jbd2: clear BH_Delay & BH_Unwritten in journal_unmap_buffer
authorEric Sandeen <sandeen@redhat.com>
Mon, 20 Feb 2012 22:53:01 +0000 (17:53 -0500)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 2 Apr 2012 16:27:18 +0000 (09:27 -0700)
commite4f7c341d6c6a78830dc918b23157ccfc399d397
tree67b44500f4449defb925bc452ac8a278f13b6cc3
parentfb9ea4df52f6114571e1e03f13f38d4bfc4b3587
jbd2: clear BH_Delay & BH_Unwritten in journal_unmap_buffer

commit 15291164b22a357cb211b618adfef4fa82fc0de3 upstream.

journal_unmap_buffer()'s zap_buffer: code clears a lot of buffer head
state ala discard_buffer(), but does not touch _Delay or _Unwritten as
discard_buffer() does.

This can be problematic in some areas of the ext4 code which assume
that if they have found a buffer marked unwritten or delay, then it's
a live one.  Perhaps those spots should check whether it is mapped
as well, but if jbd2 is going to tear down a buffer, let's really
tear it down completely.

Without this I get some fsx failures on sub-page-block filesystems
up until v3.2, at which point 4e96b2dbbf1d7e81f22047a50f862555a6cb87cb
and 189e868fa8fdca702eb9db9d8afc46b5cb9144c9 make the failures go
away, because buried within that large change is some more flag
clearing.  I still think it's worth doing in jbd2, since
->invalidatepage leads here directly, and it's the right place
to clear away these flags.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/jbd2/transaction.c