xfs
[Top] [All Lists]

[PATCH 102/102] xfs: handle EOF correctly in xfs_vm_writepage

To: xfs@xxxxxxxxxxx
Subject: [PATCH 102/102] xfs: handle EOF correctly in xfs_vm_writepage
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Thu, 23 Aug 2012 15:03:00 +1000
In-reply-to: <1345698180-13612-1-git-send-email-david@xxxxxxxxxxxxx>
References: <1345698180-13612-1-git-send-email-david@xxxxxxxxxxxxx>
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>

Upstream commit: 6b7a03f03a2f8b1629133e35729eba4727fae3cc

We need to zero out part of a page which beyond EOF before setting uptodate,
otherwise, mapread or write will see non-zero data beyond EOF.

Based on the code in fs/buffer.c and the following ext4 commit:

  ext4: handle EOF correctly in ext4_bio_write_page()

And yes, I wish we had a good test case for it.

Signed-off-by: Christoph Hellwig <hch@xxxxxx>
Reviewed-by: Mark Tinguely <tinguely@xxxxxxx>
Signed-off-by: Ben Myers <bpm@xxxxxxx>
---
 fs/xfs/linux-2.6/xfs_aops.c |   19 +++++++++++++++++--
 1 file changed, 17 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index 42eaf00..8c2ed74 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -978,11 +978,26 @@ xfs_vm_writepage(
        end_index = offset >> PAGE_CACHE_SHIFT;
        last_index = (offset - 1) >> PAGE_CACHE_SHIFT;
        if (page->index >= end_index) {
-               if ((page->index >= end_index + 1) ||
-                   !(i_size_read(inode) & (PAGE_CACHE_SIZE - 1))) {
+               unsigned offset_into_page = offset & (PAGE_CACHE_SIZE - 1);
+
+               /*
+                * Just skip the page if it is fully outside i_size, e.g. due
+                * to a truncate operation that is in progress.
+                */
+               if (page->index >= end_index + 1 || offset_into_page == 0) {
                        unlock_page(page);
                        return 0;
                }
+
+               /*
+                * The page straddles i_size.  It must be zeroed out on each
+                * and every writepage invocation because it may be mmapped.
+                * "A file is mapped in multiples of the page size.  For a file
+                * that is not a multiple of the  page size, the remaining
+                * memory is zeroed when mapped, and writes to that region are
+                * not written out to the file."
+                */
+               zero_user_segment(page, offset_into_page, PAGE_CACHE_SIZE);
        }
 
        end_offset = min_t(unsigned long long,
-- 
1.7.10

<Prev in Thread] Current Thread [Next in Thread>