]> err.no Git - linux-2.6/commitdiff
x86: unify fault_32|64.c by ifdef'd function bodies
authorHarvey Harrison <harvey.harrison@gmail.com>
Wed, 30 Jan 2008 12:34:10 +0000 (13:34 +0100)
committerIngo Molnar <mingo@elte.hu>
Wed, 30 Jan 2008 12:34:10 +0000 (13:34 +0100)
It's about time to get on with unifying these files, elimination
of the ugly ifdefs can occur in the unified file.

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
arch/x86/mm/fault_32.c
arch/x86/mm/fault_64.c

index 28ea3d3ec8f860b56074777394d8d14fe1c0fbab..7d9ecbbba745870f2b52fd2ad562fbc36d171154 100644 (file)
@@ -173,8 +173,17 @@ static void force_sig_info_fault(int si_signo, int si_code,
        force_sig_info(si_signo, &info, tsk);
 }
 
+#ifdef CONFIG_X86_64
+static int bad_address(void *p)
+{
+       unsigned long dummy;
+       return probe_kernel_address((unsigned long *)p, dummy);
+}
+#endif
+
 void dump_pagetable(unsigned long address)
 {
+#ifdef CONFIG_X86_32
        __typeof__(pte_val(__pte(0))) page;
 
        page = read_cr3();
@@ -209,8 +218,42 @@ void dump_pagetable(unsigned long address)
        }
 
        printk("\n");
+#else /* CONFIG_X86_64 */
+       pgd_t *pgd;
+       pud_t *pud;
+       pmd_t *pmd;
+       pte_t *pte;
+
+       pgd = (pgd_t *)read_cr3();
+
+       pgd = __va((unsigned long)pgd & PHYSICAL_PAGE_MASK);
+       pgd += pgd_index(address);
+       if (bad_address(pgd)) goto bad;
+       printk("PGD %lx ", pgd_val(*pgd));
+       if (!pgd_present(*pgd)) goto ret;
+
+       pud = pud_offset(pgd, address);
+       if (bad_address(pud)) goto bad;
+       printk("PUD %lx ", pud_val(*pud));
+       if (!pud_present(*pud)) goto ret;
+
+       pmd = pmd_offset(pud, address);
+       if (bad_address(pmd)) goto bad;
+       printk("PMD %lx ", pmd_val(*pmd));
+       if (!pmd_present(*pmd) || pmd_large(*pmd)) goto ret;
+
+       pte = pte_offset_kernel(pmd, address);
+       if (bad_address(pte)) goto bad;
+       printk("PTE %lx", pte_val(*pte));
+ret:
+       printk("\n");
+       return;
+bad:
+       printk("BAD\n");
+#endif
 }
 
+#ifdef CONFIG_X86_32
 static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
 {
        unsigned index = pgd_index(address);
@@ -246,6 +289,7 @@ static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
                BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
        return pmd_k;
 }
+#endif
 
 #ifdef CONFIG_X86_64
 static const char errata93_warning[] =
@@ -326,6 +370,7 @@ static int is_f00f_bug(struct pt_regs *regs, unsigned long address)
 static void show_fault_oops(struct pt_regs *regs, unsigned long error_code,
                            unsigned long address)
 {
+#ifdef CONFIG_X86_32
        if (!oops_may_print())
                return;
 
@@ -350,8 +395,40 @@ static void show_fault_oops(struct pt_regs *regs, unsigned long error_code,
        printk(KERN_ALERT "IP:");
        printk_address(regs->ip, 1);
        dump_pagetable(address);
+#else /* CONFIG_X86_64 */
+       printk(KERN_ALERT "BUG: unable to handle kernel ");
+       if (address < PAGE_SIZE)
+               printk(KERN_CONT "NULL pointer dereference");
+       else
+               printk(KERN_CONT "paging request");
+       printk(KERN_CONT " at %016lx\n", address);
+
+       printk(KERN_ALERT "IP:");
+       printk_address(regs->ip, 1);
+       dump_pagetable(address);
+#endif
 }
 
+#ifdef CONFIG_X86_64
+static noinline void pgtable_bad(unsigned long address, struct pt_regs *regs,
+                                unsigned long error_code)
+{
+       unsigned long flags = oops_begin();
+       struct task_struct *tsk;
+
+       printk(KERN_ALERT "%s: Corrupted page table at address %lx\n",
+              current->comm, address);
+       dump_pagetable(address);
+       tsk = current;
+       tsk->thread.cr2 = address;
+       tsk->thread.trap_no = 14;
+       tsk->thread.error_code = error_code;
+       if (__die("Bad pagetable", regs, error_code))
+               regs = NULL;
+       oops_end(flags, regs, SIGKILL);
+}
+#endif
+
 /*
  * Handle a fault on the vmalloc or module mapping area
  *
@@ -706,6 +783,7 @@ do_sigbus:
 
 void vmalloc_sync_all(void)
 {
+#ifdef CONFIG_X86_32
        /*
         * Note that races in the updates of insync and start aren't
         * problematic: insync can only get set bits added, and updates to
@@ -740,4 +818,42 @@ void vmalloc_sync_all(void)
                if (address == start && test_bit(pgd_index(address), insync))
                        start = address + PGDIR_SIZE;
        }
+#else /* CONFIG_X86_64 */
+       /*
+        * Note that races in the updates of insync and start aren't
+        * problematic: insync can only get set bits added, and updates to
+        * start are only improving performance (without affecting correctness
+        * if undone).
+        */
+       static DECLARE_BITMAP(insync, PTRS_PER_PGD);
+       static unsigned long start = VMALLOC_START & PGDIR_MASK;
+       unsigned long address;
+
+       for (address = start; address <= VMALLOC_END; address += PGDIR_SIZE) {
+               if (!test_bit(pgd_index(address), insync)) {
+                       const pgd_t *pgd_ref = pgd_offset_k(address);
+                       struct page *page;
+
+                       if (pgd_none(*pgd_ref))
+                               continue;
+                       spin_lock(&pgd_lock);
+                       list_for_each_entry(page, &pgd_list, lru) {
+                               pgd_t *pgd;
+                               pgd = (pgd_t *)page_address(page) + pgd_index(address);
+                               if (pgd_none(*pgd))
+                                       set_pgd(pgd, *pgd_ref);
+                               else
+                                       BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref));
+                       }
+                       spin_unlock(&pgd_lock);
+                       set_bit(pgd_index(address), insync);
+               }
+               if (address == start)
+                       start = address + PGDIR_SIZE;
+       }
+       /* Check that there is no need to do the same for the modules area. */
+       BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL));
+       BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) ==
+                               (__START_KERNEL & PGDIR_MASK)));
+#endif
 }
index e12c34ba4ba13b1a62438d607459f583f306e199..edca689c62d52d549e0894e2dd2a3b83e9a97d18 100644 (file)
@@ -176,14 +176,52 @@ static void force_sig_info_fault(int si_signo, int si_code,
        force_sig_info(si_signo, &info, tsk);
 }
 
+#ifdef CONFIG_X86_64
 static int bad_address(void *p)
 {
        unsigned long dummy;
        return probe_kernel_address((unsigned long *)p, dummy);
 }
+#endif
 
 void dump_pagetable(unsigned long address)
 {
+#ifdef CONFIG_X86_32
+       __typeof__(pte_val(__pte(0))) page;
+
+       page = read_cr3();
+       page = ((__typeof__(page) *) __va(page))[address >> PGDIR_SHIFT];
+#ifdef CONFIG_X86_PAE
+       printk("*pdpt = %016Lx ", page);
+       if ((page >> PAGE_SHIFT) < max_low_pfn
+           && page & _PAGE_PRESENT) {
+               page &= PAGE_MASK;
+               page = ((__typeof__(page) *) __va(page))[(address >> PMD_SHIFT)
+                                                        & (PTRS_PER_PMD - 1)];
+               printk(KERN_CONT "*pde = %016Lx ", page);
+               page &= ~_PAGE_NX;
+       }
+#else
+       printk("*pde = %08lx ", page);
+#endif
+
+       /*
+        * We must not directly access the pte in the highpte
+        * case if the page table is located in highmem.
+        * And let's rather not kmap-atomic the pte, just in case
+        * it's allocated already.
+        */
+       if ((page >> PAGE_SHIFT) < max_low_pfn
+           && (page & _PAGE_PRESENT)
+           && !(page & _PAGE_PSE)) {
+               page &= PAGE_MASK;
+               page = ((__typeof__(page) *) __va(page))[(address >> PAGE_SHIFT)
+                                                        & (PTRS_PER_PTE - 1)];
+               printk("*pte = %0*Lx ", sizeof(page)*2, (u64)page);
+       }
+
+       printk("\n");
+#else /* CONFIG_X86_64 */
        pgd_t *pgd;
        pud_t *pud;
        pmd_t *pmd;
@@ -215,7 +253,46 @@ ret:
        return;
 bad:
        printk("BAD\n");
+#endif
+}
+
+#ifdef CONFIG_X86_32
+static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
+{
+       unsigned index = pgd_index(address);
+       pgd_t *pgd_k;
+       pud_t *pud, *pud_k;
+       pmd_t *pmd, *pmd_k;
+
+       pgd += index;
+       pgd_k = init_mm.pgd + index;
+
+       if (!pgd_present(*pgd_k))
+               return NULL;
+
+       /*
+        * set_pgd(pgd, *pgd_k); here would be useless on PAE
+        * and redundant with the set_pmd() on non-PAE. As would
+        * set_pud.
+        */
+
+       pud = pud_offset(pgd, address);
+       pud_k = pud_offset(pgd_k, address);
+       if (!pud_present(*pud_k))
+               return NULL;
+
+       pmd = pmd_offset(pud, address);
+       pmd_k = pmd_offset(pud_k, address);
+       if (!pmd_present(*pmd_k))
+               return NULL;
+       if (!pmd_present(*pmd)) {
+               set_pmd(pmd, *pmd_k);
+               arch_flush_lazy_mmu_mode();
+       } else
+               BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
+       return pmd_k;
 }
+#endif
 
 #ifdef CONFIG_X86_64
 static const char errata93_warning[] =
@@ -296,6 +373,32 @@ static int is_f00f_bug(struct pt_regs *regs, unsigned long address)
 static void show_fault_oops(struct pt_regs *regs, unsigned long error_code,
                            unsigned long address)
 {
+#ifdef CONFIG_X86_32
+       if (!oops_may_print())
+               return;
+
+#ifdef CONFIG_X86_PAE
+       if (error_code & PF_INSTR) {
+               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);
+       }
+#endif
+       printk(KERN_ALERT "BUG: unable to handle kernel ");
+       if (address < PAGE_SIZE)
+               printk(KERN_CONT "NULL pointer dereference");
+       else
+               printk(KERN_CONT "paging request");
+       printk(KERN_CONT " at %08lx\n", address);
+
+       printk(KERN_ALERT "IP:");
+       printk_address(regs->ip, 1);
+       dump_pagetable(address);
+#else /* CONFIG_X86_64 */
        printk(KERN_ALERT "BUG: unable to handle kernel ");
        if (address < PAGE_SIZE)
                printk(KERN_CONT "NULL pointer dereference");
@@ -306,8 +409,10 @@ static void show_fault_oops(struct pt_regs *regs, unsigned long error_code,
        printk(KERN_ALERT "IP:");
        printk_address(regs->ip, 1);
        dump_pagetable(address);
+#endif
 }
 
+#ifdef CONFIG_X86_64
 static noinline void pgtable_bad(unsigned long address, struct pt_regs *regs,
                                 unsigned long error_code)
 {
@@ -325,6 +430,7 @@ static noinline void pgtable_bad(unsigned long address, struct pt_regs *regs,
                regs = NULL;
        oops_end(flags, regs, SIGKILL);
 }
+#endif
 
 /*
  * Handle a fault on the vmalloc area
@@ -590,12 +696,15 @@ bad_area:
 bad_area_nosemaphore:
        /* User mode accesses just cause a SIGSEGV */
        if (error_code & PF_USER) {
-
                /*
                 * It's possible to have interrupts off here.
                 */
                local_irq_enable();
 
+               /*
+                * Valid to do another page fault here because this one came
+                * from user space.
+                */
                if (is_prefetch(regs, address, error_code))
                        return;
 
@@ -696,6 +805,42 @@ LIST_HEAD(pgd_list);
 
 void vmalloc_sync_all(void)
 {
+#ifdef CONFIG_X86_32
+       /*
+        * Note that races in the updates of insync and start aren't
+        * problematic: insync can only get set bits added, and updates to
+        * start are only improving performance (without affecting correctness
+        * if undone).
+        */
+       static DECLARE_BITMAP(insync, PTRS_PER_PGD);
+       static unsigned long start = TASK_SIZE;
+       unsigned long address;
+
+       if (SHARED_KERNEL_PMD)
+               return;
+
+       BUILD_BUG_ON(TASK_SIZE & ~PGDIR_MASK);
+       for (address = start; address >= TASK_SIZE; address += PGDIR_SIZE) {
+               if (!test_bit(pgd_index(address), insync)) {
+                       unsigned long flags;
+                       struct page *page;
+
+                       spin_lock_irqsave(&pgd_lock, flags);
+                       for (page = pgd_list; page; page =
+                                       (struct page *)page->index)
+                               if (!vmalloc_sync_one(page_address(page),
+                                                               address)) {
+                                       BUG_ON(page != pgd_list);
+                                       break;
+                               }
+                       spin_unlock_irqrestore(&pgd_lock, flags);
+                       if (!page)
+                               set_bit(pgd_index(address), insync);
+               }
+               if (address == start && test_bit(pgd_index(address), insync))
+                       start = address + PGDIR_SIZE;
+       }
+#else /* CONFIG_X86_64 */
        /*
         * Note that races in the updates of insync and start aren't
         * problematic: insync can only get set bits added, and updates to
@@ -732,4 +877,5 @@ void vmalloc_sync_all(void)
        BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL));
        BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) ==
                                (__START_KERNEL & PGDIR_MASK)));
+#endif
 }