8,0 + /* * This starts the data section. */ +.data + #ifdef CONFIG_X86_PAE -.section ".data.page_aligned","wa" - /* Page-aligned for the benefit of paravirt? */ - .align PAGE_SIZE_asm +.section .swapper_pg_dir,"a",@progbits ENTRY(swapper_pg_dir) .long pa(swapper_pg_pmd+PGD_IDENT_ATTR),0 /* low identity map */ # if KPMDS == 3 @@ -651,11 +709,12 @@ ENTRY(swapper_pg_dir) .data ENTRY(stack_start) - .long init_thread_union+THREAD_SIZE + .long init_thread_union+THREAD_SIZE-8 .long __BOOT_DS ready: .byte 0 +.section .rodata,"a",@progbits early_recursion_flag: .long 0 @@ -691,7 +750,7 @@ 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: @@ -702,7 +761,7 @@ idt_descr: .word 0 # 32 bit align gdt_desc.address ENTRY(early_gdt_descr) .word GDT_ENTRIES*8-1 - .long per_cpu__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 @@ -711,5 +770,59 @@ 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 */ + .quad 0x0000000000000000 /* 0x20 unused */ + .quad 0x0000000000000000 /* 0x28 unused */ + .quad 0x0000000000000000 /* 0x33 TLS entry 1 */ + .quad 0x0000000000000000 /* 0x3b TLS entry 2 */ + .quad 0x0000000000000000 /* 0x43 TLS entry 3 */ + .quad 0x0000000000000000 /* 0x4b reserved */ + .quad 0x0000000000000000 /* 0x53 reserved */ + .quad 0x0000000000000000 /* 0x5b reserved */ + + .quad 0x00cf9b000000ffff /* 0x60 kernel 4GB code at 0x00000000 */ + .quad 0x00cf93000000ffff /* 0x68 kernel 4GB data at 0x00000000 */ + .quad 0x00cffb000000ffff /* 0x73 user 4GB code at 0x00000000 */ + .quad 0x00cff3000000ffff /* 0x7b user 4GB data at 0x00000000 */ + + .quad 0x0000000000000000 /* 0x80 TSS descriptor */ + .quad 0x0000000000000000 /* 0x88 LDT descriptor */ + + /* + * Segments used for calling PnP BIOS have byte granularity. + * The code segments and data segments have fixed 64k limits, + * the transfer segment sizes are set at run time. + */ + .quad 0x00409b000000ffff /* 0x90 32-bit code */ + .quad 0x00009b000000ffff /* 0x98 16-bit code */ + .quad 0x000093000000ffff /* 0xa0 16-bit data */ + .quad 0x0000930000000000 /* 0xa8 16-bit data */ + .quad 0x0000930000000000 /* 0xb0 16-bit data */ + + /* + * The APM segments have byte granularity and their bases + * are set at run time. All have 64k limits. + */ + .quad 0x00409b000000ffff /* 0xb8 APM CS code */ + .quad 0x00009b000000ffff /* 0xc0 APM CS 16 code (16 bit) */ + .quad 0x004093000000ffff /* 0xc8 APM DS data */ + + .quad 0x00c0930000000000 /* 0xd0 - ESPFIX SS */ + .quad 0x0040930000000000 /* 0xd8 - PERCPU */ + .quad 0x0000000000000000 /* 0xe0 - PCIBIOS_CS */ + .quad 0x0000000000000000 /* 0xe8 - PCIBIOS_DS */ + .quad 0x0000000000000000 /* 0xf0 - unused */ + .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 -urNp linux-2.6.28.8/arch/x86/kernel/head64.c linux-2.6.28.8/arch/x86/kernel/head64.c --- linux-2.6.28.8/arch/x86/kernel/head64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/head64.c 2009-02-21 09:37:48.000000000 -0500 @@ -93,6 +93,8 @@ void __init x86_64_start_kernel(char * r /* clear bss before set_intr_gate with early_idt_handler */ clear_bss(); + x86_64_init_pda(); + /* Make NULL pointers segfault */ zap_identity_mappings(); @@ -111,8 +113,6 @@ void __init x86_64_start_kernel(char * r if (console_loglevel == 10) early_printk("Kernel alive\n"); - x86_64_init_pda(); - x86_64_start_reservations(real_mode_data); } diff -urNp linux-2.6.28.8/arch/x86/kernel/head_64.S linux-2.6.28.8/arch/x86/kernel/head_64.S --- linux-2.6.28.8/arch/x86/kernel/head_64.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/head_64.S 2009-02-21 09:37:48.000000000 -0500 @@ -38,6 +38,10 @@ 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_VMEMMAP_START = pgd_index(VMEMMAP_START) +L3_VMEMMAP_START = pud_index(VMEMMAP_START) .text .section .text.head @@ -85,35 +89,22 @@ startup_64: */ addq %rbp, init_level4_pgt + 0(%rip) addq %rbp, init_level4_pgt + (L4_PAGE_OFFSET*8)(%rip) + addq %rbp, init_level4_pgt + (L4_VMALLOC_START*8)(%rip) + addq %rbp, init_level4_pgt + (L4_VMEMMAP_START*8)(%rip) addq %rbp, init_level4_pgt + (L4_START_KERNEL*8)(%rip) addq %rbp, level3_ident_pgt + 0(%rip) + addq %rbp, level3_ident_pgt + 8(%rip) + addq %rbp, level3_ident_pgt + 16(%rip) + addq %rbp, level3_ident_pgt + 24(%rip) - addq %rbp, level3_kernel_pgt + (510*8)(%rip) - addq %rbp, level3_kernel_pgt + (511*8)(%rip) + addq %rbp, level3_vmemmap_pgt + (L3_VMEMMAP_START*8)(%rip) - addq %rbp, level2_fixmap_pgt + (506*8)(%rip) + addq %rbp, level3_kernel_pgt + (L3_START_KERNEL*8)(%rip) + addq %rbp, level3_kernel_pgt + (L3_START_KERNEL*8+8)(%rip) - /* Add an Identity mapping if I am above 1G */ - leaq _text(%rip), %rdi - andq $PMD_PAGE_MASK, %rdi - - movq %rdi, %rax - shrq $PUD_SHIFT, %rax - andq $(PTRS_PER_PUD - 1), %rax - jz ident_complete - - leaq (level2_spare_pgt - __START_KERNEL_map + _KERNPG_TABLE)(%rbp), %rdx - leaq level3_ident_pgt(%rip), %rbx - movq %rdx, 0(%rbx, %rax, 8) - - movq %rdi, %rax - shrq $PMD_SHIFT, %rax - andq $(PTRS_PER_PMD - 1), %rax - leaq __PAGE_KERNEL_IDENT_LARGE_EXEC(%rdi), %rdx - leaq level2_spare_pgt(%rip), %rbx - movq %rdx, 0(%rbx, %rax, 8) -ident_complete: + addq %rbp, level2_fixmap_pgt + (506*8)(%rip) + addq %rbp, level2_fixmap_pgt + (507*8)(%rip) /* * Fixup the kernel text+data virtual addresses. Note that @@ -187,6 +178,10 @@ ENTRY(secondary_startup_64) btl $20,%edi /* No Execute supported? */ jnc 1f btsl $_EFER_NX, %eax + leaq init_level4_pgt(%rip), %rdi + btsq $_PAGE_BIT_NX, 8*L4_PAGE_OFFSET(%rdi) + btsq $_PAGE_BIT_NX, 8*L4_VMALLOC_START(%rdi) + btsq $_PAGE_BIT_NX, 8*L4_VMEMMAP_START(%rdi) 1: wrmsr /* Make changes effective */ /* Setup cr0 */ @@ -257,16 +252,16 @@ ENTRY(secondary_startup_64) .align 8 ENTRY(initial_code) .quad x86_64_start_kernel - __FINITDATA ENTRY(stack_start) .quad init_thread_union+THREAD_SIZE-8 .word 0 + __FINITDATA bad_address: jmp bad_address - .section ".init.text","ax" + __INIT #ifdef CONFIG_EARLY_PRINTK .globl early_idt_handlers early_idt_handlers: @@ -311,18 +306,23 @@ ENTRY(early_idt_handler) #endif /* EARLY_PRINTK */ 1: hlt jmp 1b + .previous #ifdef CONFIG_EARLY_PRINTK + __INITDATA early_recursion_flag: .long 0 + .previous + .section .rodata,"a",@progbits early_idt_msg: .asciz "PANIC: early exception %02lx rip %lx:%lx error %lx cr2 %lx\n" early_idt_ripmsg: .asciz "RIP %s\n" -#endif /* CONFIG_EARLY_PRINTK */ .previous +#endif /* CONFIG_EARLY_PRINTK */ + .section .rodata,"a",@progbits .balign PAGE_SIZE #define NEXT_PAGE(name) \ @@ -347,13 +347,27 @@ 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_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 NEXT_PAGE(level3_ident_pgt) .quad level2_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .fill 511,8,0 + .quad level2_ident_pgt + PAGE_SIZE - __START_KERNEL_map + _KERNPG_TABLE + .quad level2_ident_pgt + 2*PAGE_SIZE - __START_KERNEL_map + _KERNPG_TABLE + .quad level2_ident_pgt + 3*PAGE_SIZE - __START_KERNEL_map + _KERNPG_TABLE + .fill 508,8,0 + +NEXT_PAGE(level3_vmalloc_pgt) + .fill 512,8,0 + +NEXT_PAGE(level3_vmemmap_pgt) + .fill L3_VMEMMAP_START,8,0 + .quad level2_vmemmap_pgt - __START_KERNEL_map + _KERNPG_TABLE NEXT_PAGE(level3_kernel_pgt) .fill L3_START_KERNEL,8,0 @@ -361,20 +375,27 @@ NEXT_PAGE(level3_kernel_pgt) .quad level2_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE .quad level2_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE +NEXT_PAGE(level2_vmemmap_pgt) + .fill 512,8,0 + NEXT_PAGE(level2_fixmap_pgt) .fill 506,8,0 .quad level1_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE - /* 8MB reserved for vsyscalls + a 2MB hole = 4 + 1 entries */ - .fill 5,8,0 + .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(level2_ident_pgt) - /* Since I easily can, map the first 1G. +NEXT_PAGE(level1_vsyscall_pgt) + .fill 512,8,0 + + /* Since I easily can, map the first 4G. * Don't set NX because code runs from these pages. */ - PMDS(0, __PAGE_KERNEL_IDENT_LARGE_EXEC, PTRS_PER_PMD) +NEXT_PAGE(level2_ident_pgt) + PMDS(0, __PAGE_KERNEL_IDENT_LARGE_EXEC, 4*PTRS_PER_PMD) NEXT_PAGE(level2_kernel_pgt) /* @@ -387,32 +408,48 @@ NEXT_PAGE(level2_kernel_pgt) * If you want to increase this then increase MODULES_VADDR * too.) */ - PMDS(0, __PAGE_KERNEL_LARGE_EXEC, - KERNEL_IMAGE_SIZE/PMD_SIZE) - -NEXT_PAGE(level2_spare_pgt) - .fill 512, 8, 0 + PMDS(0, __PAGE_KERNEL_LARGE_EXEC, KERNEL_IMAGE_SIZE/PMD_SIZE) #undef PMDS #undef NEXT_PAGE - .data + .align PAGE_SIZE +ENTRY(cpu_gdt_table) + .rept NR_CPUS + .quad 0x0000000000000000 /* NULL descriptor */ + .quad 0x00cf9b000000ffff /* __KERNEL32_CS */ + .quad 0x00af9b000000ffff /* __KERNEL_CS */ + .quad 0x00cf93000000ffff /* __KERNEL_DS */ + .quad 0x00cffb000000ffff /* __USER32_CS */ + .quad 0x00cff3000000ffff /* __USER_DS, __USER32_DS */ + .quad 0x00affb000000ffff /* __USER_CS */ + .quad 0x0 /* unused */ + .quad 0,0 /* TSS */ + .quad 0,0 /* LDT */ + .quad 0,0,0 /* three TLS descriptors */ + .quad 0x0000f40000000000 /* node/CPU stored in limit */ + /* asm/segment.h:GDT_ENTRIES must match this */ + + /* zero the remaining page */ + .fill PAGE_SIZE / 8 - GDT_ENTRIES,8,0 + .endr + .align 16 .globl early_gdt_descr early_gdt_descr: .word GDT_ENTRIES*8-1 - .quad per_cpu__gdt_page + .quad cpu_gdt_table ENTRY(phys_base) /* This must match the first entry in level2_kernel_pgt */ .quad 0x0000000000000000 #include "../../x86/xen/xen-head.S" - - .section .bss, "aw", @nobits + + .section .rodata,"a",@progbits .align L1_CACHE_BYTES ENTRY(idt_table) - .skip 256 * 16 + .fill 512,8,0 .section .bss.page_aligned, "aw", @nobits .align PAGE_SIZE diff -urNp linux-2.6.28.8/arch/x86/kernel/i386_ksyms_32.c linux-2.6.28.8/arch/x86/kernel/i386_ksyms_32.c --- linux-2.6.28.8/arch/x86/kernel/i386_ksyms_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/i386_ksyms_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -10,8 +10,12 @@ EXPORT_SYMBOL(mcount); #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); @@ -26,3 +30,7 @@ EXPORT_SYMBOL(strstr); EXPORT_SYMBOL(csum_partial); EXPORT_SYMBOL(empty_zero_page); + +#ifdef CONFIG_PAX_KERNEXEC +EXPORT_SYMBOL(KERNEL_TEXT_OFFSET); +#endif diff -urNp linux-2.6.28.8/arch/x86/kernel/init_task.c linux-2.6.28.8/arch/x86/kernel/init_task.c --- linux-2.6.28.8/arch/x86/kernel/init_task.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/init_task.c 2009-02-21 09:37:48.000000000 -0500 @@ -42,5 +42,5 @@ EXPORT_SYMBOL(init_task); * section. Since TSS's are completely CPU-local, we want them * on exact cacheline boundaries, to eliminate cacheline ping-pong. */ -DEFINE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss) = INIT_TSS; - +struct tss_struct init_tss[NR_CPUS] ____cacheline_internodealigned_in_smp = { [0 ... NR_CPUS-1] = INIT_TSS }; +EXPORT_SYMBOL(init_tss); diff -urNp linux-2.6.28.8/arch/x86/kernel/ioport.c linux-2.6.28.8/arch/x86/kernel/ioport.c --- linux-2.6.28.8/arch/x86/kernel/ioport.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/ioport.c 2009-02-21 09:37:48.000000000 -0500 @@ -6,6 +6,7 @@ #include #include #include +#include #include #include #include @@ -41,6 +42,12 @@ asmlinkage long sys_ioperm(unsigned long if ((from + num <= from) || (from + num > IO_BITMAP_BITS)) return -EINVAL; +#ifdef CONFIG_GRKERNSEC_IO + if (turn_on) { + gr_handle_ioperm(); + return -EPERM; + } +#endif if (turn_on && !capable(CAP_SYS_RAWIO)) return -EPERM; @@ -67,7 +74,7 @@ asmlinkage long sys_ioperm(unsigned long * because the ->io_bitmap_max value must match the bitmap * contents: */ - tss = &per_cpu(init_tss, get_cpu()); + tss = init_tss + get_cpu(); set_bitmap(t->io_bitmap_ptr, from, num, !turn_on); @@ -122,8 +129,13 @@ static int do_iopl(unsigned int level, s return -EINVAL; /* Trying to gain more privileges? */ if (level > old) { +#ifdef CONFIG_GRKERNSEC_IO + gr_handle_iopl(); + return -EPERM; +#else if (!capable(CAP_SYS_RAWIO)) return -EPERM; +#endif } regs->flags = (regs->flags & ~X86_EFLAGS_IOPL) | (level << 12); diff -urNp linux-2.6.28.8/arch/x86/kernel/irq_32.c linux-2.6.28.8/arch/x86/kernel/irq_32.c --- linux-2.6.28.8/arch/x86/kernel/irq_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/irq_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -93,7 +93,7 @@ execute_on_irq_stack(int overflow, struc return 0; /* build the stack frame on the IRQ stack */ - isp = (u32 *) ((char*)irqctx + sizeof(*irqctx)); + isp = (u32 *) ((char*)irqctx + sizeof(*irqctx) - 8); irqctx->tinfo.task = curctx->tinfo.task; irqctx->tinfo.previous_esp = current_stack_pointer; @@ -174,7 +174,7 @@ asmlinkage void do_softirq(void) irqctx->tinfo.previous_esp = current_stack_pointer; /* build the stack frame on the softirq stack */ - isp = (u32*) ((char*)irqctx + sizeof(*irqctx)); + isp = (u32*) ((char*)irqctx + sizeof(*irqctx) - 8); call_on_stack(__do_softirq, isp); /* diff -urNp linux-2.6.28.8/arch/x86/kernel/kprobes.c linux-2.6.28.8/arch/x86/kernel/kprobes.c --- linux-2.6.28.8/arch/x86/kernel/kprobes.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/kprobes.c 2009-02-21 09:37:48.000000000 -0500 @@ -166,9 +166,24 @@ static void __kprobes set_jmp_op(void *f char op; s32 raddr; } __attribute__((packed)) * jop; - jop = (struct __arch_jmp_op *)from; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + + jop = (struct __arch_jmp_op *)(ktla_ktva(from)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + jop->raddr = (s32)((long)(to) - ((long)(from) + 5)); jop->op = RELATIVEJUMP_INSTRUCTION; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } /* @@ -342,16 +357,29 @@ static void __kprobes fix_riprel(struct static void __kprobes arch_copy_kprobe(struct kprobe *p) { - memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + + memcpy(p->ainsn.insn, ktla_ktva(p->addr), MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif fix_riprel(p); - if (can_boost(p->addr)) + if (can_boost(ktla_ktva(p->addr))) p->ainsn.boostable = 0; else p->ainsn.boostable = -1; - p->opcode = *p->addr; + p->opcode = *(ktla_ktva(p->addr)); } int __kprobes arch_prepare_kprobe(struct kprobe *p) @@ -428,7 +456,7 @@ static void __kprobes prepare_singlestep if (p->opcode == BREAKPOINT_INSTRUCTION) regs->ip = (unsigned long)p->addr; else - regs->ip = (unsigned long)p->ainsn.insn; + regs->ip = ktva_ktla((unsigned long)p->ainsn.insn); } void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri, @@ -449,7 +477,7 @@ static void __kprobes setup_singlestep(s if (p->ainsn.boostable == 1 && !p->post_handler) { /* Boost up -- we can execute copied instructions directly */ reset_current_kprobe(); - regs->ip = (unsigned long)p->ainsn.insn; + regs->ip = ktva_ktla((unsigned long)p->ainsn.insn); preempt_enable_no_resched(); return; } @@ -519,7 +547,7 @@ static int __kprobes kprobe_handler(stru struct kprobe_ctlblk *kcb; addr = (kprobe_opcode_t *)(regs->ip - sizeof(kprobe_opcode_t)); - if (*addr != BREAKPOINT_INSTRUCTION) { + 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 @@ -770,7 +798,7 @@ static void __kprobes resume_execution(s 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; @@ -953,7 +981,7 @@ int __kprobes kprobe_exceptions_notify(s struct die_args *args = data; int ret = NOTIFY_DONE; - if (args->regs && user_mode_vm(args->regs)) + if (args->regs && user_mode(args->regs)) return ret; switch (val) { diff -urNp linux-2.6.28.8/arch/x86/kernel/ldt.c linux-2.6.28.8/arch/x86/kernel/ldt.c --- linux-2.6.28.8/arch/x86/kernel/ldt.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/ldt.c 2009-02-21 09:37:48.000000000 -0500 @@ -66,13 +66,13 @@ static int alloc_ldt(mm_context_t *pc, i if (reload) { #ifdef CONFIG_SMP preempt_disable(); - load_LDT(pc); + load_LDT_nolock(pc); if (!cpus_equal(current->mm->cpu_vm_mask, cpumask_of_cpu(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 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 retval = copy_ldt(&mm->context, &old_mm->context); mutex_unlock(&old_mm->context.lock); } + + if (tsk == current) { + mm->context.vdso = ~0UL; + +#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, u } } +#ifdef CONFIG_PAX_SEGMEXEC + if ((mm->pax_flags & MF_PAX_SEGMEXEC) && (ldt_info.contents & MODIFY_LDT_CONTENTS_CODE)) { + error = -EINVAL; + goto out_unlock; + } +#endif + fill_ldt(&ldt, &ldt_info); if (oldmode) ldt.avl = 0; diff -urNp linux-2.6.28.8/arch/x86/kernel/machine_kexec_32.c linux-2.6.28.8/arch/x86/kernel/machine_kexec_32.c --- linux-2.6.28.8/arch/x86/kernel/machine_kexec_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/machine_kexec_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -34,7 +34,7 @@ static u32 kexec_pmd1[1024] PAGE_ALIGNED static u32 kexec_pte0[1024] PAGE_ALIGNED; static u32 kexec_pte1[1024] PAGE_ALIGNED; -static void set_idt(void *newidt, __u16 limit) +static void set_idt(struct desc_struct *newidt, __u16 limit) { struct desc_ptr curidt; @@ -46,7 +46,7 @@ static void set_idt(void *newidt, __u16 } -static void set_gdt(void *newgdt, __u16 limit) +static void set_gdt(struct desc_struct *newgdt, __u16 limit) { struct desc_ptr curgdt; @@ -145,7 +145,7 @@ 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 -urNp linux-2.6.28.8/arch/x86/kernel/module_32.c linux-2.6.28.8/arch/x86/kernel/module_32.c --- linux-2.6.28.8/arch/x86/kernel/module_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/module_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -23,6 +23,9 @@ #include #include +#include +#include + #if 0 #define DEBUGP printk #else @@ -33,9 +36,31 @@ void *module_alloc(unsigned long size) { if (size == 0) return NULL; + +#ifdef CONFIG_PAX_KERNEXEC + return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO, PAGE_KERNEL); +#else return vmalloc_exec(size); +#endif + } +#ifdef CONFIG_PAX_KERNEXEC +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_VADDR, (unsigned long)&MODULES_END); + if (area) + return area->addr; + + return NULL; +} +EXPORT_SYMBOL(module_alloc_exec); +#endif /* Free memory returned from module_alloc */ void module_free(struct module *mod, void *module_region) @@ -45,6 +70,45 @@ void module_free(struct module *mod, voi table entries. */ } +#ifdef CONFIG_PAX_KERNEXEC +void module_free_exec(struct module *mod, void *module_region) +{ + struct vm_struct **p, *tmp; + + if (!module_region) + return; + + if ((PAGE_SIZE-1) & (unsigned long)module_region) { + printk(KERN_ERR "Trying to module_free_exec() bad address (%p)\n", module_region); + WARN_ON(1); + return; + } + + write_lock(&vmlist_lock); + for (p = &vmlist; (tmp = *p) != NULL; p = &tmp->next) + if (tmp->addr == module_region) + break; + + if (tmp) { + unsigned long cr0; + + pax_open_kernel(cr0); + memset(tmp->addr, 0xCC, tmp->size); + pax_close_kernel(cr0); + + *p = tmp->next; + kfree(tmp); + } + write_unlock(&vmlist_lock); + + if (!tmp) { + printk(KERN_ERR "Trying to module_free_exec() nonexistent vm area (%p)\n", + module_region); + WARN_ON(1); + } +} +#endif + /* We don't need anything special. */ int module_frob_arch_sections(Elf_Ehdr *hdr, Elf_Shdr *sechdrs, @@ -63,14 +127,20 @@ 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; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif 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 @@ -78,12 +148,32 @@ int apply_relocate(Elf32_Shdr *sechdrs, switch (ELF32_R_TYPE(rel[i].r_info)) { case R_386_32: + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + /* We add the value into the location given */ - *location += sym->st_value; + *plocation += sym->st_value; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + break; case R_386_PC32: + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + /* Add the value, subtract its postition */ - *location += sym->st_value - (uint32_t)location; + *plocation += sym->st_value - location; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + break; default: printk(KERN_ERR "module %s: Unknown relocation: %u\n", diff -urNp linux-2.6.28.8/arch/x86/kernel/module_64.c linux-2.6.28.8/arch/x86/kernel/module_64.c --- linux-2.6.28.8/arch/x86/kernel/module_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/module_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -40,7 +40,7 @@ void module_free(struct module *mod, voi table entries. */ } -void *module_alloc(unsigned long size) +static void *__module_alloc(unsigned long size, pgprot_t prot) { struct vm_struct *area; @@ -54,8 +54,31 @@ void *module_alloc(unsigned long size) if (!area) return NULL; - return __vmalloc_area(area, GFP_KERNEL, PAGE_KERNEL_EXEC); + return __vmalloc_area(area, GFP_KERNEL | __GFP_ZERO, prot); +} + +#ifdef CONFIG_PAX_KERNEXEC +void *module_alloc(unsigned long size) +{ + return __module_alloc(size, PAGE_KERNEL); +} + +void module_free_exec(struct module *mod, void *module_region) +{ + module_free(mod, module_region); +} + +void *module_alloc_exec(unsigned long size) +{ + return __module_alloc(size, PAGE_KERNEL_RX); } +#else +void *module_alloc(unsigned long size) +{ + return __module_alloc(size, PAGE_KERNEL_EXEC); +} +#endif + #endif /* We don't need anything special. */ @@ -77,7 +100,11 @@ int apply_relocate_add(Elf64_Shdr *sechd Elf64_Rela *rel = (void *)sechdrs[relsec].sh_addr; Elf64_Sym *sym; void *loc; - u64 val; + u64 val; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif DEBUGP("Applying relocate section %u to %u\n", relsec, sechdrs[relsec].sh_info); @@ -101,21 +128,61 @@ int apply_relocate_add(Elf64_Shdr *sechd case R_X86_64_NONE: break; case R_X86_64_64: + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + *(u64 *)loc = val; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + break; case R_X86_64_32: + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + *(u32 *)loc = val; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + if (val != *(u32 *)loc) goto overflow; break; case R_X86_64_32S: + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + *(s32 *)loc = val; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + if ((s64)val != *(s32 *)loc) goto overflow; break; case R_X86_64_PC32: val -= (u64)loc; + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + *(u32 *)loc = val; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + #if 0 if ((s64)val != *(s32 *)loc) goto overflow; diff -urNp linux-2.6.28.8/arch/x86/kernel/paravirt.c linux-2.6.28.8/arch/x86/kernel/paravirt.c --- linux-2.6.28.8/arch/x86/kernel/paravirt.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/paravirt.c 2009-03-07 10:34:42.000000000 -0500 @@ -44,7 +44,7 @@ void _paravirt_nop(void) { } -static void __init default_banner(void) +static void default_banner(void) { printk(KERN_INFO "Booting paravirtualized kernel on %s\n", pv_info.name); @@ -164,7 +164,7 @@ unsigned paravirt_patch_insns(void *insn 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; } @@ -292,21 +292,21 @@ void arch_flush_lazy_cpu_mode(void) preempt_enable(); } -struct pv_info pv_info = { +struct pv_info pv_info __read_only = { .name = "bare hardware", .paravirt_enabled = 0, .kernel_rpl = 0, .shared_kernel_pmd = 1, /* Only used when CONFIG_X86_PAE is set */ }; -struct pv_init_ops pv_init_ops = { +struct pv_init_ops pv_init_ops __read_only = { .patch = native_patch, .banner = default_banner, .arch_setup = paravirt_nop, .memory_setup = machine_specific_memory_setup, }; -struct pv_time_ops pv_time_ops = { +struct pv_time_ops pv_time_ops __read_only = { .time_init = hpet_time_init, .get_wallclock = native_get_wallclock, .set_wallclock = native_set_wallclock, @@ -314,7 +314,7 @@ struct pv_time_ops pv_time_ops = { .get_tsc_khz = native_calibrate_tsc, }; -struct pv_irq_ops pv_irq_ops = { +struct pv_irq_ops pv_irq_ops __read_only = { .init_IRQ = native_init_IRQ, .save_fl = native_save_fl, .restore_fl = native_restore_fl, @@ -327,7 +327,7 @@ struct pv_irq_ops pv_irq_ops = { #endif }; -struct pv_cpu_ops pv_cpu_ops = { +struct pv_cpu_ops pv_cpu_ops __read_only = { .cpuid = native_cpuid, .get_debugreg = native_get_debugreg, .set_debugreg = native_set_debugreg, @@ -389,7 +389,7 @@ struct pv_cpu_ops pv_cpu_ops = { }, }; -struct pv_apic_ops pv_apic_ops = { +struct pv_apic_ops pv_apic_ops __read_only = { #ifdef CONFIG_X86_LOCAL_APIC .setup_boot_clock = setup_boot_APIC_clock, .setup_secondary_clock = setup_secondary_APIC_clock, @@ -397,7 +397,7 @@ struct pv_apic_ops pv_apic_ops = { #endif }; -struct pv_mmu_ops pv_mmu_ops = { +struct pv_mmu_ops pv_mmu_ops __read_only = { #ifndef CONFIG_X86_64 .pagetable_setup_start = native_pagetable_setup_start, .pagetable_setup_done = native_pagetable_setup_done, diff -urNp linux-2.6.28.8/arch/x86/kernel/paravirt-spinlocks.c linux-2.6.28.8/arch/x86/kernel/paravirt-spinlocks.c --- linux-2.6.28.8/arch/x86/kernel/paravirt-spinlocks.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/paravirt-spinlocks.c 2009-02-21 09:37:48.000000000 -0500 @@ -13,7 +13,7 @@ default_spin_lock_flags(raw_spinlock_t * __raw_spin_lock(lock); } -struct pv_lock_ops pv_lock_ops = { +struct pv_lock_ops pv_lock_ops __read_only = { #ifdef CONFIG_SMP .spin_is_locked = __ticket_spin_is_locked, .spin_is_contended = __ticket_spin_is_contended, diff -urNp linux-2.6.28.8/arch/x86/kernel/process_32.c linux-2.6.28.8/arch/x86/kernel/process_32.c --- linux-2.6.28.8/arch/x86/kernel/process_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/process_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -65,8 +65,10 @@ asmlinkage void ret_from_fork(void) __as DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task; EXPORT_PER_CPU_SYMBOL(current_task); +#ifdef CONFIG_SMP DEFINE_PER_CPU(int, cpu_number); EXPORT_PER_CPU_SYMBOL(cpu_number); +#endif /* * Return saved PC of a blocked thread. @@ -74,6 +76,7 @@ EXPORT_PER_CPU_SYMBOL(cpu_number); unsigned long thread_saved_pc(struct task_struct *tsk) { return ((unsigned long *)tsk->thread.sp)[3]; +//XXX return tsk->thread.eip; } #ifndef CONFIG_SMP @@ -131,7 +134,7 @@ void __show_regs(struct pt_regs *regs, i unsigned short ss, gs; const char *board; - if (user_mode_vm(regs)) { + if (user_mode(regs)) { sp = regs->sp; ss = regs->ss & 0xffff; savesegment(gs, gs); @@ -212,8 +215,8 @@ int kernel_thread(int (*fn)(void *), voi regs.bx = (unsigned long) fn; regs.dx = (unsigned long) arg; - regs.ds = __USER_DS; - regs.es = __USER_DS; + regs.ds = __KERNEL_DS; + regs.es = __KERNEL_DS; regs.fs = __KERNEL_PERCPU; regs.orig_ax = -1; regs.ip = (unsigned long) kernel_thread_helper; @@ -235,7 +238,7 @@ void exit_thread(void) struct task_struct *tsk = current; struct thread_struct *t = &tsk->thread; int cpu = get_cpu(); - struct tss_struct *tss = &per_cpu(init_tss, cpu); + struct tss_struct *tss = init_tss + cpu; kfree(t->io_bitmap_ptr); t->io_bitmap_ptr = NULL; @@ -264,6 +267,7 @@ void flush_thread(void) { struct task_struct *tsk = current; + loadsegment(gs, 0); tsk->thread.debugreg0 = 0; tsk->thread.debugreg1 = 0; tsk->thread.debugreg2 = 0; @@ -303,7 +307,7 @@ int copy_thread(int nr, unsigned long cl struct task_struct *tsk; int err; - childregs = task_pt_regs(p); + childregs = task_stack_page(p) + THREAD_SIZE - sizeof(struct pt_regs) - 8; *childregs = *regs; childregs->ax = 0; childregs->sp = sp; @@ -332,6 +336,7 @@ int copy_thread(int nr, unsigned long cl * Set a new TLS for the child thread? */ if (clone_flags & CLONE_SETTLS) +//XXX needs set_fs()? err = do_set_thread_area(p, -1, (struct user_desc __user *)childregs->si, 0); @@ -553,7 +558,7 @@ struct task_struct * __switch_to(struct 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; /* never put a printk in __switch_to... printk() calls wake_up*() indirectly */ @@ -581,6 +586,11 @@ struct task_struct * __switch_to(struct */ savesegment(gs, prev->gs); +#ifdef CONFIG_PAX_MEMORY_UDEREF + if (!segment_eq(task_thread_info(prev_p)->addr_limit, task_thread_info(next_p)->addr_limit)) + __set_fs(task_thread_info(next_p)->addr_limit, cpu); +#endif + /* * Load the per-thread Thread-Local Storage descriptor. */ @@ -719,15 +729,27 @@ unsigned long get_wchan(struct task_stru return 0; } -unsigned long arch_align_stack(unsigned long sp) +#ifdef CONFIG_PAX_RANDKSTACK +asmlinkage void pax_randomize_kstack(void) { - 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; + + rdtscl(time); + + /* P4 seems to return a 0 LSB, ignore it */ +#ifdef CONFIG_MPENTIUM4 + time &= 0x1EUL; + time <<= 2; +#else + time &= 0xFUL; + time <<= 3; +#endif + + thread->sp0 ^= time; + load_sp0(init_tss + smp_processor_id(), thread); } +#endif diff -urNp linux-2.6.28.8/arch/x86/kernel/process_64.c linux-2.6.28.8/arch/x86/kernel/process_64.c --- linux-2.6.28.8/arch/x86/kernel/process_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/process_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -109,6 +109,8 @@ static inline void play_dead(void) void cpu_idle(void) { current_thread_info()->status |= TS_POLLING; + current->stack_canary = pax_get_random_long(); + write_pda(stack_canary, current->stack_canary); /* endless idle loop with no priority at all */ while (1) { tick_nohz_stop_sched_tick(1); @@ -223,7 +225,7 @@ void exit_thread(void) struct thread_struct *t = &me->thread; if (me->thread.io_bitmap_ptr) { - struct tss_struct *tss = &per_cpu(init_tss, get_cpu()); + struct tss_struct *tss = init_tss + get_cpu(); kfree(t->io_bitmap_ptr); t->io_bitmap_ptr = NULL; @@ -558,7 +560,7 @@ __switch_to(struct task_struct *prev_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; /* we're going to use this soon, after a few expensive things */ @@ -647,7 +649,6 @@ __switch_to(struct task_struct *prev_p, (unsigned long)task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET); #ifdef CONFIG_CC_STACKPROTECTOR - write_pda(stack_canary, next_p->stack_canary); /* * Build time only check to make sure the stack_canary is at * offset 40 in the pda; this is a gcc ABI requirement @@ -746,12 +747,11 @@ unsigned long get_wchan(struct task_stru 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-8-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-8-sizeof(u64)) return 0; ip = *(u64 *)(fp+8); if (!in_sched_functions(ip)) @@ -860,16 +860,3 @@ long sys_arch_prctl(int code, unsigned l { return do_arch_prctl(current, code, addr); } - -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; -} - -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; -} diff -urNp linux-2.6.28.8/arch/x86/kernel/ptrace.c linux-2.6.28.8/arch/x86/kernel/ptrace.c --- linux-2.6.28.8/arch/x86/kernel/ptrace.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/ptrace.c 2009-03-07 10:29:51.000000000 -0500 @@ -1502,7 +1502,7 @@ void send_sigtrap(struct task_struct *ts info.si_code = si_code; /* User-mode ip? */ - info.si_addr = user_mode_vm(regs) ? (void __user *) regs->ip : NULL; + info.si_addr = user_mode(regs) ? (void __user *) regs->ip : NULL; /* Send us the fake SIGTRAP */ force_sig_info(SIGTRAP, &info, tsk); diff -urNp linux-2.6.28.8/arch/x86/kernel/reboot.c linux-2.6.28.8/arch/x86/kernel/reboot.c --- linux-2.6.28.8/arch/x86/kernel/reboot.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/reboot.c 2009-03-07 10:32:37.000000000 -0500 @@ -28,7 +28,7 @@ void (*pm_power_off)(void); EXPORT_SYMBOL(pm_power_off); static const struct desc_ptr no_idt = {}; -static int reboot_mode; +static unsigned short reboot_mode; enum reboot_type reboot_type = BOOT_KBD; int reboot_force; @@ -210,7 +210,7 @@ static struct dmi_system_id __initdata r DMI_MATCH(DMI_PRODUCT_NAME, "Dell XPS710"), }, }, - { } + { NULL, NULL, {{0, {0}}}, NULL} }; static int __init reboot_init(void) @@ -226,12 +226,12 @@ core_initcall(reboot_init); controller to pulse the CPU reset line, which is more thorough, but doesn't work with at least one type of 486 motherboard. It is easy to stop this code working; hence the copious comments. */ -static const unsigned long long -real_mode_gdt_entries [3] = +static struct desc_struct +real_mode_gdt_entries [3] __read_only = { - 0x0000000000000000ULL, /* Null descriptor */ - 0x00009b000000ffffULL, /* 16-bit real-mode 64k code at 0x00000000 */ - 0x000093000100ffffULL /* 16-bit real-mode 64k data at 0x00000100 */ + {{{0x00000000, 0x00000000}}}, /* Null descriptor */ + {{{0x0000ffff, 0x00009b00}}}, /* 16-bit real-mode 64k code at 0x00000000 */ + {{{0x0100ffff, 0x00009300}}} /* 16-bit real-mode 64k data at 0x00000100 */ }; static const struct desc_ptr @@ -280,7 +280,7 @@ static const unsigned char jump_to_bios * specified by the code and length parameters. * We assume that length will aways be less that 100! */ -void machine_real_restart(const unsigned char *code, int length) +void machine_real_restart(const unsigned char *code, unsigned int length) { local_irq_disable(); @@ -300,8 +300,8 @@ void machine_real_restart(const unsigned /* Remap the kernel at virtual address zero, as well as offset zero from the kernel segment. This assumes the kernel segment starts at virtual address PAGE_OFFSET. */ - memcpy(swapper_pg_dir, swapper_pg_dir + KERNEL_PGD_BOUNDARY, - sizeof(swapper_pg_dir [0]) * KERNEL_PGD_PTRS); + clone_pgd_range(swapper_pg_dir, swapper_pg_dir + KERNEL_PGD_BOUNDARY, + min_t(unsigned long, KERNEL_PGD_PTRS, KERNEL_PGD_BOUNDARY)); /* * Use `swapper_pg_dir' as our page directory. @@ -313,16 +313,15 @@ void machine_real_restart(const unsigned boot)". This seems like a fairly standard thing that gets set by REBOOT.COM programs, and the previous reset routine did this too. */ - *((unsigned short *)0x472) = reboot_mode; + *(unsigned short *)(__va(0x472)) = reboot_mode; /* For the switch to real mode, copy some code to low memory. It has to be in the first 64k because it is running in 16-bit mode, and it has to have the same physical and virtual address, because it turns off paging. Copy it near the end of the first page, out of the way of BIOS variables. */ - memcpy((void *)(0x1000 - sizeof(real_mode_switch) - 100), - real_mode_switch, sizeof (real_mode_switch)); - memcpy((void *)(0x1000 - 100), code, length); + memcpy(__va(0x1000 - sizeof (real_mode_switch) - 100), real_mode_switch, sizeof (real_mode_switch)); + memcpy(__va(0x1000 - 100), code, length); /* Set up the IDT for real mode. */ load_idt(&real_mode_idt); diff -urNp linux-2.6.28.8/arch/x86/kernel/setup.c linux-2.6.28.8/arch/x86/kernel/setup.c --- linux-2.6.28.8/arch/x86/kernel/setup.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/setup.c 2009-02-21 09:37:48.000000000 -0500 @@ -738,6 +738,7 @@ void start_periodic_check_for_corruption } #endif +#ifdef CONFIG_X86_RESERVE_LOW_64K static int __init dmi_low_memory_corruption(const struct dmi_system_id *d) { printk(KERN_NOTICE @@ -749,6 +750,7 @@ static int __init dmi_low_memory_corrupt return 0; } +#endif /* List of systems that have known low memory corruption BIOS problems */ static struct dmi_system_id __initdata bad_bios_dmi_table[] = { @@ -845,8 +847,8 @@ 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; #ifdef CONFIG_X86_32 init_mm.brk = init_pg_tables_end + PAGE_OFFSET; @@ -854,9 +856,9 @@ void __init setup_arch(char **cmdline_p) init_mm.brk = (unsigned long) &_end; #endif - code_resource.start = virt_to_phys(_text); - code_resource.end = virt_to_phys(_etext)-1; - data_resource.start = virt_to_phys(_etext); + code_resource.start = virt_to_phys(ktla_ktva(_text)); + code_resource.end = virt_to_phys(ktla_ktva(_etext))-1; + data_resource.start = virt_to_phys(_data); data_resource.end = virt_to_phys(_edata)-1; bss_resource.start = virt_to_phys(&__bss_start); bss_resource.end = virt_to_phys(&__bss_stop)-1; diff -urNp linux-2.6.28.8/arch/x86/kernel/setup_percpu.c linux-2.6.28.8/arch/x86/kernel/setup_percpu.c --- linux-2.6.28.8/arch/x86/kernel/setup_percpu.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/setup_percpu.c 2009-02-21 09:37:48.000000000 -0500 @@ -179,7 +179,11 @@ void __init setup_per_cpu_areas(void) cpu, node, __pa(ptr)); } #endif +#ifdef CONFIG_X86_32 + __per_cpu_offset[cpu] = ptr - __per_cpu_start; +#else per_cpu_offset(cpu) = ptr - __per_cpu_start; +#endif memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); } diff -urNp linux-2.6.28.8/arch/x86/kernel/signal_32.c linux-2.6.28.8/arch/x86/kernel/signal_32.c --- linux-2.6.28.8/arch/x86/kernel/signal_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/signal_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -367,9 +367,9 @@ __setup_frame(int sig, struct k_sigactio } if (current->mm->context.vdso) - restorer = VDSO32_SYMBOL(current->mm->context.vdso, sigreturn); + restorer = (void __user *)VDSO32_SYMBOL(current->mm->context.vdso, sigreturn); else - restorer = &frame->retcode; + restorer = (void __user *)&frame->retcode; if (ka->sa.sa_flags & SA_RESTORER) restorer = ka->sa.sa_restorer; @@ -442,7 +442,7 @@ static int __setup_rt_frame(int sig, str return -EFAULT; /* Set up to return from userspace. */ - restorer = VDSO32_SYMBOL(current->mm->context.vdso, rt_sigreturn); + restorer = (void __user *)VDSO32_SYMBOL(current->mm->context.vdso, rt_sigreturn); if (ka->sa.sa_flags & SA_RESTORER) restorer = ka->sa.sa_restorer; err |= __put_user(restorer, &frame->pretcode); @@ -612,7 +612,7 @@ static void do_signal(struct pt_regs *re * X86_32: vm86 regs switched out by assembly code before reaching * here, so testing against kernel CS suffices. */ - if (!user_mode(regs)) + if (!user_mode_novm(regs)) return; if (current_thread_info()->status & TS_RESTORE_SIGMASK) diff -urNp linux-2.6.28.8/arch/x86/kernel/signal_64.c linux-2.6.28.8/arch/x86/kernel/signal_64.c --- linux-2.6.28.8/arch/x86/kernel/signal_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/signal_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -239,8 +239,8 @@ static int __setup_rt_frame(int sig, str err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, set->sig[0], me); err |= __put_user(fp, &frame->uc.uc_mcontext.fpstate); if (sizeof(*set) == 16) { - __put_user(set->sig[0], &frame->uc.uc_sigmask.sig[0]); - __put_user(set->sig[1], &frame->uc.uc_sigmask.sig[1]); + err |= __put_user(set->sig[0], &frame->uc.uc_sigmask.sig[0]); + err |= __put_user(set->sig[1], &frame->uc.uc_sigmask.sig[1]); } else err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); diff -urNp linux-2.6.28.8/arch/x86/kernel/smpboot.c linux-2.6.28.8/arch/x86/kernel/smpboot.c --- linux-2.6.28.8/arch/x86/kernel/smpboot.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/smpboot.c 2009-02-21 09:37:48.000000000 -0500 @@ -814,6 +814,11 @@ static int __cpuinit do_boot_cpu(int api .cpu = cpu, .done = COMPLETION_INITIALIZER_ONSTACK(c_idle.done), }; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + INIT_WORK(&c_idle.work, do_fork_idle); #ifdef CONFIG_X86_64 @@ -864,7 +869,17 @@ do_rest: cpu_pda(cpu)->pcurrent = c_idle.idle; clear_tsk_thread_flag(c_idle.idle, TIF_FORK); #endif + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + initial_code = (unsigned long)start_secondary; stack_start.sp = (void *) c_idle.idle->thread.sp; diff -urNp linux-2.6.28.8/arch/x86/kernel/smpcommon.c linux-2.6.28.8/arch/x86/kernel/smpcommon.c --- linux-2.6.28.8/arch/x86/kernel/smpcommon.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/smpcommon.c 2009-02-21 09:37:48.000000000 -0500 @@ -3,9 +3,10 @@ */ #include #include +#include #ifdef CONFIG_X86_32 -DEFINE_PER_CPU(unsigned long, this_cpu_off); +DEFINE_PER_CPU(unsigned long, this_cpu_off) = (unsigned long)__per_cpu_start; EXPORT_PER_CPU_SYMBOL(this_cpu_off); /* @@ -15,16 +16,19 @@ EXPORT_PER_CPU_SYMBOL(this_cpu_off); */ __cpuinit void init_gdt(int cpu) { - struct desc_struct gdt; + struct desc_struct d, *gdt = get_cpu_gdt_table(cpu); + unsigned long base, limit; - pack_descriptor(&gdt, __per_cpu_offset[cpu], 0xFFFFF, - 0x2 | DESCTYPE_S, 0x8); - gdt.s = 1; + base = per_cpu_offset(cpu); + limit = PERCPU_ENOUGH_ROOM - 1; + if (limit < 64*1024) + pack_descriptor(&d, base, limit, 0x80 | DESCTYPE_S | 0x3, 0x4); + else + pack_descriptor(&d, base, limit >> PAGE_SHIFT, 0x80 | DESCTYPE_S | 0x3, 0xC); - write_gdt_entry(get_cpu_gdt_table(cpu), - GDT_ENTRY_PERCPU, &gdt, DESCTYPE_S); + write_gdt_entry(gdt, GDT_ENTRY_PERCPU, &d, DESCTYPE_S); - per_cpu(this_cpu_off, cpu) = __per_cpu_offset[cpu]; + per_cpu(this_cpu_off, cpu) = base; per_cpu(cpu_number, cpu) = cpu; } #endif diff -urNp linux-2.6.28.8/arch/x86/kernel/step.c linux-2.6.28.8/arch/x86/kernel/step.c --- linux-2.6.28.8/arch/x86/kernel/step.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/step.c 2009-02-21 09:37:48.000000000 -0500 @@ -23,22 +23,20 @@ unsigned long convert_ip_to_linear(struc * and APM bios ones we just ignore here. */ if ((seg & SEGMENT_TI_MASK) == SEGMENT_LDT) { - u32 *desc; + 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)) - addr = -1L; /* bogus selector, access would fault */ + if (unlikely(seg >= child->mm->context.size)) + addr = -EINVAL; else { - desc = child->mm->context.ldt + seg; - base = ((desc[0] >> 16) | - ((desc[1] & 0xff) << 16) | - (desc[1] & 0xff000000)); + desc = &child->mm->context.ldt[seg]; + base = (desc->a >> 16) | ((desc->b & 0xff) << 16) | (desc->b & 0xff000000); /* 16-bit code segment? */ - if (!((desc[1] >> 22) & 1)) + if (!((desc->b >> 22) & 1)) addr &= 0xffff; addr += base; } @@ -54,6 +52,9 @@ static int is_setting_trap_flag(struct t 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]) { @@ -75,7 +76,7 @@ static int is_setting_trap_flag(struct t #ifdef CONFIG_X86_64 case 0x40 ... 0x4f: - if (regs->cs != __USER_CS) + if ((regs->cs & 0xffff) != __USER_CS) /* 32-bit mode: register increment */ return 0; /* 64-bit mode: REX prefix */ diff -urNp linux-2.6.28.8/arch/x86/kernel/syscall_table_32.S linux-2.6.28.8/arch/x86/kernel/syscall_table_32.S --- linux-2.6.28.8/arch/x86/kernel/syscall_table_32.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/syscall_table_32.S 2009-02-21 09:37:48.000000000 -0500 @@ -1,3 +1,4 @@ +.section .rodata,"a",@progbits ENTRY(sys_call_table) .long sys_restart_syscall /* 0 - old "setup()" system call, used for restarting */ .long sys_exit diff -urNp linux-2.6.28.8/arch/x86/kernel/sys_i386_32.c linux-2.6.28.8/arch/x86/kernel/sys_i386_32.c --- linux-2.6.28.8/arch/x86/kernel/sys_i386_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/sys_i386_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -24,6 +24,21 @@ #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 (len > pax_task_size || addr > pax_task_size - len) + return -EINVAL; + + return 0; +} + asmlinkage long sys_mmap2(unsigned long addr, unsigned long len, unsigned long prot, unsigned long flags, unsigned long fd, unsigned long pgoff) @@ -83,6 +98,205 @@ out: return err; } +unsigned long +arch_get_unmapped_area(struct file *filp, unsigned long addr, + unsigned long len, unsigned long pgoff, unsigned long flags) +{ + struct mm_struct *mm = current->mm; + struct vm_area_struct *vma; + unsigned long start_addr, pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + 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); + vma = find_vma(mm, addr); + if (pax_task_size - len >= addr && + (!vma || addr + len <= vma->vm_start)) + return addr; + } + if (len > mm->cached_hole_size) { + start_addr = addr = mm->free_area_cache; + } else { + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; + } + +#ifdef CONFIG_PAX_PAGEEXEC + if (!nx_enabled && (mm->pax_flags & MF_PAX_PAGEEXEC) && (flags & MAP_EXECUTABLE) && start_addr >= mm->mmap_base) { + start_addr = 0x00110000UL; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + start_addr += mm->delta_mmap & 0x03FFF000UL; +#endif + + if (mm->start_brk <= start_addr && start_addr < mm->mmap_base) + start_addr = addr = mm->mmap_base; + else + addr = start_addr; + } +#endif + +full_search: + for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { + /* At this point: (!vma || addr < vma->vm_end). */ + if (pax_task_size - len < addr) { + /* + * Start a new search - just in case we missed + * some holes. + */ + if (start_addr != mm->mmap_base) { + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; + goto full_search; + } + return -ENOMEM; + } + if (!vma || addr + len <= vma->vm_start) { + /* + * Remember the place where we stopped the search: + */ + mm->free_area_cache = addr + len; + return addr; + } + if (addr + mm->cached_hole_size < vma->vm_start) + mm->cached_hole_size = vma->vm_start - addr; + addr = vma->vm_end; + if (mm->start_brk <= addr && addr < mm->mmap_base) { + start_addr = addr = mm->mmap_base; + mm->cached_hole_size = 0; + goto full_search; + } + } +} + +unsigned long +arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, + const unsigned long len, const unsigned long pgoff, + const unsigned long flags) +{ + struct vm_area_struct *vma; + struct mm_struct *mm = current->mm; + unsigned long base = mm->mmap_base, addr = addr0, pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + /* 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 (!nx_enabled && (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); + vma = find_vma(mm, addr); + if (pax_task_size - len >= addr && + (!vma || addr + len <= vma->vm_start)) + return addr; + } + + /* check if free_area_cache is useful for us */ + if (len <= mm->cached_hole_size) { + mm->cached_hole_size = 0; + mm->free_area_cache = mm->mmap_base; + } + + /* either no address requested or can't fit in requested address hole */ + addr = mm->free_area_cache; + + /* make sure it can fit in the remaining address space */ + if (addr > len) { + vma = find_vma(mm, addr-len); + if (!vma || addr <= vma->vm_start) + /* remember the address as a hint for next time */ + return (mm->free_area_cache = addr-len); + } + + if (mm->mmap_base < len) + goto bottomup; + + addr = mm->mmap_base-len; + + do { + /* + * Lookup failure means no vma is above this address, + * else if new region fits below vma->vm_start, + * return with success: + */ + vma = find_vma(mm, addr); + if (!vma || addr+len <= vma->vm_start) + /* remember the address as a hint for next time */ + return (mm->free_area_cache = addr); + + /* remember the largest hole we saw so far */ + if (addr + mm->cached_hole_size < vma->vm_start) + mm->cached_hole_size = vma->vm_start - addr; + + /* try just below the current vma->vm_start */ + addr = vma->vm_start-len; + } while (len < vma->vm_start); + +bottomup: + /* + * A failed mmap() very likely causes application failure, + * so fall back to the bottom-up function here. This scenario + * can happen with large stack limits and large mmap() + * allocations. + */ + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + mm->mmap_base = SEGMEXEC_TASK_UNMAPPED_BASE; + else +#endif + + mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + + mm->free_area_cache = mm->mmap_base; + mm->cached_hole_size = ~0UL; + addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags); + /* + * Restore the topdown base: + */ + mm->mmap_base = base; + mm->free_area_cache = base; + mm->cached_hole_size = ~0UL; + + return addr; +} struct sel_arg_struct { unsigned long n; diff -urNp linux-2.6.28.8/arch/x86/kernel/sys_x86_64.c linux-2.6.28.8/arch/x86/kernel/sys_x86_64.c --- linux-2.6.28.8/arch/x86/kernel/sys_x86_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/sys_x86_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -47,8 +47,8 @@ out: return error; } -static void find_start_end(unsigned long flags, unsigned long *begin, - unsigned long *end) +static void find_start_end(struct mm_struct *mm, unsigned long flags, + unsigned long *begin, unsigned long *end) { if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) { unsigned long new_begin; @@ -67,7 +67,7 @@ static void find_start_end(unsigned long *begin = new_begin; } } else { - *begin = TASK_UNMAPPED_BASE; + *begin = mm->mmap_base; *end = TASK_SIZE; } } @@ -84,11 +84,15 @@ arch_get_unmapped_area(struct file *filp 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); @@ -143,7 +147,7 @@ arch_get_unmapped_area_topdown(struct fi { struct vm_area_struct *vma; struct mm_struct *mm = current->mm; - unsigned long addr = addr0; + unsigned long base = mm->mmap_base, addr = addr0; /* requested length too big for entire address space */ if (len > TASK_SIZE) @@ -156,6 +160,10 @@ arch_get_unmapped_area_topdown(struct fi if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) goto bottomup; +#ifdef CONFIG_PAX_RANDMMAP + if (!(mm->pax_flags & MF_PAX_RANDMMAP)) +#endif + /* requesting a specific address */ if (addr) { addr = PAGE_ALIGN(addr); @@ -213,13 +221,21 @@ bottomup: * can happen with large stack limits and large mmap() * allocations. */ + mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + + mm->free_area_cache = mm->mmap_base; mm->cached_hole_size = ~0UL; - mm->free_area_cache = TASK_UNMAPPED_BASE; addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags); /* * Restore the topdown base: */ - mm->free_area_cache = mm->mmap_base; + mm->mmap_base = base; + mm->free_area_cache = base; mm->cached_hole_size = ~0UL; return addr; diff -urNp linux-2.6.28.8/arch/x86/kernel/time_32.c linux-2.6.28.8/arch/x86/kernel/time_32.c --- linux-2.6.28.8/arch/x86/kernel/time_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/time_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -47,22 +47,32 @@ unsigned long profile_pc(struct pt_regs unsigned long pc = instruction_pointer(regs); #ifdef CONFIG_SMP - 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 *)®s->sp; /* Return address is either directly at stack pointer 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 } #endif + + if (!user_mode(regs)) + pc = ktla_ktva(pc); + return pc; } EXPORT_SYMBOL(profile_pc); diff -urNp linux-2.6.28.8/arch/x86/kernel/time_64.c linux-2.6.28.8/arch/x86/kernel/time_64.c --- linux-2.6.28.8/arch/x86/kernel/time_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/time_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -34,7 +34,7 @@ unsigned long profile_pc(struct pt_regs /* Assume the lock function has either no stack frame or a copy of flags from PUSHF Eflags always has bits 22 and up cleared unlike kernel addresses. */ - 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)); #else diff -urNp linux-2.6.28.8/arch/x86/kernel/tlb_32.c linux-2.6.28.8/arch/x86/kernel/tlb_32.c --- linux-2.6.28.8/arch/x86/kernel/tlb_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/tlb_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -5,7 +5,7 @@ #include DEFINE_PER_CPU(struct tlb_state, cpu_tlbstate) - ____cacheline_aligned = { &init_mm, 0, }; + ____cacheline_aligned = { &init_mm, 0, {0} }; /* must come after the send_IPI functions above for inlining */ #include diff -urNp linux-2.6.28.8/arch/x86/kernel/tls.c linux-2.6.28.8/arch/x86/kernel/tls.c --- linux-2.6.28.8/arch/x86/kernel/tls.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/tls.c 2009-02-21 09:37:48.000000000 -0500 @@ -85,6 +85,11 @@ int do_set_thread_area(struct task_struc if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX) return -EINVAL; +#ifdef CONFIG_PAX_SEGMEXEC + if ((p->mm->pax_flags & MF_PAX_SEGMEXEC) && (info.contents & MODIFY_LDT_CONTENTS_CODE)) + return -EINVAL; +#endif + set_tls_desc(p, idx, &info, 1); return 0; diff -urNp linux-2.6.28.8/arch/x86/kernel/traps.c linux-2.6.28.8/arch/x86/kernel/traps.c --- linux-2.6.28.8/arch/x86/kernel/traps.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/traps.c 2009-03-07 10:29:51.000000000 -0500 @@ -79,14 +79,6 @@ asmlinkage int system_call(void); /* Do we ignore FPU interrupts ? */ char ignore_fpu_irq; - -/* - * The IDT has to be page-aligned to simplify the Pentium - * F0 0F bug workaround.. We have a special link segment - * for this. - */ -gate_desc idt_table[256] - __attribute__((__section__(".data.idt"))) = { { { { 0, 0 } } }, }; #endif static int ignore_nmis; @@ -121,7 +113,7 @@ static inline void preempt_conditional_c static inline void die_if_kernel(const char *str, struct pt_regs *regs, long err) { - if (!user_mode_vm(regs)) + if (!user_mode(regs)) die(str, regs, err); } @@ -138,7 +130,7 @@ static int lazy_iobitmap_copy(void) int cpu; cpu = get_cpu(); - tss = &per_cpu(init_tss, cpu); + tss = init_tss + cpu; thread = ¤t->thread; if (tss->x86_tss.io_bitmap_base == INVALID_IO_BITMAP_OFFSET_LAZY && @@ -174,7 +166,7 @@ do_trap(int trapnr, int signr, char *str struct task_struct *tsk = current; #ifdef CONFIG_X86_32 - if (regs->flags & X86_VM_MASK) { + if (v8086_mode(regs)) { /* * traps 0, 1, 3, 4, and 5 should be forwarded to vm86. * On nmi (interrupt 2), do_trap should not be called. @@ -185,7 +177,7 @@ do_trap(int trapnr, int signr, char *str } #endif - if (!user_mode(regs)) + if (!user_mode_novm(regs)) goto kernel_trap; #ifdef CONFIG_X86_32 @@ -227,6 +219,12 @@ kernel_trap: tsk->thread.trap_no = trapnr; die(str, regs, error_code); } + +#ifdef CONFIG_PAX_REFCOUNT + if (trapnr == 4) + pax_report_refcount_overflow(regs); +#endif + return; #ifdef CONFIG_X86_32 @@ -318,14 +316,30 @@ do_general_protection(struct pt_regs *re return; } - if (regs->flags & X86_VM_MASK) + if (v8086_mode(regs)) goto gp_in_vm86; #endif tsk = current; - if (!user_mode(regs)) + if (!user_mode_novm(regs)) goto gp_in_kernel; +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) + if (!nx_enabled && tsk->mm && (tsk->mm->pax_flags & MF_PAX_PAGEEXEC)) { + struct mm_struct *mm = tsk->mm; + unsigned long limit; + + down_write(&mm->mmap_sem); + limit = mm->context.user_cs_limit; + if (limit < TASK_SIZE) { + track_exec_limit(mm, limit, TASK_SIZE, VM_EXEC); + up_write(&mm->mmap_sem); + return; + } + up_write(&mm->mmap_sem); + } +#endif + tsk->thread.error_code = error_code; tsk->thread.trap_no = 13; @@ -358,6 +372,13 @@ gp_in_kernel: if (notify_die(DIE_GPF, "general protection fault", regs, error_code, 13, SIGSEGV) == NOTIFY_STOP) return; + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + if ((regs->cs & 0xFFFF) == __KERNEL_CS) + die("PAX: suspicious general protection fault", regs, error_code); + else +#endif + die("general protection fault", regs, error_code); } @@ -604,7 +625,7 @@ dotraplinkage void __kprobes do_debug(st } #ifdef CONFIG_X86_32 - if (regs->flags & X86_VM_MASK) + if (v8086_mode(regs)) goto debug_vm86; #endif @@ -616,7 +637,7 @@ dotraplinkage void __kprobes do_debug(st * kernel space (but re-enable TF when returning to user mode). */ if (condition & DR_STEP) { - if (!user_mode(regs)) + if (!user_mode_novm(regs)) goto clear_TF_reenable; } @@ -808,7 +829,7 @@ do_simd_coprocessor_error(struct pt_regs * Handle strange cache flush from user space exception * in all other cases. This is undocumented behaviour. */ - if (regs->flags & X86_VM_MASK) { + if (v8086_mode(regs)) { handle_vm86_fault((struct kernel_vm86_regs *)regs, error_code); return; } @@ -837,19 +858,14 @@ do_spurious_interrupt_bug(struct pt_regs #ifdef CONFIG_X86_32 unsigned long patch_espfix_desc(unsigned long uesp, unsigned long kesp) { - struct desc_struct *gdt = get_cpu_gdt_table(smp_processor_id()); unsigned long base = (kesp - uesp) & -THREAD_SIZE; unsigned long new_kesp = kesp - base; unsigned long lim_pages = (new_kesp | (THREAD_SIZE - 1)) >> PAGE_SHIFT; - __u64 desc = *(__u64 *)&gdt[GDT_ENTRY_ESPFIX_SS]; + struct desc_struct ss; /* Set up base for espfix segment */ - desc &= 0x00f0ff0000000000ULL; - desc |= ((((__u64)base) << 16) & 0x000000ffffff0000ULL) | - ((((__u64)base) << 32) & 0xff00000000000000ULL) | - ((((__u64)lim_pages) << 32) & 0x000f000000000000ULL) | - (lim_pages & 0xffff); - *(__u64 *)&gdt[GDT_ENTRY_ESPFIX_SS] = desc; + pack_descriptor(&ss, base, lim_pages, 0x93, 0xC); + write_gdt_entry(get_cpu_gdt_table(smp_processor_id()), GDT_ENTRY_ESPFIX_SS, &ss, DESCTYPE_S); return new_kesp; } diff -urNp linux-2.6.28.8/arch/x86/kernel/tsc.c linux-2.6.28.8/arch/x86/kernel/tsc.c --- linux-2.6.28.8/arch/x86/kernel/tsc.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/tsc.c 2009-02-21 09:37:48.000000000 -0500 @@ -728,7 +728,7 @@ static struct dmi_system_id __initdata b DMI_MATCH(DMI_BOARD_NAME, "2635FA0"), }, }, - {} + { NULL, NULL, {{0, {0}}}, NULL} }; /* diff -urNp linux-2.6.28.8/arch/x86/kernel/vm86_32.c linux-2.6.28.8/arch/x86/kernel/vm86_32.c --- linux-2.6.28.8/arch/x86/kernel/vm86_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/vm86_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -148,7 +148,7 @@ struct pt_regs *save_v86_state(struct ke 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); @@ -325,7 +325,7 @@ static void do_sys_vm86(struct kernel_vm tsk->thread.saved_fs = info->regs32->fs; savesegment(gs, tsk->thread.saved_gs); - 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; diff -urNp linux-2.6.28.8/arch/x86/kernel/vmi_32.c linux-2.6.28.8/arch/x86/kernel/vmi_32.c --- linux-2.6.28.8/arch/x86/kernel/vmi_32.c 2009-02-08 00:54:27.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/vmi_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -102,18 +102,43 @@ static unsigned patch_internal(int call, { u64 reloc; struct vmi_relocation_info *const rel = (struct vmi_relocation_info *)&reloc; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + reloc = call_vrom_long_func(vmi_rom, get_reloc, call); switch(rel->type) { case VMI_RELOCATION_CALL_REL: BUG_ON(len < 5); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + *(char *)insnbuf = MNEM_CALL; patch_offset(insnbuf, ip, (unsigned long)rel->eip); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + return 5; case VMI_RELOCATION_JUMP_REL: BUG_ON(len < 5); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + *(char *)insnbuf = MNEM_JMP; patch_offset(insnbuf, ip, (unsigned long)rel->eip); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + return 5; case VMI_RELOCATION_NOP: @@ -526,14 +551,14 @@ static void vmi_set_pud(pud_t *pudp, pud static void vmi_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) { - const pte_t pte = { .pte = 0 }; + const pte_t pte = __pte(0ULL); vmi_check_page_type(__pa(ptep) >> PAGE_SHIFT, VMI_PAGE_PTE); vmi_ops.set_pte(pte, ptep, vmi_flags_addr(mm, addr, VMI_PAGE_PT, 0)); } static void vmi_pmd_clear(pmd_t *pmd) { - const pte_t pte = { .pte = 0 }; + const pte_t pte = __pte(0ULL); vmi_check_page_type(__pa(pmd) >> PAGE_SHIFT, VMI_PAGE_PMD); vmi_ops.set_pte(pte, (pte_t *)pmd, VMI_PAGE_PD); } @@ -562,8 +587,8 @@ vmi_startup_ipi_hook(int phys_apicid, un ap.ss = __KERNEL_DS; ap.esp = (unsigned long) start_esp; - ap.ds = __USER_DS; - ap.es = __USER_DS; + ap.ds = __KERNEL_DS; + ap.es = __KERNEL_DS; ap.fs = __KERNEL_PERCPU; ap.gs = 0; @@ -758,12 +783,20 @@ static inline int __init activate_vmi(vo u64 reloc; const struct vmi_relocation_info *rel = (struct vmi_relocation_info *)&reloc; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + if (call_vrom_func(vmi_rom, vmi_init) != 0) { printk(KERN_ERR "VMI ROM failed to initialize!"); return 0; } savesegment(cs, kernel_cs); +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + pv_info.paravirt_enabled = 1; pv_info.kernel_rpl = kernel_cs & SEGMENT_RPL_MASK; pv_info.name = "vmi"; @@ -954,6 +987,10 @@ static inline int __init activate_vmi(vo para_fill(pv_irq_ops.safe_halt, Halt); +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + /* * Alternative instruction rewriting doesn't happen soon enough * to convert VMI_IRET to a call instead of a jump; so we have diff -urNp linux-2.6.28.8/arch/x86/kernel/vmlinux_32.lds.S linux-2.6.28.8/arch/x86/kernel/vmlinux_32.lds.S --- linux-2.6.28.8/arch/x86/kernel/vmlinux_32.lds.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/vmlinux_32.lds.S 2009-03-07 10:35:39.000000000 -0500 @@ -15,6 +15,20 @@ #include #include #include +#include + +#ifdef CONFIG_X86_PAE +#define PMD_SHIFT 21 +#else +#define PMD_SHIFT 22 +#endif +#define PMD_SIZE (1 << PMD_SHIFT) + +#ifdef CONFIG_PAX_KERNEXEC +#define __KERNEL_TEXT_OFFSET (__PAGE_OFFSET + (((____LOAD_PHYSICAL_ADDR + 2*(PMD_SIZE - 1)) - 1) & ~(PMD_SIZE - 1))) +#else +#define __KERNEL_TEXT_OFFSET 0 +#endif OUTPUT_FORMAT("elf32-i386", "elf32-i386", "elf32-i386") OUTPUT_ARCH(i386) @@ -22,81 +36,23 @@ ENTRY(phys_startup_32) jiffies = jiffies_64; PHDRS { - text PT_LOAD FLAGS(5); /* R_E */ - data PT_LOAD FLAGS(7); /* RWE */ - note PT_NOTE FLAGS(0); /* ___ */ + initdata PT_LOAD FLAGS(6); /* RW_ */ + percpu PT_LOAD FLAGS(6); /* RW_ */ + inittext PT_LOAD FLAGS(5); /* R_E */ + text PT_LOAD FLAGS(5); /* R_E */ + rodata PT_LOAD FLAGS(4); /* R__ */ + data PT_LOAD FLAGS(6); /* RW_ */ + note PT_NOTE FLAGS(0); /* ___ */ } SECTIONS { - . = LOAD_OFFSET + LOAD_PHYSICAL_ADDR; - phys_startup_32 = startup_32 - LOAD_OFFSET; - - .text.head : AT(ADDR(.text.head) - LOAD_OFFSET) { - _text = .; /* Text and read-only data */ - *(.text.head) - } :text = 0x9090 - - /* read-only */ - .text : AT(ADDR(.text) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); /* not really needed, already page aligned */ - *(.text.page_aligned) - TEXT_TEXT - SCHED_TEXT - LOCK_TEXT - KPROBES_TEXT - *(.fixup) - *(.gnu.warning) - _etext = .; /* End of text section */ - } :text = 0x9090 - - NOTES :text :note - - . = ALIGN(16); /* Exception table */ - __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { - __start___ex_table = .; - *(__ex_table) - __stop___ex_table = .; - } :text = 0x9090 - - RODATA - - /* writeable */ - . = ALIGN(PAGE_SIZE); - .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Data */ - DATA_DATA - CONSTRUCTORS - } :data - - . = ALIGN(PAGE_SIZE); - .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { - __nosave_begin = .; - *(.data.nosave) - . = ALIGN(PAGE_SIZE); - __nosave_end = .; - } + . = LOAD_OFFSET + ____LOAD_PHYSICAL_ADDR; - . = ALIGN(PAGE_SIZE); - .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { - *(.data.page_aligned) - *(.data.idt) - } - - . = ALIGN(32); - .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { - *(.data.cacheline_aligned) - } - - /* rarely changed data like cpu maps */ - . = ALIGN(32); - .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) { - *(.data.read_mostly) - _edata = .; /* End of data section */ - } - - . = ALIGN(THREAD_SIZE); /* init_task */ - .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { - *(.data.init_task) - } + .text.startup : AT(ADDR(.text.startup) - LOAD_OFFSET) { + __LOAD_PHYSICAL_ADDR = . - LOAD_OFFSET; + phys_startup_32 = startup_32 - LOAD_OFFSET + __KERNEL_TEXT_OFFSET; + *(.text.startup) + } :initdata /* might get freed after init */ . = ALIGN(PAGE_SIZE); @@ -114,14 +70,8 @@ SECTIONS . = ALIGN(PAGE_SIZE); /* will be freed after init */ - . = ALIGN(PAGE_SIZE); /* Init code and data */ - .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) { - __init_begin = .; - _sinittext = .; - INIT_TEXT - _einittext = .; - } .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { + __init_begin = .; INIT_DATA } . = ALIGN(16); @@ -161,11 +111,6 @@ SECTIONS *(.parainstructions) __parainstructions_end = .; } - /* .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 } @@ -178,18 +123,138 @@ SECTIONS } #endif . = ALIGN(PAGE_SIZE); - .data.percpu : AT(ADDR(.data.percpu) - LOAD_OFFSET) { - __per_cpu_start = .; - *(.data.percpu.page_aligned) + per_cpu_start = .; + .data.percpu (0) : AT(ADDR(.data.percpu) - LOAD_OFFSET + per_cpu_start) { + __per_cpu_start = . + per_cpu_start; + LONG(0) *(.data.percpu) *(.data.percpu.shared_aligned) - __per_cpu_end = .; - } + . = ALIGN(PAGE_SIZE); + *(.data.percpu.page_aligned) + __per_cpu_end = . + per_cpu_start; + } :percpu + . += per_cpu_start; . = ALIGN(PAGE_SIZE); /* freed after init ends here */ + . = ALIGN(PAGE_SIZE); /* Init code and data */ + .init.text (. - __KERNEL_TEXT_OFFSET) : AT(ADDR(.init.text) - LOAD_OFFSET + __KERNEL_TEXT_OFFSET) { + _sinittext = .; + INIT_TEXT + _einittext = .; + } :inittext + + /* .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 + } + + .filler : AT(ADDR(.filler) - LOAD_OFFSET + __KERNEL_TEXT_OFFSET) { + BYTE(0) + . = ALIGN(2*PMD_SIZE) - 1; + } + + /* freed after init ends here */ + + .text.head : AT(ADDR(.text.head) - LOAD_OFFSET + __KERNEL_TEXT_OFFSET) { + __init_end = . + __KERNEL_TEXT_OFFSET; + KERNEL_TEXT_OFFSET = . + __KERNEL_TEXT_OFFSET; + _text = .; /* Text and read-only data */ + *(.text.head) + } :text = 0x9090 + + /* read-only */ + .text : AT(ADDR(.text) - LOAD_OFFSET + __KERNEL_TEXT_OFFSET) { + . = ALIGN(PAGE_SIZE); /* not really needed, already page aligned */ + *(.text.page_aligned) + TEXT_TEXT + SCHED_TEXT + LOCK_TEXT + KPROBES_TEXT + *(.fixup) + *(.gnu.warning) + _etext = .; /* End of text section */ + } :text = 0x9090 + + . += __KERNEL_TEXT_OFFSET; + + . = ALIGN(4096); + NOTES :rodata :note + + . = ALIGN(16); /* Exception table */ + __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { + __start___ex_table = .; + *(__ex_table) + __stop___ex_table = .; + } :rodata + + RO_DATA(PAGE_SIZE) + + . = ALIGN(PAGE_SIZE); + .rodata.page_aligned : AT(ADDR(.rodata.page_aligned) - LOAD_OFFSET) { + *(.idt) + . = ALIGN(PAGE_SIZE); + *(.empty_zero_page) + *(.swapper_pg_pmd) + *(.swapper_pg_dir) + +#if defined(CONFIG_PAX_KERNEXEC) && !defined(CONFIG_MODULES) + . = ALIGN(PMD_SIZE); +#endif + + } + +#if defined(CONFIG_PAX_KERNEXEC) && defined(CONFIG_MODULES) + . = ALIGN(PAGE_SIZE); + .module.text : AT(ADDR(.module.text) - LOAD_OFFSET) { + MODULES_VADDR = .; + BYTE(0) + . += (6 * 1024 * 1024); + . = ALIGN(PMD_SIZE); + MODULES_END = . - 1; + } +#endif + + /* writeable */ + . = ALIGN(PAGE_SIZE); + .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Data */ + _data = .; + DATA_DATA + CONSTRUCTORS + } :data + + . = ALIGN(PAGE_SIZE); + .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { + __nosave_begin = .; + *(.data.nosave) + . = ALIGN(PAGE_SIZE); + __nosave_end = .; + } + + . = ALIGN(PAGE_SIZE); + .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { + *(.data.page_aligned) + } + + . = ALIGN(32); + .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { + *(.data.cacheline_aligned) + } + + /* rarely changed data like cpu maps */ + . = ALIGN(32); + .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) { + *(.data.read_mostly) + _edata = .; /* End of data section */ + } + + . = ALIGN(THREAD_SIZE); /* init_task */ + .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { + *(.data.init_task) + } + .bss : AT(ADDR(.bss) - LOAD_OFFSET) { - __init_end = .; __bss_start = .; /* BSS */ *(.bss.page_aligned) *(.bss) diff -urNp linux-2.6.28.8/arch/x86/kernel/vmlinux_64.lds.S linux-2.6.28.8/arch/x86/kernel/vmlinux_64.lds.S --- linux-2.6.28.8/arch/x86/kernel/vmlinux_64.lds.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/vmlinux_64.lds.S 2009-02-21 09:37:48.000000000 -0500 @@ -16,7 +16,7 @@ jiffies_64 = jiffies; _proxy_pda = 1; PHDRS { text PT_LOAD FLAGS(5); /* R_E */ - data PT_LOAD FLAGS(7); /* RWE */ + data PT_LOAD FLAGS(6); /* RW_ */ user PT_LOAD FLAGS(7); /* RWE */ data.init PT_LOAD FLAGS(7); /* RWE */ note PT_NOTE FLAGS(0); /* ___ */ @@ -49,17 +49,20 @@ SECTIONS __stop___ex_table = .; } :text = 0x9090 - RODATA + RO_DATA(PAGE_SIZE) +#ifdef CONFIG_PAX_KERNEXEC + . = ALIGN(2*1024*1024); /* Align data segment to PMD size boundary */ +#else . = ALIGN(PAGE_SIZE); /* Align data segment to page size boundary */ +#endif /* Data */ + _data = .; .data : AT(ADDR(.data) - LOAD_OFFSET) { DATA_DATA CONSTRUCTORS } :data - _edata = .; /* End of data section */ - . = ALIGN(PAGE_SIZE); . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { @@ -70,9 +73,27 @@ SECTIONS *(.data.read_mostly) } + . = ALIGN(THREAD_SIZE); /* init_task */ + .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { + *(.data.init_task) + } + + . = ALIGN(PAGE_SIZE); + .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { + *(.data.page_aligned) + } + + . = ALIGN(PAGE_SIZE); + __nosave_begin = .; + .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { *(.data.nosave) } + . = ALIGN(PAGE_SIZE); + __nosave_end = .; + + _edata = .; /* End of data section */ + #define VSYSCALL_ADDR (-10*1024*1024) -#define VSYSCALL_PHYS_ADDR ((LOADADDR(.data.read_mostly) + SIZEOF(.data.read_mostly) + 4095) & ~(4095)) -#define VSYSCALL_VIRT_ADDR ((ADDR(.data.read_mostly) + SIZEOF(.data.read_mostly) + 4095) & ~(4095)) +#define VSYSCALL_PHYS_ADDR ((LOADADDR(.data_nosave) + SIZEOF(.data_nosave) + 4095) & ~(4095)) +#define VSYSCALL_VIRT_ADDR ((ADDR(.data_nosave) + SIZEOF(.data_nosave) + 4095) & ~(4095)) #define VLOAD_OFFSET (VSYSCALL_ADDR - VSYSCALL_PHYS_ADDR) #define VLOAD(x) (ADDR(x) - VLOAD_OFFSET) @@ -120,23 +141,13 @@ SECTIONS #undef VVIRT_OFFSET #undef VVIRT - . = ALIGN(THREAD_SIZE); /* init_task */ - .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { - *(.data.init_task) - }:data.init - - . = ALIGN(PAGE_SIZE); - .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { - *(.data.page_aligned) - } - /* might get freed after init */ . = ALIGN(PAGE_SIZE); __smp_alt_begin = .; __smp_locks = .; .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { *(.smp_locks) - } + } :data.init __smp_locks_end = .; . = ALIGN(PAGE_SIZE); __smp_alt_end = .; @@ -212,16 +223,11 @@ SECTIONS . = ALIGN(PAGE_SIZE); __init_end = .; - . = ALIGN(PAGE_SIZE); - __nosave_begin = .; - .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { *(.data.nosave) } - . = ALIGN(PAGE_SIZE); - __nosave_end = .; - __bss_start = .; /* BSS */ .bss : AT(ADDR(.bss) - LOAD_OFFSET) { *(.bss.page_aligned) *(.bss) + . = ALIGN(2*1024*1024); } __bss_stop = .; diff -urNp linux-2.6.28.8/arch/x86/kernel/vsyscall_64.c linux-2.6.28.8/arch/x86/kernel/vsyscall_64.c --- linux-2.6.28.8/arch/x86/kernel/vsyscall_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kernel/vsyscall_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -236,13 +236,13 @@ static ctl_table kernel_table2[] = { .data = &vsyscall_gtod_data.sysctl_enabled, .maxlen = sizeof(int), .mode = 0644, .proc_handler = vsyscall_sysctl_change }, - {} + { 0, NULL, NULL, 0, 0, NULL, NULL, NULL, NULL, NULL, NULL } }; static ctl_table kernel_root_table2[] = { { .ctl_name = CTL_KERN, .procname = "kernel", .mode = 0555, .child = kernel_table2 }, - {} + { 0, NULL, NULL, 0, 0, NULL, NULL, NULL, NULL, NULL, NULL } }; #endif diff -urNp linux-2.6.28.8/arch/x86/kvm/svm.c linux-2.6.28.8/arch/x86/kvm/svm.c --- linux-2.6.28.8/arch/x86/kvm/svm.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kvm/svm.c 2009-02-21 09:37:48.000000000 -0500 @@ -1505,7 +1505,19 @@ static void reload_tss(struct kvm_vcpu * int cpu = raw_smp_processor_id(); struct svm_cpu_data *svm_data = per_cpu(svm_data, cpu); + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + svm_data->tss_desc->type = 9; /* available 32/64-bit TSS */ + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + load_TR_desc(); } @@ -1912,7 +1924,7 @@ static int get_npt_level(void) #endif } -static struct kvm_x86_ops svm_x86_ops = { +static const struct kvm_x86_ops svm_x86_ops = { .cpu_has_kvm_support = has_svm, .disabled_by_bios = is_disabled, .hardware_setup = svm_hardware_setup, diff -urNp linux-2.6.28.8/arch/x86/kvm/vmx.c linux-2.6.28.8/arch/x86/kvm/vmx.c --- linux-2.6.28.8/arch/x86/kvm/vmx.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kvm/vmx.c 2009-02-21 09:37:48.000000000 -0500 @@ -122,7 +122,7 @@ static struct vmcs_config { u32 vmentry_ctrl; } vmcs_config; -struct vmx_capability { +static struct vmx_capability { u32 ept; u32 vpid; } vmx_capability; @@ -491,9 +491,23 @@ static void reload_tss(void) struct descriptor_table gdt; struct desc_struct *descs; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + kvm_get_gdt(&gdt); descs = (void *)gdt.base; + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + descs[GDT_ENTRY_TSS].type = 9; /* available TSS */ + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + load_TR_desc(); } @@ -2164,7 +2178,7 @@ static int vmx_vcpu_setup(struct vcpu_vm vmcs_writel(HOST_IDTR_BASE, dt.base); /* 22.2.4 */ asm("mov $.Lkvm_vmx_return, %0" : "=r"(kvm_vmx_return)); - vmcs_writel(HOST_RIP, kvm_vmx_return); /* 22.2.5 */ + vmcs_writel(HOST_RIP, ktla_ktva(kvm_vmx_return)); /* 22.2.5 */ vmcs_write32(VM_EXIT_MSR_STORE_COUNT, 0); vmcs_write32(VM_EXIT_MSR_LOAD_COUNT, 0); vmcs_write32(VM_ENTRY_MSR_LOAD_COUNT, 0); @@ -3267,6 +3281,12 @@ static void vmx_vcpu_run(struct kvm_vcpu "jmp .Lkvm_vmx_return \n\t" ".Llaunched: " __ex(ASM_VMX_VMRESUME) "\n\t" ".Lkvm_vmx_return: " + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + "ljmp %[cs],$.Lkvm_vmx_return2\n\t" + ".Lkvm_vmx_return2: " +#endif + /* Save guest registers, load host registers, keep flags */ "xchg %0, (%%"R"sp) \n\t" "mov %%"R"ax, %c[rax](%0) \n\t" @@ -3313,6 +3333,11 @@ static void vmx_vcpu_run(struct kvm_vcpu [r15]"i"(offsetof(struct vcpu_vmx, vcpu.arch.regs[VCPU_REGS_R15])), #endif [cr2]"i"(offsetof(struct vcpu_vmx, vcpu.arch.cr2)) + +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + ,[cs]"i"(__KERNEL_CS) +#endif + : "cc", "memory" , R"bx", R"di", R"si" #ifdef CONFIG_X86_64 @@ -3331,7 +3356,7 @@ static void vmx_vcpu_run(struct kvm_vcpu (vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & (GUEST_INTR_STATE_STI | GUEST_INTR_STATE_MOV_SS)) == 0; - asm("mov %0, %%ds; mov %0, %%es" : : "r"(__USER_DS)); + asm("mov %0, %%ds; mov %0, %%es" : : "r"(__KERNEL_DS)); vmx->launched = 1; intr_info = vmcs_read32(VM_EXIT_INTR_INFO); @@ -3455,7 +3480,7 @@ static int get_ept_level(void) return VMX_EPT_DEFAULT_GAW + 1; } -static struct kvm_x86_ops vmx_x86_ops = { +static const struct kvm_x86_ops vmx_x86_ops = { .cpu_has_kvm_support = cpu_has_kvm_support, .disabled_by_bios = vmx_disabled_by_bios, .hardware_setup = hardware_setup, diff -urNp linux-2.6.28.8/arch/x86/kvm/x86.c linux-2.6.28.8/arch/x86/kvm/x86.c --- linux-2.6.28.8/arch/x86/kvm/x86.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/kvm/x86.c 2009-02-21 09:37:48.000000000 -0500 @@ -68,41 +68,41 @@ static u64 __read_mostly efer_reserved_b static int kvm_dev_ioctl_get_supported_cpuid(struct kvm_cpuid2 *cpuid, struct kvm_cpuid_entry2 __user *entries); -struct kvm_x86_ops *kvm_x86_ops; +const struct kvm_x86_ops *kvm_x86_ops; EXPORT_SYMBOL_GPL(kvm_x86_ops); struct kvm_stats_debugfs_item debugfs_entries[] = { - { "pf_fixed", VCPU_STAT(pf_fixed) }, - { "pf_guest", VCPU_STAT(pf_guest) }, - { "tlb_flush", VCPU_STAT(tlb_flush) }, - { "invlpg", VCPU_STAT(invlpg) }, - { "exits", VCPU_STAT(exits) }, - { "io_exits", VCPU_STAT(io_exits) }, - { "mmio_exits", VCPU_STAT(mmio_exits) }, - { "signal_exits", VCPU_STAT(signal_exits) }, - { "irq_window", VCPU_STAT(irq_window_exits) }, - { "nmi_window", VCPU_STAT(nmi_window_exits) }, - { "halt_exits", VCPU_STAT(halt_exits) }, - { "halt_wakeup", VCPU_STAT(halt_wakeup) }, - { "hypercalls", VCPU_STAT(hypercalls) }, - { "request_irq", VCPU_STAT(request_irq_exits) }, - { "irq_exits", VCPU_STAT(irq_exits) }, - { "host_state_reload", VCPU_STAT(host_state_reload) }, - { "efer_reload", VCPU_STAT(efer_reload) }, - { "fpu_reload", VCPU_STAT(fpu_reload) }, - { "insn_emulation", VCPU_STAT(insn_emulation) }, - { "insn_emulation_fail", VCPU_STAT(insn_emulation_fail) }, - { "irq_injections", VCPU_STAT(irq_injections) }, - { "mmu_shadow_zapped", VM_STAT(mmu_shadow_zapped) }, - { "mmu_pte_write", VM_STAT(mmu_pte_write) }, - { "mmu_pte_updated", VM_STAT(mmu_pte_updated) }, - { "mmu_pde_zapped", VM_STAT(mmu_pde_zapped) }, - { "mmu_flooded", VM_STAT(mmu_flooded) }, - { "mmu_recycled", VM_STAT(mmu_recycled) }, - { "mmu_cache_miss", VM_STAT(mmu_cache_miss) }, - { "mmu_unsync", VM_STAT(mmu_unsync) }, - { "remote_tlb_flush", VM_STAT(remote_tlb_flush) }, - { "largepages", VM_STAT(lpages) }, + { "pf_fixed", VCPU_STAT(pf_fixed), NULL }, + { "pf_guest", VCPU_STAT(pf_guest), NULL }, + { "tlb_flush", VCPU_STAT(tlb_flush), NULL }, + { "invlpg", VCPU_STAT(invlpg), NULL }, + { "exits", VCPU_STAT(exits), NULL }, + { "io_exits", VCPU_STAT(io_exits), NULL }, + { "mmio_exits", VCPU_STAT(mmio_exits), NULL }, + { "signal_exits", VCPU_STAT(signal_exits), NULL }, + { "irq_window", VCPU_STAT(irq_window_exits), NULL }, + { "nmi_window", VCPU_STAT(nmi_window_exits), NULL }, + { "halt_exits", VCPU_STAT(halt_exits), NULL }, + { "halt_wakeup", VCPU_STAT(halt_wakeup), NULL }, + { "hypercalls", VCPU_STAT(hypercalls), NULL }, + { "request_irq", VCPU_STAT(request_irq_exits), NULL }, + { "irq_exits", VCPU_STAT(irq_exits), NULL }, + { "host_state_reload", VCPU_STAT(host_state_reload), NULL }, + { "efer_reload", VCPU_STAT(efer_reload), NULL }, + { "fpu_reload", VCPU_STAT(fpu_reload), NULL }, + { "insn_emulation", VCPU_STAT(insn_emulation), NULL }, + { "insn_emulation_fail", VCPU_STAT(insn_emulation_fail), NULL }, + { "irq_injections", VCPU_STAT(irq_injections), NULL }, + { "mmu_shadow_zapped", VM_STAT(mmu_shadow_zapped), NULL }, + { "mmu_pte_write", VM_STAT(mmu_pte_write), NULL }, + { "mmu_pte_updated", VM_STAT(mmu_pte_updated), NULL }, + { "mmu_pde_zapped", VM_STAT(mmu_pde_zapped), NULL }, + { "mmu_flooded", VM_STAT(mmu_flooded), NULL }, + { "mmu_recycled", VM_STAT(mmu_recycled), NULL }, + { "mmu_cache_miss", VM_STAT(mmu_cache_miss), NULL }, + { "mmu_unsync", VM_STAT(mmu_unsync), NULL }, + { "remote_tlb_flush", VM_STAT(remote_tlb_flush), NULL }, + { "largepages", VM_STAT(lpages), NULL }, { NULL } }; @@ -1304,7 +1304,7 @@ static int kvm_vcpu_ioctl_set_lapic(stru static int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq) { - if (irq->irq < 0 || irq->irq >= 256) + if (irq->irq >= 256) return -EINVAL; if (irqchip_in_kernel(vcpu->kvm)) return -ENXIO; @@ -2509,10 +2509,10 @@ int kvm_emulate_pio_string(struct kvm_vc } EXPORT_SYMBOL_GPL(kvm_emulate_pio_string); -int kvm_arch_init(void *opaque) +int kvm_arch_init(const void *opaque) { int r; - struct kvm_x86_ops *ops = (struct kvm_x86_ops *)opaque; + const struct kvm_x86_ops *ops = (const struct kvm_x86_ops *)opaque; if (kvm_x86_ops) { printk(KERN_ERR "kvm: already loaded the other module\n"); diff -urNp linux-2.6.28.8/arch/x86/lib/checksum_32.S linux-2.6.28.8/arch/x86/lib/checksum_32.S --- linux-2.6.28.8/arch/x86/lib/checksum_32.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/checksum_32.S 2009-02-21 09:37:48.000000000 -0500 @@ -28,7 +28,8 @@ #include #include #include - +#include + /* * computes a partial checksum, e.g. for TCP/UDP fragments */ @@ -304,9 +305,22 @@ unsigned int csum_partial_copy_generic ( #define ARGBASE 16 #define FP 12 - -ENTRY(csum_partial_copy_generic) + +ENTRY(csum_partial_copy_generic_to_user) CFI_STARTPROC + pushl $(__USER_DS) + CFI_ADJUST_CFA_OFFSET 4 + popl %es + CFI_ADJUST_CFA_OFFSET -4 + jmp csum_partial_copy_generic + +ENTRY(csum_partial_copy_generic_from_user) + pushl $(__USER_DS) + CFI_ADJUST_CFA_OFFSET 4 + popl %ds + CFI_ADJUST_CFA_OFFSET -4 + +ENTRY(csum_partial_copy_generic) subl $4,%esp CFI_ADJUST_CFA_OFFSET 4 pushl %edi @@ -331,7 +345,7 @@ 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 @@ -343,30 +357,30 @@ 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 @@ -380,7 +394,7 @@ 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 @@ -392,12 +406,12 @@ 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: @@ -408,7 +422,7 @@ 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 @@ -421,11 +435,19 @@ DST( movb %cl, (%edi) ) 6002: movl ARGBASE+24(%esp), %ebx # dst_err_ptr - movl $-EFAULT,(%ebx) + movl $-EFAULT,%ss:(%ebx) jmp 5000b .previous + pushl %ss + CFI_ADJUST_CFA_OFFSET 4 + popl %ds + CFI_ADJUST_CFA_OFFSET -4 + pushl %ss + CFI_ADJUST_CFA_OFFSET 4 + popl %es + CFI_ADJUST_CFA_OFFSET -4 popl %ebx CFI_ADJUST_CFA_OFFSET -4 CFI_RESTORE ebx @@ -439,26 +461,41 @@ DST( movb %cl, (%edi) ) CFI_ADJUST_CFA_OFFSET -4 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 + pushl $(__USER_DS) + CFI_ADJUST_CFA_OFFSET 4 + popl %es + CFI_ADJUST_CFA_OFFSET -4 + jmp csum_partial_copy_generic + +ENTRY(csum_partial_copy_generic_from_user) + pushl $(__USER_DS) + CFI_ADJUST_CFA_OFFSET 4 + popl %ds + CFI_ADJUST_CFA_OFFSET -4 + +ENTRY(csum_partial_copy_generic) pushl %ebx CFI_ADJUST_CFA_OFFSET 4 CFI_REL_OFFSET ebx, 0 @@ -482,7 +519,7 @@ 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 @@ -503,19 +540,19 @@ 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 @@ -523,10 +560,18 @@ 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 + pushl %ss + CFI_ADJUST_CFA_OFFSET 4 + popl %ds + CFI_ADJUST_CFA_OFFSET -4 + pushl %ss + CFI_ADJUST_CFA_OFFSET 4 + popl %es + CFI_ADJUST_CFA_OFFSET -4 popl %esi CFI_ADJUST_CFA_OFFSET -4 CFI_RESTORE esi @@ -538,7 +583,7 @@ 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 -urNp linux-2.6.28.8/arch/x86/lib/clear_page_64.S linux-2.6.28.8/arch/x86/lib/clear_page_64.S --- linux-2.6.28.8/arch/x86/lib/clear_page_64.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/clear_page_64.S 2009-02-21 09:37:48.000000000 -0500 @@ -44,7 +44,7 @@ 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: diff -urNp linux-2.6.28.8/arch/x86/lib/copy_page_64.S linux-2.6.28.8/arch/x86/lib/copy_page_64.S --- linux-2.6.28.8/arch/x86/lib/copy_page_64.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/copy_page_64.S 2009-02-21 09:37:48.000000000 -0500 @@ -104,7 +104,7 @@ ENDPROC(copy_page) #include - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 1: .byte 0xeb /* jmp */ .byte (copy_page_c - copy_page) - (2f - 1b) /* offset */ 2: diff -urNp linux-2.6.28.8/arch/x86/lib/copy_user_64.S linux-2.6.28.8/arch/x86/lib/copy_user_64.S --- linux-2.6.28.8/arch/x86/lib/copy_user_64.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/copy_user_64.S 2009-02-21 09:37:48.000000000 -0500 @@ -21,7 +21,7 @@ .byte 0xe9 /* 32bit jump */ .long \orig-1f /* by default jump to orig */ 1: - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 2: .byte 0xe9 /* near jump with 32bit immediate */ .long \alt-1b /* offset */ /* or alternatively to alt */ .previous @@ -106,6 +106,8 @@ ENDPROC(__copy_from_user_inatomic) ENTRY(bad_from_user) bad_from_user: CFI_STARTPROC + testl %edx,%edx + js bad_to_user movl %edx,%ecx xorl %eax,%eax rep diff -urNp linux-2.6.28.8/arch/x86/lib/getuser.S linux-2.6.28.8/arch/x86/lib/getuser.S --- linux-2.6.28.8/arch/x86/lib/getuser.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/getuser.S 2009-02-21 09:37:48.000000000 -0500 @@ -33,6 +33,7 @@ #include #include #include +#include .text ENTRY(__get_user_1) @@ -40,7 +41,19 @@ ENTRY(__get_user_1) GET_THREAD_INFO(%_ASM_DX) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user + +#ifdef CONFIG_X86_32 + pushl $(__USER_DS) + popl %ds +#endif + 1: movzb (%_ASM_AX),%edx + +#ifdef CONFIG_X86_32 + pushl %ss + pop %ds +#endif + xor %eax,%eax ret CFI_ENDPROC @@ -53,7 +66,19 @@ ENTRY(__get_user_2) GET_THREAD_INFO(%_ASM_DX) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user + +#ifdef CONFIG_X86_32 + pushl $(__USER_DS) + popl %ds +#endif + 2: movzwl -1(%_ASM_AX),%edx + +#ifdef CONFIG_X86_32 + pushl %ss + pop %ds +#endif + xor %eax,%eax ret CFI_ENDPROC @@ -66,7 +91,19 @@ ENTRY(__get_user_4) GET_THREAD_INFO(%_ASM_DX) cmp TI_addr_limit(%_ASM_DX),%_ASM_AX jae bad_get_user + +#ifdef CONFIG_X86_32 + pushl $(__USER_DS) + popl %ds +#endif + 3: mov -3(%_ASM_AX),%edx + +#ifdef CONFIG_X86_32 + pushl %ss + pop %ds +#endif + xor %eax,%eax ret CFI_ENDPROC @@ -89,6 +126,12 @@ ENDPROC(__get_user_8) bad_get_user: CFI_STARTPROC + +#ifdef CONFIG_X86_32 + pushl %ss + pop %ds +#endif + xor %edx,%edx mov $(-EFAULT),%_ASM_AX ret diff -urNp linux-2.6.28.8/arch/x86/lib/memcpy_64.S linux-2.6.28.8/arch/x86/lib/memcpy_64.S --- linux-2.6.28.8/arch/x86/lib/memcpy_64.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/memcpy_64.S 2009-02-21 09:37:48.000000000 -0500 @@ -114,7 +114,7 @@ ENDPROC(__memcpy) /* Some CPUs run faster using the string copy instructions. It is also a lot simpler. Use this when possible */ - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 1: .byte 0xeb /* jmp */ .byte (memcpy_c - memcpy) - (2f - 1b) /* offset */ 2: diff -urNp linux-2.6.28.8/arch/x86/lib/memset_64.S linux-2.6.28.8/arch/x86/lib/memset_64.S --- linux-2.6.28.8/arch/x86/lib/memset_64.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/memset_64.S 2009-02-21 09:37:48.000000000 -0500 @@ -118,7 +118,7 @@ ENDPROC(__memset) #include - .section .altinstr_replacement,"ax" + .section .altinstr_replacement,"a" 1: .byte 0xeb /* jmp */ .byte (memset_c - memset) - (2f - 1b) /* offset */ 2: diff -urNp linux-2.6.28.8/arch/x86/lib/mmx_32.c linux-2.6.28.8/arch/x86/lib/mmx_32.c --- linux-2.6.28.8/arch/x86/lib/mmx_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/mmx_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -29,6 +29,7 @@ void *_mmx_memcpy(void *to, const void * { 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 * 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, voi * 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 -urNp linux-2.6.28.8/arch/x86/lib/putuser.S linux-2.6.28.8/arch/x86/lib/putuser.S --- linux-2.6.28.8/arch/x86/lib/putuser.S 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/putuser.S 2009-02-21 09:37:48.000000000 -0500 @@ -15,6 +15,7 @@ #include #include #include +#include /* @@ -39,7 +40,19 @@ ENTRY(__put_user_1) ENTER cmp TI_addr_limit(%_ASM_BX),%_ASM_CX jae bad_put_user + +#ifdef CONFIG_X86_32 + pushl $(__USER_DS) + popl %ds +#endif + 1: movb %al,(%_ASM_CX) + +#ifdef CONFIG_X86_32 + pushl %ss + popl %ds +#endif + xor %eax,%eax EXIT ENDPROC(__put_user_1) @@ -50,7 +63,19 @@ ENTRY(__put_user_2) sub $1,%_ASM_BX cmp %_ASM_BX,%_ASM_CX jae bad_put_user + +#ifdef CONFIG_X86_32 + pushl $(__USER_DS) + popl %ds +#endif + 2: movw %ax,(%_ASM_CX) + +#ifdef CONFIG_X86_32 + pushl %ss + popl %ds +#endif + xor %eax,%eax EXIT ENDPROC(__put_user_2) @@ -61,7 +86,19 @@ ENTRY(__put_user_4) sub $3,%_ASM_BX cmp %_ASM_BX,%_ASM_CX jae bad_put_user + +#ifdef CONFIG_X86_32 + pushl $(__USER_DS) + popl %ds +#endif + 3: movl %eax,(%_ASM_CX) + +#ifdef CONFIG_X86_32 + pushl %ss + popl %ds +#endif + xor %eax,%eax EXIT ENDPROC(__put_user_4) @@ -72,16 +109,34 @@ ENTRY(__put_user_8) sub $7,%_ASM_BX cmp %_ASM_BX,%_ASM_CX jae bad_put_user + +#ifdef CONFIG_X86_32 + pushl $(__USER_DS) + popl %ds +#endif + 4: mov %_ASM_AX,(%_ASM_CX) #ifdef CONFIG_X86_32 5: movl %edx,4(%_ASM_CX) #endif + +#ifdef CONFIG_X86_32 + pushl %ss + popl %ds +#endif + xor %eax,%eax EXIT ENDPROC(__put_user_8) bad_put_user: CFI_STARTPROC + +#ifdef CONFIG_X86_32 + pushl %ss + popl %ds +#endif + movl $-EFAULT,%eax EXIT END(bad_put_user) diff -urNp linux-2.6.28.8/arch/x86/lib/usercopy_32.c linux-2.6.28.8/arch/x86/lib/usercopy_32.c --- linux-2.6.28.8/arch/x86/lib/usercopy_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/lib/usercopy_32.c 2009-02-21 09:39:27.000000000 -0500 @@ -36,31 +36,38 @@ static inline int __movsl_is_ok(unsigned * Copy a null terminated string from userspace. */ -#define __do_strncpy_from_user(dst, src, count, res) \ -do { \ - int __d0, __d1, __d2; \ - might_sleep(); \ - __asm__ __volatile__( \ - " testl %1,%1\n" \ - " jz 2f\n" \ - "0: lodsb\n" \ - " stosb\n" \ - " testb %%al,%%al\n" \ - " jz 1f\n" \ - " decl %1\n" \ - " jnz 0b\n" \ - "1: subl %1,%0\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "3: movl %5,%0\n" \ - " jmp 2b\n" \ - ".previous\n" \ - _ASM_EXTABLE(0b,3b) \ - : "=&d"(res), "=&c"(count), "=&a" (__d0), "=&S" (__d1), \ - "=&D" (__d2) \ - : "i"(-EFAULT), "0"(count), "1"(count), "3"(src), "4"(dst) \ - : "memory"); \ -} while (0) +static long __do_strncpy_from_user(char *dst, const char __user *src, long count) +{ + int __d0, __d1, __d2; + long res = -EFAULT; + + might_sleep(); + __asm__ __volatile__( + " movw %w10,%%ds\n" + " testl %1,%1\n" + " jz 2f\n" + "0: lodsb\n" + " stosb\n" + " testb %%al,%%al\n" + " jz 1f\n" + " decl %1\n" + " jnz 0b\n" + "1: subl %1,%0\n" + "2:\n" + " pushl %%ss\n" + " popl %%ds\n" + ".section .fixup,\"ax\"\n" + "3: movl %5,%0\n" + " jmp 2b\n" + ".previous\n" + _ASM_EXTABLE(0b,3b) + : "=&d"(res), "=&c"(count), "=&a" (__d0), "=&S" (__d1), + "=&D" (__d2) + : "i"(-EFAULT), "0"(count), "1"(count), "3"(src), "4"(dst), + "r"(__USER_DS) + : "memory"); + return res; +} /** * __strncpy_from_user: - Copy a NUL terminated string from userspace, with less checking. @@ -85,9 +92,7 @@ do { \ long __strncpy_from_user(char *dst, const char __user *src, long count) { - long res; - __do_strncpy_from_user(dst, src, count, res); - return res; + return __do_strncpy_from_user(dst, src, count); } EXPORT_SYMBOL(__strncpy_from_user); @@ -114,7 +119,7 @@ strncpy_from_user(char *dst, const char { long res = -EFAULT; if (access_ok(VERIFY_READ, src, 1)) - __do_strncpy_from_user(dst, src, count, res); + res = __do_strncpy_from_user(dst, src, count); return res; } EXPORT_SYMBOL(strncpy_from_user); @@ -123,24 +128,30 @@ EXPORT_SYMBOL(strncpy_from_user); * Zero Userspace */ -#define __do_clear_user(addr,size) \ -do { \ - int __d0; \ - might_sleep(); \ - __asm__ __volatile__( \ - "0: rep; stosl\n" \ - " movl %2,%0\n" \ - "1: rep; stosb\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "3: lea 0(%2,%0,4),%0\n" \ - " jmp 2b\n" \ - ".previous\n" \ - _ASM_EXTABLE(0b,3b) \ - _ASM_EXTABLE(1b,2b) \ - : "=&c"(size), "=&D" (__d0) \ - : "r"(size & 3), "0"(size / 4), "1"(addr), "a"(0)); \ -} while (0) +static unsigned long __do_clear_user(void __user *addr, unsigned long size) +{ + int __d0; + + might_sleep(); + __asm__ __volatile__( + " movw %w6,%%es\n" + "0: rep; stosl\n" + " movl %2,%0\n" + "1: rep; stosb\n" + "2:\n" + " pushl %%ss\n" + " popl %%es\n" + ".section .fixup,\"ax\"\n" + "3: lea 0(%2,%0,4),%0\n" + " jmp 2b\n" + ".previous\n" + _ASM_EXTABLE(0b,3b) + _ASM_EXTABLE(1b,2b) + : "=&c"(size), "=&D" (__d0) + : "r"(size & 3), "0"(size / 4), "1"(addr), "a"(0), + "r"(__USER_DS)); + return size; +} /** * clear_user: - Zero a block of memory in user space. @@ -157,7 +168,7 @@ clear_user(void __user *to, unsigned lon { might_sleep(); if (access_ok(VERIFY_WRITE, to, n)) - __do_clear_user(to, n); + n = __do_clear_user(to, n); return n; } EXPORT_SYMBOL(clear_user); @@ -176,8 +187,7 @@ EXPORT_SYMBOL(clear_user); unsigned long __clear_user(void __user *to, unsigned long n) { - __do_clear_user(to, n); - return n; + return __do_clear_user(to, n); } EXPORT_SYMBOL(__clear_user); @@ -200,14 +210,17 @@ long strnlen_user(const char __user *s, might_sleep(); __asm__ __volatile__( + " movw %w8,%%es\n" " testl %0, %0\n" " jz 3f\n" - " andl %0,%%ecx\n" + " movl %0,%%ecx\n" "0: repne; scasb\n" " setne %%al\n" " subl %%ecx,%0\n" " addl %0,%%eax\n" "1:\n" + " pushl %%ss\n" + " popl %%es\n" ".section .fixup,\"ax\"\n" "2: xorl %%eax,%%eax\n" " jmp 1b\n" @@ -219,7 +232,7 @@ long strnlen_user(const char __user *s, " .long 0b,2b\n" ".previous" :"=&r" (n), "=&D" (s), "=&a" (res), "=&c" (tmp) - :"0" (n), "1" (s), "2" (0), "3" (mask) + :"0" (n), "1" (s), "2" (0), "3" (mask), "r" (__USER_DS) :"cc"); return res & mask; } @@ -227,10 +240,11 @@ EXPORT_SYMBOL(strnlen_user); #ifdef CONFIG_X86_INTEL_USERCOPY static unsigned long -__copy_user_intel(void __user *to, const void *from, unsigned long size) +__generic_copy_to_user_intel(void __user *to, const void *from, unsigned long size) { int d0, d1; __asm__ __volatile__( + " movw %w6, %%es\n" " .align 2,0x90\n" "1: movl 32(%4), %%eax\n" " cmpl $67, %0\n" @@ -239,36 +253,36 @@ __copy_user_intel(void __user *to, const " .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: movl %%eax, %%es:0(%3)\n" + "6: movl %%edx, %%es: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: movl %%eax, %%es:8(%3)\n" + "10: movl %%edx, %%es: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: movl %%eax, %%es:16(%3)\n" + "14: movl %%edx, %%es: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: movl %%eax, %%es:24(%3)\n" + "18: movl %%edx, %%es: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: movl %%eax, %%es:32(%3)\n" + "22: movl %%edx, %%es: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: movl %%eax, %%es:40(%3)\n" + "26: movl %%edx, %%es: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: movl %%eax, %%es:48(%3)\n" + "30: movl %%edx, %%es: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: movl %%eax, %%es:56(%3)\n" + "34: movl %%edx, %%es:60(%3)\n" " addl $-64, %0\n" " addl $64, %4\n" " addl $64, %3\n" @@ -282,6 +296,8 @@ __copy_user_intel(void __user *to, const "36: movl %%eax, %0\n" "37: rep; movsb\n" "100:\n" + " pushl %%ss\n" + " popl %%es\n" ".section .fixup,\"ax\"\n" "101: lea 0(%%eax,%0,4),%0\n" " jmp 100b\n" @@ -328,7 +344,117 @@ __copy_user_intel(void __user *to, const " .long 99b,101b\n" ".previous" : "=&c"(size), "=&D" (d0), "=&S" (d1) - : "1"(to), "2"(from), "0"(size) + : "1"(to), "2"(from), "0"(size), "r"(__USER_DS) + : "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__( + " movw %w6, %%ds\n" + " .align 2,0x90\n" + "1: movl 32(%4), %%eax\n" + " cmpl $67, %0\n" + " jbe 3f\n" + "2: movl 64(%4), %%eax\n" + " .align 2,0x90\n" + "3: movl 0(%4), %%eax\n" + "4: movl 4(%4), %%edx\n" + "5: movl %%eax, %%es:0(%3)\n" + "6: movl %%edx, %%es:4(%3)\n" + "7: movl 8(%4), %%eax\n" + "8: movl 12(%4),%%edx\n" + "9: movl %%eax, %%es:8(%3)\n" + "10: movl %%edx, %%es:12(%3)\n" + "11: movl 16(%4), %%eax\n" + "12: movl 20(%4), %%edx\n" + "13: movl %%eax, %%es:16(%3)\n" + "14: movl %%edx, %%es:20(%3)\n" + "15: movl 24(%4), %%eax\n" + "16: movl 28(%4), %%edx\n" + "17: movl %%eax, %%es:24(%3)\n" + "18: movl %%edx, %%es:28(%3)\n" + "19: movl 32(%4), %%eax\n" + "20: movl 36(%4), %%edx\n" + "21: movl %%eax, %%es:32(%3)\n" + "22: movl %%edx, %%es:36(%3)\n" + "23: movl 40(%4), %%eax\n" + "24: movl 44(%4), %%edx\n" + "25: movl %%eax, %%es:40(%3)\n" + "26: movl %%edx, %%es:44(%3)\n" + "27: movl 48(%4), %%eax\n" + "28: movl 52(%4), %%edx\n" + "29: movl %%eax, %%es:48(%3)\n" + "30: movl %%edx, %%es:52(%3)\n" + "31: movl 56(%4), %%eax\n" + "32: movl 60(%4), %%edx\n" + "33: movl %%eax, %%es:56(%3)\n" + "34: movl %%edx, %%es: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; movsl\n" + "36: movl %%eax, %0\n" + "37: rep; movsb\n" + "100:\n" + " pushl %%ss\n" + " popl %%ds\n" + ".section .fixup,\"ax\"\n" + "101: lea 0(%%eax,%0,4),%0\n" + " jmp 100b\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .align 4\n" + " .long 1b,100b\n" + " .long 2b,100b\n" + " .long 3b,100b\n" + " .long 4b,100b\n" + " .long 5b,100b\n" + " .long 6b,100b\n" + " .long 7b,100b\n" + " .long 8b,100b\n" + " .long 9b,100b\n" + " .long 10b,100b\n" + " .long 11b,100b\n" + " .long 12b,100b\n" + " .long 13b,100b\n" + " .long 14b,100b\n" + " .long 15b,100b\n" + " .long 16b,100b\n" + " .long 17b,100b\n" + " .long 18b,100b\n" + " .long 19b,100b\n" + " .long 20b,100b\n" + " .long 21b,100b\n" + " .long 22b,100b\n" + " .long 23b,100b\n" + " .long 24b,100b\n" + " .long 25b,100b\n" + " .long 26b,100b\n" + " .long 27b,100b\n" + " .long 28b,100b\n" + " .long 29b,100b\n" + " .long 30b,100b\n" + " .long 31b,100b\n" + " .long 32b,100b\n" + " .long 33b,100b\n" + " .long 34b,100b\n" + " .long 35b,100b\n" + " .long 36b,100b\n" + " .long 37b,100b\n" + " .long 99b,101b\n" + ".previous" + : "=&c"(size), "=&D" (d0), "=&S" (d1) + : "1"(to), "2"(from), "0"(size), "r"(__USER_DS) : "eax", "edx", "memory"); return size; } @@ -338,6 +464,7 @@ __copy_user_zeroing_intel(void *to, cons { int d0, d1; __asm__ __volatile__( + " movw %w6, %%ds\n" " .align 2,0x90\n" "0: movl 32(%4), %%eax\n" " cmpl $67, %0\n" @@ -346,36 +473,36 @@ __copy_user_zeroing_intel(void *to, cons " .align 2,0x90\n" "2: movl 0(%4), %%eax\n" "21: movl 4(%4), %%edx\n" - " movl %%eax, 0(%3)\n" - " movl %%edx, 4(%3)\n" + " movl %%eax, %%es:0(%3)\n" + " movl %%edx, %%es:4(%3)\n" "3: movl 8(%4), %%eax\n" "31: movl 12(%4),%%edx\n" - " movl %%eax, 8(%3)\n" - " movl %%edx, 12(%3)\n" + " movl %%eax, %%es:8(%3)\n" + " movl %%edx, %%es:12(%3)\n" "4: movl 16(%4), %%eax\n" "41: movl 20(%4), %%edx\n" - " movl %%eax, 16(%3)\n" - " movl %%edx, 20(%3)\n" + " movl %%eax, %%es:16(%3)\n" + " movl %%edx, %%es:20(%3)\n" "10: movl 24(%4), %%eax\n" "51: movl 28(%4), %%edx\n" - " movl %%eax, 24(%3)\n" - " movl %%edx, 28(%3)\n" + " movl %%eax, %%es:24(%3)\n" + " movl %%edx, %%es:28(%3)\n" "11: movl 32(%4), %%eax\n" "61: movl 36(%4), %%edx\n" - " movl %%eax, 32(%3)\n" - " movl %%edx, 36(%3)\n" + " movl %%eax, %%es:32(%3)\n" + " movl %%edx, %%es:36(%3)\n" "12: movl 40(%4), %%eax\n" "71: movl 44(%4), %%edx\n" - " movl %%eax, 40(%3)\n" - " movl %%edx, 44(%3)\n" + " movl %%eax, %%es:40(%3)\n" + " movl %%edx, %%es:44(%3)\n" "13: movl 48(%4), %%eax\n" "81: movl 52(%4), %%edx\n" - " movl %%eax, 48(%3)\n" - " movl %%edx, 52(%3)\n" + " movl %%eax, %%es:48(%3)\n" + " movl %%edx, %%es:52(%3)\n" "14: movl 56(%4), %%eax\n" "91: movl 60(%4), %%edx\n" - " movl %%eax, 56(%3)\n" - " movl %%edx, 60(%3)\n" + " movl %%eax, %%es:56(%3)\n" + " movl %%edx, %%es:60(%3)\n" " addl $-64, %0\n" " addl $64, %4\n" " addl $64, %3\n" @@ -389,6 +516,8 @@ __copy_user_zeroing_intel(void *to, cons " movl %%eax,%0\n" "7: rep; movsb\n" "8:\n" + " pushl %%ss\n" + " popl %%ds\n" ".section .fixup,\"ax\"\n" "9: lea 0(%%eax,%0,4),%0\n" "16: pushl %0\n" @@ -423,7 +552,7 @@ __copy_user_zeroing_intel(void *to, cons " .long 7b,16b\n" ".previous" : "=&c"(size), "=&D" (d0), "=&S" (d1) - : "1"(to), "2"(from), "0"(size) + : "1"(to), "2"(from), "0"(size), "r"(__USER_DS) : "eax", "edx", "memory"); return size; } @@ -439,6 +568,7 @@ static unsigned long __copy_user_zeroing int d0, d1; __asm__ __volatile__( + " movw %w6, %%ds\n" " .align 2,0x90\n" "0: movl 32(%4), %%eax\n" " cmpl $67, %0\n" @@ -447,36 +577,36 @@ static unsigned long __copy_user_zeroing " .align 2,0x90\n" "2: movl 0(%4), %%eax\n" "21: movl 4(%4), %%edx\n" - " movnti %%eax, 0(%3)\n" - " movnti %%edx, 4(%3)\n" + " movnti %%eax, %%es:0(%3)\n" + " movnti %%edx, %%es:4(%3)\n" "3: movl 8(%4), %%eax\n" "31: movl 12(%4),%%edx\n" - " movnti %%eax, 8(%3)\n" - " movnti %%edx, 12(%3)\n" + " movnti %%eax, %%es:8(%3)\n" + " movnti %%edx, %%es:12(%3)\n" "4: movl 16(%4), %%eax\n" "41: movl 20(%4), %%edx\n" - " movnti %%eax, 16(%3)\n" - " movnti %%edx, 20(%3)\n" + " movnti %%eax, %%es:16(%3)\n" + " movnti %%edx, %%es:20(%3)\n" "10: movl 24(%4), %%eax\n" "51: movl 28(%4), %%edx\n" - " movnti %%eax, 24(%3)\n" - " movnti %%edx, 28(%3)\n" + " movnti %%eax, %%es:24(%3)\n" + " movnti %%edx, %%es:28(%3)\n" "11: movl 32(%4), %%eax\n" "61: movl 36(%4), %%edx\n" - " movnti %%eax, 32(%3)\n" - " movnti %%edx, 36(%3)\n" + " movnti %%eax, %%es:32(%3)\n" + " movnti %%edx, %%es:36(%3)\n" "12: movl 40(%4), %%eax\n" "71: movl 44(%4), %%edx\n" - " movnti %%eax, 40(%3)\n" - " movnti %%edx, 44(%3)\n" + " movnti %%eax, %%es:40(%3)\n" + " movnti %%edx, %%es:44(%3)\n" "13: movl 48(%4), %%eax\n" "81: movl 52(%4), %%edx\n" - " movnti %%eax, 48(%3)\n" - " movnti %%edx, 52(%3)\n" + " movnti %%eax, %%es:48(%3)\n" + " movnti %%edx, %%es:52(%3)\n" "14: movl 56(%4), %%eax\n" "91: movl 60(%4), %%edx\n" - " movnti %%eax, 56(%3)\n" - " movnti %%edx, 60(%3)\n" + " movnti %%eax, %%es:56(%3)\n" + " movnti %%edx, %%es:60(%3)\n" " addl $-64, %0\n" " addl $64, %4\n" " addl $64, %3\n" @@ -491,6 +621,8 @@ static unsigned long __copy_user_zeroing " movl %%eax,%0\n" "7: rep; movsb\n" "8:\n" + " pushl %%ss\n" + " popl %%ds\n" ".section .fixup,\"ax\"\n" "9: lea 0(%%eax,%0,4),%0\n" "16: pushl %0\n" @@ -525,7 +657,7 @@ static unsigned long __copy_user_zeroing " .long 7b,16b\n" ".previous" : "=&c"(size), "=&D" (d0), "=&S" (d1) - : "1"(to), "2"(from), "0"(size) + : "1"(to), "2"(from), "0"(size), "r"(__USER_DS) : "eax", "edx", "memory"); return size; } @@ -536,6 +668,7 @@ static unsigned long __copy_user_intel_n int d0, d1; __asm__ __volatile__( + " movw %w6, %%ds\n" " .align 2,0x90\n" "0: movl 32(%4), %%eax\n" " cmpl $67, %0\n" @@ -544,36 +677,36 @@ static unsigned long __copy_user_intel_n " .align 2,0x90\n" "2: movl 0(%4), %%eax\n" "21: movl 4(%4), %%edx\n" - " movnti %%eax, 0(%3)\n" - " movnti %%edx, 4(%3)\n" + " movnti %%eax, %%es:0(%3)\n" + " movnti %%edx, %%es:4(%3)\n" "3: movl 8(%4), %%eax\n" "31: movl 12(%4),%%edx\n" - " movnti %%eax, 8(%3)\n" - " movnti %%edx, 12(%3)\n" + " movnti %%eax, %%es:8(%3)\n" + " movnti %%edx, %%es:12(%3)\n" "4: movl 16(%4), %%eax\n" "41: movl 20(%4), %%edx\n" - " movnti %%eax, 16(%3)\n" - " movnti %%edx, 20(%3)\n" + " movnti %%eax, %%es:16(%3)\n" + " movnti %%edx, %%es:20(%3)\n" "10: movl 24(%4), %%eax\n" "51: movl 28(%4), %%edx\n" - " movnti %%eax, 24(%3)\n" - " movnti %%edx, 28(%3)\n" + " movnti %%eax, %%es:24(%3)\n" + " movnti %%edx, %%es:28(%3)\n" "11: movl 32(%4), %%eax\n" "61: movl 36(%4), %%edx\n" - " movnti %%eax, 32(%3)\n" - " movnti %%edx, 36(%3)\n" + " movnti %%eax, %%es:32(%3)\n" + " movnti %%edx, %%es:36(%3)\n" "12: movl 40(%4), %%eax\n" "71: movl 44(%4), %%edx\n" - " movnti %%eax, 40(%3)\n" - " movnti %%edx, 44(%3)\n" + " movnti %%eax, %%es:40(%3)\n" + " movnti %%edx, %%es:44(%3)\n" "13: movl 48(%4), %%eax\n" "81: movl 52(%4), %%edx\n" - " movnti %%eax, 48(%3)\n" - " movnti %%edx, 52(%3)\n" + " movnti %%eax, %%es:48(%3)\n" + " movnti %%edx, %%es:52(%3)\n" "14: movl 56(%4), %%eax\n" "91: movl 60(%4), %%edx\n" - " movnti %%eax, 56(%3)\n" - " movnti %%edx, 60(%3)\n" + " movnti %%eax, %%es:56(%3)\n" + " movnti %%edx, %%es:60(%3)\n" " addl $-64, %0\n" " addl $64, %4\n" " addl $64, %3\n" @@ -588,6 +721,8 @@ static unsigned long __copy_user_intel_n " movl %%eax,%0\n" "7: rep; movsb\n" "8:\n" + " pushl %%ss\n" + " popl %%ds\n" ".section .fixup,\"ax\"\n" "9: lea 0(%%eax,%0,4),%0\n" "16: jmp 8b\n" @@ -616,7 +751,7 @@ static unsigned long __copy_user_intel_n " .long 7b,16b\n" ".previous" : "=&c"(size), "=&D" (d0), "=&S" (d1) - : "1"(to), "2"(from), "0"(size) + : "1"(to), "2"(from), "0"(size), "r"(__USER_DS) : "eax", "edx", "memory"); return size; } @@ -629,90 +764,146 @@ static unsigned long __copy_user_intel_n */ 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) \ -do { \ - int __d0, __d1, __d2; \ - __asm__ __volatile__( \ - " 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" \ - " movl %3,%0\n" \ - " shrl $2,%0\n" \ - " andl $3,%3\n" \ - " .align 2,0x90\n" \ - "0: rep; movsl\n" \ - " movl %3,%0\n" \ - "1: rep; movsb\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "5: addl %3,%0\n" \ - " jmp 2b\n" \ - "3: lea 0(%3,%0,4),%0\n" \ - " jmp 2b\n" \ - ".previous\n" \ - ".section __ex_table,\"a\"\n" \ - " .align 4\n" \ - " .long 4b,5b\n" \ - " .long 0b,3b\n" \ - " .long 1b,2b\n" \ - ".previous" \ - : "=&c"(size), "=&D" (__d0), "=&S" (__d1), "=r"(__d2) \ - : "3"(size), "0"(size), "1"(to), "2"(from) \ - : "memory"); \ -} while (0) - -#define __copy_user_zeroing(to, from, size) \ -do { \ - int __d0, __d1, __d2; \ - __asm__ __volatile__( \ - " 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" \ - " movl %3,%0\n" \ - " shrl $2,%0\n" \ - " andl $3,%3\n" \ - " .align 2,0x90\n" \ - "0: rep; movsl\n" \ - " movl %3,%0\n" \ - "1: rep; movsb\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "5: addl %3,%0\n" \ - " jmp 6f\n" \ - "3: lea 0(%3,%0,4),%0\n" \ - "6: pushl %0\n" \ - " pushl %%eax\n" \ - " xorl %%eax,%%eax\n" \ - " rep; stosb\n" \ - " popl %%eax\n" \ - " popl %0\n" \ - " jmp 2b\n" \ - ".previous\n" \ - ".section __ex_table,\"a\"\n" \ - " .align 4\n" \ - " .long 4b,5b\n" \ - " .long 0b,3b\n" \ - " .long 1b,6b\n" \ - ".previous" \ - : "=&c"(size), "=&D" (__d0), "=&S" (__d1), "=r"(__d2) \ - : "3"(size), "0"(size), "1"(to), "2"(from) \ - : "memory"); \ -} while (0) +static unsigned long +__generic_copy_to_user(void __user *to, const void *from, unsigned long size) +{ + int __d0, __d1, __d2; + + __asm__ __volatile__( + " movw %w8,%%es\n" + " 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" + " movl %3,%0\n" + " shrl $2,%0\n" + " andl $3,%3\n" + " .align 2,0x90\n" + "0: rep; movsl\n" + " movl %3,%0\n" + "1: rep; movsb\n" + "2:\n" + " pushl %%ss\n" + " popl %%es\n" + ".section .fixup,\"ax\"\n" + "5: addl %3,%0\n" + " jmp 2b\n" + "3: lea 0(%3,%0,4),%0\n" + " jmp 2b\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .align 4\n" + " .long 4b,5b\n" + " .long 0b,3b\n" + " .long 1b,2b\n" + ".previous" + : "=&c"(size), "=&D" (__d0), "=&S" (__d1), "=r"(__d2) + : "3"(size), "0"(size), "1"(to), "2"(from), "r"(__USER_DS) + : "memory"); + return size; +} + +static unsigned long +__generic_copy_from_user(void *to, const void __user *from, unsigned long size) +{ + int __d0, __d1, __d2; + + __asm__ __volatile__( + " movw %w8,%%ds\n" + " 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" + " movl %3,%0\n" + " shrl $2,%0\n" + " andl $3,%3\n" + " .align 2,0x90\n" + "0: rep; movsl\n" + " movl %3,%0\n" + "1: rep; movsb\n" + "2:\n" + " pushl %%ss\n" + " popl %%ds\n" + ".section .fixup,\"ax\"\n" + "5: addl %3,%0\n" + " jmp 2b\n" + "3: lea 0(%3,%0,4),%0\n" + " jmp 2b\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .align 4\n" + " .long 4b,5b\n" + " .long 0b,3b\n" + " .long 1b,2b\n" + ".previous" + : "=&c"(size), "=&D" (__d0), "=&S" (__d1), "=r"(__d2) + : "3"(size), "0"(size), "1"(to), "2"(from), "r"(__USER_DS) + : "memory"); + return size; +} + +static unsigned long +__copy_user_zeroing(void *to, const void __user *from, unsigned long size) +{ + int __d0, __d1, __d2; + + __asm__ __volatile__( + " movw %w8,%%ds\n" + " 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" + " movl %3,%0\n" + " shrl $2,%0\n" + " andl $3,%3\n" + " .align 2,0x90\n" + "0: rep; movsl\n" + " movl %3,%0\n" + "1: rep; movsb\n" + "2:\n" + " pushl %%ss\n" + " popl %%ds\n" + ".section .fixup,\"ax\"\n" + "5: addl %3,%0\n" + " jmp 6f\n" + "3: lea 0(%3,%0,4),%0\n" + "6: pushl %0\n" + " pushl %%eax\n" + " xorl %%eax,%%eax\n" + " rep; stosb\n" + " popl %%eax\n" + " popl %0\n" + " jmp 2b\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .align 4\n" + " .long 4b,5b\n" + " .long 0b,3b\n" + " .long 1b,6b\n" + ".previous" + : "=&c"(size), "=&D" (__d0), "=&S" (__d1), "=r"(__d2) + : "3"(size), "0"(size), "1"(to), "2"(from), "r"(__USER_DS) + : "memory"); + return size; +} unsigned long __copy_to_user_ll(void __user *to, const void *from, unsigned long n) @@ -775,9 +966,9 @@ survive: } #endif if (movsl_is_ok(to, from, n)) - __copy_user(to, from, n); + n = __generic_copy_to_user(to, from, n); else - n = __copy_user_intel(to, from, n); + n = __generic_copy_to_user_intel(to, from, n); return n; } EXPORT_SYMBOL(__copy_to_user_ll); @@ -786,7 +977,7 @@ unsigned long __copy_from_user_ll(void * unsigned long n) { if (movsl_is_ok(to, from, n)) - __copy_user_zeroing(to, from, n); + n = __copy_user_zeroing(to, from, n); else n = __copy_user_zeroing_intel(to, from, n); return n; @@ -797,10 +988,9 @@ unsigned long __copy_from_user_ll_nozero unsigned long n) { if (movsl_is_ok(to, from, n)) - __copy_user(to, from, n); + n = __generic_copy_from_user(to, from, n); else - n = __copy_user_intel((void __user *)to, - (const void *)from, n); + n = __generic_copy_from_user_intel(to, from, n); return n; } EXPORT_SYMBOL(__copy_from_user_ll_nozero); @@ -812,9 +1002,9 @@ unsigned long __copy_from_user_ll_nocach if (n > 64 && cpu_has_xmm2) n = __copy_user_zeroing_intel_nocache(to, from, n); else - __copy_user_zeroing(to, from, n); + n = __copy_user_zeroing(to, from, n); #else - __copy_user_zeroing(to, from, n); + n = __copy_user_zeroing(to, from, n); #endif return n; } @@ -827,9 +1017,9 @@ unsigned long __copy_from_user_ll_nocach if (n > 64 && cpu_has_xmm2) n = __copy_user_intel_nocache(to, from, n); else - __copy_user(to, from, n); + n = __generic_copy_from_user(to, from, n); #else - __copy_user(to, from, n); + n = __generic_copy_from_user(to, from, n); #endif return n; } @@ -878,8 +1068,35 @@ copy_from_user(void *to, const void __us { if (access_ok(VERIFY_READ, from, n)) n = __copy_from_user(to, from, n); - else + else if ((long)n > 0) memset(to, 0, n); return n; } EXPORT_SYMBOL(copy_from_user); + +#ifdef CONFIG_PAX_MEMORY_UDEREF +void __set_fs(mm_segment_t x, int cpu) +{ + unsigned long limit = x.seg; + struct desc_struct d; + + current_thread_info()->addr_limit = x; + if (likely(limit)) + limit = (limit - 1UL) >> PAGE_SHIFT; + pack_descriptor(&d, 0UL, limit, 0xF3, 0xC); + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_DEFAULT_USER_DS, &d, DESCTYPE_S); +} + +void set_fs(mm_segment_t x) +{ + __set_fs(x, get_cpu()); + put_cpu_no_resched(); +} +#else +void set_fs(mm_segment_t x) +{ + current_thread_info()->addr_limit = x; +} +#endif + +EXPORT_SYMBOL(set_fs); diff -urNp linux-2.6.28.8/arch/x86/mach-voyager/voyager_basic.c linux-2.6.28.8/arch/x86/mach-voyager/voyager_basic.c --- linux-2.6.28.8/arch/x86/mach-voyager/voyager_basic.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mach-voyager/voyager_basic.c 2009-02-21 09:37:48.000000000 -0500 @@ -123,7 +123,7 @@ int __init voyager_memory_detect(int reg __u8 cmos[4]; ClickMap_t *map; unsigned long map_addr; - unsigned long old; + pte_t old; if (region >= CLICK_ENTRIES) { printk("Voyager: Illegal ClickMap region %d\n", region); @@ -138,7 +138,7 @@ int __init voyager_memory_detect(int reg /* steal page 0 for this */ old = pg0[0]; - pg0[0] = ((map_addr & PAGE_MASK) | _PAGE_RW | _PAGE_PRESENT); + pg0[0] = __pte((map_addr & PAGE_MASK) | _PAGE_RW | _PAGE_PRESENT); local_flush_tlb(); /* now clear everything out but page 0 */ map = (ClickMap_t *) (map_addr & (~PAGE_MASK)); diff -urNp linux-2.6.28.8/arch/x86/mach-voyager/voyager_smp.c linux-2.6.28.8/arch/x86/mach-voyager/voyager_smp.c --- linux-2.6.28.8/arch/x86/mach-voyager/voyager_smp.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mach-voyager/voyager_smp.c 2009-02-21 09:37:48.000000000 -0500 @@ -521,6 +521,10 @@ static void __init do_boot_cpu(__u8 cpu) __u32 *hijack_vector; __u32 start_phys_address = setup_trampoline(); +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + /* There's a clever trick to this: The linux trampoline is * compiled to begin at absolute location zero, so make the * address zero but have the data segment selector compensate @@ -540,7 +544,17 @@ static void __init do_boot_cpu(__u8 cpu) init_gdt(cpu); per_cpu(current_task, cpu) = idle; - early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + + early_gdt_descr.address = get_cpu_gdt_table(cpu); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + irq_ctx_init(cpu); /* Note: Don't modify initial ss override */ @@ -1154,7 +1168,7 @@ void smp_local_timer_interrupt(void) per_cpu(prof_counter, cpu); } - update_process_times(user_mode_vm(get_irq_regs())); + update_process_times(user_mode(get_irq_regs())); } if (((1 << cpu) & voyager_extended_vic_processors) == 0) diff -urNp linux-2.6.28.8/arch/x86/Makefile linux-2.6.28.8/arch/x86/Makefile --- linux-2.6.28.8/arch/x86/Makefile 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/Makefile 2009-02-21 09:37:48.000000000 -0500 @@ -232,3 +232,12 @@ endef CLEAN_FILES += arch/x86/boot/fdimage \ arch/x86/boot/image.iso \ arch/x86/boot/mtools.conf + +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 -urNp linux-2.6.28.8/arch/x86/mm/extable.c linux-2.6.28.8/arch/x86/mm/extable.c --- linux-2.6.28.8/arch/x86/mm/extable.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/extable.c 2009-02-21 09:37:48.000000000 -0500 @@ -1,14 +1,62 @@ #include #include +#include #include +/* + * The exception table needs to be sorted so that the binary + * search that we use to find entries in it works properly. + * This is used both for the kernel exception table and for + * the exception tables of modules that get loaded. + */ +static int cmp_ex(const void *a, const void *b) +{ + const struct exception_table_entry *x = a, *y = b; + + /* avoid overflow */ + if (x->insn > y->insn) + return 1; + if (x->insn < y->insn) + return -1; + return 0; +} + +static void swap_ex(void *a, void *b, int size) +{ + struct exception_table_entry t, *x = a, *y = b; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + + t = *x; + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + + *x = *y; + *y = t; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + +} + +void sort_extable(struct exception_table_entry *start, + struct exception_table_entry *finish) +{ + sort(start, finish - start, sizeof(struct exception_table_entry), + cmp_ex, swap_ex); +} int fixup_exception(struct pt_regs *regs) { const struct exception_table_entry *fixup; #ifdef CONFIG_PNPBIOS - if (unlikely(SEGMENT_IS_PNP_CODE(regs->cs))) { + if (unlikely(!v8086_mode(regs) && SEGMENT_IS_PNP_CODE(regs->cs))) { extern u32 pnp_bios_fault_eip, pnp_bios_fault_esp; extern u32 pnp_bios_is_utter_crap; pnp_bios_is_utter_crap = 1; diff -urNp linux-2.6.28.8/arch/x86/mm/fault.c linux-2.6.28.8/arch/x86/mm/fault.c --- linux-2.6.28.8/arch/x86/mm/fault.c 2009-02-07 16:10:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/fault.c 2009-02-21 09:37:48.000000000 -0500 @@ -26,6 +26,8 @@ #include #include #include +#include +#include #include #include @@ -67,7 +69,7 @@ static inline int notify_page_fault(stru int ret = 0; /* kprobe_running() needs smp_processor_id() */ - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { preempt_disable(); if (kprobe_running() && kprobe_fault_handler(regs, 14)) ret = 1; @@ -265,6 +267,30 @@ bad: #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 + #ifdef CONFIG_X86_32 static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address) { @@ -351,7 +377,7 @@ static int is_errata93(struct pt_regs *r static int is_errata100(struct pt_regs *regs, unsigned long address) { #ifdef CONFIG_X86_64 - if ((regs->cs == __USER32_CS || (regs->cs & (1<<2))) && + if ((regs->cs == __USER32_CS || (regs->cs & SEGMENT_LDT)) && (address >> 32)) return 1; #endif @@ -386,14 +412,31 @@ static void show_fault_oops(struct pt_re #endif #ifdef CONFIG_X86_PAE - if (error_code & PF_INSTR) { + if (nx_enabled && (error_code & PF_INSTR)) { unsigned int level; pte_t *pte = lookup_address(address, &level); if (pte && pte_present(*pte) && !pte_exec(*pte)) printk(KERN_CRIT "kernel tried to execute " "NX-protected page - exploit attempt? " - "(uid: %d)\n", current->uid); + "(uid: %d, task: %s, pid: %d)\n", + current->uid, current->comm, task_pid_nr(current)); + } +#endif + +#ifdef CONFIG_PAX_KERNEXEC +#ifdef CONFIG_MODULES + if (init_mm.start_code <= address && address < (unsigned long)MODULES_END) +#else + if (init_mm.start_code <= address && address < init_mm.end_code) +#endif + { + if (current->signal->curr_ip) + printk(KERN_ERR "PAX: From %u.%u.%u.%u: %s:%d, uid/euid: %u/%u, attempted to modify kernel code\n", + NIPQUAD(current->signal->curr_ip), current->comm, task_pid_nr(current), current->uid, current->euid); + else + printk(KERN_ERR "PAX: %s:%d, uid/euid: %u/%u, attempted to modify kernel code\n", + current->comm, task_pid_nr(current), current->uid, current->euid); } #endif @@ -585,20 +628,26 @@ void __kprobes do_page_fault(struct pt_r struct task_struct *tsk; struct mm_struct *mm; struct vm_area_struct *vma; - unsigned long address; int write, si_code; int fault; #ifdef CONFIG_X86_64 unsigned long flags; #endif +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) + pte_t *pte; + pmd_t *pmd; + spinlock_t *ptl; + unsigned char pte_mask; +#endif + + /* get the address */ + const unsigned long address = read_cr2(); + tsk = current; mm = tsk->mm; prefetchw(&mm->mmap_sem); - /* get the address */ - address = read_cr2(); - si_code = SEGV_MAPERR; if (unlikely(kmmio_fault(regs, address))) @@ -651,7 +700,7 @@ void __kprobes do_page_fault(struct pt_r * User-mode registers count as a user access even for any * potential system fault or CPU buglet. */ - if (user_mode_vm(regs)) { + if (user_mode(regs)) { local_irq_enable(); error_code |= PF_USER; } else if (regs->flags & X86_EFLAGS_IF) @@ -667,7 +716,7 @@ void __kprobes do_page_fault(struct pt_r * atomic region then we must not take the fault. */ if (unlikely(in_atomic() || !mm)) - goto bad_area_nosemaphore; + goto bad_area_nopax; again: /* @@ -689,10 +738,104 @@ again: if (!down_read_trylock(&mm->mmap_sem)) { if ((error_code & PF_USER) == 0 && !search_exception_tables(regs->ip)) - goto bad_area_nosemaphore; + goto bad_area_nopax; down_read(&mm->mmap_sem); } +#if defined(CONFIG_X86_32) && defined(CONFIG_PAX_PAGEEXEC) + if (nx_enabled || (error_code & (PF_PROT|PF_USER)) != (PF_PROT|PF_USER) || v8086_mode(regs) || + !(mm->pax_flags & MF_PAX_PAGEEXEC)) + goto not_pax_fault; + + /* 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; + } +#endif + + pax_report_fault(regs, (void *)regs->ip, (void *)regs->sp); + do_group_exit(SIGKILL); + } + + pmd = pax_get_pmd(mm, address); + if (unlikely(!pmd)) + goto not_pax_fault; + + pte = pte_offset_map_lock(mm, pmd, address, &ptl); + if (unlikely(!(pte_val(*pte) & _PAGE_PRESENT) || pte_user(*pte))) { + pte_unmap_unlock(pte, ptl); + goto not_pax_fault; + } + + if (unlikely((error_code & PF_WRITE) && !pte_write(*pte))) { + /* write attempt to a protected page in user mode */ + pte_unmap_unlock(pte, ptl); + goto not_pax_fault; + } + +#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; + } + + pte_mask = _PAGE_ACCESSED | _PAGE_USER | ((error_code & PF_WRITE) << (_PAGE_BIT_DIRTY-1)); + + /* + * PaX: fill DTLB with user rights and retry + */ + __asm__ __volatile__ ( +#ifdef CONFIG_PAX_MEMORY_UDEREF + "movw %w4,%%es\n" +#endif + "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 + "testb $0,%%es:(%0)\n" + "xorb %3,(%1)\n" +#ifdef CONFIG_PAX_MEMORY_UDEREF + "pushl %%ss\n" + "popl %%es\n" +#endif + : + : "r" (address), "r" (pte), "q" (pte_mask), "i" (_PAGE_USER), "r" (__USER_DS) + : "memory", "cc"); + pte_unmap_unlock(pte, ptl); + up_read(&mm->mmap_sem); + return; + +not_pax_fault: +#endif + vma = find_vma(mm, address); if (!vma) goto bad_area; @@ -700,16 +843,20 @@ again: goto good_area; if (!(vma->vm_flags & VM_GROWSDOWN)) goto bad_area; - 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 (address + 65536 + 32 * sizeof(unsigned long) < regs->sp) - goto bad_area; - } + /* + * 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 (address + 65536 + 32 * sizeof(unsigned long) < task_pt_regs(tsk)->sp) + goto bad_area; + +#ifdef CONFIG_PAX_SEGMEXEC + if ((mm->pax_flags & MF_PAX_SEGMEXEC) && vma->vm_end - SEGMEXEC_TASK_SIZE - 1 < address - SEGMEXEC_TASK_SIZE - 1) + goto bad_area; +#endif + if (expand_stack(vma, address)) goto bad_area; /* @@ -719,6 +866,8 @@ again: good_area: si_code = SEGV_ACCERR; write = 0; + if (nx_enabled && (error_code & PF_INSTR) && !(vma->vm_flags & VM_EXEC)) + goto bad_area; switch (error_code & (PF_PROT|PF_WRITE)) { default: /* 3: write, present */ /* fall through */ @@ -773,6 +922,54 @@ bad_area: up_read(&mm->mmap_sem); bad_area_nosemaphore: + +#if defined(CONFIG_PAX_PAGEEXEC) || defined(CONFIG_PAX_SEGMEXEC) + if (mm && (error_code & PF_USER)) { + unsigned long ip = regs->ip; + + if (v8086_mode(regs)) + ip = ((regs->cs & 0xffff) << 4) + (regs->ip & 0xffff); + + /* + * It's possible to have interrupts off here. + */ + local_irq_enable(); + +#ifdef CONFIG_PAX_PAGEEXEC + if ((mm->pax_flags & MF_PAX_PAGEEXEC) && + ((nx_enabled && (error_code & PF_INSTR)) || (!(error_code & (PF_PROT | PF_WRITE)) && regs->ip == 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 + +#ifdef CONFIG_PAX_SEGMEXEC + if ((mm->pax_flags & MF_PAX_SEGMEXEC) && !(error_code & (PF_PROT | PF_WRITE)) && (regs->ip + SEGMEXEC_TASK_SIZE == 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 + + } +#endif + +bad_area_nopax: /* User mode accesses just cause a SIGSEGV */ if (error_code & PF_USER) { /* @@ -851,7 +1048,7 @@ no_context: #ifdef CONFIG_X86_32 die("Oops", regs, error_code); bust_spinlocks(0); - do_exit(SIGKILL); + do_group_exit(SIGKILL); #else if (__die("Oops", regs, error_code)) regs = NULL; @@ -944,3 +1141,174 @@ void vmalloc_sync_all(void) } #endif } + +#ifdef CONFIG_PAX_EMUTRAMP +static int pax_handle_fetch_fault_32(struct pt_regs *regs) +{ + int err; + + 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: 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(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 __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 / sizeof(long); i++) { + unsigned long c; + if (get_user(c, (unsigned long __user *)sp+i)) +#ifdef CONFIG_X86_32 + printk(KERN_CONT "???????? "); +#else + printk(KERN_CONT "???????????????? "); +#endif + else + printk(KERN_CONT "%0*lx ", 2 * (int)sizeof(long), c); + } + printk("\n"); +} +#endif diff -urNp linux-2.6.28.8/arch/x86/mm/highmem_32.c linux-2.6.28.8/arch/x86/mm/highmem_32.c --- linux-2.6.28.8/arch/x86/mm/highmem_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/highmem_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -74,6 +74,10 @@ void *kmap_atomic_prot(struct page *page enum fixed_addresses idx; unsigned long vaddr; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ pagefault_disable(); @@ -85,7 +89,17 @@ void *kmap_atomic_prot(struct page *page idx = type + KM_TYPE_NR*smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); BUG_ON(!pte_none(*(kmap_pte-idx))); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + set_pte(kmap_pte-idx, mk_pte(page, prot)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + arch_flush_lazy_mmu_mode(); return (void *)vaddr; @@ -101,15 +115,29 @@ void kunmap_atomic(void *kvaddr, enum km unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK; enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id(); +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + /* * Force other mappings to Oops if they'll try to access this pte * without first remap it. Keeping stale mappings around is a bad idea * also, in case the page changes cacheability attributes or becomes * a protected page in a hypervisor. */ - if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx)) + if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx)) { + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + kpte_clear_flush(kmap_pte-idx, vaddr); - else { + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + + } else { #ifdef CONFIG_DEBUG_HIGHMEM BUG_ON(vaddr < PAGE_OFFSET); BUG_ON(vaddr >= (unsigned long)high_memory); @@ -128,11 +156,25 @@ void *kmap_atomic_pfn(unsigned long pfn, enum fixed_addresses idx; unsigned long vaddr; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + pagefault_disable(); idx = type + KM_TYPE_NR*smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + set_pte(kmap_pte-idx, pfn_pte(pfn, kmap_prot)); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + arch_flush_lazy_mmu_mode(); return (void*) vaddr; diff -urNp linux-2.6.28.8/arch/x86/mm/hugetlbpage.c linux-2.6.28.8/arch/x86/mm/hugetlbpage.c --- linux-2.6.28.8/arch/x86/mm/hugetlbpage.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/hugetlbpage.c 2009-02-21 09:37:48.000000000 -0500 @@ -263,13 +263,18 @@ static unsigned long hugetlb_get_unmappe struct hstate *h = hstate_file(file); struct mm_struct *mm = current->mm; struct vm_area_struct *vma; - unsigned long start_addr; + unsigned long start_addr, pax_task_size = TASK_SIZE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif if (len > mm->cached_hole_size) { - start_addr = mm->free_area_cache; + start_addr = mm->free_area_cache; } else { - start_addr = TASK_UNMAPPED_BASE; - mm->cached_hole_size = 0; + start_addr = mm->mmap_base; + mm->cached_hole_size = 0; } full_search: @@ -277,13 +282,13 @@ full_search: for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { /* At this point: (!vma || addr < vma->vm_end). */ - if (TASK_SIZE - len < addr) { + if (pax_task_size - len < addr) { /* * Start a new search - just in case we missed * some holes. */ - if (start_addr != TASK_UNMAPPED_BASE) { - start_addr = TASK_UNMAPPED_BASE; + if (start_addr != mm->mmap_base) { + start_addr = mm->mmap_base; mm->cached_hole_size = 0; goto full_search; } @@ -306,9 +311,8 @@ static unsigned long hugetlb_get_unmappe struct hstate *h = hstate_file(file); struct mm_struct *mm = current->mm; struct vm_area_struct *vma, *prev_vma; - unsigned long base = mm->mmap_base, addr = addr0; + unsigned long base = mm->mmap_base, addr; unsigned long largest_hole = mm->cached_hole_size; - int first_time = 1; /* don't allow allocations above current base */ if (mm->free_area_cache > base) @@ -318,7 +322,7 @@ static unsigned long hugetlb_get_unmappe largest_hole = 0; mm->free_area_cache = base; } -try_again: + /* make sure it can fit in the remaining address space */ if (mm->free_area_cache < len) goto fail; @@ -360,22 +364,26 @@ try_again: fail: /* - * if hint left us with no space for the requested - * mapping then try again: - */ - if (first_time) { - mm->free_area_cache = base; - largest_hole = 0; - first_time = 0; - goto try_again; - } - /* * A failed mmap() very likely causes application failure, * so fall back to the bottom-up function here. This scenario * can happen with large stack limits and large mmap() * allocations. */ - mm->free_area_cache = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + mm->mmap_base = SEGMEXEC_TASK_UNMAPPED_BASE; + else +#endif + + mm->mmap_base = TASK_UNMAPPED_BASE; + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + + mm->free_area_cache = mm->mmap_base; mm->cached_hole_size = ~0UL; addr = hugetlb_get_unmapped_area_bottomup(file, addr0, len, pgoff, flags); @@ -383,6 +391,7 @@ fail: /* * Restore the topdown base: */ + mm->mmap_base = base; mm->free_area_cache = base; mm->cached_hole_size = ~0UL; @@ -396,10 +405,17 @@ hugetlb_get_unmapped_area(struct file *f struct hstate *h = hstate_file(file); struct mm_struct *mm = current->mm; struct vm_area_struct *vma; + unsigned long pax_task_size = TASK_SIZE; if (len & ~huge_page_mask(h)) return -EINVAL; - if (len > TASK_SIZE) + +#ifdef CONFIG_PAX_SEGMEXEC + if (mm->pax_flags & MF_PAX_SEGMEXEC) + pax_task_size = SEGMEXEC_TASK_SIZE; +#endif + + if (len > pax_task_size) return -ENOMEM; if (flags & MAP_FIXED) { @@ -411,7 +427,7 @@ hugetlb_get_unmapped_area(struct file *f if (addr) { addr = ALIGN(addr, huge_page_size(h)); vma = find_vma(mm, addr); - if (TASK_SIZE - len >= addr && + if (pax_task_size - len >= addr && (!vma || addr + len <= vma->vm_start)) return addr; } diff -urNp linux-2.6.28.8/arch/x86/mm/init_32.c linux-2.6.28.8/arch/x86/mm/init_32.c --- linux-2.6.28.8/arch/x86/mm/init_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/init_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -49,6 +49,7 @@ #include #include #include +#include unsigned int __VMALLOC_RESERVE = 128 << 20; @@ -82,35 +83,6 @@ static __init void *alloc_low_page(unsig } /* - * 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 - unsigned long phys; - if (!(pgd_val(*pgd) & _PAGE_PRESENT)) { - if (after_init_bootmem) - pmd_table = (pmd_t *)alloc_bootmem_low_pages(PAGE_SIZE); - else - pmd_table = (pmd_t *)alloc_low_page(&phys); - 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)); - } -#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: */ @@ -132,7 +104,11 @@ static pte_t * __init one_page_table_ini } 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)); } @@ -154,6 +130,7 @@ page_table_range_init(unsigned long star int pgd_idx, pmd_idx; unsigned long vaddr; pgd_t *pgd; + pud_t *pud; pmd_t *pmd; vaddr = start; @@ -162,8 +139,13 @@ page_table_range_init(unsigned long star 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++) { one_page_table_init(pmd); @@ -174,11 +156,23 @@ page_table_range_init(unsigned long star } } -static inline int is_kernel_text(unsigned long addr) +static inline int is_kernel_text(unsigned long start, unsigned long end) { - if (addr >= PAGE_OFFSET && addr <= (unsigned long)__init_end) - return 1; - return 0; + unsigned long etext; + +#if defined(CONFIG_MODULES) && defined(CONFIG_PAX_KERNEXEC) + etext = ktva_ktla((unsigned long)&MODULES_END); +#else + etext = (unsigned long)&_etext; +#endif + + if ((start > ktla_ktva(etext) || + end <= ktla_ktva((unsigned long)_stext)) && + (start > ktla_ktva((unsigned long)_einittext) || + end <= ktla_ktva((unsigned long)_sinittext)) && + (start > (unsigned long)__va(0xfffff) || end <= (unsigned long)__va(0xc0000))) + return 0; + return 1; } /* @@ -191,9 +185,10 @@ static void __init kernel_physical_mappi unsigned long end_pfn, int use_pse) { - 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; @@ -223,8 +218,13 @@ repeat: pfn = start_pfn; pgd_idx = pgd_index((pfn<> PAGE_SHIFT); +#endif if (pfn >= end_pfn) continue; @@ -236,14 +236,13 @@ 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 @@ -253,11 +252,7 @@ repeat: __pgprot(PTE_IDENT_ATTR | _PAGE_PSE); - addr2 = (pfn + PTRS_PER_PTE-1) * PAGE_SIZE + - PAGE_OFFSET + PAGE_SIZE-1; - - if (is_kernel_text(addr) || - is_kernel_text(addr2)) + if (is_kernel_text(address, address + PMD_SIZE)) prot = PAGE_KERNEL_LARGE_EXEC; pages_2m++; @@ -274,7 +269,7 @@ repeat: pte_ofs = pte_index((pfn<> PAGE_SHIFT) <= pagenr && pagenr < (ISA_END_ADDRESS >> PAGE_SHIFT)) return 1; if (!page_is_ram(pagenr)) return 1; @@ -460,7 +457,7 @@ void __init native_pagetable_setup_start pud = pud_offset(pgd, va); pmd = pmd_offset(pud, va); - if (!pmd_present(*pmd)) + if (!pmd_present(*pmd) || pmd_huge(*pmd)) break; pte = pte_offset_kernel(pmd, va); @@ -512,9 +509,7 @@ static void __init early_ioremap_page_ta static void __init pagetable_init(void) { - pgd_t *pgd_base = swapper_pg_dir; - - permanent_kmaps_init(pgd_base); + permanent_kmaps_init(swapper_pg_dir); } #ifdef CONFIG_ACPI_SLEEP @@ -522,12 +517,12 @@ static void __init pagetable_init(void) * ACPI suspend needs this for resume, because things like the intel-agp * driver might have split up a kernel 4MB mapping. */ -char swsusp_pg_dir[PAGE_SIZE] +pgd_t swsusp_pg_dir[PTRS_PER_PGD] __attribute__ ((aligned(PAGE_SIZE))); static inline void save_pg_dir(void) { - memcpy(swsusp_pg_dir, swapper_pg_dir, PAGE_SIZE); + clone_pgd_range(swsusp_pg_dir, swapper_pg_dir, PTRS_PER_PGD); } #else /* !CONFIG_ACPI_SLEEP */ static inline void save_pg_dir(void) @@ -557,13 +552,11 @@ void zap_low_mappings(void) int nx_enabled; -pteval_t __supported_pte_mask __read_mostly = ~(_PAGE_NX | _PAGE_GLOBAL | _PAGE_IOMAP); +pteval_t __supported_pte_mask __read_only = ~(_PAGE_NX | _PAGE_GLOBAL | _PAGE_IOMAP); EXPORT_SYMBOL_GPL(__supported_pte_mask); #ifdef CONFIG_X86_PAE -static int disable_nx __initdata; - /* * noexec = on|off * @@ -572,40 +565,33 @@ static int disable_nx __initdata; * on Enable * off Disable */ +#if !defined(CONFIG_PAX_PAGEEXEC) static int __init noexec_setup(char *str) { if (!str || !strcmp(str, "on")) { - if (cpu_has_nx) { - __supported_pte_mask |= _PAGE_NX; - disable_nx = 0; - } + if (cpu_has_nx) + nx_enabled = 1; } else { - if (!strcmp(str, "off")) { - disable_nx = 1; - __supported_pte_mask &= ~_PAGE_NX; - } else { + if (!strcmp(str, "off")) + nx_enabled = 0; + else return -EINVAL; - } } return 0; } early_param("noexec", noexec_setup); +#endif static void __init set_nx(void) { - unsigned int v[4], l, h; - - if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) { - cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]); + if (!nx_enabled && cpu_has_nx) { + unsigned l, h; - if ((v[3] & (1 << 20)) && !disable_nx) { - rdmsr(MSR_EFER, l, h); - l |= EFER_NX; - wrmsr(MSR_EFER, l, h); - nx_enabled = 1; - __supported_pte_mask |= _PAGE_NX; - } + __supported_pte_mask &= ~_PAGE_NX; + rdmsr(MSR_EFER, l, h); + l &= ~EFER_NX; + wrmsr(MSR_EFER, l, h); } } #endif @@ -988,7 +974,7 @@ void __init mem_init(void) set_highmem_pages_init(); codesize = (unsigned long) &_etext - (unsigned long) &_text; - datasize = (unsigned long) &_edata - (unsigned long) &_etext; + datasize = (unsigned long) &_edata - (unsigned long) &_data; initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin; kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT); @@ -1034,10 +1020,10 @@ void __init mem_init(void) ((unsigned long)&__init_end - (unsigned long)&__init_begin) >> 10, - (unsigned long)&_etext, (unsigned long)&_edata, - ((unsigned long)&_edata - (unsigned long)&_etext) >> 10, + (unsigned long)&_data, (unsigned long)&_edata, + ((unsigned long)&_edata - (unsigned long)&_data) >> 10, - (unsigned long)&_text, (unsigned long)&_etext, + ktla_ktva((unsigned long)&_text), ktla_ktva((unsigned long)&_etext), ((unsigned long)&_etext - (unsigned long)&_text) >> 10); #ifdef CONFIG_HIGHMEM @@ -1166,6 +1152,46 @@ void free_init_pages(char *what, unsigne void free_initmem(void) { + +#ifdef CONFIG_PAX_KERNEXEC + /* PaX: limit KERNEL_CS to actual size */ + unsigned long addr, limit; + struct desc_struct d; + int cpu; + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + +#ifdef CONFIG_MODULES + limit = ktva_ktla((unsigned long)&MODULES_END); +#else + limit = (unsigned long)&_etext; +#endif + limit = (limit - 1UL) >> PAGE_SHIFT; + + for (cpu = 0; cpu < NR_CPUS; cpu++) { + pack_descriptor(&d, get_desc_base(&get_cpu_gdt_table(cpu)[GDT_ENTRY_KERNEL_CS]), limit, 0x9B, 0xC); + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_KERNEL_CS, &d, DESCTYPE_S); + } + + /* PaX: make KERNEL_CS read-only */ + for (addr = ktla_ktva((unsigned long)&_text); addr < (unsigned long)&_data; 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 + 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 + flush_tlb_all(); +#endif + free_init_pages("unused kernel memory", (unsigned long)(&__init_begin), (unsigned long)(&__init_end)); diff -urNp linux-2.6.28.8/arch/x86/mm/init_64.c linux-2.6.28.8/arch/x86/mm/init_64.c --- linux-2.6.28.8/arch/x86/mm/init_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/init_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -175,6 +175,10 @@ set_pte_vaddr_pud(pud_t *pud_page, unsig pmd_t *pmd; pte_t *pte; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + pud = pud_page + pud_index(vaddr); if (pud_none(*pud)) { pmd = (pmd_t *) spp_getpage(); @@ -196,8 +200,17 @@ set_pte_vaddr_pud(pud_t *pud_page, unsig } pte = pte_offset_kernel(pmd, vaddr); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + set_pte(pte, new_pte); +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + /* * It's enough to flush this one mapping. * (PGE mappings get flushed as well) @@ -238,14 +251,12 @@ static void __init __init_extra_mapping( 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)); @@ -886,7 +897,9 @@ EXPORT_SYMBOL_GPL(memory_add_physaddr_to */ int devmem_is_allowed(unsigned long pagenr) { - if (pagenr <= 256) + if (!pagenr) + return 1; + if ((ISA_START_ADDRESS >> PAGE_SHIFT) <= pagenr && pagenr < (ISA_END_ADDRESS >> PAGE_SHIFT)) return 1; if (!page_is_ram(pagenr)) return 1; @@ -977,6 +990,39 @@ void free_init_pages(char *what, unsigne void free_initmem(void) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long addr, end; + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + + /* 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 ((unsigned long)_text <= addr && addr < (unsigned long)_data) + 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 ((unsigned long)__va(__pa(_text)) <= addr && addr < (unsigned long)__va(__pa(_data))) + set_pmd(pmd, __pmd(pmd_val(*pmd) & ~_PAGE_RW)); + else + set_pmd(pmd, __pmd(pmd_val(*pmd) | (_PAGE_NX & __supported_pte_mask))); + } + + flush_tlb_all(); +#endif + free_init_pages("unused kernel memory", (unsigned long)(&__init_begin), (unsigned long)(&__init_end)); @@ -1149,7 +1195,7 @@ int in_gate_area_no_task(unsigned long a const char *arch_vma_name(struct vm_area_struct *vma) { - if (vma->vm_mm && vma->vm_start == (long)vma->vm_mm->context.vdso) + if (vma->vm_mm && vma->vm_start == vma->vm_mm->context.vdso) return "[vdso]"; if (vma == &gate_vma) return "[vsyscall]"; diff -urNp linux-2.6.28.8/arch/x86/mm/ioremap.c linux-2.6.28.8/arch/x86/mm/ioremap.c --- linux-2.6.28.8/arch/x86/mm/ioremap.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/ioremap.c 2009-02-21 09:37:48.000000000 -0500 @@ -114,8 +114,8 @@ int page_is_ram(unsigned long pagenr) * Second special case: Some BIOSen report the PC BIOS * area (640->1Mb) as ram even though it is not. */ - if (pagenr >= (BIOS_BEGIN >> PAGE_SHIFT) && - pagenr < (BIOS_END >> PAGE_SHIFT)) + if (pagenr >= (ISA_START_ADDRESS >> PAGE_SHIFT) && + pagenr < (ISA_END_ADDRESS >> PAGE_SHIFT)) return 0; for (i = 0; i < e820.nr_map; i++) { @@ -293,6 +293,8 @@ static void __iomem *__ioremap_caller(re break; } + prot = canon_pgprot(prot); + /* * Ok, go for it.. */ @@ -508,7 +510,7 @@ static int __init early_ioremap_debug_se early_param("early_ioremap_debug", early_ioremap_debug_setup); static __initdata int after_paging_init; -static pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] __page_aligned_bss; +static __initdata pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] __aligned(PAGE_SIZE); static inline pmd_t * __init early_ioremap_pmd(unsigned long addr) { @@ -523,7 +525,11 @@ static inline pmd_t * __init early_iorem static inline pte_t * __init early_ioremap_pte(unsigned long addr) { +#ifdef CONFIG_X86_32 return &bm_pte[pte_index(addr)]; +#else + return &level1_fixmap_pgt[pte_index(addr)]; +#endif } void __init early_ioremap_init(void) @@ -534,8 +540,10 @@ void __init early_ioremap_init(void) printk(KERN_INFO "early_ioremap_init()\n"); pmd = early_ioremap_pmd(fix_to_virt(FIX_BTMAP_BEGIN)); +#ifdef CONFIG_X86_32 memset(bm_pte, 0, sizeof(bm_pte)); pmd_populate_kernel(&init_mm, pmd, bm_pte); +#endif /* * The boot-ioremap range spans multiple pmds, for which diff -urNp linux-2.6.28.8/arch/x86/mm/mmap.c linux-2.6.28.8/arch/x86/mm/mmap.c --- linux-2.6.28.8/arch/x86/mm/mmap.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/mmap.c 2009-02-21 09:37:48.000000000 -0500 @@ -36,7 +36,7 @@ * Leave an at least ~128 MB hole. */ #define MIN_GAP (128*1024*1024) -#define MAX_GAP (TASK_SIZE/6*5) +#define MAX_GAP (pax_task_size/6*5) /* * True on X86_32 or when emulating IA32 on X86_64 @@ -81,27 +81,40 @@ 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 = current->signal->rlim[RLIMIT_STACK].rlim_cur; + 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(); } @@ -112,11 +125,23 @@ static unsigned long mmap_legacy_base(vo void arch_pick_mmap_layout(struct mm_struct *mm) { if (mmap_is_legacy()) { - mm->mmap_base = mmap_legacy_base(); + mm->mmap_base = mmap_legacy_base(mm); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base += mm->delta_mmap; +#endif + mm->get_unmapped_area = arch_get_unmapped_area; mm->unmap_area = arch_unmap_area; } else { - mm->mmap_base = mmap_base(); + mm->mmap_base = mmap_base(mm); + +#ifdef CONFIG_PAX_RANDMMAP + if (mm->pax_flags & MF_PAX_RANDMMAP) + mm->mmap_base -= mm->delta_mmap + mm->delta_stack; +#endif + mm->get_unmapped_area = arch_get_unmapped_area_topdown; mm->unmap_area = arch_unmap_area_topdown; } diff -urNp linux-2.6.28.8/arch/x86/mm/numa_32.c linux-2.6.28.8/arch/x86/mm/numa_32.c --- linux-2.6.28.8/arch/x86/mm/numa_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/numa_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -98,7 +98,6 @@ unsigned long node_memmap_size_bytes(int } #endif -extern unsigned long find_max_low_pfn(void); extern unsigned long highend_pfn, highstart_pfn; #define LARGE_PAGE_BYTES (PTRS_PER_PTE * PAGE_SIZE) diff -urNp linux-2.6.28.8/arch/x86/mm/pageattr.c linux-2.6.28.8/arch/x86/mm/pageattr.c --- linux-2.6.28.8/arch/x86/mm/pageattr.c 2009-02-20 22:26:38.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/pageattr.c 2009-02-21 09:37:48.000000000 -0500 @@ -20,6 +20,7 @@ #include #include #include +#include /* * The current flushing context - we pass it instead of 5 arguments: @@ -259,7 +260,7 @@ static inline pgprot_t static_protection * 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)) + if (within(address, ktla_ktva((unsigned long)_text), ktla_ktva((unsigned long)_etext))) pgprot_val(forbidden) |= _PAGE_NX; /* @@ -321,8 +322,20 @@ EXPORT_SYMBOL_GPL(lookup_address); */ static void __set_pmd_pte(pte_t *kpte, unsigned long address, pte_t pte) { + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; + + pax_open_kernel(cr0); +#endif + /* change init_mm */ set_pte_atomic(kpte, pte); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + #ifdef CONFIG_X86_32 if (!SHARED_KERNEL_PMD) { struct page *page; diff -urNp linux-2.6.28.8/arch/x86/mm/pageattr-test.c linux-2.6.28.8/arch/x86/mm/pageattr-test.c --- linux-2.6.28.8/arch/x86/mm/pageattr-test.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/pageattr-test.c 2009-03-07 10:35:39.000000000 -0500 @@ -36,7 +36,7 @@ enum { static int pte_testbit(pte_t pte) { - return pte_flags(pte) & _PAGE_UNUSED1; + return pte_flags(pte) & _PAGE_CPA_TEST; } struct split_state { diff -urNp linux-2.6.28.8/arch/x86/mm/pat.c linux-2.6.28.8/arch/x86/mm/pat.c --- linux-2.6.28.8/arch/x86/mm/pat.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/pat.c 2009-03-07 14:05:58.000000000 -0500 @@ -491,7 +491,7 @@ pgprot_t phys_mem_access_prot(struct fil return vma_prot; } -#ifdef CONFIG_STRICT_DEVMEM +#ifndef CONFIG_STRICT_DEVMEM /* This check is done in drivers/char/mem.c in case of STRICT_DEVMEM*/ static inline int range_is_allowed(unsigned long pfn, unsigned long size) { diff -urNp linux-2.6.28.8/arch/x86/mm/pgtable_32.c linux-2.6.28.8/arch/x86/mm/pgtable_32.c --- linux-2.6.28.8/arch/x86/mm/pgtable_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/mm/pgtable_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -31,6 +31,10 @@ void set_pte_vaddr(unsigned long vaddr, pmd_t *pmd; pte_t *pte; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + pgd = swapper_pg_dir + pgd_index(vaddr); if (pgd_none(*pgd)) { BUG(); @@ -47,11 +51,20 @@ void set_pte_vaddr(unsigned long vaddr, return; } pte = pte_offset_kernel(pmd, vaddr); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + if (pte_val(pteval)) set_pte_present(&init_mm, vaddr, pte, pteval); else pte_clear(&init_mm, vaddr, pte); +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + /* * It's enough to flush this one mapping. * (PGE mappings get flushed as well) diff -urNp linux-2.6.28.8/arch/x86/oprofile/backtrace.c linux-2.6.28.8/arch/x86/oprofile/backtrace.c --- linux-2.6.28.8/arch/x86/oprofile/backtrace.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/oprofile/backtrace.c 2009-02-21 09:37:48.000000000 -0500 @@ -37,7 +37,7 @@ static void backtrace_address(void *data unsigned int *depth = data; if ((*depth)--) - oprofile_add_trace(addr); + oprofile_add_trace(ktla_ktva(addr)); } static struct stacktrace_ops backtrace_ops = { @@ -78,7 +78,7 @@ x86_backtrace(struct pt_regs * const reg struct frame_head *head = (struct frame_head *)frame_pointer(regs); unsigned long stack = kernel_trap_sp(regs); - if (!user_mode_vm(regs)) { + if (!user_mode(regs)) { if (depth) dump_trace(NULL, regs, (unsigned long *)stack, 0, &backtrace_ops, &depth); diff -urNp linux-2.6.28.8/arch/x86/oprofile/op_model_p4.c linux-2.6.28.8/arch/x86/oprofile/op_model_p4.c --- linux-2.6.28.8/arch/x86/oprofile/op_model_p4.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/oprofile/op_model_p4.c 2009-02-21 09:37:48.000000000 -0500 @@ -48,7 +48,7 @@ static inline void setup_num_counters(vo #endif } -static int inline addr_increment(void) +static inline int addr_increment(void) { #ifdef CONFIG_SMP return smp_num_siblings == 2 ? 2 : 1; diff -urNp linux-2.6.28.8/arch/x86/pci/common.c linux-2.6.28.8/arch/x86/pci/common.c --- linux-2.6.28.8/arch/x86/pci/common.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/pci/common.c 2009-02-21 09:37:48.000000000 -0500 @@ -362,7 +362,7 @@ static struct dmi_system_id __devinitdat DMI_MATCH(DMI_PRODUCT_NAME, "ProLiant DL585 G2"), }, }, - {} + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL} }; void __init dmi_check_pciprobe(void) diff -urNp linux-2.6.28.8/arch/x86/pci/fixup.c linux-2.6.28.8/arch/x86/pci/fixup.c --- linux-2.6.28.8/arch/x86/pci/fixup.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/pci/fixup.c 2009-02-21 09:37:48.000000000 -0500 @@ -365,7 +365,7 @@ static struct dmi_system_id __devinitdat DMI_MATCH(DMI_PRODUCT_NAME, "MS-6702E"), }, }, - {} + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; /* @@ -436,7 +436,7 @@ static struct dmi_system_id __devinitdat DMI_MATCH(DMI_PRODUCT_VERSION, "PSA40U"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; static void __devinit pci_pre_fixup_toshiba_ohci1394(struct pci_dev *dev) diff -urNp linux-2.6.28.8/arch/x86/pci/irq.c linux-2.6.28.8/arch/x86/pci/irq.c --- linux-2.6.28.8/arch/x86/pci/irq.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/pci/irq.c 2009-02-21 09:37:48.000000000 -0500 @@ -544,7 +544,7 @@ static __init int intel_router_probe(str static struct pci_device_id __initdata pirq_440gx[] = { { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82443GX_0) }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82443GX_2) }, - { }, + { PCI_DEVICE(0, 0) } }; /* 440GX has a proprietary PIRQ router -- don't use it */ @@ -1148,7 +1148,7 @@ static struct dmi_system_id __initdata p DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate 360"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; int __init pcibios_irq_init(void) diff -urNp linux-2.6.28.8/arch/x86/pci/pcbios.c linux-2.6.28.8/arch/x86/pci/pcbios.c --- linux-2.6.28.8/arch/x86/pci/pcbios.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/pci/pcbios.c 2009-02-21 09:37:48.000000000 -0500 @@ -57,50 +57,120 @@ union bios32 { static struct { unsigned long address; unsigned short segment; -} bios32_indirect = { 0, __KERNEL_CS }; +} bios32_indirect __read_only = { 0, __PCIBIOS_CS }; /* * Returns the entry point for the given service, NULL on error */ -static unsigned long bios32_service(unsigned long service) +static unsigned long __devinit bios32_service(unsigned long service) { unsigned char return_code; /* %al */ unsigned long address; /* %ebx */ unsigned long length; /* %ecx */ unsigned long entry; /* %edx */ unsigned long flags; + struct desc_struct d, *gdt; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif local_irq_save(flags); - __asm__("lcall *(%%edi); cld" + + gdt = get_cpu_gdt_table(smp_processor_id()); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + + pack_descriptor(&d, 0UL, 0xFFFFFUL, 0x9B, 0xC); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_CS, &d, DESCTYPE_S); + pack_descriptor(&d, 0UL, 0xFFFFFUL, 0x93, 0xC); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_DS, &d, DESCTYPE_S); + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + + __asm__("movw %w7, %%ds; lcall *(%%edi); push %%ss; pop %%ds; cld" : "=a" (return_code), "=b" (address), "=c" (length), "=d" (entry) : "0" (service), "1" (0), - "D" (&bios32_indirect)); + "D" (&bios32_indirect), + "r"(__PCIBIOS_DS) + : "memory"); + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + + gdt[GDT_ENTRY_PCIBIOS_CS].a = 0; + gdt[GDT_ENTRY_PCIBIOS_CS].b = 0; + gdt[GDT_ENTRY_PCIBIOS_DS].a = 0; + gdt[GDT_ENTRY_PCIBIOS_DS].b = 0; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + local_irq_restore(flags); switch (return_code) { - case 0: - return address + entry; - case 0x80: /* Not present */ - printk(KERN_WARNING "bios32_service(0x%lx): not present\n", service); - return 0; - default: /* Shouldn't happen */ - printk(KERN_WARNING "bios32_service(0x%lx): returned 0x%x -- BIOS bug!\n", - service, return_code); + case 0: { + int cpu; + unsigned char flags; + + printk(KERN_INFO "bios32_service: base:%08lx length:%08lx entry:%08lx\n", address, length, entry); + if (address >= 0xFFFF0 || length > 0x100000 - address || length <= entry) { + printk(KERN_WARNING "bios32_service: not valid\n"); return 0; + } + address = address + PAGE_OFFSET; + length += 16UL; /* some BIOSs underreport this... */ + flags = 4; + if (length >= 64*1024*1024) { + length >>= PAGE_SHIFT; + flags |= 8; + } + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + + for (cpu = 0; cpu < NR_CPUS; cpu++) { + gdt = get_cpu_gdt_table(cpu); + pack_descriptor(&d, address, length, 0x9b, flags); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_CS, &d, DESCTYPE_S); + pack_descriptor(&d, address, length, 0x93, flags); + write_gdt_entry(gdt, GDT_ENTRY_PCIBIOS_DS, &d, DESCTYPE_S); + } + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + + return entry; + } + case 0x80: /* Not present */ + printk(KERN_WARNING "bios32_service(0x%lx): not present\n", service); + return 0; + default: /* Shouldn't happen */ + printk(KERN_WARNING "bios32_service(0x%lx): returned 0x%x -- BIOS bug!\n", + service, return_code); + return 0; } } static struct { unsigned long address; unsigned short segment; -} pci_indirect = { 0, __KERNEL_CS }; +} pci_indirect __read_only = { 0, __PCIBIOS_CS }; -static int pci_bios_present; +static int pci_bios_present __read_only; static int __devinit check_pcibios(void) { @@ -109,11 +179,13 @@ static int __devinit check_pcibios(void) unsigned long flags, pcibios_entry; if ((pcibios_entry = bios32_service(PCI_SERVICE))) { - pci_indirect.address = pcibios_entry + PAGE_OFFSET; + pci_indirect.address = pcibios_entry; local_irq_save(flags); - __asm__( - "lcall *(%%edi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%edi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -122,7 +194,8 @@ static int __devinit check_pcibios(void) "=b" (ebx), "=c" (ecx) : "1" (PCIBIOS_PCI_BIOS_PRESENT), - "D" (&pci_indirect) + "D" (&pci_indirect), + "r" (__PCIBIOS_DS) : "memory"); local_irq_restore(flags); @@ -166,7 +239,10 @@ static int pci_bios_read(unsigned int se switch (len) { case 1: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -175,7 +251,8 @@ static int pci_bios_read(unsigned int se : "1" (PCIBIOS_READ_CONFIG_BYTE), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); /* * Zero-extend the result beyond 8 bits, do not trust the * BIOS having done it: @@ -183,7 +260,10 @@ static int pci_bios_read(unsigned int se *value &= 0xff; break; case 2: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -192,7 +272,8 @@ static int pci_bios_read(unsigned int se : "1" (PCIBIOS_READ_CONFIG_WORD), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); /* * Zero-extend the result beyond 16 bits, do not trust the * BIOS having done it: @@ -200,7 +281,10 @@ static int pci_bios_read(unsigned int se *value &= 0xffff; break; case 4: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -209,7 +293,8 @@ static int pci_bios_read(unsigned int se : "1" (PCIBIOS_READ_CONFIG_DWORD), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; } @@ -232,7 +317,10 @@ static int pci_bios_write(unsigned int s switch (len) { case 1: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -241,10 +329,14 @@ static int pci_bios_write(unsigned int s "c" (value), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; case 2: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -253,10 +345,14 @@ static int pci_bios_write(unsigned int s "c" (value), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; case 4: - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w6, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n\t" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -265,7 +361,8 @@ static int pci_bios_write(unsigned int s "c" (value), "b" (bx), "D" ((long)reg), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); break; } @@ -369,10 +466,13 @@ struct irq_routing_table * pcibios_get_i DBG("PCI: Fetching IRQ routing table... "); __asm__("push %%es\n\t" + "movw %w8, %%ds\n\t" "push %%ds\n\t" "pop %%es\n\t" - "lcall *(%%esi); cld\n\t" + "lcall *%%ss:(%%esi); cld\n\t" "pop %%es\n\t" + "push %%ss\n\t" + "pop %%ds\n" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -383,7 +483,8 @@ struct irq_routing_table * pcibios_get_i "1" (0), "D" ((long) &opt), "S" (&pci_indirect), - "m" (opt) + "m" (opt), + "r" (__PCIBIOS_DS) : "memory"); DBG("OK ret=%d, size=%d, map=%x\n", ret, opt.size, map); if (ret & 0xff00) @@ -407,7 +508,10 @@ int pcibios_set_irq_routing(struct pci_d { int ret; - __asm__("lcall *(%%esi); cld\n\t" + __asm__("movw %w5, %%ds\n\t" + "lcall *%%ss:(%%esi); cld\n\t" + "push %%ss\n\t" + "pop %%ds\n" "jc 1f\n\t" "xor %%ah, %%ah\n" "1:" @@ -415,7 +519,8 @@ int pcibios_set_irq_routing(struct pci_d : "0" (PCIBIOS_SET_PCI_HW_INT), "b" ((dev->bus->number << 8) | dev->devfn), "c" ((irq << 8) | (pin + 10)), - "S" (&pci_indirect)); + "S" (&pci_indirect), + "r" (__PCIBIOS_DS)); return !(ret & 0xff00); } EXPORT_SYMBOL(pcibios_set_irq_routing); diff -urNp linux-2.6.28.8/arch/x86/power/cpu_32.c linux-2.6.28.8/arch/x86/power/cpu_32.c --- linux-2.6.28.8/arch/x86/power/cpu_32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/power/cpu_32.c 2009-02-21 09:37:48.000000000 -0500 @@ -67,7 +67,7 @@ static void do_fpu_end(void) static void fix_processor_context(void) { int cpu = smp_processor_id(); - struct tss_struct *t = &per_cpu(init_tss, cpu); + struct tss_struct *t = init_tss + cpu; set_tss_desc(cpu, t); /* * This just modifies memory; should not be diff -urNp linux-2.6.28.8/arch/x86/power/cpu_64.c linux-2.6.28.8/arch/x86/power/cpu_64.c --- linux-2.6.28.8/arch/x86/power/cpu_64.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/power/cpu_64.c 2009-02-21 09:37:48.000000000 -0500 @@ -143,7 +143,11 @@ void restore_processor_state(void) static void fix_processor_context(void) { int cpu = smp_processor_id(); - struct tss_struct *t = &per_cpu(init_tss, cpu); + struct tss_struct *t = init_tss + cpu; + +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif /* * This just modifies memory; should not be necessary. But... This @@ -152,8 +156,16 @@ static void fix_processor_context(void) */ set_tss_desc(cpu, t); +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + get_cpu_gdt_table(cpu)[GDT_ENTRY_TSS].type = 9; +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + syscall_init(); /* This sets MSR_*STAR and related */ load_TR_desc(); /* This does ltr */ load_LDT(¤t->active_mm->context); /* This does lldt */ diff -urNp linux-2.6.28.8/arch/x86/vdso/vdso32-setup.c linux-2.6.28.8/arch/x86/vdso/vdso32-setup.c --- linux-2.6.28.8/arch/x86/vdso/vdso32-setup.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/vdso/vdso32-setup.c 2009-02-21 09:37:48.000000000 -0500 @@ -226,7 +226,7 @@ static inline void map_compat_vdso(int m void enable_sep_cpu(void) { int cpu = get_cpu(); - struct tss_struct *tss = &per_cpu(init_tss, cpu); + struct tss_struct *tss = init_tss + cpu; if (!boot_cpu_has(X86_FEATURE_SEP)) { put_cpu(); @@ -249,7 +249,7 @@ static int __init gate_vma_init(void) gate_vma.vm_start = FIXADDR_USER_START; gate_vma.vm_end = FIXADDR_USER_END; gate_vma.vm_flags = VM_READ | VM_MAYREAD | VM_EXEC | VM_MAYEXEC; - gate_vma.vm_page_prot = __P101; + gate_vma.vm_page_prot = vm_get_page_prot(gate_vma.vm_flags); /* * Make sure the vDSO gets into every core dump. * Dumping its contents makes post-mortem fully interpretable later @@ -331,7 +331,7 @@ int arch_setup_additional_pages(struct l if (compat) addr = VDSO_HIGH_BASE; else { - addr = get_unmapped_area(NULL, 0, PAGE_SIZE, 0, 0); + addr = get_unmapped_area(NULL, 0, PAGE_SIZE, 0, MAP_EXECUTABLE); if (IS_ERR_VALUE(addr)) { ret = addr; goto up_fail; @@ -358,7 +358,7 @@ int arch_setup_additional_pages(struct l goto up_fail; } - current->mm->context.vdso = (void *)addr; + current->mm->context.vdso = addr; current_thread_info()->sysenter_return = VDSO32_SYMBOL(addr, SYSENTER_RETURN); @@ -384,7 +384,7 @@ static ctl_table abi_table2[] = { .mode = 0644, .proc_handler = proc_dointvec }, - {} + { 0, NULL, NULL, 0, 0, NULL, NULL, NULL, NULL, NULL, NULL } }; static ctl_table abi_root_table2[] = { @@ -394,7 +394,7 @@ static ctl_table abi_root_table2[] = { .mode = 0555, .child = abi_table2 }, - {} + { 0, NULL, NULL, 0, 0, NULL, NULL, NULL, NULL, NULL, NULL } }; static __init int ia32_binfmt_init(void) @@ -409,8 +409,14 @@ __initcall(ia32_binfmt_init); const char *arch_vma_name(struct vm_area_struct *vma) { - if (vma->vm_mm && vma->vm_start == (long)vma->vm_mm->context.vdso) + if (vma->vm_mm && vma->vm_start == vma->vm_mm->context.vdso) return "[vdso]"; + +#ifdef CONFIG_PAX_SEGMEXEC + if (vma->vm_mm && vma->vm_mirror && vma->vm_mirror->vm_start == vma->vm_mm->context.vdso) + return "[vdso]"; +#endif + return NULL; } @@ -419,7 +425,7 @@ struct vm_area_struct *get_gate_vma(stru struct mm_struct *mm = tsk->mm; /* Check to see if this task was created in compat vdso mode */ - if (mm && mm->context.vdso == (void *)VDSO_HIGH_BASE) + if (mm && mm->context.vdso == VDSO_HIGH_BASE) return &gate_vma; return NULL; } diff -urNp linux-2.6.28.8/arch/x86/vdso/vma.c linux-2.6.28.8/arch/x86/vdso/vma.c --- linux-2.6.28.8/arch/x86/vdso/vma.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/vdso/vma.c 2009-02-21 09:37:48.000000000 -0500 @@ -123,7 +123,7 @@ int arch_setup_additional_pages(struct l if (ret) goto up_fail; - current->mm->context.vdso = (void *)addr; + current->mm->context.vdso = addr; up_fail: up_write(&mm->mmap_sem); return ret; diff -urNp linux-2.6.28.8/arch/x86/xen/enlighten.c linux-2.6.28.8/arch/x86/xen/enlighten.c --- linux-2.6.28.8/arch/x86/xen/enlighten.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/arch/x86/xen/enlighten.c 2009-03-07 10:29:51.000000000 -0500 @@ -318,7 +318,7 @@ static void xen_set_ldt(const void *addr static void xen_load_gdt(const struct desc_ptr *dtr) { unsigned long *frames; - unsigned long va = dtr->address; + unsigned long va = (unsigned long)dtr->address; unsigned int size = dtr->size + 1; unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE; int f; @@ -333,7 +333,7 @@ static void xen_load_gdt(const struct de mcs = xen_mc_entry(sizeof(*frames) * pages); frames = mcs.args; - for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) { + for (f = 0; va < (unsigned long)dtr->address + size; va += PAGE_SIZE, f++) { frames[f] = virt_to_mfn(va); make_lowmem_page_readonly((void *)va); } @@ -441,7 +441,7 @@ static void xen_write_idt_entry(gate_des preempt_disable(); - start = __get_cpu_var(idt_desc).address; + start = (unsigned long)__get_cpu_var(idt_desc).address; end = start + __get_cpu_var(idt_desc).size + 1; xen_mc_flush(); @@ -1526,6 +1526,8 @@ static __init pgd_t *xen_setup_kernel_pa convert_pfn_mfn(init_level4_pgt); convert_pfn_mfn(level3_ident_pgt); convert_pfn_mfn(level3_kernel_pgt); + convert_pfn_mfn(level3_vmalloc_pgt); + convert_pfn_mfn(level3_vmemmap_pgt); l3 = m2v(pgd[pgd_index(__START_KERNEL_map)].pgd); l2 = m2v(l3[pud_index(__START_KERNEL_map)].pud); @@ -1544,9 +1546,12 @@ static __init pgd_t *xen_setup_kernel_pa set_page_prot(init_level4_pgt, PAGE_KERNEL_RO); set_page_prot(level3_ident_pgt, PAGE_KERNEL_RO); set_page_prot(level3_kernel_pgt, PAGE_KERNEL_RO); + set_page_prot(level3_vmalloc_pgt, PAGE_KERNEL_RO); + set_page_prot(level3_vmemmap_pgt, PAGE_KERNEL_RO); set_page_prot(level3_user_vsyscall, PAGE_KERNEL_RO); set_page_prot(level2_kernel_pgt, PAGE_KERNEL_RO); set_page_prot(level2_fixmap_pgt, PAGE_KERNEL_RO); + set_page_prot(level1_fixmap_pgt, PAGE_KERNEL_RO); /* Pin down new L4 */ pin_pagetable_pfn(MMUEXT_PIN_L4_TABLE, diff -urNp linux-2.6.28.8/arch/x86/xen/smp.c linux-2.6.28.8/arch/x86/xen/smp.c --- linux-2.6.28.8/arch/x86/xen/smp.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/arch/x86/xen/smp.c 2009-02-21 09:37:48.000000000 -0500 @@ -171,11 +171,6 @@ static void __init xen_smp_prepare_boot_ { BUG_ON(smp_processor_id() != 0); native_smp_prepare_boot_cpu(); - - /* We've switched to the "real" per-cpu gdt, so make sure the - old memory can be recycled */ - make_lowmem_page_readwrite(&per_cpu_var(gdt_page)); - xen_setup_vcpu_info_placement(); } @@ -231,8 +226,8 @@ cpu_initialize_context(unsigned int cpu, gdt = get_cpu_gdt_table(cpu); ctxt->flags = VGCF_IN_KERNEL; - ctxt->user_regs.ds = __USER_DS; - ctxt->user_regs.es = __USER_DS; + ctxt->user_regs.ds = __KERNEL_DS; + ctxt->user_regs.es = __KERNEL_DS; ctxt->user_regs.ss = __KERNEL_DS; #ifdef CONFIG_X86_32 ctxt->user_regs.fs = __KERNEL_PERCPU; diff -urNp linux-2.6.28.8/crypto/async_tx/async_tx.c linux-2.6.28.8/crypto/async_tx/async_tx.c --- linux-2.6.28.8/crypto/async_tx/async_tx.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/crypto/async_tx/async_tx.c 2009-02-21 09:37:48.000000000 -0500 @@ -358,8 +358,8 @@ async_tx_init(void) err: printk(KERN_ERR "async_tx: initialization failure\n"); - while (--cap >= 0) - free_percpu(channel_table[cap]); + while (cap) + free_percpu(channel_table[--cap]); return 1; } diff -urNp linux-2.6.28.8/crypto/lrw.c linux-2.6.28.8/crypto/lrw.c --- linux-2.6.28.8/crypto/lrw.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/crypto/lrw.c 2009-02-21 09:37:48.000000000 -0500 @@ -54,7 +54,7 @@ static int setkey(struct crypto_tfm *par struct priv *ctx = crypto_tfm_ctx(parent); struct crypto_cipher *child = ctx->child; int err, i; - be128 tmp = { 0 }; + be128 tmp = { 0, 0 }; int bsize = crypto_cipher_blocksize(child); crypto_cipher_clear_flags(child, CRYPTO_TFM_REQ_MASK); diff -urNp linux-2.6.28.8/Documentation/dontdiff linux-2.6.28.8/Documentation/dontdiff --- linux-2.6.28.8/Documentation/dontdiff 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/Documentation/dontdiff 2009-02-21 09:37:48.000000000 -0500 @@ -3,6 +3,7 @@ *.bin *.cpio *.csp +*.dbg *.dsp *.dvi *.elf @@ -49,6 +50,10 @@ 53c700_d.h CVS ChangeSet +GPATH +GRTAGS +GSYMS +GTAGS Image Kerntypes Module.markers @@ -62,7 +67,6 @@ aic7*reg_print.c* aic7*seq.h* aicasm aicdb.h* -asm asm-offsets.h asm_offsets.h autoconf.h* @@ -77,6 +81,7 @@ btfixupprep build bvmlinux bzImage* +capflags.c classlist.h* comp*.log compile.h* @@ -188,12 +193,15 @@ version.h* vmlinux vmlinux-* vmlinux.aout +vmlinux.bin.all vmlinux.lds +vmlinux.relocs vsyscall.lds vsyscall_32.lds wanxlfw.inc uImage unifdef +utsrelease.h wakeup.bin wakeup.elf wakeup.lds diff -urNp linux-2.6.28.8/drivers/acpi/blacklist.c linux-2.6.28.8/drivers/acpi/blacklist.c --- linux-2.6.28.8/drivers/acpi/blacklist.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/acpi/blacklist.c 2009-02-21 09:37:48.000000000 -0500 @@ -71,7 +71,7 @@ static struct acpi_blacklist_item acpi_b {"IBM ", "TP600E ", 0x00000105, ACPI_SIG_DSDT, less_than_or_equal, "Incorrect _ADR", 1}, - {""} + {"", "", 0, 0, 0, all_versions, 0} }; #if CONFIG_ACPI_BLACKLIST_YEAR diff -urNp linux-2.6.28.8/drivers/acpi/osl.c linux-2.6.28.8/drivers/acpi/osl.c --- linux-2.6.28.8/drivers/acpi/osl.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/acpi/osl.c 2009-02-21 09:37:48.000000000 -0500 @@ -483,6 +483,8 @@ acpi_os_read_memory(acpi_physical_addres void __iomem *virt_addr; virt_addr = ioremap(phys_addr, width); + if (!virt_addr) + return AE_NO_MEMORY; if (!value) value = &dummy; @@ -511,6 +513,8 @@ acpi_os_write_memory(acpi_physical_addre void __iomem *virt_addr; virt_addr = ioremap(phys_addr, width); + if (!virt_addr) + return AE_NO_MEMORY; switch (width) { case 8: diff -urNp linux-2.6.28.8/drivers/acpi/processor_core.c linux-2.6.28.8/drivers/acpi/processor_core.c --- linux-2.6.28.8/drivers/acpi/processor_core.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/acpi/processor_core.c 2009-02-21 09:37:48.000000000 -0500 @@ -678,7 +678,7 @@ static int __cpuinit acpi_processor_star return 0; } - BUG_ON((pr->id >= nr_cpu_ids) || (pr->id < 0)); + BUG_ON(pr->id >= nr_cpu_ids); /* * Buggy BIOS check diff -urNp linux-2.6.28.8/drivers/acpi/processor_idle.c linux-2.6.28.8/drivers/acpi/processor_idle.c --- linux-2.6.28.8/drivers/acpi/processor_idle.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/acpi/processor_idle.c 2009-02-21 09:37:48.000000000 -0500 @@ -181,7 +181,7 @@ static struct dmi_system_id __cpuinitdat DMI_MATCH(DMI_BIOS_VENDOR,"Phoenix Technologies LTD"), DMI_MATCH(DMI_BIOS_VERSION,"SHE845M0.86C.0013.D.0302131307")}, (void *)2}, - {}, + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL}, }; static inline u32 ticks_elapsed(u32 t1, u32 t2) diff -urNp linux-2.6.28.8/drivers/acpi/tables/tbfadt.c linux-2.6.28.8/drivers/acpi/tables/tbfadt.c --- linux-2.6.28.8/drivers/acpi/tables/tbfadt.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/acpi/tables/tbfadt.c 2009-02-21 09:37:48.000000000 -0500 @@ -48,7 +48,7 @@ ACPI_MODULE_NAME("tbfadt") /* Local prototypes */ -static void inline +static inline void acpi_tb_init_generic_address(struct acpi_generic_address *generic_address, u8 byte_width, u64 address); @@ -122,7 +122,7 @@ static struct acpi_fadt_info fadt_info_t * ******************************************************************************/ -static void inline +static inline void acpi_tb_init_generic_address(struct acpi_generic_address *generic_address, u8 byte_width, u64 address) { diff -urNp linux-2.6.28.8/drivers/ata/ahci.c linux-2.6.28.8/drivers/ata/ahci.c --- linux-2.6.28.8/drivers/ata/ahci.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ata/ahci.c 2009-02-21 09:37:48.000000000 -0500 @@ -606,7 +606,7 @@ static const struct pci_device_id ahci_p { PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_STORAGE_SATA_AHCI, 0xffffff, board_ahci }, - { } /* terminate list */ + { 0, 0, 0, 0, 0, 0, 0 } /* terminate list */ }; diff -urNp linux-2.6.28.8/drivers/ata/ata_piix.c linux-2.6.28.8/drivers/ata/ata_piix.c --- linux-2.6.28.8/drivers/ata/ata_piix.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ata/ata_piix.c 2009-02-21 09:37:48.000000000 -0500 @@ -289,7 +289,7 @@ static const struct pci_device_id piix_p { 0x8086, 0x3b2d, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich8_2port_sata }, /* SATA Controller IDE (PCH) */ { 0x8086, 0x3b2e, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich8_sata }, - { } /* terminate list */ + { 0, 0, 0, 0, 0, 0, 0 } /* terminate list */ }; static struct pci_driver piix_pci_driver = { @@ -593,7 +593,7 @@ static const struct ich_laptop ich_lapto { 0x266F, 0x1025, 0x0066 }, /* ICH6 on ACER Aspire 1694WLMi */ { 0x2653, 0x1043, 0x82D8 }, /* ICH6M on Asus Eee 701 */ /* end marker */ - { 0, } + { 0, 0, 0 } }; /** @@ -1052,7 +1052,7 @@ static int piix_broken_suspend(void) }, }, - { } /* terminate list */ + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } /* terminate list */ }; static const char *oemstrs[] = { "Tecra M3,", diff -urNp linux-2.6.28.8/drivers/ata/libata-core.c linux-2.6.28.8/drivers/ata/libata-core.c --- linux-2.6.28.8/drivers/ata/libata-core.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/drivers/ata/libata-core.c 2009-03-07 10:29:51.000000000 -0500 @@ -807,7 +807,7 @@ static const struct ata_xfer_ent { { ATA_SHIFT_PIO, ATA_NR_PIO_MODES, XFER_PIO_0 }, { ATA_SHIFT_MWDMA, ATA_NR_MWDMA_MODES, XFER_MW_DMA_0 }, { ATA_SHIFT_UDMA, ATA_NR_UDMA_MODES, XFER_UDMA_0 }, - { -1, }, + { -1, 0, 0 } }; /** @@ -2983,7 +2983,7 @@ static const struct ata_timing ata_timin { XFER_UDMA_5, 0, 0, 0, 0, 0, 0, 0, 20 }, { XFER_UDMA_6, 0, 0, 0, 0, 0, 0, 0, 15 }, - { 0xFF } + { 0xFF, 0, 0, 0, 0, 0, 0, 0, 0 } }; #define ENOUGH(v, unit) (((v)-1)/(unit)+1) @@ -4149,7 +4149,7 @@ static const struct ata_blacklist_entry { "MTRON MSP-SATA*", NULL, ATA_HORKAGE_BRIDGE_OK, }, /* End Marker */ - { } + { NULL, NULL, 0 } }; static int strn_pattern_cmp(const char *patt, const char *name, int wildchar) diff -urNp linux-2.6.28.8/drivers/char/agp/frontend.c linux-2.6.28.8/drivers/char/agp/frontend.c --- linux-2.6.28.8/drivers/char/agp/frontend.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/agp/frontend.c 2009-02-21 09:37:48.000000000 -0500 @@ -824,7 +824,7 @@ static int agpioc_reserve_wrap(struct ag if (copy_from_user(&reserve, arg, sizeof(struct agp_region))) return -EFAULT; - if ((unsigned) reserve.seg_count >= ~0U/sizeof(struct agp_segment)) + if ((unsigned) reserve.seg_count >= ~0U/sizeof(struct agp_segment_priv)) return -EFAULT; client = agp_find_client_by_pid(reserve.pid); diff -urNp linux-2.6.28.8/drivers/char/agp/intel-agp.c linux-2.6.28.8/drivers/char/agp/intel-agp.c --- linux-2.6.28.8/drivers/char/agp/intel-agp.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/drivers/char/agp/intel-agp.c 2009-03-07 10:29:51.000000000 -0500 @@ -2369,7 +2369,7 @@ static struct pci_device_id agp_intel_pc ID(PCI_DEVICE_ID_INTEL_Q45_HB), ID(PCI_DEVICE_ID_INTEL_G45_HB), ID(PCI_DEVICE_ID_INTEL_G41_HB), - { } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, agp_intel_pci_table); diff -urNp linux-2.6.28.8/drivers/char/hpet.c linux-2.6.28.8/drivers/char/hpet.c --- linux-2.6.28.8/drivers/char/hpet.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/hpet.c 2009-02-21 09:37:48.000000000 -0500 @@ -975,7 +975,7 @@ static struct acpi_driver hpet_acpi_driv }, }; -static struct miscdevice hpet_misc = { HPET_MINOR, "hpet", &hpet_fops }; +static struct miscdevice hpet_misc = { HPET_MINOR, "hpet", &hpet_fops, {NULL, NULL}, NULL, NULL }; static int __init hpet_init(void) { diff -urNp linux-2.6.28.8/drivers/char/keyboard.c linux-2.6.28.8/drivers/char/keyboard.c --- linux-2.6.28.8/drivers/char/keyboard.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/keyboard.c 2009-02-21 09:37:48.000000000 -0500 @@ -635,6 +635,16 @@ static void k_spec(struct vc_data *vc, u kbd->kbdmode == VC_MEDIUMRAW) && value != KVAL(K_SAK)) return; /* SAK is allowed even in raw mode */ + +#if defined(CONFIG_GRKERNSEC_PROC) || defined(CONFIG_GRKERNSEC_PROC_MEMMAP) + { + void *func = fn_handler[value]; + if (func == fn_show_state || func == fn_show_ptregs || + func == fn_show_mem) + return; + } +#endif + fn_handler[value](vc); } @@ -1388,7 +1398,7 @@ static const struct input_device_id kbd_ .evbit = { BIT_MASK(EV_SND) }, }, - { }, /* Terminating entry */ + { 0 }, /* Terminating entry */ }; MODULE_DEVICE_TABLE(input, kbd_ids); diff -urNp linux-2.6.28.8/drivers/char/mem.c linux-2.6.28.8/drivers/char/mem.c --- linux-2.6.28.8/drivers/char/mem.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/mem.c 2009-02-21 09:37:48.000000000 -0500 @@ -18,6 +18,7 @@ #include #include #include +#include #include #include #include @@ -35,6 +36,10 @@ # include #endif +#ifdef CONFIG_GRKERNSEC +extern struct file_operations grsec_fops; +#endif + /* * Architectures vary in how they handle caching for addresses * outside of main memory. @@ -192,6 +197,11 @@ static ssize_t write_mem(struct file * f if (!valid_phys_addr_range(p, count)) return -EFAULT; +#ifdef CONFIG_GRKERNSEC_KMEM + gr_handle_mem_write(); + return -EPERM; +#endif + written = 0; #ifdef __ARCH_HAS_NO_PAGE_ZERO_MAPPED @@ -350,6 +360,11 @@ static int mmap_mem(struct file * file, &vma->vm_page_prot)) return -EINVAL; +#ifdef CONFIG_GRKERNSEC_KMEM + if (gr_handle_mem_mmap(vma->vm_pgoff << PAGE_SHIFT, vma)) + return -EPERM; +#endif + vma->vm_page_prot = phys_mem_access_prot(file, vma->vm_pgoff, size, vma->vm_page_prot); @@ -588,6 +603,11 @@ static ssize_t write_kmem(struct file * ssize_t written; char * kbuf; /* k-addr because vwrite() takes vmlist_lock rwlock */ +#ifdef CONFIG_GRKERNSEC_KMEM + gr_handle_kmem_write(); + return -EPERM; +#endif + if (p < (unsigned long) high_memory) { wrote = count; @@ -791,6 +811,16 @@ static loff_t memory_lseek(struct file * static int open_port(struct inode * inode, struct file * filp) { +#ifdef CONFIG_GRKERNSEC_KMEM + gr_handle_open_port(); + return -EPERM; +#endif + + return capable(CAP_SYS_RAWIO) ? 0 : -EPERM; +} + +static int open_mem(struct inode * inode, struct file * filp) +{ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM; } @@ -798,7 +828,6 @@ static int open_port(struct inode * inod #define full_lseek null_lseek #define write_zero write_null #define read_full read_zero -#define open_mem open_port #define open_kmem open_mem #define open_oldmem open_mem @@ -938,6 +967,11 @@ static int memory_open(struct inode * in filp->f_op = &oldmem_fops; break; #endif +#ifdef CONFIG_GRKERNSEC + case 13: + filp->f_op = &grsec_fops; + break; +#endif default: unlock_kernel(); return -ENXIO; @@ -974,6 +1008,9 @@ static const struct { #ifdef CONFIG_CRASH_DUMP {12,"oldmem", S_IRUSR | S_IWUSR | S_IRGRP, &oldmem_fops}, #endif +#ifdef CONFIG_GRKERNSEC + {13,"grsec", S_IRUSR | S_IWUGO, &grsec_fops}, +#endif }; static struct class *mem_class; diff -urNp linux-2.6.28.8/drivers/char/nvram.c linux-2.6.28.8/drivers/char/nvram.c --- linux-2.6.28.8/drivers/char/nvram.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/nvram.c 2009-02-21 09:37:48.000000000 -0500 @@ -433,7 +433,10 @@ static const struct file_operations nvra static struct miscdevice nvram_dev = { NVRAM_MINOR, "nvram", - &nvram_fops + &nvram_fops, + {NULL, NULL}, + NULL, + NULL }; static int __init diff -urNp linux-2.6.28.8/drivers/char/random.c linux-2.6.28.8/drivers/char/random.c --- linux-2.6.28.8/drivers/char/random.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/random.c 2009-02-21 09:37:48.000000000 -0500 @@ -249,8 +249,13 @@ /* * Configuration information */ +#ifdef CONFIG_GRKERNSEC_RANDNET +#define INPUT_POOL_WORDS 512 +#define OUTPUT_POOL_WORDS 128 +#else #define INPUT_POOL_WORDS 128 #define OUTPUT_POOL_WORDS 32 +#endif #define SEC_XFER_SIZE 512 /* @@ -287,10 +292,17 @@ static struct poolinfo { int poolwords; int tap1, tap2, tap3, tap4, tap5; } poolinfo_table[] = { +#ifdef CONFIG_GRKERNSEC_RANDNET + /* x^512 + x^411 + x^308 + x^208 +x^104 + x + 1 -- 225 */ + { 512, 411, 308, 208, 104, 1 }, + /* x^128 + x^103 + x^76 + x^51 + x^25 + x + 1 -- 105 */ + { 128, 103, 76, 51, 25, 1 }, +#else /* x^128 + x^103 + x^76 + x^51 +x^25 + x + 1 -- 105 */ { 128, 103, 76, 51, 25, 1 }, /* x^32 + x^26 + x^20 + x^14 + x^7 + x + 1 -- 15 */ { 32, 26, 20, 14, 7, 1 }, +#endif #if 0 /* x^2048 + x^1638 + x^1231 + x^819 + x^411 + x + 1 -- 115 */ { 2048, 1638, 1231, 819, 411, 1 }, @@ -1185,7 +1197,7 @@ EXPORT_SYMBOL(generate_random_uuid); #include static int min_read_thresh = 8, min_write_thresh; -static int max_read_thresh = INPUT_POOL_WORDS * 32; +static int max_read_thresh = OUTPUT_POOL_WORDS * 32; static int max_write_thresh = INPUT_POOL_WORDS * 32; static char sysctl_bootid[16]; diff -urNp linux-2.6.28.8/drivers/char/tpm/tpm.c linux-2.6.28.8/drivers/char/tpm/tpm.c --- linux-2.6.28.8/drivers/char/tpm/tpm.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/tpm/tpm.c 2009-02-21 09:37:48.000000000 -0500 @@ -1035,7 +1035,7 @@ ssize_t tpm_write(struct file *file, con mutex_lock(&chip->buffer_mutex); - if (in_size > TPM_BUFSIZE) + if (in_size > (unsigned int)TPM_BUFSIZE) in_size = TPM_BUFSIZE; if (copy_from_user diff -urNp linux-2.6.28.8/drivers/char/vt_ioctl.c linux-2.6.28.8/drivers/char/vt_ioctl.c --- linux-2.6.28.8/drivers/char/vt_ioctl.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/char/vt_ioctl.c 2009-02-21 09:37:48.000000000 -0500 @@ -96,6 +96,12 @@ do_kdsk_ioctl(int cmd, struct kbentry __ case KDSKBENT: if (!perm) return -EPERM; + +#ifdef CONFIG_GRKERNSEC + if (!capable(CAP_SYS_TTY_CONFIG)) + return -EPERM; +#endif + if (!i && v == K_NOSUCHMAP) { /* deallocate map */ key_map = key_maps[s]; @@ -236,6 +242,13 @@ do_kdgkb_ioctl(int cmd, struct kbsentry goto reterr; } +#ifdef CONFIG_GRKERNSEC + if (!capable(CAP_SYS_TTY_CONFIG)) { + ret = -EPERM; + goto reterr; + } +#endif + q = func_table[i]; first_free = funcbufptr + (funcbufsize - funcbufleft); for (j = i+1; j < MAX_NR_FUNC && !func_table[j]; j++) diff -urNp linux-2.6.28.8/drivers/edac/edac_core.h linux-2.6.28.8/drivers/edac/edac_core.h --- linux-2.6.28.8/drivers/edac/edac_core.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/edac/edac_core.h 2009-02-21 09:37:48.000000000 -0500 @@ -85,11 +85,11 @@ extern int edac_debug_level; #else /* !CONFIG_EDAC_DEBUG */ -#define debugf0( ... ) -#define debugf1( ... ) -#define debugf2( ... ) -#define debugf3( ... ) -#define debugf4( ... ) +#define debugf0( ... ) do {} while (0) +#define debugf1( ... ) do {} while (0) +#define debugf2( ... ) do {} while (0) +#define debugf3( ... ) do {} while (0) +#define debugf4( ... ) do {} while (0) #endif /* !CONFIG_EDAC_DEBUG */ diff -urNp linux-2.6.28.8/drivers/firmware/dmi_scan.c linux-2.6.28.8/drivers/firmware/dmi_scan.c --- linux-2.6.28.8/drivers/firmware/dmi_scan.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/firmware/dmi_scan.c 2009-02-21 09:37:48.000000000 -0500 @@ -389,11 +389,6 @@ void __init dmi_scan_machine(void) } } else { - /* - * no iounmap() for that ioremap(); it would be a no-op, but - * it's so early in setup that sucker gets confused into doing - * what it shouldn't if we actually call it. - */ p = dmi_ioremap(0xF0000, 0x10000); if (p == NULL) goto error; diff -urNp linux-2.6.28.8/drivers/hwmon/fscpos.c linux-2.6.28.8/drivers/hwmon/fscpos.c --- linux-2.6.28.8/drivers/hwmon/fscpos.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/hwmon/fscpos.c 2009-02-21 09:37:48.000000000 -0500 @@ -240,7 +240,6 @@ static ssize_t set_pwm(struct i2c_client unsigned long v = simple_strtoul(buf, NULL, 10); /* Range: 0..255 */ - if (v < 0) v = 0; if (v > 255) v = 255; mutex_lock(&data->update_lock); diff -urNp linux-2.6.28.8/drivers/hwmon/k8temp.c linux-2.6.28.8/drivers/hwmon/k8temp.c --- linux-2.6.28.8/drivers/hwmon/k8temp.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/hwmon/k8temp.c 2009-02-21 09:37:48.000000000 -0500 @@ -130,7 +130,7 @@ static DEVICE_ATTR(name, S_IRUGO, show_n static struct pci_device_id k8temp_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC) }, - { 0 }, + { 0, 0, 0, 0, 0, 0, 0 }, }; MODULE_DEVICE_TABLE(pci, k8temp_ids); diff -urNp linux-2.6.28.8/drivers/hwmon/sis5595.c linux-2.6.28.8/drivers/hwmon/sis5595.c --- linux-2.6.28.8/drivers/hwmon/sis5595.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/hwmon/sis5595.c 2009-02-21 09:37:48.000000000 -0500 @@ -698,7 +698,7 @@ static struct sis5595_data *sis5595_upda static struct pci_device_id sis5595_pci_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_SI, PCI_DEVICE_ID_SI_503) }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, sis5595_pci_ids); diff -urNp linux-2.6.28.8/drivers/hwmon/via686a.c linux-2.6.28.8/drivers/hwmon/via686a.c --- linux-2.6.28.8/drivers/hwmon/via686a.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/hwmon/via686a.c 2009-02-21 09:37:48.000000000 -0500 @@ -768,7 +768,7 @@ static struct via686a_data *via686a_upda static struct pci_device_id via686a_pci_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_82C686_4) }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, via686a_pci_ids); diff -urNp linux-2.6.28.8/drivers/hwmon/vt8231.c linux-2.6.28.8/drivers/hwmon/vt8231.c --- linux-2.6.28.8/drivers/hwmon/vt8231.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/hwmon/vt8231.c 2009-02-21 09:37:48.000000000 -0500 @@ -698,7 +698,7 @@ static struct platform_driver vt8231_dri static struct pci_device_id vt8231_pci_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8231_4) }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, vt8231_pci_ids); diff -urNp linux-2.6.28.8/drivers/hwmon/w83791d.c linux-2.6.28.8/drivers/hwmon/w83791d.c --- linux-2.6.28.8/drivers/hwmon/w83791d.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/hwmon/w83791d.c 2009-02-21 09:37:48.000000000 -0500 @@ -327,8 +327,8 @@ static int w83791d_detect(struct i2c_cli struct i2c_board_info *info); static int w83791d_remove(struct i2c_client *client); -static int w83791d_read(struct i2c_client *client, u8 register); -static int w83791d_write(struct i2c_client *client, u8 register, u8 value); +static int w83791d_read(struct i2c_client *client, u8 reg); +static int w83791d_write(struct i2c_client *client, u8 reg, u8 value); static struct w83791d_data *w83791d_update_device(struct device *dev); #ifdef DEBUG diff -urNp linux-2.6.28.8/drivers/i2c/busses/i2c-i801.c linux-2.6.28.8/drivers/i2c/busses/i2c-i801.c --- linux-2.6.28.8/drivers/i2c/busses/i2c-i801.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/i2c/busses/i2c-i801.c 2009-02-21 09:37:48.000000000 -0500 @@ -578,7 +578,7 @@ static struct pci_device_id i801_ids[] = { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH10_4) }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH10_5) }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_PCH_SMBUS) }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE (pci, i801_ids); diff -urNp linux-2.6.28.8/drivers/i2c/busses/i2c-piix4.c linux-2.6.28.8/drivers/i2c/busses/i2c-piix4.c --- linux-2.6.28.8/drivers/i2c/busses/i2c-piix4.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/i2c/busses/i2c-piix4.c 2009-02-21 09:37:48.000000000 -0500 @@ -123,7 +123,7 @@ static struct dmi_system_id __devinitdat .ident = "IBM", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "IBM"), }, }, - { }, + { NULL, NULL, {DMI_MATCH(DMI_NONE, NULL)}, NULL }, }; static int __devinit piix4_setup(struct pci_dev *PIIX4_dev, @@ -424,7 +424,7 @@ static struct pci_device_id piix4_ids[] PCI_DEVICE_ID_SERVERWORKS_CSB6) }, { PCI_DEVICE(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_HT1000SB) }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE (pci, piix4_ids); diff -urNp linux-2.6.28.8/drivers/i2c/busses/i2c-sis630.c linux-2.6.28.8/drivers/i2c/busses/i2c-sis630.c --- linux-2.6.28.8/drivers/i2c/busses/i2c-sis630.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/i2c/busses/i2c-sis630.c 2009-02-21 09:37:48.000000000 -0500 @@ -472,7 +472,7 @@ static struct i2c_adapter sis630_adapter static struct pci_device_id sis630_ids[] __devinitdata = { { PCI_DEVICE(PCI_VENDOR_ID_SI, PCI_DEVICE_ID_SI_503) }, { PCI_DEVICE(PCI_VENDOR_ID_SI, PCI_DEVICE_ID_SI_LPC) }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE (pci, sis630_ids); diff -urNp linux-2.6.28.8/drivers/i2c/busses/i2c-sis96x.c linux-2.6.28.8/drivers/i2c/busses/i2c-sis96x.c --- linux-2.6.28.8/drivers/i2c/busses/i2c-sis96x.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/i2c/busses/i2c-sis96x.c 2009-02-21 09:37:48.000000000 -0500 @@ -248,7 +248,7 @@ static struct i2c_adapter sis96x_adapter static struct pci_device_id sis96x_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_SI, PCI_DEVICE_ID_SI_SMBUS) }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE (pci, sis96x_ids); diff -urNp linux-2.6.28.8/drivers/ieee1394/dv1394.c linux-2.6.28.8/drivers/ieee1394/dv1394.c --- linux-2.6.28.8/drivers/ieee1394/dv1394.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ieee1394/dv1394.c 2009-02-21 09:37:48.000000000 -0500 @@ -739,7 +739,7 @@ static void frame_prepare(struct video_c based upon DIF section and sequence */ -static void inline +static inline void frame_put_packet (struct frame *f, struct packet *p) { int section_type = p->data[0] >> 5; /* section type is in bits 5 - 7 */ @@ -2177,7 +2177,7 @@ static struct ieee1394_device_id dv1394_ .specifier_id = AVC_UNIT_SPEC_ID_ENTRY & 0xffffff, .version = AVC_SW_VERSION_ENTRY & 0xffffff }, - { } + { 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(ieee1394, dv1394_id_table); diff -urNp linux-2.6.28.8/drivers/ieee1394/eth1394.c linux-2.6.28.8/drivers/ieee1394/eth1394.c --- linux-2.6.28.8/drivers/ieee1394/eth1394.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ieee1394/eth1394.c 2009-02-21 09:37:48.000000000 -0500 @@ -451,7 +451,7 @@ static struct ieee1394_device_id eth1394 .specifier_id = ETHER1394_GASP_SPECIFIER_ID, .version = ETHER1394_GASP_VERSION, }, - {} + { 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(ieee1394, eth1394_id_table); diff -urNp linux-2.6.28.8/drivers/ieee1394/hosts.c linux-2.6.28.8/drivers/ieee1394/hosts.c --- linux-2.6.28.8/drivers/ieee1394/hosts.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ieee1394/hosts.c 2009-02-21 09:37:48.000000000 -0500 @@ -78,6 +78,7 @@ static int dummy_isoctl(struct hpsb_iso } static struct hpsb_host_driver dummy_driver = { + .name = "dummy", .transmit_packet = dummy_transmit_packet, .devctl = dummy_devctl, .isoctl = dummy_isoctl diff -urNp linux-2.6.28.8/drivers/ieee1394/ohci1394.c linux-2.6.28.8/drivers/ieee1394/ohci1394.c --- linux-2.6.28.8/drivers/ieee1394/ohci1394.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ieee1394/ohci1394.c 2009-02-21 09:37:48.000000000 -0500 @@ -147,9 +147,9 @@ printk(level "%s: " fmt "\n" , OHCI1394_ printk(level "%s: fw-host%d: " fmt "\n" , OHCI1394_DRIVER_NAME, ohci->host->id , ## args) /* Module Parameters */ -static int phys_dma = 1; +static int phys_dma; module_param(phys_dma, int, 0444); -MODULE_PARM_DESC(phys_dma, "Enable physical DMA (default = 1)."); +MODULE_PARM_DESC(phys_dma, "Enable physical DMA (default = 0)."); static void dma_trm_tasklet(unsigned long data); static void dma_trm_reset(struct dma_trm_ctx *d); @@ -3437,7 +3437,7 @@ static struct pci_device_id ohci1394_pci .subvendor = PCI_ANY_ID, .subdevice = PCI_ANY_ID, }, - { 0, }, + { 0, 0, 0, 0, 0, 0, 0 }, }; MODULE_DEVICE_TABLE(pci, ohci1394_pci_tbl); diff -urNp linux-2.6.28.8/drivers/ieee1394/raw1394.c linux-2.6.28.8/drivers/ieee1394/raw1394.c --- linux-2.6.28.8/drivers/ieee1394/raw1394.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ieee1394/raw1394.c 2009-02-21 09:37:48.000000000 -0500 @@ -2995,7 +2995,7 @@ static struct ieee1394_device_id raw1394 .match_flags = IEEE1394_MATCH_SPECIFIER_ID | IEEE1394_MATCH_VERSION, .specifier_id = CAMERA_UNIT_SPEC_ID_ENTRY & 0xffffff, .version = (CAMERA_SW_VERSION_ENTRY + 2) & 0xffffff}, - {} + { 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(ieee1394, raw1394_id_table); diff -urNp linux-2.6.28.8/drivers/ieee1394/sbp2.c linux-2.6.28.8/drivers/ieee1394/sbp2.c --- linux-2.6.28.8/drivers/ieee1394/sbp2.c 2009-02-07 16:10:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ieee1394/sbp2.c 2009-02-21 09:37:48.000000000 -0500 @@ -290,7 +290,7 @@ static struct ieee1394_device_id sbp2_id .match_flags = IEEE1394_MATCH_SPECIFIER_ID | IEEE1394_MATCH_VERSION, .specifier_id = SBP2_UNIT_SPEC_ID_ENTRY & 0xffffff, .version = SBP2_SW_VERSION_ENTRY & 0xffffff}, - {} + { 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(ieee1394, sbp2_id_table); @@ -2110,7 +2110,7 @@ MODULE_DESCRIPTION("IEEE-1394 SBP-2 prot MODULE_SUPPORTED_DEVICE(SBP2_DEVICE_NAME); MODULE_LICENSE("GPL"); -static int sbp2_module_init(void) +static int __init sbp2_module_init(void) { int ret; diff -urNp linux-2.6.28.8/drivers/ieee1394/video1394.c linux-2.6.28.8/drivers/ieee1394/video1394.c --- linux-2.6.28.8/drivers/ieee1394/video1394.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/ieee1394/video1394.c 2009-02-21 09:37:48.000000000 -0500 @@ -1310,7 +1310,7 @@ static struct ieee1394_device_id video13 .specifier_id = CAMERA_UNIT_SPEC_ID_ENTRY & 0xffffff, .version = (CAMERA_SW_VERSION_ENTRY + 2) & 0xffffff }, - { } + { 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(ieee1394, video1394_id_table); diff -urNp linux-2.6.28.8/drivers/input/keyboard/atkbd.c linux-2.6.28.8/drivers/input/keyboard/atkbd.c --- linux-2.6.28.8/drivers/input/keyboard/atkbd.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/input/keyboard/atkbd.c 2009-02-21 09:37:48.000000000 -0500 @@ -1164,7 +1164,7 @@ static struct serio_device_id atkbd_seri .id = SERIO_ANY, .extra = SERIO_ANY, }, - { 0 } + { 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(serio, atkbd_serio_ids); diff -urNp linux-2.6.28.8/drivers/input/mouse/lifebook.c linux-2.6.28.8/drivers/input/mouse/lifebook.c --- linux-2.6.28.8/drivers/input/mouse/lifebook.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/input/mouse/lifebook.c 2009-02-21 09:37:48.000000000 -0500 @@ -110,7 +110,7 @@ static const struct dmi_system_id lifebo DMI_MATCH(DMI_PRODUCT_NAME, "LifeBook B142"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL} }; static psmouse_ret_t lifebook_process_byte(struct psmouse *psmouse) diff -urNp linux-2.6.28.8/drivers/input/mouse/psmouse-base.c linux-2.6.28.8/drivers/input/mouse/psmouse-base.c --- linux-2.6.28.8/drivers/input/mouse/psmouse-base.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/input/mouse/psmouse-base.c 2009-02-21 09:37:48.000000000 -0500 @@ -1378,7 +1378,7 @@ static struct serio_device_id psmouse_se .id = SERIO_ANY, .extra = SERIO_ANY, }, - { 0 } + { 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(serio, psmouse_serio_ids); diff -urNp linux-2.6.28.8/drivers/input/mouse/synaptics.c linux-2.6.28.8/drivers/input/mouse/synaptics.c --- linux-2.6.28.8/drivers/input/mouse/synaptics.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/input/mouse/synaptics.c 2009-02-21 09:37:48.000000000 -0500 @@ -417,7 +417,7 @@ static void synaptics_process_packet(str break; case 2: if (SYN_MODEL_PEN(priv->model_id)) - ; /* Nothing, treat a pen as a single finger */ + break; /* Nothing, treat a pen as a single finger */ break; case 4 ... 15: if (SYN_CAP_PALMDETECT(priv->capabilities)) @@ -624,7 +624,7 @@ static const struct dmi_system_id toshib DMI_MATCH(DMI_PRODUCT_NAME, "PORTEGE M300"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; #endif diff -urNp linux-2.6.28.8/drivers/input/mousedev.c linux-2.6.28.8/drivers/input/mousedev.c --- linux-2.6.28.8/drivers/input/mousedev.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/input/mousedev.c 2009-02-21 09:37:48.000000000 -0500 @@ -1063,7 +1063,7 @@ static struct input_handler mousedev_han #ifdef CONFIG_INPUT_MOUSEDEV_PSAUX static struct miscdevice psaux_mouse = { - PSMOUSE_MINOR, "psaux", &mousedev_fops + PSMOUSE_MINOR, "psaux", &mousedev_fops, {NULL, NULL}, NULL, NULL }; static int psaux_registered; #endif diff -urNp linux-2.6.28.8/drivers/input/serio/i8042-x86ia64io.h linux-2.6.28.8/drivers/input/serio/i8042-x86ia64io.h --- linux-2.6.28.8/drivers/input/serio/i8042-x86ia64io.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/input/serio/i8042-x86ia64io.h 2009-02-21 09:37:48.000000000 -0500 @@ -143,7 +143,7 @@ static struct dmi_system_id __initdata i DMI_MATCH(DMI_PRODUCT_VERSION, "M606"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; /* @@ -351,7 +351,7 @@ static struct dmi_system_id __initdata i DMI_MATCH(DMI_PRODUCT_NAME, "HEL80I"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; #ifdef CONFIG_PNP @@ -363,7 +363,7 @@ static struct dmi_system_id __initdata i DMI_MATCH(DMI_BOARD_VENDOR, "Intel Corporation"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; #endif @@ -430,7 +430,7 @@ static struct dmi_system_id __initdata i DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate 4280"), }, }, - { } + { NULL, NULL, {DMI_MATCH(DMI_NONE, {0})}, NULL } }; #endif /* CONFIG_X86 */ diff -urNp linux-2.6.28.8/drivers/input/serio/serio_raw.c linux-2.6.28.8/drivers/input/serio/serio_raw.c --- linux-2.6.28.8/drivers/input/serio/serio_raw.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/input/serio/serio_raw.c 2009-02-21 09:37:48.000000000 -0500 @@ -378,7 +378,7 @@ static struct serio_device_id serio_raw_ .id = SERIO_ANY, .extra = SERIO_ANY, }, - { 0 } + { 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(serio, serio_raw_serio_ids); diff -urNp linux-2.6.28.8/drivers/lguest/core.c linux-2.6.28.8/drivers/lguest/core.c --- linux-2.6.28.8/drivers/lguest/core.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/lguest/core.c 2009-03-07 10:28:24.000000000 -0500 @@ -80,9 +80,17 @@ static __init int map_switcher(void) * (SWITCHER_ADDR). We might not get it in theory, but in practice * it's worked so far. The end address needs +1 because __get_vm_area * allocates an extra guard page, so we need space for that. */ + +#if defined(CONFIG_MODULES) && defined(CONFIG_X86_32) && defined(CONFIG_PAX_KERNEXEC) + switcher_vma = __get_vm_area(TOTAL_SWITCHER_PAGES * PAGE_SIZE, + VM_ALLOC | VM_KERNEXEC, SWITCHER_ADDR, SWITCHER_ADDR + + (TOTAL_SWITCHER_PAGES+1) * PAGE_SIZE); +#else switcher_vma = __get_vm_area(TOTAL_SWITCHER_PAGES * PAGE_SIZE, VM_ALLOC, SWITCHER_ADDR, SWITCHER_ADDR + (TOTAL_SWITCHER_PAGES+1) * PAGE_SIZE); +#endif + if (!switcher_vma) { err = -ENOMEM; printk("lguest: could not map switcher pages high\n"); diff -urNp linux-2.6.28.8/drivers/md/bitmap.c linux-2.6.28.8/drivers/md/bitmap.c --- linux-2.6.28.8/drivers/md/bitmap.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/md/bitmap.c 2009-02-21 09:37:48.000000000 -0500 @@ -57,7 +57,7 @@ # if DEBUG > 0 # define PRINTK(x...) printk(KERN_DEBUG x) # else -# define PRINTK(x...) +# define PRINTK(x...) do {} while (0) # endif #endif diff -urNp linux-2.6.28.8/drivers/mtd/devices/doc2000.c linux-2.6.28.8/drivers/mtd/devices/doc2000.c --- linux-2.6.28.8/drivers/mtd/devices/doc2000.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/mtd/devices/doc2000.c 2009-02-21 09:37:48.000000000 -0500 @@ -777,7 +777,7 @@ static int doc_write(struct mtd_info *mt /* The ECC will not be calculated correctly if less than 512 is written */ /* DBB- - if (len != 0x200 && eccbuf) + if (len != 0x200) printk(KERN_WARNING "ECC needs a full sector write (adr: %lx size %lx)\n", (long) to, (long) len); diff -urNp linux-2.6.28.8/drivers/mtd/devices/doc2001.c linux-2.6.28.8/drivers/mtd/devices/doc2001.c --- linux-2.6.28.8/drivers/mtd/devices/doc2001.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/mtd/devices/doc2001.c 2009-02-21 09:37:48.000000000 -0500 @@ -396,6 +396,8 @@ static int doc_read (struct mtd_info *mt /* Don't allow read past end of device */ if (from >= this->totlen) return -EINVAL; + if (!len) + return -EINVAL; /* Don't allow a single read to cross a 512-byte block boundary */ if (from + len > ((from | 0x1ff) + 1)) diff -urNp linux-2.6.28.8/drivers/mtd/devices/slram.c linux-2.6.28.8/drivers/mtd/devices/slram.c --- linux-2.6.28.8/drivers/mtd/devices/slram.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/mtd/devices/slram.c 2009-02-21 09:37:48.000000000 -0500 @@ -273,7 +273,7 @@ static int parse_cmdline(char *devname, } T("slram: devname=%s, devstart=0x%lx, devlength=0x%lx\n", devname, devstart, devlength); - if ((devstart < 0) || (devlength < 0) || (devlength % SLRAM_BLK_SZ != 0)) { + if (devlength % SLRAM_BLK_SZ != 0) { E("slram: Illegal start / length parameter.\n"); return(-EINVAL); } diff -urNp linux-2.6.28.8/drivers/mtd/ubi/build.c linux-2.6.28.8/drivers/mtd/ubi/build.c --- linux-2.6.28.8/drivers/mtd/ubi/build.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/mtd/ubi/build.c 2009-02-21 09:37:48.000000000 -0500 @@ -1104,7 +1104,7 @@ static int __init bytes_str_to_int(const unsigned long result; result = simple_strtoul(str, &endp, 0); - if (str == endp || result < 0) { + if (str == endp) { printk(KERN_ERR "UBI error: incorrect bytes count: \"%s\"\n", str); return -EINVAL; diff -urNp linux-2.6.28.8/drivers/net/eepro100.c linux-2.6.28.8/drivers/net/eepro100.c --- linux-2.6.28.8/drivers/net/eepro100.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/net/eepro100.c 2009-02-21 09:37:48.000000000 -0500 @@ -47,7 +47,7 @@ static int rxdmacount /* = 0 */; # define rx_align(skb) skb_reserve((skb), 2) # define RxFD_ALIGNMENT __attribute__ ((aligned (2), packed)) #else -# define rx_align(skb) +# define rx_align(skb) do {} while (0) # define RxFD_ALIGNMENT #endif @@ -2334,33 +2334,33 @@ static void __devexit eepro100_remove_on } static struct pci_device_id eepro100_pci_tbl[] = { - { PCI_VENDOR_ID_INTEL, 0x1229, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1209, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1029, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1030, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1031, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1032, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1033, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1034, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1035, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1036, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1037, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1038, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1039, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x103A, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x103B, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x103C, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x103D, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x103E, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1050, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1059, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x1227, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x2449, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x2459, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x245D, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x5200, PCI_ANY_ID, PCI_ANY_ID, }, - { PCI_VENDOR_ID_INTEL, 0x5201, PCI_ANY_ID, PCI_ANY_ID, }, - { 0,} + { PCI_VENDOR_ID_INTEL, 0x1229, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1209, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1029, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1030, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1031, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1032, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1033, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1034, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1035, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1036, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1037, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1038, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1039, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x103A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x103B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x103C, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x103D, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x103E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1050, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1059, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x1227, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x2449, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x2459, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x245D, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x5200, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { PCI_VENDOR_ID_INTEL, 0x5201, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 }, + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, eepro100_pci_tbl); diff -urNp linux-2.6.28.8/drivers/net/irda/vlsi_ir.c linux-2.6.28.8/drivers/net/irda/vlsi_ir.c --- linux-2.6.28.8/drivers/net/irda/vlsi_ir.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/net/irda/vlsi_ir.c 2009-02-21 09:37:48.000000000 -0500 @@ -906,13 +906,12 @@ static int vlsi_hard_start_xmit(struct s /* no race - tx-ring already empty */ vlsi_set_baud(idev, iobase); netif_wake_queue(ndev); - } - else - ; + } else { /* keep the speed change pending like it would * for any len>0 packet. tx completion interrupt * will apply it when the tx ring becomes empty. */ + } spin_unlock_irqrestore(&idev->lock, flags); dev_kfree_skb_any(skb); return 0; diff -urNp linux-2.6.28.8/drivers/net/pcnet32.c linux-2.6.28.8/drivers/net/pcnet32.c --- linux-2.6.28.8/drivers/net/pcnet32.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/net/pcnet32.c 2009-02-21 09:37:48.000000000 -0500 @@ -78,7 +78,7 @@ static int cards_found; /* * VLB I/O addresses */ -static unsigned int pcnet32_portlist[] __initdata = +static unsigned int pcnet32_portlist[] __devinitdata = { 0x300, 0x320, 0x340, 0x360, 0 }; static int pcnet32_debug = 0; diff -urNp linux-2.6.28.8/drivers/net/tg3.h linux-2.6.28.8/drivers/net/tg3.h --- linux-2.6.28.8/drivers/net/tg3.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/net/tg3.h 2009-02-21 09:37:48.000000000 -0500 @@ -102,6 +102,7 @@ #define CHIPREV_ID_5750_A0 0x4000 #define CHIPREV_ID_5750_A1 0x4001 #define CHIPREV_ID_5750_A3 0x4003 +#define CHIPREV_ID_5750_C1 0x4201 #define CHIPREV_ID_5750_C2 0x4202 #define CHIPREV_ID_5752_A0_HW 0x5000 #define CHIPREV_ID_5752_A0 0x6000 diff -urNp linux-2.6.28.8/drivers/net/wireless/iwlwifi/iwl3945-base.c linux-2.6.28.8/drivers/net/wireless/iwlwifi/iwl3945-base.c --- linux-2.6.28.8/drivers/net/wireless/iwlwifi/iwl3945-base.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/net/wireless/iwlwifi/iwl3945-base.c 2009-02-21 09:37:48.000000000 -0500 @@ -785,7 +785,7 @@ static int iwl3945_send_cmd_sync(struct IWL_ERROR("Error: Response NULL in '%s'\n", get_cmd_string(cmd->id)); ret = -EIO; - goto out; + goto cancel; } ret = 0; diff -urNp linux-2.6.28.8/drivers/pci/hotplug/cpqphp_nvram.c linux-2.6.28.8/drivers/pci/hotplug/cpqphp_nvram.c --- linux-2.6.28.8/drivers/pci/hotplug/cpqphp_nvram.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pci/hotplug/cpqphp_nvram.c 2009-02-21 09:37:48.000000000 -0500 @@ -425,9 +425,13 @@ static u32 store_HRT (void __iomem *rom_ void compaq_nvram_init (void __iomem *rom_start) { + +#ifndef CONFIG_PAX_KERNEXEC if (rom_start) { compaq_int15_entry_point = (rom_start + ROM_INT15_PHY_ADDR - ROM_PHY_ADDR); } +#endif + dbg("int15 entry = %p\n", compaq_int15_entry_point); /* initialize our int15 lock */ diff -urNp linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv.c linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv.c --- linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv.c 2009-02-21 09:37:48.000000000 -0500 @@ -59,7 +59,7 @@ static struct pcie_port_service_id aer_i .port_type = PCIE_RC_PORT, .service_type = PCIE_PORT_SERVICE_AER, }, - { /* end: all zeroes */ } + { 0, 0, 0, 0, 0, 0, 0, 0, 0 } }; static struct pci_error_handlers aer_error_handlers = { diff -urNp linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv_core.c linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv_core.c --- linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv_core.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/drivers/pci/pcie/aer/aerdrv_core.c 2009-03-07 10:29:51.000000000 -0500 @@ -667,7 +667,7 @@ static void aer_isr_one_error(struct pci struct aer_err_source *e_src) { struct device *s_device; - struct aer_err_info e_info = {0, 0, 0,}; + struct aer_err_info e_info = {0, 0, 0, {0, 0, 0, 0}}; int i; u16 id; diff -urNp linux-2.6.28.8/drivers/pci/pcie/portdrv_pci.c linux-2.6.28.8/drivers/pci/pcie/portdrv_pci.c --- linux-2.6.28.8/drivers/pci/pcie/portdrv_pci.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/drivers/pci/pcie/portdrv_pci.c 2009-03-07 10:29:51.000000000 -0500 @@ -263,7 +263,7 @@ static void pcie_portdrv_err_resume(stru static const struct pci_device_id port_pci_ids[] = { { /* handle any PCI-Express port */ PCI_DEVICE_CLASS(((PCI_CLASS_BRIDGE_PCI << 8) | 0x00), ~0), - }, { /* end: all zeroes */ } + }, { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, port_pci_ids); diff -urNp linux-2.6.28.8/drivers/pci/proc.c linux-2.6.28.8/drivers/pci/proc.c --- linux-2.6.28.8/drivers/pci/proc.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pci/proc.c 2009-02-21 09:37:48.000000000 -0500 @@ -470,7 +470,16 @@ static const struct file_operations proc static int __init pci_proc_init(void) { struct pci_dev *dev = NULL; + +#ifdef CONFIG_GRKERNSEC_PROC_ADD +#ifdef CONFIG_GRKERNSEC_PROC_USER + proc_bus_pci_dir = proc_mkdir_mode("bus/pci", S_IRUSR | S_IXUSR, NULL); +#elif defined(CONFIG_GRKERNSEC_PROC_USERGROUP) + proc_bus_pci_dir = proc_mkdir_mode("bus/pci", S_IRUSR | S_IXUSR | S_IRGRP | S_IXGRP, NULL); +#endif +#else proc_bus_pci_dir = proc_mkdir("bus/pci", NULL); +#endif proc_create("devices", 0, proc_bus_pci_dir, &proc_bus_pci_dev_operations); proc_initialized = 1; diff -urNp linux-2.6.28.8/drivers/pcmcia/ti113x.h linux-2.6.28.8/drivers/pcmcia/ti113x.h --- linux-2.6.28.8/drivers/pcmcia/ti113x.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pcmcia/ti113x.h 2009-02-21 09:37:48.000000000 -0500 @@ -903,7 +903,7 @@ static struct pci_device_id ene_tune_tbl DEVID(PCI_VENDOR_ID_MOTOROLA, 0x3410, 0xECC0, PCI_ANY_ID, ENE_TEST_C9_TLTENABLE | ENE_TEST_C9_PFENABLE, ENE_TEST_C9_TLTENABLE), - {} + { 0, 0, 0, 0, 0, 0, 0 } }; static void ene_tune_bridge(struct pcmcia_socket *sock, struct pci_bus *bus) diff -urNp linux-2.6.28.8/drivers/pcmcia/yenta_socket.c linux-2.6.28.8/drivers/pcmcia/yenta_socket.c --- linux-2.6.28.8/drivers/pcmcia/yenta_socket.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pcmcia/yenta_socket.c 2009-02-21 09:37:48.000000000 -0500 @@ -1366,7 +1366,7 @@ static struct pci_device_id yenta_table /* match any cardbus bridge */ CB_ID(PCI_ANY_ID, PCI_ANY_ID, DEFAULT), - { /* all zeroes */ } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, yenta_table); diff -urNp linux-2.6.28.8/drivers/pnp/pnpbios/bioscalls.c linux-2.6.28.8/drivers/pnp/pnpbios/bioscalls.c --- linux-2.6.28.8/drivers/pnp/pnpbios/bioscalls.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pnp/pnpbios/bioscalls.c 2009-02-21 09:37:48.000000000 -0500 @@ -60,7 +60,7 @@ set_base(gdt[(selname) >> 3], (u32)(addr set_limit(gdt[(selname) >> 3], size); \ } while(0) -static struct desc_struct bad_bios_desc; +static struct desc_struct bad_bios_desc __read_only; /* * At some point we want to use this stack frame pointer to unwind @@ -87,6 +87,10 @@ static inline u16 call_pnp_bios(u16 func struct desc_struct save_desc_40; int cpu; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + /* * PnP BIOSes are generally not terribly re-entrant. * Also, don't rely on them to save everything correctly. @@ -96,8 +100,17 @@ static inline u16 call_pnp_bios(u16 func cpu = get_cpu(); save_desc_40 = get_cpu_gdt_table(cpu)[0x40 / 8]; + +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + get_cpu_gdt_table(cpu)[0x40 / 8] = bad_bios_desc; +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + /* On some boxes IRQ's during PnP BIOS calls are deadly. */ spin_lock_irqsave(&pnp_bios_lock, flags); @@ -134,7 +147,16 @@ static inline u16 call_pnp_bios(u16 func :"memory"); spin_unlock_irqrestore(&pnp_bios_lock, flags); +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + get_cpu_gdt_table(cpu)[0x40 / 8] = save_desc_40; + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + put_cpu(); /* If we get here and this is set then the PnP BIOS faulted on us. */ @@ -468,16 +490,24 @@ int pnp_bios_read_escd(char *data, u32 n return status; } -void pnpbios_calls_init(union pnp_bios_install_struct *header) +void __init pnpbios_calls_init(union pnp_bios_install_struct *header) { int i; +#ifdef CONFIG_PAX_KERNEXEC + unsigned long cr0; +#endif + spin_lock_init(&pnp_bios_lock); pnp_bios_callpoint.offset = header->fields.pm16offset; pnp_bios_callpoint.segment = PNP_CS16; +#ifdef CONFIG_PAX_KERNEXEC + pax_open_kernel(cr0); +#endif + bad_bios_desc.a = 0; - bad_bios_desc.b = 0x00409200; + bad_bios_desc.b = 0x00409300; set_base(bad_bios_desc, __va((unsigned long)0x40 << 4)); _set_limit((char *)&bad_bios_desc, 4095 - (0x40 << 4)); @@ -491,4 +521,9 @@ void pnpbios_calls_init(union pnp_bios_i set_base(gdt[GDT_ENTRY_PNPBIOS_DS], __va(header->fields.pm16dseg)); } + +#ifdef CONFIG_PAX_KERNEXEC + pax_close_kernel(cr0); +#endif + } diff -urNp linux-2.6.28.8/drivers/pnp/quirks.c linux-2.6.28.8/drivers/pnp/quirks.c --- linux-2.6.28.8/drivers/pnp/quirks.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pnp/quirks.c 2009-02-21 09:37:48.000000000 -0500 @@ -327,7 +327,7 @@ static struct pnp_fixup pnp_fixups[] = { /* PnP resources that might overlap PCI BARs */ {"PNP0c01", quirk_system_pci_resources}, {"PNP0c02", quirk_system_pci_resources}, - {""} + {"", NULL} }; void pnp_fixup_device(struct pnp_dev *dev) diff -urNp linux-2.6.28.8/drivers/pnp/resource.c linux-2.6.28.8/drivers/pnp/resource.c --- linux-2.6.28.8/drivers/pnp/resource.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/pnp/resource.c 2009-02-21 09:37:48.000000000 -0500 @@ -355,7 +355,7 @@ int pnp_check_irq(struct pnp_dev *dev, s return 1; /* check if the resource is valid */ - if (*irq < 0 || *irq > 15) + if (*irq > 15) return 0; /* check if the resource is reserved */ @@ -419,7 +419,7 @@ int pnp_check_dma(struct pnp_dev *dev, s return 1; /* check if the resource is valid */ - if (*dma < 0 || *dma == 4 || *dma > 7) + if (*dma == 4 || *dma > 7) return 0; /* check if the resource is reserved */ diff -urNp linux-2.6.28.8/drivers/scsi/scsi_logging.h linux-2.6.28.8/drivers/scsi/scsi_logging.h --- linux-2.6.28.8/drivers/scsi/scsi_logging.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/scsi/scsi_logging.h 2009-02-21 09:37:48.000000000 -0500 @@ -51,7 +51,7 @@ do { \ } while (0); \ } while (0) #else -#define SCSI_CHECK_LOGGING(SHIFT, BITS, LEVEL, CMD) +#define SCSI_CHECK_LOGGING(SHIFT, BITS, LEVEL, CMD) do {} while (0) #endif /* CONFIG_SCSI_LOGGING */ /* diff -urNp linux-2.6.28.8/drivers/serial/8250_pci.c linux-2.6.28.8/drivers/serial/8250_pci.c --- linux-2.6.28.8/drivers/serial/8250_pci.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/drivers/serial/8250_pci.c 2009-03-07 10:29:51.000000000 -0500 @@ -3138,7 +3138,7 @@ static struct pci_device_id serial_pci_t PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_COMMUNICATION_MULTISERIAL << 8, 0xffff00, pbn_default }, - { 0, } + { 0, 0, 0, 0, 0, 0, 0 } }; static struct pci_driver serial_pci_driver = { diff -urNp linux-2.6.28.8/drivers/usb/class/cdc-acm.c linux-2.6.28.8/drivers/usb/class/cdc-acm.c --- linux-2.6.28.8/drivers/usb/class/cdc-acm.c 2009-03-07 10:24:49.000000000 -0500 +++ linux-2.6.28.8/drivers/usb/class/cdc-acm.c 2009-03-07 10:29:51.000000000 -0500 @@ -1388,7 +1388,7 @@ static struct usb_device_id acm_ids[] = USB_CDC_ACM_PROTO_AT_CDMA) }, /* NOTE: COMM/ACM/0xff is likely MSFT RNDIS ... NOT a modem!! */ - { } + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE (usb, acm_ids); diff -urNp linux-2.6.28.8/drivers/usb/class/usblp.c linux-2.6.28.8/drivers/usb/class/usblp.c --- linux-2.6.28.8/drivers/usb/class/usblp.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/usb/class/usblp.c 2009-02-21 09:37:48.000000000 -0500 @@ -227,7 +227,7 @@ static const struct quirk_printer_struct { 0x0409, 0xf1be, USBLP_QUIRK_BIDIR }, /* NEC Picty800 (HP OEM) */ { 0x0482, 0x0010, USBLP_QUIRK_BIDIR }, /* Kyocera Mita FS 820, by zut */ { 0x04b8, 0x0202, USBLP_QUIRK_BAD_CLASS }, /* Seiko Epson Receipt Printer M129C */ - { 0, 0 } + { 0, 0, 0 } }; static int usblp_wwait(struct usblp *usblp, int nonblock); @@ -1402,7 +1402,7 @@ static struct usb_device_id usblp_ids [] { USB_INTERFACE_INFO(7, 1, 2) }, { USB_INTERFACE_INFO(7, 1, 3) }, { USB_DEVICE(0x04b8, 0x0202) }, /* Seiko Epson Receipt Printer M129C */ - { } /* Terminating entry */ + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } /* Terminating entry */ }; MODULE_DEVICE_TABLE (usb, usblp_ids); diff -urNp linux-2.6.28.8/drivers/usb/core/hub.c linux-2.6.28.8/drivers/usb/core/hub.c --- linux-2.6.28.8/drivers/usb/core/hub.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/usb/core/hub.c 2009-02-21 09:37:48.000000000 -0500 @@ -3194,7 +3194,7 @@ static struct usb_device_id hub_id_table .bDeviceClass = USB_CLASS_HUB}, { .match_flags = USB_DEVICE_ID_MATCH_INT_CLASS, .bInterfaceClass = USB_CLASS_HUB}, - { } /* Terminating entry */ + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } /* Terminating entry */ }; MODULE_DEVICE_TABLE (usb, hub_id_table); diff -urNp linux-2.6.28.8/drivers/usb/host/ehci-pci.c linux-2.6.28.8/drivers/usb/host/ehci-pci.c --- linux-2.6.28.8/drivers/usb/host/ehci-pci.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/usb/host/ehci-pci.c 2009-02-21 09:37:48.000000000 -0500 @@ -414,7 +414,7 @@ static const struct pci_device_id pci_id PCI_DEVICE_CLASS(PCI_CLASS_SERIAL_USB_EHCI, ~0), .driver_data = (unsigned long) &ehci_pci_hc_driver, }, - { /* end: all zeroes */ } + { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, pci_ids); diff -urNp linux-2.6.28.8/drivers/usb/host/uhci-hcd.c linux-2.6.28.8/drivers/usb/host/uhci-hcd.c --- linux-2.6.28.8/drivers/usb/host/uhci-hcd.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/usb/host/uhci-hcd.c 2009-02-21 09:37:48.000000000 -0500 @@ -927,7 +927,7 @@ static const struct pci_device_id uhci_p /* handle any USB UHCI controller */ PCI_DEVICE_CLASS(PCI_CLASS_SERIAL_USB_UHCI, ~0), .driver_data = (unsigned long) &uhci_driver, - }, { /* end: all zeroes */ } + }, { 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE(pci, uhci_pci_ids); diff -urNp linux-2.6.28.8/drivers/usb/storage/debug.h linux-2.6.28.8/drivers/usb/storage/debug.h --- linux-2.6.28.8/drivers/usb/storage/debug.h 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/usb/storage/debug.h 2009-02-21 09:37:48.000000000 -0500 @@ -54,9 +54,9 @@ void usb_stor_show_sense( unsigned char #define US_DEBUGPX(x...) printk( x ) #define US_DEBUG(x) x #else -#define US_DEBUGP(x...) -#define US_DEBUGPX(x...) -#define US_DEBUG(x) +#define US_DEBUGP(x...) do {} while (0) +#define US_DEBUGPX(x...) do {} while (0) +#define US_DEBUG(x) do {} while (0) #endif #endif diff -urNp linux-2.6.28.8/drivers/usb/storage/usb.c linux-2.6.28.8/drivers/usb/storage/usb.c --- linux-2.6.28.8/drivers/usb/storage/usb.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/usb/storage/usb.c 2009-02-21 09:37:48.000000000 -0500 @@ -139,7 +139,7 @@ static struct usb_device_id storage_usb_ #undef COMPLIANT_DEV #undef USUAL_DEV /* Terminating entry */ - { } + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } }; MODULE_DEVICE_TABLE (usb, storage_usb_ids); @@ -182,7 +182,7 @@ static struct us_unusual_dev us_unusual_ # undef USUAL_DEV /* Terminating entry */ - { NULL } + { NULL, NULL, 0, 0, NULL } }; diff -urNp linux-2.6.28.8/drivers/uwb/wlp/messages.c linux-2.6.28.8/drivers/uwb/wlp/messages.c --- linux-2.6.28.8/drivers/uwb/wlp/messages.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/uwb/wlp/messages.c 2009-02-21 09:37:48.000000000 -0500 @@ -988,7 +988,7 @@ int wlp_parse_f0(struct wlp *wlp, struct size_t len = skb->len; size_t used; ssize_t result; - struct wlp_nonce enonce, rnonce; + struct wlp_nonce enonce = {{0}}, rnonce = {{0}}; enum wlp_assc_error assc_err; char enonce_buf[WLP_WSS_NONCE_STRSIZE]; char rnonce_buf[WLP_WSS_NONCE_STRSIZE]; diff -urNp linux-2.6.28.8/drivers/video/fbcmap.c linux-2.6.28.8/drivers/video/fbcmap.c --- linux-2.6.28.8/drivers/video/fbcmap.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/video/fbcmap.c 2009-02-21 09:37:48.000000000 -0500 @@ -250,8 +250,7 @@ int fb_set_user_cmap(struct fb_cmap_user int rc, size = cmap->len * sizeof(u16); struct fb_cmap umap; - if (cmap->start < 0 || (!info->fbops->fb_setcolreg && - !info->fbops->fb_setcmap)) + if (!info->fbops->fb_setcolreg && !info->fbops->fb_setcmap) return -EINVAL; memset(&umap, 0, sizeof(struct fb_cmap)); diff -urNp linux-2.6.28.8/drivers/video/fbmem.c linux-2.6.28.8/drivers/video/fbmem.c --- linux-2.6.28.8/drivers/video/fbmem.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/video/fbmem.c 2009-02-21 09:37:48.000000000 -0500 @@ -393,7 +393,7 @@ static void fb_do_show_logo(struct fb_in image->dx += image->width + 8; } } else if (rotate == FB_ROTATE_UD) { - for (x = 0; x < num && image->dx >= 0; x++) { + for (x = 0; x < num && (__s32)image->dx >= 0; x++) { info->fbops->fb_imageblit(info, image); image->dx -= image->width + 8; } @@ -405,7 +405,7 @@ static void fb_do_show_logo(struct fb_in image->dy += image->height + 8; } } else if (rotate == FB_ROTATE_CCW) { - for (x = 0; x < num && image->dy >= 0; x++) { + for (x = 0; x < num && (__s32)image->dy >= 0; x++) { info->fbops->fb_imageblit(info, image); image->dy -= image->height + 8; } @@ -1090,7 +1090,7 @@ static long do_fb_ioctl(struct fb_info * ret = -EINVAL; break; } - if (con2fb.framebuffer < 0 || con2fb.framebuffer >= FB_MAX) { + if (con2fb.framebuffer >= FB_MAX) { ret = -EINVAL; break; } diff -urNp linux-2.6.28.8/drivers/video/fbmon.c linux-2.6.28.8/drivers/video/fbmon.c --- linux-2.6.28.8/drivers/video/fbmon.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/video/fbmon.c 2009-02-21 09:37:48.000000000 -0500 @@ -45,7 +45,7 @@ #ifdef DEBUG #define DPRINTK(fmt, args...) printk(fmt,## args) #else -#define DPRINTK(fmt, args...) +#define DPRINTK(fmt, args...) do {} while (0) #endif #define FBMON_FIX_HEADER 1 diff -urNp linux-2.6.28.8/drivers/video/i810/i810_accel.c linux-2.6.28.8/drivers/video/i810/i810_accel.c --- linux-2.6.28.8/drivers/video/i810/i810_accel.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/video/i810/i810_accel.c 2009-02-21 09:37:48.000000000 -0500 @@ -73,6 +73,7 @@ static inline int wait_for_space(struct } } printk("ringbuffer lockup!!!\n"); + printk("head:%u tail:%u iring.size:%u space:%u\n", head, tail, par->iring.size, space); i810_report_error(mmio); par->dev_flags |= LOCKUP; info->pixmap.scan_align = 1; diff -urNp linux-2.6.28.8/drivers/video/i810/i810_main.c linux-2.6.28.8/drivers/video/i810/i810_main.c --- linux-2.6.28.8/drivers/video/i810/i810_main.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/video/i810/i810_main.c 2009-03-07 14:10:58.000000000 -0500 @@ -120,7 +120,7 @@ static struct pci_device_id i810fb_pci_t PCI_ANY_ID, PCI_ANY_ID, 0, 0, 4 }, { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82815_CGC, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 5 }, - { 0 }, + { 0, 0, 0, 0, 0, 0, 0 }, }; static struct pci_driver i810fb_driver = { diff -urNp linux-2.6.28.8/drivers/video/modedb.c linux-2.6.28.8/drivers/video/modedb.c --- linux-2.6.28.8/drivers/video/modedb.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/video/modedb.c 2009-02-21 09:37:48.000000000 -0500 @@ -38,232 +38,232 @@ static const struct fb_videomode modedb[ { /* 640x400 @ 70 Hz, 31.5 kHz hsync */ NULL, 70, 640, 400, 39721, 40, 24, 39, 9, 96, 2, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 640x480 @ 60 Hz, 31.5 kHz hsync */ NULL, 60, 640, 480, 39721, 40, 24, 32, 11, 96, 2, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 800x600 @ 56 Hz, 35.15 kHz hsync */ NULL, 56, 800, 600, 27777, 128, 24, 22, 1, 72, 2, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1024x768 @ 87 Hz interlaced, 35.5 kHz hsync */ NULL, 87, 1024, 768, 22271, 56, 24, 33, 8, 160, 8, - 0, FB_VMODE_INTERLACED + 0, FB_VMODE_INTERLACED, FB_MODE_IS_UNKNOWN }, { /* 640x400 @ 85 Hz, 37.86 kHz hsync */ NULL, 85, 640, 400, 31746, 96, 32, 41, 1, 64, 3, - FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 640x480 @ 72 Hz, 36.5 kHz hsync */ NULL, 72, 640, 480, 31746, 144, 40, 30, 8, 40, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 640x480 @ 75 Hz, 37.50 kHz hsync */ NULL, 75, 640, 480, 31746, 120, 16, 16, 1, 64, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 800x600 @ 60 Hz, 37.8 kHz hsync */ NULL, 60, 800, 600, 25000, 88, 40, 23, 1, 128, 4, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 640x480 @ 85 Hz, 43.27 kHz hsync */ NULL, 85, 640, 480, 27777, 80, 56, 25, 1, 56, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1152x864 @ 89 Hz interlaced, 44 kHz hsync */ NULL, 89, 1152, 864, 15384, 96, 16, 110, 1, 216, 10, - 0, FB_VMODE_INTERLACED + 0, FB_VMODE_INTERLACED, FB_MODE_IS_UNKNOWN }, { /* 800x600 @ 72 Hz, 48.0 kHz hsync */ NULL, 72, 800, 600, 20000, 64, 56, 23, 37, 120, 6, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1024x768 @ 60 Hz, 48.4 kHz hsync */ NULL, 60, 1024, 768, 15384, 168, 8, 29, 3, 144, 6, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 640x480 @ 100 Hz, 53.01 kHz hsync */ NULL, 100, 640, 480, 21834, 96, 32, 36, 8, 96, 6, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1152x864 @ 60 Hz, 53.5 kHz hsync */ NULL, 60, 1152, 864, 11123, 208, 64, 16, 4, 256, 8, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 800x600 @ 85 Hz, 55.84 kHz hsync */ NULL, 85, 800, 600, 16460, 160, 64, 36, 16, 64, 5, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1024x768 @ 70 Hz, 56.5 kHz hsync */ NULL, 70, 1024, 768, 13333, 144, 24, 29, 3, 136, 6, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x1024 @ 87 Hz interlaced, 51 kHz hsync */ NULL, 87, 1280, 1024, 12500, 56, 16, 128, 1, 216, 12, - 0, FB_VMODE_INTERLACED + 0, FB_VMODE_INTERLACED, FB_MODE_IS_UNKNOWN }, { /* 800x600 @ 100 Hz, 64.02 kHz hsync */ NULL, 100, 800, 600, 14357, 160, 64, 30, 4, 64, 6, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1024x768 @ 76 Hz, 62.5 kHz hsync */ NULL, 76, 1024, 768, 11764, 208, 8, 36, 16, 120, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1152x864 @ 70 Hz, 62.4 kHz hsync */ NULL, 70, 1152, 864, 10869, 106, 56, 20, 1, 160, 10, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x1024 @ 61 Hz, 64.2 kHz hsync */ NULL, 61, 1280, 1024, 9090, 200, 48, 26, 1, 184, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1400x1050 @ 60Hz, 63.9 kHz hsync */ NULL, 60, 1400, 1050, 9259, 136, 40, 13, 1, 112, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1400x1050 @ 75,107 Hz, 82,392 kHz +hsync +vsync*/ NULL, 75, 1400, 1050, 7190, 120, 56, 23, 10, 112, 13, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1400x1050 @ 60 Hz, ? kHz +hsync +vsync*/ NULL, 60, 1400, 1050, 9259, 128, 40, 12, 0, 112, 3, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1024x768 @ 85 Hz, 70.24 kHz hsync */ NULL, 85, 1024, 768, 10111, 192, 32, 34, 14, 160, 6, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1152x864 @ 78 Hz, 70.8 kHz hsync */ NULL, 78, 1152, 864, 9090, 228, 88, 32, 0, 84, 12, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x1024 @ 70 Hz, 74.59 kHz hsync */ NULL, 70, 1280, 1024, 7905, 224, 32, 28, 8, 160, 8, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1600x1200 @ 60Hz, 75.00 kHz hsync */ NULL, 60, 1600, 1200, 6172, 304, 64, 46, 1, 192, 3, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1152x864 @ 84 Hz, 76.0 kHz hsync */ NULL, 84, 1152, 864, 7407, 184, 312, 32, 0, 128, 12, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x1024 @ 74 Hz, 78.85 kHz hsync */ NULL, 74, 1280, 1024, 7407, 256, 32, 34, 3, 144, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1024x768 @ 100Hz, 80.21 kHz hsync */ NULL, 100, 1024, 768, 8658, 192, 32, 21, 3, 192, 10, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x1024 @ 76 Hz, 81.13 kHz hsync */ NULL, 76, 1280, 1024, 7407, 248, 32, 34, 3, 104, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1600x1200 @ 70 Hz, 87.50 kHz hsync */ NULL, 70, 1600, 1200, 5291, 304, 64, 46, 1, 192, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1152x864 @ 100 Hz, 89.62 kHz hsync */ NULL, 100, 1152, 864, 7264, 224, 32, 17, 2, 128, 19, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x1024 @ 85 Hz, 91.15 kHz hsync */ NULL, 85, 1280, 1024, 6349, 224, 64, 44, 1, 160, 3, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1600x1200 @ 75 Hz, 93.75 kHz hsync */ NULL, 75, 1600, 1200, 4938, 304, 64, 46, 1, 192, 3, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1680x1050 @ 60 Hz, 65.191 kHz hsync */ NULL, 60, 1680, 1050, 6848, 280, 104, 30, 3, 176, 6, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1600x1200 @ 85 Hz, 105.77 kHz hsync */ NULL, 85, 1600, 1200, 4545, 272, 16, 37, 4, 192, 3, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x1024 @ 100 Hz, 107.16 kHz hsync */ NULL, 100, 1280, 1024, 5502, 256, 32, 26, 7, 128, 15, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1800x1440 @ 64Hz, 96.15 kHz hsync */ NULL, 64, 1800, 1440, 4347, 304, 96, 46, 1, 192, 3, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1800x1440 @ 70Hz, 104.52 kHz hsync */ NULL, 70, 1800, 1440, 4000, 304, 96, 46, 1, 192, 3, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 512x384 @ 78 Hz, 31.50 kHz hsync */ NULL, 78, 512, 384, 49603, 48, 16, 16, 1, 64, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 512x384 @ 85 Hz, 34.38 kHz hsync */ NULL, 85, 512, 384, 45454, 48, 16, 16, 1, 64, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 320x200 @ 70 Hz, 31.5 kHz hsync, 8:5 aspect ratio */ NULL, 70, 320, 200, 79440, 16, 16, 20, 4, 48, 1, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 320x240 @ 60 Hz, 31.5 kHz hsync, 4:3 aspect ratio */ NULL, 60, 320, 240, 79440, 16, 16, 16, 5, 48, 1, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 320x240 @ 72 Hz, 36.5 kHz hsync */ NULL, 72, 320, 240, 63492, 16, 16, 16, 4, 48, 2, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 400x300 @ 56 Hz, 35.2 kHz hsync, 4:3 aspect ratio */ NULL, 56, 400, 300, 55555, 64, 16, 10, 1, 32, 1, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 400x300 @ 60 Hz, 37.8 kHz hsync */ NULL, 60, 400, 300, 50000, 48, 16, 11, 1, 64, 2, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 400x300 @ 72 Hz, 48.0 kHz hsync */ NULL, 72, 400, 300, 40000, 32, 24, 11, 19, 64, 3, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 480x300 @ 56 Hz, 35.2 kHz hsync, 8:5 aspect ratio */ NULL, 56, 480, 300, 46176, 80, 16, 10, 1, 40, 1, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 480x300 @ 60 Hz, 37.8 kHz hsync */ NULL, 60, 480, 300, 41858, 56, 16, 11, 1, 80, 2, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 480x300 @ 63 Hz, 39.6 kHz hsync */ NULL, 63, 480, 300, 40000, 56, 16, 11, 1, 80, 2, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 480x300 @ 72 Hz, 48.0 kHz hsync */ NULL, 72, 480, 300, 33386, 40, 24, 11, 19, 80, 3, - 0, FB_VMODE_DOUBLE + 0, FB_VMODE_DOUBLE, FB_MODE_IS_UNKNOWN }, { /* 1920x1200 @ 60 Hz, 74.5 Khz hsync */ NULL, 60, 1920, 1200, 5177, 128, 336, 1, 38, 208, 3, FB_SYNC_HOR_HIGH_ACT | FB_SYNC_VERT_HIGH_ACT, - FB_VMODE_NONINTERLACED + FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1152x768, 60 Hz, PowerBook G4 Titanium I and II */ NULL, 60, 1152, 768, 14047, 158, 26, 29, 3, 136, 6, - FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED + FB_SYNC_HOR_HIGH_ACT|FB_SYNC_VERT_HIGH_ACT, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1366x768, 60 Hz, 47.403 kHz hsync, WXGA 16:9 aspect ratio */ NULL, 60, 1366, 768, 13806, 120, 10, 14, 3, 32, 5, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, { /* 1280x800, 60 Hz, 47.403 kHz hsync, WXGA 16:10 aspect ratio */ NULL, 60, 1280, 800, 12048, 200, 64, 24, 1, 136, 3, - 0, FB_VMODE_NONINTERLACED + 0, FB_VMODE_NONINTERLACED, FB_MODE_IS_UNKNOWN }, }; diff -urNp linux-2.6.28.8/drivers/video/uvesafb.c linux-2.6.28.8/drivers/video/uvesafb.c --- linux-2.6.28.8/drivers/video/uvesafb.c 2009-02-06 16:47:45.000000000 -0500 +++ linux-2.6.28.8/drivers/video/uvesafb.c 2009-02-21 09:37:48.000000000 -0500 @@ -18,6 +18,7 @@ #include #include #include +#include #include