calculation of pgoff in do_linear_fault() uses mixed units
authorDean Nelson <dcn@sgi.com>
Tue, 16 Oct 2007 08:24:45 +0000 (01:24 -0700)
committerLinus Torvalds <torvalds@woody.linux-foundation.org>
Tue, 16 Oct 2007 16:42:53 +0000 (09:42 -0700)
The calculation of pgoff in do_linear_fault() should use PAGE_SHIFT and not
PAGE_CACHE_SHIFT since vma->vm_pgoff is in units of PAGE_SIZE and not
PAGE_CACHE_SIZE.  At the moment linux/pagemap.h has PAGE_CACHE_SHIFT
defined as PAGE_SHIFT, but should that ever change this calculation would
break.

Signed-off-by: Dean Nelson <dcn@sgi.com>
Acked-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/memory.c

index 2a8430844b6dc5b2a4ef4c4a28a4723c5f5937cc..b3c42f0f65c29188aafa5c8e25262a610b4a2441 100644 (file)
@@ -2361,7 +2361,7 @@ static int do_linear_fault(struct mm_struct *mm, struct vm_area_struct *vma,
                int write_access, pte_t orig_pte)
 {
        pgoff_t pgoff = (((address & PAGE_MASK)
-                       - vma->vm_start) >> PAGE_CACHE_SHIFT) + vma->vm_pgoff;
+                       - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
        unsigned int flags = (write_access ? FAULT_FLAG_WRITE : 0);
 
        pte_unmap(page_table);