From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933023Ab1ESJQo (ORCPT ); Thu, 19 May 2011 05:16:44 -0400 Received: from mx1.redhat.com ([209.132.183.28]:24874 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932764Ab1ESJKL (ORCPT ); Thu, 19 May 2011 05:10:11 -0400 From: Steven Whitehouse To: linux-kernel@vger.kernel.org, cluster-devel@redhat.com Cc: Steven Whitehouse Subject: [PATCH 07/32] GFS2: Use filemap_fdatawrite() to write back the AIL Date: Thu, 19 May 2011 09:47:02 +0100 Message-Id: <1305794847-3291-8-git-send-email-swhiteho@redhat.com> In-Reply-To: <1305794847-3291-1-git-send-email-swhiteho@redhat.com> References: <1305794847-3291-1-git-send-email-swhiteho@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org In order to ensure that the mapping stats (and thus the bdi) are correctly updated, this patch changes the AIL writeback to use the filemap_datawrite function. This helps prevent stalls in balance_dirty_pages() due to large amounts of dirty metadata when there is little or no dirty data around. Signed-off-by: Steven Whitehouse diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index 5b102c1..3ebafa1 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -91,6 +91,7 @@ static void gfs2_ail1_start_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai) __releases(&sdp->sd_ail_lock) __acquires(&sdp->sd_ail_lock) { + struct gfs2_glock *gl = NULL; struct gfs2_bufdata *bd, *s; struct buffer_head *bh; int retry; @@ -113,19 +114,13 @@ __acquires(&sdp->sd_ail_lock) if (!buffer_dirty(bh)) continue; - + if (gl == bd->bd_gl) + continue; + gl = bd->bd_gl; list_move(&bd->bd_ail_st_list, &ai->ai_ail1_list); - get_bh(bh); spin_unlock(&sdp->sd_ail_lock); - lock_buffer(bh); - if (test_clear_buffer_dirty(bh)) { - bh->b_end_io = end_buffer_write_sync; - submit_bh(WRITE_SYNC, bh); - } else { - unlock_buffer(bh); - brelse(bh); - } + filemap_fdatawrite(gfs2_glock2aspace(gl)); spin_lock(&sdp->sd_ail_lock); retry = 1; -- 1.7.4