Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757526AbYB0S5b (ORCPT ); Wed, 27 Feb 2008 13:57:31 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1757241AbYB0S5P (ORCPT ); Wed, 27 Feb 2008 13:57:15 -0500 Received: from mx1.redhat.com ([66.187.233.31]:43133 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757220AbYB0S5N (ORCPT ); Wed, 27 Feb 2008 13:57:13 -0500 From: Josef Bacik To: linux-kernel@vger.kernel.org Subject: [PATCH 1/2] fix the way jbd/jbd2 clears the b_modified flag Date: Wed, 27 Feb 2008 13:46:09 -0500 User-Agent: KMail/1.9.6 (enterprise 0.20071204.744707) Cc: linux-ext4@vger.kernel.org, Jan Kara MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200802271346.10298.jbacik@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5086 Lines: 145 Hello, Currently at the start of a journal commit we loop through all of the buffers on the committing transaction and clear the b_modified flag (the flag that is set when a transaction modifies the buffer) under the j_list_lock. The problem is that everywhere else this flag is modified only under the jbd lock buffer flag, so it will race with a running transaction who could potentially set it, and have it unset by the committing transaction. This is also a big waste, you can have several thousands of buffers that you are clearing the modified flag on when you may not need to. This patch removes this code and instead clears the b_modified flag upon entering do_get_write_access/journal_get_create_access, so if that transaction does indeed use the buffer then it will be accounted for properly, and if it does not then we know we didn't use it. That will be important for the next patch in this series. Tested thoroughly by myself using postmark/iozone/bonnie++. Thank you, Signed-off-by: Josef Bacik Index: linux-2.6/fs/jbd/commit.c =================================================================== --- linux-2.6.orig/fs/jbd/commit.c +++ linux-2.6/fs/jbd/commit.c @@ -407,22 +407,6 @@ void journal_commit_transaction(journal_ jbd_debug (3, "JBD: commit phase 2\n"); /* - * First, drop modified flag: all accesses to the buffers - * will be tracked for a new trasaction only -bzzz - */ - spin_lock(&journal->j_list_lock); - if (commit_transaction->t_buffers) { - new_jh = jh = commit_transaction->t_buffers->b_tnext; - do { - J_ASSERT_JH(new_jh, new_jh->b_modified == 1 || - new_jh->b_modified == 0); - new_jh->b_modified = 0; - new_jh = new_jh->b_tnext; - } while (new_jh != jh); - } - spin_unlock(&journal->j_list_lock); - - /* * Now start flushing things to disk, in the order they appear * on the transaction lists. Data blocks go first. */ Index: linux-2.6/fs/jbd/transaction.c =================================================================== --- linux-2.6.orig/fs/jbd/transaction.c +++ linux-2.6/fs/jbd/transaction.c @@ -609,6 +609,12 @@ repeat: goto done; /* + * this is the first time this transaction is touching this buffer, + * reset the modified flag + */ + jh->b_modified = 0; + + /* * If there is already a copy-out version of this buffer, then we don't * need to make another one */ @@ -820,9 +826,16 @@ int journal_get_create_access(handle_t * if (jh->b_transaction == NULL) { jh->b_transaction = transaction; + + /* first access by this transaction */ + jh->b_modified = 0; + JBUFFER_TRACE(jh, "file as BJ_Reserved"); __journal_file_buffer(jh, transaction, BJ_Reserved); } else if (jh->b_transaction == journal->j_committing_transaction) { + /* first access by this transaction */ + jh->b_modified = 0; + JBUFFER_TRACE(jh, "set next transaction"); jh->b_next_transaction = transaction; } Index: linux-2.6/fs/jbd2/commit.c =================================================================== --- linux-2.6.orig/fs/jbd2/commit.c +++ linux-2.6/fs/jbd2/commit.c @@ -520,22 +520,6 @@ void jbd2_journal_commit_transaction(jou jbd_debug (3, "JBD: commit phase 2\n"); /* - * First, drop modified flag: all accesses to the buffers - * will be tracked for a new trasaction only -bzzz - */ - spin_lock(&journal->j_list_lock); - if (commit_transaction->t_buffers) { - new_jh = jh = commit_transaction->t_buffers->b_tnext; - do { - J_ASSERT_JH(new_jh, new_jh->b_modified == 1 || - new_jh->b_modified == 0); - new_jh->b_modified = 0; - new_jh = new_jh->b_tnext; - } while (new_jh != jh); - } - spin_unlock(&journal->j_list_lock); - - /* * Now start flushing things to disk, in the order they appear * on the transaction lists. Data blocks go first. */ Index: linux-2.6/fs/jbd2/transaction.c =================================================================== --- linux-2.6.orig/fs/jbd2/transaction.c +++ linux-2.6/fs/jbd2/transaction.c @@ -618,6 +618,12 @@ repeat: goto done; /* + * this is the first time this transaction is touching this buffer, + * reset the modified flag + */ + jh->b_modified = 0; + + /* * If there is already a copy-out version of this buffer, then we don't * need to make another one */ @@ -829,9 +835,16 @@ int jbd2_journal_get_create_access(handl if (jh->b_transaction == NULL) { jh->b_transaction = transaction; + + /* first access by this transaction */ + jh->b_modified = 0; + JBUFFER_TRACE(jh, "file as BJ_Reserved"); __jbd2_journal_file_buffer(jh, transaction, BJ_Reserved); } else if (jh->b_transaction == journal->j_committing_transaction) { + /* first access by this transaction */ + jh->b_modified = 0; + JBUFFER_TRACE(jh, "set next transaction"); jh->b_next_transaction = transaction; } -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/