tree 0f0bf8deaeabc2d14fbded203392ec5bf7dc37ad
parent 56d1254917d9f301a8e24155cd3f2236e642cb7d
author Qu Fuping <qufuping@ercist.iscas.ac.cn> 1121441768 -0500
committer Dave Kleikamp <shaggy@austin.ibm.com> 1121441768 -0500

JFS: fsync wrong behavior when I/O failure occurs

This is half of a patch that Qu Fuping submitted in April.  The first part
was applied to fs/mpage.c in 2.6.12-rc4.

jfs_fsync should return error, but it doesn't wait for the metadata page to
be uptodate, e.g.:
jfs_fsync->jfs_commit_inode->txCommit->diWrite->read_metapage->
__get_metapage->read_cache_page reads a page from disk. Because read is
async, when read_cache_page: err = filler(data, page), filler will not
return error, it just submits I/O request and returns. So, page is not
uptodate.  Checking only if(IS_ERROR(mp->page)) is not enough, we should
add "|| !PageUptodate(mp->page)"

Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>