From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hannes Reinecke Subject: Re: [PATCH 3/3] megasas: LSI Megaraid SAS emulation Date: Mon, 04 Jul 2011 08:13:53 +0200 Message-ID: <4E115A21.8010003@suse.de> References: <1309534555-22178-1-git-send-email-hare@suse.de> <1309534555-22178-2-git-send-email-hare@suse.de> <1309534555-22178-3-git-send-email-hare@suse.de> <1309534555-22178-4-git-send-email-hare@suse.de> <4E0F2215.4040307@suse.de> <4E107E66.3030001@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Alexander Graf , qemu-devel@nongnu.org, Stefan Haynoczi , kvm@vger.kernel.org To: Paolo Bonzini Return-path: Received: from cantor2.suse.de ([195.135.220.15]:58749 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753507Ab1GDGNz (ORCPT ); Mon, 4 Jul 2011 02:13:55 -0400 In-Reply-To: <4E107E66.3030001@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 07/03/2011 04:36 PM, Paolo Bonzini wrote: > On 07/02/2011 03:50 PM, Hannes Reinecke wrote: >> (And no, I will not getting into another dog-fight with Paul B. here= =2E >> Virtio can do without bounce buffers. AHCI can. So I fail to see why >> SCSI has to rely on bounce buffers.) > > I agree, but I do see why a SCSI device might prefer to rely on > bounce buffers for non-I/O commands. This is why in my last RFC > series for vmw_pvscsi I let the device choose whether to force a > bounce buffer or get an external iovec from the HBA. > Yes, sure, for non-I/O commands it's perfectly okay. Most of which will be emulated anyway. It's bounce buffers for I/O which kills performance. But I seem to have missed your last RFC (I'm not reading qemu-devel=20 on a regular basis ...). Care to send me a pointer to it? Cheers, Hannes --=20 Dr. Hannes Reinecke zSeries & Storage hare@suse.de +49 911 74053 688 SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 N=FCrnberg GF: J. Hawn, J. Guild, F. Imend=F6rffer, HRB 16746 (AG N=FCrnberg)