From: akpm@osdl.org Subject: + jbd-wait-for-already-submitted-t_sync_datalist-buffer.patch added to -mm tree Date: Wed, 29 Nov 2006 14:32:40 -0800 Message-ID: <200611292232.kATMWevr013719@shell0.pdx.osdl.net> Cc: hifumi.hisashi@oss.ntt.co.jp, jack@ucw.cz, linux-ext4@vger.kernel.org, sct@redhat.com Return-path: Received: from smtp.osdl.org ([65.172.181.25]:15817 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S1758889AbWK2WdE (ORCPT ); Wed, 29 Nov 2006 17:33:04 -0500 To: mm-commits@vger.kernel.org Sender: linux-ext4-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org The patch titled jbd: wait for already submitted t_sync_datalist buffer to complete has been added to the -mm tree. Its filename is jbd-wait-for-already-submitted-t_sync_datalist-buffer.patch See http://www.zip.com.au/~akpm/linux/patches/stuff/added-to-mm.txt to find out what to do about this ------------------------------------------------------ Subject: jbd: wait for already submitted t_sync_datalist buffer to complete From: Hisashi Hifumi In the current jbd code, if a buffer on BJ_SyncData list is dirty and not locked, the buffer is refiled to BJ_Locked list, submitted to the IO and waited for IO completion. But the fsstress test showed the case that when a buffer was already submitted to the IO just before the buffer_dirty(bh) check, the buffer was not waited for IO completion. Following patch solves this problem. If it is assumed that a buffer is submitted to the IO before the buffer_dirty(bh) check and still being written to disk, this buffer is refiled to BJ_Locked list. Signed-off-by: Hisashi Hifumi Cc: Jan Kara Cc: "Stephen C. Tweedie" Cc: Signed-off-by: Andrew Morton --- fs/jbd/commit.c | 8 ++++++-- 1 files changed, 6 insertions(+), 2 deletions(-) diff -puN fs/jbd/commit.c~jbd-wait-for-already-submitted-t_sync_datalist-buffer fs/jbd/commit.c --- a/fs/jbd/commit.c~jbd-wait-for-already-submitted-t_sync_datalist-buffer +++ a/fs/jbd/commit.c @@ -248,8 +248,12 @@ write_out_data: bufs = 0; goto write_out_data; } - } - else { + } else if (!locked && buffer_locked(bh)) { + __journal_file_buffer(jh, commit_transaction, + BJ_Locked); + jbd_unlock_bh_state(bh); + put_bh(bh); + } else { BUFFER_TRACE(bh, "writeout complete: unfile"); __journal_unfile_buffer(jh); jbd_unlock_bh_state(bh); _ Patches currently in -mm which might be from hifumi.hisashi@oss.ntt.co.jp are jbd-wait-for-already-submitted-t_sync_datalist-buffer.patch