From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Nicholas A. Bellinger" Subject: Re: [PATCH] scsi, mptsas : drop scsi_host lock when calling mptsas_qcmd Date: Thu, 16 Sep 2010 14:34:07 -0700 Message-ID: <1284672847.13344.97.camel@haakon2.linux-iscsi.org> References: <1284666254.7280.54.camel@schen9-DESK> <1284670136.13344.93.camel@haakon2.linux-iscsi.org> <20100916212530.GA22051@gargoyle.ger.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from smtp104.sbc.mail.re3.yahoo.com ([66.196.96.80]:24132 "HELO smtp104.sbc.mail.re3.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1752677Ab0IPViR (ORCPT ); Thu, 16 Sep 2010 17:38:17 -0400 In-Reply-To: <20100916212530.GA22051@gargoyle.ger.corp.intel.com> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Andi Kleen Cc: Tim Chen , Eric Moore , linux-scsi@vger.kernel.org, vasu.dev@intel.com, willy@linux.intel.com, Mike Christie , Hannes Reinecke On Thu, 2010-09-16 at 23:25 +0200, Andi Kleen wrote: > > I asked James about getting Vasu's unlocked_qcmds=1 patch merged, but he > > convinced me that doing conditional locking while is very simple, is not > > the proper way for getting this resolved in mainline code. I think in > > the end this will require a longer sit down to do a wholesale conversion > > of all existing SCSI LLD drivers, and identifing the broken ones that > > still need a struct Scsi_Host->host_lock'ed SHT->queuecommand() for > > whatever strange & legacy reasons. > > The standard way to do that would be to first move the lock down > into the drivers (similar to how it has been done with the BKL). > This would be a fairly mechanic mindless patch. Lots of typing, > but not really a lot of real code review needed. > > Then next step the drivers who know they don't want it can remove it. Well, the main bit is finding those drives that actually call some form of struct Scsi_Host->host_lock unlock() -> do_work() -> lock() from within their own ->queuecomamnd() caller, and avoid running into the big ugly deadlock that would happen here.. I know that Open-iSCSi and my own iSCSI Initiator do this "optimization", but I am not sure which other LLDs also do this as well.. Best, --nab