Commit 22b31eec63e5f2e219a3ee15f456897272bc73e8

Authored by Hugh Dickins
Committed by Linus Torvalds
1 parent 3dc147414c

badpage: vm_normal_page use print_bad_pte

print_bad_pte() is so far being called only when zap_pte_range() finds
negative page_mapcount, or there's a fault on a pte_file where it does not
belong.  That's weak coverage when we suspect pagetable corruption.

Originally, it was called when vm_normal_page() found an invalid pfn: but
pfn_valid is expensive on some architectures and configurations, so 2.6.24
put that under CONFIG_DEBUG_VM (which doesn't help in the field), then
2.6.26 replaced it by a VM_BUG_ON (likewise).

Reinstate the print_bad_pte() in vm_normal_page(), but use a cheaper test
than pfn_valid(): memmap_init_zone() (used in bootup and hotplug) keep a
__read_mostly note of the highest_memmap_pfn, vm_normal_page() then check
pfn against that.  We could call this pfn_plausible() or pfn_sane(), but I
doubt we'll need it elsewhere: of course it's not reliable, but gives much
stronger pagetable validation on many boxes.

Also use print_bad_pte() when the pte_special bit is found outside a
VM_PFNMAP or VM_MIXEDMAP area, instead of VM_BUG_ON.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Showing 3 changed files with 15 additions and 10 deletions Side-by-side Diff

... ... @@ -49,6 +49,7 @@
49 49 /*
50 50 * in mm/page_alloc.c
51 51 */
  52 +extern unsigned long highest_memmap_pfn;
52 53 extern void __free_pages_bootmem(struct page *page, unsigned int order);
53 54  
54 55 /*
... ... @@ -467,21 +467,18 @@
467 467 struct page *vm_normal_page(struct vm_area_struct *vma, unsigned long addr,
468 468 pte_t pte)
469 469 {
470   - unsigned long pfn;
  470 + unsigned long pfn = pte_pfn(pte);
471 471  
472 472 if (HAVE_PTE_SPECIAL) {
473   - if (likely(!pte_special(pte))) {
474   - VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
475   - return pte_page(pte);
476   - }
477   - VM_BUG_ON(!(vma->vm_flags & (VM_PFNMAP | VM_MIXEDMAP)));
  473 + if (likely(!pte_special(pte)))
  474 + goto check_pfn;
  475 + if (!(vma->vm_flags & (VM_PFNMAP | VM_MIXEDMAP)))
  476 + print_bad_pte(vma, addr, pte, NULL);
478 477 return NULL;
479 478 }
480 479  
481 480 /* !HAVE_PTE_SPECIAL case follows: */
482 481  
483   - pfn = pte_pfn(pte);
484   -
485 482 if (unlikely(vma->vm_flags & (VM_PFNMAP|VM_MIXEDMAP))) {
486 483 if (vma->vm_flags & VM_MIXEDMAP) {
487 484 if (!pfn_valid(pfn))
488 485  
... ... @@ -497,11 +494,14 @@
497 494 }
498 495 }
499 496  
500   - VM_BUG_ON(!pfn_valid(pfn));
  497 +check_pfn:
  498 + if (unlikely(pfn > highest_memmap_pfn)) {
  499 + print_bad_pte(vma, addr, pte, NULL);
  500 + return NULL;
  501 + }
501 502  
502 503 /*
503 504 * NOTE! We still have PageReserved() pages in the page tables.
504   - *
505 505 * eg. VDSO mappings can cause them to exist.
506 506 */
507 507 out:
... ... @@ -69,6 +69,7 @@
69 69  
70 70 unsigned long totalram_pages __read_mostly;
71 71 unsigned long totalreserve_pages __read_mostly;
  72 +unsigned long highest_memmap_pfn __read_mostly;
72 73 int percpu_pagelist_fraction;
73 74  
74 75 #ifdef CONFIG_HUGETLB_PAGE_SIZE_VARIABLE
... ... @@ -2596,6 +2597,9 @@
2596 2597 unsigned long end_pfn = start_pfn + size;
2597 2598 unsigned long pfn;
2598 2599 struct zone *z;
  2600 +
  2601 + if (highest_memmap_pfn < end_pfn - 1)
  2602 + highest_memmap_pfn = end_pfn - 1;
2599 2603  
2600 2604 z = &NODE_DATA(nid)->node_zones[zone];
2601 2605 for (pfn = start_pfn; pfn < end_pfn; pfn++) {