From: Matthias Prager Subject: Re: mpt2sas + raid10 goes boom Date: Tue, 09 Apr 2013 14:45:48 +0200 Message-ID: <51640D7C.9060106@matthiasprager.de> References: <5162C25C.3090707@matthiasprager.de> <5162F827.7010603@matthiasprager.de> <5163E941.1050907@matthiasprager.de> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Cc: Matthias Prager , linux-ext4@vger.kernel.org, Chris Dunlop To: Baruch Even , linux-scsi , linux-raid Return-path: Received: from dd15408.kasserver.com ([85.13.136.168]:54666 "EHLO dd15408.kasserver.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932926Ab3DIMp5 (ORCPT ); Tue, 9 Apr 2013 08:45:57 -0400 In-Reply-To: <5163E941.1050907@matthiasprager.de> Sender: linux-ext4-owner@vger.kernel.org List-ID: Hello everyone, an update: I was able to reproduce the problem on my testing machine (at least sort of) and confirmed that c8dc9c6 md: raid1,10: Handle REQ_WRITE_SAME flag in write bios fixes things. Also applied c8dc9c6 to the main system's 3.8.6 kernel. Working without any issues. One interesting detail: on the testing system's 9240-8i lsi controller (megaraid_sas instead of mpt2sas driver) the log messages where different and seemingly not coming from controller errors: Apr 9 12:59:17 kernel: bio too big device sdb (4096 > 256) Apr 9 12:59:17 kernel: md12: WRITE SAME failed. Manually zeroing. Whereas the lsi 9211-8i gave these messages: [ 2772.726292] mpt2sas0: log_info(0x31120320): originator(PL), code(0x12), sub_code(0x0320) [ 2772.726296] mpt2sas0: log_info(0x31120320): originator(PL), code(0x12), sub_code(0x0320) [ 2772.940873] mpt2sas0: log_info(0x31120320): originator(PL), code(0x12), sub_code(0x0320) [ 2773.205568] mpt2sas0: log_info(0x31120320): originator(PL), code(0x12), sub_code(0x0320) [ 2773.953718] mpt2sas0: log_info(0x31120320): originator(PL), code(0x12), sub_code(0x0320) [ 2774.203121] mpt2sas0: log_info(0x31120320): originator(PL), code(0x12), sub_code(0x0320) [ 2774.452462] mpt2sas0: log_info(0x31120320): originator(PL), code(0x12), sub_code(0x0320) [ 2774.452476] sd 0:0:4:0: [sde] Unhandled error code [ 2774.452479] sd 0:0:4:0: [sde] [ 2774.452480] Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK [ 2774.452482] sd 0:0:4:0: [sde] CDB: [ 2774.452483] Write(10): 2a 00 a9 84 24 08 00 00 08 00 [ 2774.452491] end_request: I/O error, dev sde, sector 2844009480 [ 2774.452636] md3: WRITE SAME failed. Manually zeroing. The former one looks a lot more friendly to me. @lsi guys: this reveals a bug in the mpt2sas driver and/or the 9211&co series firmware, correct? Thanks everyone for your help so far, Matthias