Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751434AbWIFQPs (ORCPT ); Wed, 6 Sep 2006 12:15:48 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751412AbWIFQPs (ORCPT ); Wed, 6 Sep 2006 12:15:48 -0400 Received: from e31.co.us.ibm.com ([32.97.110.149]:34537 "EHLO e31.co.us.ibm.com") by vger.kernel.org with ESMTP id S1751385AbWIFQPq (ORCPT ); Wed, 6 Sep 2006 12:15:46 -0400 Subject: Re: [RFC][PATCH] set_page_buffer_dirty should skip unmapped buffers From: Badari Pulavarty To: Jan Kara Cc: Andrew Morton , Anton Altaparmakov , sct@redhat.com, linux-fsdevel , lkml , ext4 In-Reply-To: <20060906153449.GC18281@atrey.karlin.mff.cuni.cz> References: <1157125829.30578.6.camel@dyn9047017100.beaverton.ibm.com> <1157128342.30578.14.camel@dyn9047017100.beaverton.ibm.com> <20060901101801.7845bca2.akpm@osdl.org> <1157472702.23501.12.camel@dyn9047017100.beaverton.ibm.com> <20060906124719.GA11868@atrey.karlin.mff.cuni.cz> <1157555559.23501.25.camel@dyn9047017100.beaverton.ibm.com> <20060906153449.GC18281@atrey.karlin.mff.cuni.cz> Content-Type: text/plain Date: Wed, 06 Sep 2006 09:19:05 -0700 Message-Id: <1157559545.23501.30.camel@dyn9047017100.beaverton.ibm.com> Mime-Version: 1.0 X-Mailer: Evolution 2.0.4 (2.0.4-4) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2549 Lines: 58 On Wed, 2006-09-06 at 17:34 +0200, Jan Kara wrote: > > On Wed, 2006-09-06 at 14:47 +0200, Jan Kara wrote: > > > > > > Andrew, what should we do ? Do you suggest handling this in jbd > > > > itself (like this patch) ? > > > Actually that part of commit code needs rewrite anyway (and after that > > > rewrite you get rid of ll_rw_block()) because of other problems - the > > > code assumes that whenever buffer is locked, it is being written to disk > > > which is not true... I have some preliminary patches for that but they > > > are not very nice and so far I didn't have enough time to find a nice > > > solution. > > > > Are you okay with current not-so-elegant fix ? > Actually I don't quite understand how it can happen what you describe > (so probably I missed something). How it can happen that some buffers > are unmapped while we are committing them? journal_unmap_buffers() > checks whether we are not committing truncated buffers and if so, it > does not do anything to such buffers... > Bye > Honza Yep. I spent lot of time trying to understand - why they are not getting skipped :( But my debug clearly shows that we are clearing the buffer, while we haven't actually submitted to ll_rw_block() code. (I added "track" flag to bh and set it in journal_commit_transaction() when we add them to wbuf[] and clear it in ll_rw_block() after submit. I checked for this flag in journal_unmap_buffer() while clearing the buffer). Here is what my debug shows: buffer is tracked bh ffff8101686ea850 size 1024 Call Trace: [] show_trace+0xb5/0x370 [] dump_stack+0x15/0x20 [] journal_invalidatepage+0x314/0x3b0 [] ext3_invalidatepage+0x38/0x40 [] do_invalidatepage+0x20/0x30 [] truncate_complete_page+0x23/0x50 [] truncate_inode_pages_range+0xcd/0x300 [] truncate_inode_pages+0x10/0x20 [] vmtruncate+0x5f/0x100 [] inode_setattr+0x30/0x140 [] ext3_setattr+0x1bb/0x230 [] notify_change+0x15e/0x320 [] do_truncate+0x53/0x80 [] sys_ftruncate+0xf8/0x130 [] system_call+0x7e/0x83 Thanks, Badari - 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/