From: Eric Sandeen Subject: [PATCH] jbd2: don't wake kjournald unnecessarily Date: Tue, 18 Dec 2012 20:59:54 -0600 Message-ID: <50D12DAA.9050700@redhat.com> References: <50D0A1FD.7040203@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Dave Wysochanski To: ext4 development , Jan Kara Return-path: Received: from mx1.redhat.com ([209.132.183.28]:21860 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751544Ab2LSDB2 (ORCPT ); Tue, 18 Dec 2012 22:01:28 -0500 In-Reply-To: <50D0A1FD.7040203@redhat.com> Sender: linux-ext4-owner@vger.kernel.org List-ID: Don't send an extra wakeup to kjournald in the case where we already have the proper target in j_commit_request, i.e. that transaction has already been requested for commit. commit deeeaf13 "jbd2: fix fsync() tid wraparound bug" changed the logic leading to a wakeup, but it caused some extra wakeups which were found to lead to a measurable performance regression. Signed-off-by: Eric Sandeen --- diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index dbf41f9..89970d5 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -518,7 +518,8 @@ int __jbd2_log_start_commit(journal_t *journal, tid_t target) * currently running transaction (if it exists). Otherwise, * the target tid must be an old one. */ - if (journal->j_running_transaction && + if (journal->j_commit_request != target && + journal->j_running_transaction && journal->j_running_transaction->t_tid == target) { /* * We want a new commit: OK, mark the request and wakeup the