Btrfs: mark mapping with error flag to report errors to userspace
authorLiu Bo <bo.li.liu@oracle.com>
Mon, 12 May 2014 04:47:36 +0000 (12:47 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Tue, 1 Jul 2014 03:09:46 +0000 (20:09 -0700)
commit 5dca6eea91653e9949ce6eb9e9acab6277e2f2c4 upstream.

According to commit 865ffef3797da2cac85b3354b5b6050dc9660978
(fs: fix fsync() error reporting),
it's not stable to just check error pages because pages can be
truncated or invalidated, we should also mark mapping with error
flag so that a later fsync can catch the error.

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/btrfs/extent_io.c

index 82de090909355c28364e524fe87ec5509d07ebde..2d391ce71ea40876d20caac11aa4c99a01710376 100644 (file)
@@ -2371,6 +2371,8 @@ int end_extent_writepage(struct page *page, int err, u64 start, u64 end)
        if (!uptodate) {
                ClearPageUptodate(page);
                SetPageError(page);
+               ret = ret < 0 ? ret : -EIO;
+               mapping_set_error(page->mapping, ret);
        }
        return 0;
 }