#include <asm/proto.h>
#include <asm/kdebug.h>
#include <asm-generic/sections.h>
-#include <asm/kdebug.h>
+
+/* Page fault error code bits */
+#define PF_PROT (1<<0) /* or no page found */
+#define PF_WRITE (1<<1)
+#define PF_USER (1<<2)
+#define PF_RSVD (1<<3)
+#define PF_INSTR (1<<4)
+
+#ifdef CONFIG_KPROBES
+ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
+
+/* Hook to register for page fault notifications */
+int register_page_fault_notifier(struct notifier_block *nb)
+{
+ vmalloc_sync_all();
+ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
+}
+
+int unregister_page_fault_notifier(struct notifier_block *nb)
+{
+ return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
+}
+
+static inline int notify_page_fault(enum die_val val, const char *str,
+ struct pt_regs *regs, long err, int trap, int sig)
+{
+ struct die_args args = {
+ .regs = regs,
+ .str = str,
+ .err = err,
+ .trapnr = trap,
+ .signr = sig
+ };
+ return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args);
+}
+#else
+static inline int notify_page_fault(enum die_val val, const char *str,
+ struct pt_regs *regs, long err, int trap, int sig)
+{
+ return NOTIFY_DONE;
+}
+#endif
void bust_spinlocks(int yes)
{
unsigned char *max_instr;
/* If it was a exec fault ignore */
- if (error_code & (1<<4))
+ if (error_code & PF_INSTR)
return 0;
instr = (unsigned char *)convert_rip_to_linear(current, regs);
pgd = __va((unsigned long)pgd & PHYSICAL_PAGE_MASK);
pgd += pgd_index(address);
- printk("PGD %lx ", pgd_val(*pgd));
if (bad_address(pgd)) goto bad;
+ printk("PGD %lx ", pgd_val(*pgd));
if (!pgd_present(*pgd)) goto ret;
- pud = __pud_offset_k((pud_t *)pgd_page(*pgd), address);
+ pud = pud_offset(pgd, address);
if (bad_address(pud)) goto bad;
printk("PUD %lx ", pud_val(*pud));
if (!pud_present(*pud)) goto ret;
return -1;
if (pgd_none(*pgd))
set_pgd(pgd, *pgd_ref);
+ else
+ BUG_ON(pgd_page(*pgd) != pgd_page(*pgd_ref));
/* Below here mismatches are bugs because these lower tables
are shared */
* This routine handles page faults. It determines the address,
* and the problem, and then passes it off to one of the appropriate
* routines.
- *
- * error_code:
- * bit 0 == 0 means no page found, 1 means protection fault
- * bit 1 == 0 means read, 1 means write
- * bit 2 == 0 means kernel, 1 means user-mode
- * bit 3 == 1 means use of reserved bit detected
- * bit 4 == 1 means fault was an instruction fetch
*/
asmlinkage void __kprobes do_page_fault(struct pt_regs *regs,
unsigned long error_code)
unsigned long flags;
siginfo_t info;
+ tsk = current;
+ mm = tsk->mm;
+ prefetchw(&mm->mmap_sem);
+
/* get the address */
__asm__("movq %%cr2,%0":"=r" (address));
- if (notify_die(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
- SIGSEGV) == NOTIFY_STOP)
- return;
- if (likely(regs->eflags & X86_EFLAGS_IF))
- local_irq_enable();
-
- if (unlikely(page_fault_trace))
- printk("pagefault rip:%lx rsp:%lx cs:%lu ss:%lu address %lx error %lx\n",
- regs->rip,regs->rsp,regs->cs,regs->ss,address,error_code);
-
- tsk = current;
- mm = tsk->mm;
info.si_code = SEGV_MAPERR;
* is always initialized because it's shared with the main
* kernel text. Only vmalloc may need PML4 syncups.
*/
- if (!(error_code & 0xd) &&
+ if (!(error_code & (PF_RSVD|PF_USER|PF_PROT)) &&
((address >= VMALLOC_START && address < VMALLOC_END))) {
- if (vmalloc_fault(address) < 0)
- goto bad_area_nosemaphore;
- return;
+ if (vmalloc_fault(address) >= 0)
+ return;
}
+ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
+ SIGSEGV) == NOTIFY_STOP)
+ return;
/*
* Don't take the mm semaphore here. If we fixup a prefetch
* fault we could otherwise deadlock.
goto bad_area_nosemaphore;
}
- if (unlikely(error_code & (1 << 3)))
+ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
+ SIGSEGV) == NOTIFY_STOP)
+ return;
+
+ if (likely(regs->eflags & X86_EFLAGS_IF))
+ local_irq_enable();
+
+ if (unlikely(page_fault_trace))
+ printk("pagefault rip:%lx rsp:%lx cs:%lu ss:%lu address %lx error %lx\n",
+ regs->rip,regs->rsp,regs->cs,regs->ss,address,error_code);
+
+ if (unlikely(error_code & PF_RSVD))
pgtable_bad(address, regs, error_code);
/*
* thus avoiding the deadlock.
*/
if (!down_read_trylock(&mm->mmap_sem)) {
- if ((error_code & 4) == 0 &&
+ if ((error_code & PF_USER) == 0 &&
!search_exception_tables(regs->rip))
goto bad_area_nosemaphore;
down_read(&mm->mmap_sem);
if (!(vma->vm_flags & VM_GROWSDOWN))
goto bad_area;
if (error_code & 4) {
- // XXX: align red zone size with ABI
- if (address + 128 < regs->rsp)
+ /* Allow userspace just enough access below the stack pointer
+ * to let the 'enter' instruction work.
+ */
+ if (address + 65536 + 32 * sizeof(unsigned long) < regs->rsp)
goto bad_area;
}
if (expand_stack(vma, address))
good_area:
info.si_code = SEGV_ACCERR;
write = 0;
- switch (error_code & 3) {
+ switch (error_code & (PF_PROT|PF_WRITE)) {
default: /* 3: write, present */
/* fall through */
- case 2: /* write, not present */
+ case PF_WRITE: /* write, not present */
if (!(vma->vm_flags & VM_WRITE))
goto bad_area;
write++;
break;
- case 1: /* read, present */
+ case PF_PROT: /* read, present */
goto bad_area;
- case 0: /* read, not present */
+ case 0: /* read, not present */
if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
goto bad_area;
}
bad_area_nosemaphore:
/* User mode accesses just cause a SIGSEGV */
- if (error_code & 4) {
+ if (error_code & PF_USER) {
if (is_prefetch(regs, address, error_code))
return;
up_read(&mm->mmap_sem);
/* Kernel mode? Handle exceptions or die */
- if (!(error_code & 4))
+ if (!(error_code & PF_USER))
goto no_context;
tsk->thread.cr2 = address;
return;
}
+DEFINE_SPINLOCK(pgd_lock);
+struct page *pgd_list;
+
+void vmalloc_sync_all(void)
+{
+ /* 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);
+ for (page = pgd_list; page;
+ page = (struct page *)page->index) {
+ 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(*pgd) != pgd_page(*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)));
+}
+
static int __init enable_pagefaulttrace(char *str)
{
page_fault_trace = 1;
- return 0;
+ return 1;
}
__setup("pagefaulttrace", enable_pagefaulttrace);