From: Bernd Schubert Subject: Re: kernel panic / NULL pointer dereference Date: Thu, 10 May 2012 18:43:10 +0200 Message-ID: <4FABF01E.7080303@itwm.fraunhofer.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: "linux-fsdevel@vger.kernel.org" To: unlisted-recipients:; (no To-header on input) Return-path: Received: from mailgw1.uni-kl.de ([131.246.120.220]:46767 "EHLO mailgw1.uni-kl.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750903Ab2EJQnQ (ORCPT ); Thu, 10 May 2012 12:43:16 -0400 Received: from itwm2.itwm.fhg.de (itwm2.itwm.fhg.de [131.246.191.3]) by mailgw1.uni-kl.de (8.14.3/8.14.3/Debian-5+lenny1) with ESMTP id q4AGhEhm012391 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NOT) for ; Thu, 10 May 2012 18:43:14 +0200 Received: from mail2.itwm.fhg.de ([131.246.191.79]:45773) by itwm2.itwm.fhg.de with esmtps (TLSv1:DES-CBC3-SHA:168) (/C=DE/ST=Rheinland-Pfalz/L=Kaiserslautern/O=Fraunhofer ITWM/OU=SLG/CN=mail2.itwm.fhg.de)(verified=1) (Exim 4.74 #1) id 1SSWSP-0004jq-F6 for linux-fsdevel@vger.kernel.org; Thu, 10 May 2012 18:43:13 +0200 In-Reply-To: Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On 05/10/2012 05:45 PM, Bernd Schubert wrote: > Hi all, > > I'm just playing with an SRP connected NetApp system and just got an XFS > related kernel panic. I guess it is due to large IO (32MiB). At least it > just came up after enabling 32MiB device max_sectors. > As the tests are running in a RHEL6 image and as I needed at least > 2.6.39 to get a large srp_tablsize with SRP, I simply installed the > lasted oracle uek kernel. If needed I'm going to update to a vanilla > version. > > >> May 10 17:31:49 sgi01 kernel: XFS (sdb): Mounting Filesystem >> May 10 17:31:49 sgi01 kernel: XFS (sdb): Ending clean mount >> May 10 17:33:00 sgi01 kernel: BUG: unable to handle kernel NULL >> pointer dereference at (null) >> May 10 17:33:00 sgi01 kernel: IP: [] >> xfs_alloc_ioend_bio+0x33/0x50 [xfs] Oh, there is a bio allocation path to return NULL: bvec_alloc_bs(gfp_mask, nr_iovecs, ) => NULL when nr_iovecs > BIO_MAX_PAGES bio_alloc_bioset(gfp_mask, nr_iovecs, ...) bio_alloc(GFP_NOIO, nvecs) xfs_alloc_ioend_bio() And nvecs/nr_iovecs is obtained by bio_get_nr_vecs(), which does not check for BIO_MAX_PAGES. Of course, all of that only happens with large IO sizes, which is exactly what I'm doing. As xfs_alloc_ioend_bio() is using GFP_NOIO it does not expect bio_alloc to fail, but as I'm trying to send large IOs I guess that is exactly what happens here. >May 10 17:33:00 sgi01 kernel: [] xfs_submit_ioend+0xfe/0x110 [xfs] >May 10 17:33:00 sgi01 kernel: [] xfs_vm_writepage+0x26b/0x510 [xfs] >May 10 17:33:00 sgi01 kernel: [] __writepage+0x17/0x40 >May 10 17:33:00 sgi01 kernel: [] write_cache_pages+0x246/0x520 >May 10 17:33:00 sgi01 kernel: [] ? set_page_dirty+0x70/0x70 >May 10 17:33:00 sgi01 kernel: [] generic_writepages+0x51/0x80 >May 10 17:33:00 sgi01 kernel: [] xfs_vm_writepages+0x5d/0x80 [xfs] >May 10 17:33:00 sgi01 kernel: [] do_writepages+0x21/0x40 >May 10 17:33:00 sgi01 kernel: [] writeback_single_inode+0x10e/0x270