From mboxrd@z Thu Jan 1 00:00:00 1970 From: Don Brace Subject: Re: hpsa failure with 4.3.0-rc1 Date: Wed, 30 Sep 2015 09:32:50 -0500 Message-ID: <560BF292.6090209@pmcs.com> References: <560B9202.9080703@suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail-ob0-f175.google.com ([209.85.214.175]:34363 "EHLO mail-ob0-f175.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751698AbbI3Ocw (ORCPT ); Wed, 30 Sep 2015 10:32:52 -0400 Received: by obbda8 with SMTP id da8so32885052obb.1 for ; Wed, 30 Sep 2015 07:32:51 -0700 (PDT) In-Reply-To: <560B9202.9080703@suse.de> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Hannes Reinecke , "linux-scsi@vger.kernel.org" Cc: Joerg Roedel , Borislav Petkov I reported this a little while ago: http://marc.info/?l=linux-kernel&m=144284810906949&w=2 It did clear up between rc1 and rc3. Not sure what changes were applied in-between. On 09/30/2015 02:40 AM, Hannes Reinecke wrote: > Hi all, > > trying to boot 4.3.0-rc1 on a system with hpsa results in a swiotlb > failure: > > hpsa 0000:22:00.0: Logical aborts not supported > hpsa 0000:22:00.0: HP SSD Smart Path aborts not supported > hpsa 0000:22:00.0: swiotlb buffer is full (sz: 786432 bytes) > swiotlb: coherent allocation failed for device 0000:22:00.0 size= > 786432 > CPU: 43 PID: 566 Comm: systemd-udevd Not tainted 4.3.0-rc1-default+ #358 > Hardware name: HP ProLiant BL660c Gen8, BIOS I32 08/20/2012 > ffff88181c314098 ffff8810194cba40 ffffffff8138fad2 00000000000c0 > ffff8810194cba80 ffffffff813ba3d9 ffff881000000008 0000000000000 > 220 > ffff88181c314098 00000000000c0000 ffff8810184101e8 ffff881018410 > 000 > Call Trace: > [] dump_stack+0x4b/0x69 > [] swiotlb_alloc_coherent+0x149/0x160 > [] x86_swiotlb_alloc_coherent+0x3e/0x50 > [] hpsa_init_one+0x915/0x1c90 [hpsa] > [] ? rcu_read_lock_sched_held+0x8c/0xa0 > [] local_pci_probe+0x40/0xa0 > [] ? pci_match_device+0xdb/0x100 > > 4.2.0-rc3 worked fine here. > Needless to say, enabling IOMMU makes the problem go away :-) > > Cheers, > > Hannes