From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: [PATCH RESEND 0/1] AHCI: Optimize interrupt processing Date: Mon, 29 Jul 2013 08:03:25 -0600 Message-ID: <51F6762D.1040906@kernel.dk> References: <1374215660.7397.1041.camel@haakon3.risingtidesystems.com> <1374248000.2266.20.camel@dabdike> <1374267684.7397.1058.camel@haakon3.risingtidesystems.com> <1374296162.7397.1098.camel@haakon3.risingtidesystems.com> <20130722150359.GA16564@dhcp-26-207.brq.redhat.com> <1374527436.7397.1145.camel@haakon3.risingtidesystems.com> <20130725101641.GB31994@dhcp-26-207.brq.redhat.com> <1374790082.7397.1411.camel@haakon3.risingtidesystems.com> <20130726020928.GL29296@kernel.dk> <1374873276.7397.1512.camel@haakon3.risingtidesystems.com> <20130729114653.GB20951@mtj.dyndns.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from merlin.infradead.org ([205.233.59.134]:43688 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755727Ab3G2OEb (ORCPT ); Mon, 29 Jul 2013 10:04:31 -0400 In-Reply-To: <20130729114653.GB20951@mtj.dyndns.org> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Tejun Heo Cc: "Nicholas A. Bellinger" , Alexander Gordeev , James Bottomley , Mike Christie , linux-kernel@vger.kernel.org, linux-ide@vger.kernel.org, Jeff Garzik , linux-scsi On 07/29/2013 05:46 AM, Tejun Heo wrote: > Hello, > > On Fri, Jul 26, 2013 at 02:14:36PM -0700, Nicholas A. Bellinger wrote: >> So I don't think (completely) getting rid of ata_port->qcmds[] will be >> possible, and just converting the ata_scsi_queuecmd() path to use the >> extra SHT->cmd_size pre-allocation for *ata_queued_cmd might end up >> being more trouble that it's worth. Still undecided on that part.. >> >> Tejun, do you have any thoughts + input here..? > > libata exception handling which includes probing doesn't go through > SCSI at all. It all works inside libata proper using ata_queuecmds > and only the result is exposed to SCSI. Most of those SCSI semantics > need to be emulated anyway, so this makes things a lot easier than > going through SCSI for each command. As it currently stands, it'd be > a lot of effort to try to embed ata_qc's into higher layer construct. > Given how it's used, I don't think it's a high priority task. > > One thing which would probably be worthwhile tho is getting rid of the > bitmap based qc tag allocator in libata. That one is just borderline > stupid to keep around on any setup which is supposed to be scalable. Your border might be wider than mine :-). Yes, the bitmap should definitely go. -- Jens Axboe