|
 |
32e0cc2 |
Path: news.gmane.org!not-for-mail
|
|
|
32e0cc2 |
From: Eric Sandeen <sandeen@redhat.com>
|
|
|
32e0cc2 |
Newsgroups: gmane.comp.file-systems.ext4
|
|
|
32e0cc2 |
Subject: [PATCH] jbd2: clear BH_Delay & BH_Unwritten in journal_unmap_buffer
|
|
|
32e0cc2 |
Date: Tue, 07 Feb 2012 16:07:20 -0600
|
|
|
32e0cc2 |
Lines: 42
|
|
|
32e0cc2 |
Approved: news@gmane.org
|
|
|
32e0cc2 |
Message-ID: <4F31A098.4050601@redhat.com>
|
|
|
32e0cc2 |
NNTP-Posting-Host: plane.gmane.org
|
|
|
32e0cc2 |
Mime-Version: 1.0
|
|
|
32e0cc2 |
Content-Type: text/plain; charset=ISO-8859-1
|
|
|
32e0cc2 |
Content-Transfer-Encoding: 7bit
|
|
|
32e0cc2 |
X-Trace: dough.gmane.org 1328656072 12026 80.91.229.3 (7 Feb 2012 23:07:52 GMT)
|
|
|
32e0cc2 |
X-Complaints-To: usenet@dough.gmane.org
|
|
|
32e0cc2 |
NNTP-Posting-Date: Tue, 7 Feb 2012 23:07:52 +0000 (UTC)
|
|
|
32e0cc2 |
To: ext4 development <linux-ext4@vger.kernel.org>
|
|
|
32e0cc2 |
Original-X-From: linux-ext4-owner@vger.kernel.org Wed Feb 08 00:07:52 2012
|
|
|
32e0cc2 |
Return-path: <linux-ext4-owner@vger.kernel.org>
|
|
|
32e0cc2 |
Envelope-to: gcfe-linux-ext4@plane.gmane.org
|
|
|
32e0cc2 |
Original-Received: from vger.kernel.org ([209.132.180.67])
|
|
|
32e0cc2 |
by plane.gmane.org with esmtp (Exim 4.69)
|
|
|
32e0cc2 |
(envelope-from <linux-ext4-owner@vger.kernel.org>)
|
|
|
32e0cc2 |
id 1Ruu8d-0000lK-5P
|
|
|
32e0cc2 |
for gcfe-linux-ext4@plane.gmane.org; Wed, 08 Feb 2012 00:07:51 +0100
|
|
|
32e0cc2 |
Original-Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
|
|
|
32e0cc2 |
id S1756187Ab2BGXHt (ORCPT <rfc822;gcfe-linux-ext4@m.gmane.org>);
|
|
|
32e0cc2 |
Tue, 7 Feb 2012 18:07:49 -0500
|
|
|
32e0cc2 |
Original-Received: from mx1.redhat.com ([209.132.183.28]:19432 "EHLO mx1.redhat.com"
|
|
|
32e0cc2 |
rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
|
|
|
32e0cc2 |
id S1753992Ab2BGXHs (ORCPT <rfc822;linux-ext4@vger.kernel.org>);
|
|
|
32e0cc2 |
Tue, 7 Feb 2012 18:07:48 -0500
|
|
|
32e0cc2 |
Original-Received: from int-mx01.intmail.prod.int.phx2.redhat.com (int-mx01.intmail.prod.int.phx2.redhat.com [10.5.11.11])
|
|
|
32e0cc2 |
by mx1.redhat.com (8.14.4/8.14.4) with ESMTP id q17N7dj0027622
|
|
|
32e0cc2 |
(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK)
|
|
|
32e0cc2 |
for <linux-ext4@vger.kernel.org>; Tue, 7 Feb 2012 18:07:48 -0500
|
|
|
32e0cc2 |
Original-Received: from liberator.sandeen.net (ovpn01.gateway.prod.ext.phx2.redhat.com [10.5.9.1])
|
|
|
32e0cc2 |
by int-mx01.intmail.prod.int.phx2.redhat.com (8.13.8/8.13.8) with ESMTP id q17M7Kgt001990
|
|
|
32e0cc2 |
(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO)
|
|
|
32e0cc2 |
for <linux-ext4@vger.kernel.org>; Tue, 7 Feb 2012 17:07:21 -0500
|
|
|
32e0cc2 |
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:10.0) Gecko/20120129 Thunderbird/10.0
|
|
|
32e0cc2 |
X-Enigmail-Version: 1.3.5
|
|
|
32e0cc2 |
X-Scanned-By: MIMEDefang 2.67 on 10.5.11.11
|
|
|
32e0cc2 |
Original-Sender: linux-ext4-owner@vger.kernel.org
|
|
|
32e0cc2 |
Precedence: bulk
|
|
|
32e0cc2 |
List-ID: <linux-ext4.vger.kernel.org>
|
|
|
32e0cc2 |
X-Mailing-List: linux-ext4@vger.kernel.org
|
|
|
32e0cc2 |
Xref: news.gmane.org gmane.comp.file-systems.ext4:30623
|
|
|
32e0cc2 |
Archived-At: <http://permalink.gmane.org/gmane.comp.file-systems.ext4/30623>
|
|
|
32e0cc2 |
|
|
|
32e0cc2 |
journal_unmap_buffer()'s zap_buffer: code clears a lot of buffer head
|
|
|
32e0cc2 |
state ala discard_buffer(), but does not touch _Delay or _Unwritten
|
|
|
32e0cc2 |
as discard_buffer() does.
|
|
|
32e0cc2 |
|
|
|
32e0cc2 |
This can be problematic in some areas of the ext4 code which assume
|
|
|
32e0cc2 |
that if they have found a buffer marked unwritten or delay, then it's
|
|
|
32e0cc2 |
a live one. Perhaps those spots should check whether it is mapped
|
|
|
32e0cc2 |
as well, but if jbd2 is going to tear down a buffer, let's really
|
|
|
32e0cc2 |
tear it down completely.
|
|
|
32e0cc2 |
|
|
|
32e0cc2 |
Without this I get some fsx failures on sub-page-block filesystems
|
|
|
32e0cc2 |
up until v3.2, at which point 4e96b2dbbf1d7e81f22047a50f862555a6cb87cb
|
|
|
32e0cc2 |
and 189e868fa8fdca702eb9db9d8afc46b5cb9144c9 make the failures go
|
|
|
32e0cc2 |
away, because buried within that large change is some more flag
|
|
|
32e0cc2 |
clearing. I still think it's worth doing in jbd2, since
|
|
|
32e0cc2 |
->invalidatepage leads here directly, and it's the right place
|
|
|
32e0cc2 |
to clear away these flags.
|
|
|
32e0cc2 |
|
|
|
32e0cc2 |
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
|
|
|
32e0cc2 |
Cc: stable@vger.kernel.org
|
|
|
32e0cc2 |
---
|
|
|
32e0cc2 |
|
|
|
32e0cc2 |
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
|
|
|
32e0cc2 |
index 35ae096..52653306 100644
|
|
|
32e0cc2 |
--- a/fs/jbd2/transaction.c
|
|
|
32e0cc2 |
+++ b/fs/jbd2/transaction.c
|
|
|
32e0cc2 |
@@ -1949,6 +1949,8 @@ zap_buffer_unlocked:
|
|
|
32e0cc2 |
clear_buffer_mapped(bh);
|
|
|
32e0cc2 |
clear_buffer_req(bh);
|
|
|
32e0cc2 |
clear_buffer_new(bh);
|
|
|
32e0cc2 |
+ clear_buffer_delay(bh);
|
|
|
32e0cc2 |
+ clear_buffer_unwritten(bh);
|
|
|
32e0cc2 |
bh->b_bdev = NULL;
|
|
|
32e0cc2 |
return may_free;
|
|
|
32e0cc2 |
}
|
|
|
32e0cc2 |
|
|
|
32e0cc2 |
|
|
|
32e0cc2 |
--
|
|
|
32e0cc2 |
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
|
|
|
32e0cc2 |
the body of a message to majordomo@vger.kernel.org
|
|
|
32e0cc2 |
More majordomo info at http://vger.kernel.org/majordomo-info.html
|
|
|
32e0cc2 |
|