From mboxrd@z Thu Jan 1 00:00:00 1970 From: Joel Becker Date: Tue, 8 Jul 2008 11:17:06 -0700 Subject: [Ocfs2-devel] [PATCH 3/4] ocfs2: Fixes tiny race between mount and recovery In-Reply-To: <48739F16.7000205@oracle.com> References: <1215462380-12975-1-git-send-email-sunil.mushran@oracle.com> <1215462380-12975-4-git-send-email-sunil.mushran@oracle.com> <20080707205037.GB8651@mail.oracle.com> <4872C7A4.1010507@oracle.com> <48739F16.7000205@oracle.com> Message-ID: <20080708181706.GA13557@mail.oracle.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ocfs2-devel@oss.oracle.com On Tue, Jul 08, 2008 at 10:08:38AM -0700, Sunil Mushran wrote: > Instead of using generation, we could use ij_pad. Renamed to > something appropriate. I like that, as it is specific to the journal part and leaves i_generation alone. Can we see ourselves needing any other use of that word? Joel > Sunil Mushran wrote: >> Possible Solution: >> Use the i_generation in the journal inode as a recovery counter. >> Meaning, we increment it each time we recover the journal. That >> is, from mount thread in ocfs2_journal_load() and from recovery >> thread in ocfs2_replay_journal(). In both places, we have the >> superblock lock making the writes safe. -- "When ideas fail, words come in very handy." - Goethe Joel Becker Principal Software Developer Oracle E-mail: joel.becker at oracle.com Phone: (650) 506-8127