cluster-devel.redhat.com archive mirror
 help / color / mirror / Atom feed
From: Steven Whitehouse <swhiteho@redhat.com>
To: cluster-devel.redhat.com
Subject: [Cluster-devel] [PATCH 06/19] GFS2: Fix bug trap and journaled data fsync
Date: Thu,  8 Sep 2011 16:50:15 +0100	[thread overview]
Message-ID: <1315497028-8473-7-git-send-email-swhiteho@redhat.com> (raw)
In-Reply-To: <1315497028-8473-1-git-send-email-swhiteho@redhat.com>

Journaled data requires that a complete flush of all dirty data for
the file is done, in order that the ail flush which comes after
will succeed.

Also the recently enhanced bug trap can trigger falsely in case
an ail flush from fsync races with a page read. This updates the
bug trap such that it will ignore buffers which are locked and
only trigger on dirty and/or pinned buffers when the ail flush
is run from fsync. The original bug trap is retained when ail
flush is run from ->go_sync()

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index 92c3db4..9d12286 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -590,6 +590,8 @@ static int gfs2_fsync(struct file *file, loff_t start, loff_t end,
 			mutex_unlock(&inode->i_mutex);
 			return ret;
 		}
+		if (gfs2_is_jdata(ip))
+			filemap_write_and_wait(mapping);
 		gfs2_ail_flush(ip->i_gl);
 		mutex_unlock(&inode->i_mutex);
 	}
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 99df483..0cc3ff4 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -46,7 +46,7 @@ static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh)
  * None of the buffers should be dirty, locked, or pinned.
  */
 
-static void __gfs2_ail_flush(struct gfs2_glock *gl)
+static void __gfs2_ail_flush(struct gfs2_glock *gl, unsigned long b_state)
 {
 	struct gfs2_sbd *sdp = gl->gl_sbd;
 	struct list_head *head = &gl->gl_ail_list;
@@ -60,7 +60,7 @@ static void __gfs2_ail_flush(struct gfs2_glock *gl)
 				bd_ail_gl_list);
 		bh = bd->bd_bh;
 		blocknr = bh->b_blocknr;
-		if (buffer_busy(bh))
+		if (bh->b_state & b_state)
 			gfs2_ail_error(gl, bh);
 		bh->b_private = NULL;
 		gfs2_remove_from_ail(bd); /* drops ref on bh */
@@ -99,7 +99,7 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 	BUG_ON(current->journal_info);
 	current->journal_info = &tr;
 
-	__gfs2_ail_flush(gl);
+	__gfs2_ail_flush(gl, (1ul << BH_Dirty)|(1ul << BH_Pinned)|(1ul << BH_Lock));
 
 	gfs2_trans_end(sdp);
 	gfs2_log_flush(sdp, NULL);
@@ -117,7 +117,7 @@ void gfs2_ail_flush(struct gfs2_glock *gl)
 	ret = gfs2_trans_begin(sdp, 0, revokes);
 	if (ret)
 		return;
-	__gfs2_ail_flush(gl);
+	__gfs2_ail_flush(gl, (1ul << BH_Dirty)|(1ul << BH_Pinned));
 	gfs2_trans_end(sdp);
 	gfs2_log_flush(sdp, NULL);
 }
-- 
1.7.4



  parent reply	other threads:[~2011-09-08 15:50 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-09-08 15:50 [Cluster-devel] Current -nmw tree content Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 01/19] GFS2: Clean up dir hash table reading Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 02/19] GFS2: Split data write & wait in fsync Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 03/19] GFS2: Fix bug-trap in ail flush code Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 04/19] GFS2: Make atime checks more efficient Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 05/19] GFS2: Fix inode allocation error path Steven Whitehouse
2011-09-08 15:50 ` Steven Whitehouse [this message]
2011-09-08 15:50 ` [Cluster-devel] [PATCH 07/19] GFS2: Use ->dirty_inode() Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 08/19] GFS2: Clean up gfs2_create Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 09/19] GFS2: Fix lseek after SEEK_DATA, SEEK_HOLE have been added Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 10/19] GFS2: Use rbtree for resource groups and clean up bitmap buffer ref count scheme Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 11/19] GFS2: Make resource groups "append only" during life of fs Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 12/19] GFS2: Cache the most recently used resource group in the inode Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 13/19] GFS2: Remove obsolete assert Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 14/19] GFS2: Call do_strip() directly from recursive_scan() Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 15/19] GFS2: Use cached rgrp in gfs2_rlist_add() Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 16/19] GFS2: Fix AIL flush issue during fsync Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 17/19] GFS2: Correctly set goal block after allocation Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 18/19] GFS2: Clean up ->page_mkwrite Steven Whitehouse
2011-09-08 15:50 ` [Cluster-devel] [PATCH 19/19] GFS2: Fix off-by-one in gfs2_blk2rgrpd Steven Whitehouse

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1315497028-8473-7-git-send-email-swhiteho@redhat.com \
    --to=swhiteho@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).