From: Marcelo Tosatti Subject: Re: Ordered mode rewrite patch Date: Fri, 25 Apr 2008 11:23:21 -0300 Message-ID: <20080425142321.GA6843@dmt> References: <20080408121149.GB6901@duck.suse.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Nathan Grennan , linux-ext4@vger.kernel.org To: Jan Kara Return-path: Received: from mx1.redhat.com ([66.187.233.31]:35931 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756653AbYDYOUa (ORCPT ); Fri, 25 Apr 2008 10:20:30 -0400 Content-Disposition: inline In-Reply-To: <20080408121149.GB6901@duck.suse.cz> Sender: linux-ext4-owner@vger.kernel.org List-ID: Hi Jan, On Tue, Apr 08, 2008 at 02:11:49PM +0200, Jan Kara wrote: > Hello, > > attached is a jumbo patch that reverses locking order of transaction > start and page lock in ext3 and rewrites handling of ordered data mode in > JBD and ext3. Note that the patch will break compilation of ext4 and OCFS2. > The patch survives LTP run on my test machine so it shouldn't eat your data > immediately but bugs are of course possible... > I'm very interested in any results (both positive and negative) you could > get with it :). Thanks for testing it. > > Honza > > PS: CCing also linux-ext4 list in case there are some other interested > testers. Next on my todo list is to port this for ext4... No luck for my testcase, similar fsync behaviour taking several seconds: vim D 0000000000000000 0 4088 4036 ffff81017a143da8 0000000000000086 0000000000000000 ffffffff80236808 ffff81022f272830 ffff81022f1ca730 ffff81022f272a70 0000000600000001 00000000ffffffff 0000000000000003 0000000000000000 0000000000000000 Call Trace: [] lock_timer_base+0x26/0x4b [] log_wait_commit+0x9f/0xed [] autoremove_wake_function+0x0/0x2e [] journal_stop+0x165/0x18d [] __writeback_single_inode+0x17f/0x29d [] autoremove_wake_function+0x0/0x2e [] sync_inode+0x24/0x32 [] ext3_sync_file+0x8a/0x9c [] do_fsync+0x52/0xa4 [] __do_fsync+0x23/0x36 [] system_call_after_swapgs+0x7b/0x80