From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1MsgQr-0000Pb-Tp for qemu-devel@nongnu.org; Tue, 29 Sep 2009 13:24:10 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1MsgQn-0000MK-Eh for qemu-devel@nongnu.org; Tue, 29 Sep 2009 13:24:09 -0400 Received: from [199.232.76.173] (port=59256 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1MsgQn-0000MC-3c for qemu-devel@nongnu.org; Tue, 29 Sep 2009 13:24:05 -0400 Received: from mail-qy0-f173.google.com ([209.85.221.173]:52453) by monty-python.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1MsgQm-0007qZ-Kj for qemu-devel@nongnu.org; Tue, 29 Sep 2009 13:24:04 -0400 Received: by qyk3 with SMTP id 3so4116705qyk.4 for ; Tue, 29 Sep 2009 10:24:04 -0700 (PDT) Message-ID: <4AC242B0.7060305@codemonkey.ws> Date: Tue, 29 Sep 2009 12:24:00 -0500 From: Anthony Liguori MIME-Version: 1.0 References: <20090907181436.GA8538@redhat.com> <4AA60A58.4090703@redhat.com> <20090908075831.GA9875@redhat.com> <200909212039.01126.rusty@rustcorp.com.au> <4AB7A01A.3000206@redhat.com> <4AB8992B.7070709@redhat.com> <4AB8DD53.7070806@redhat.com> <4AB8DECA.3090908@redhat.com> <4AB8E88C.4040103@redhat.com> <4AB980E6.2070203@codemonkey.ws> <4AB9AA8E.7060800@third-harmonic.com> <4AC1A49A.1010308@redhat.com> <4AC210FB.8030204@codemonkey.ws> <4AC2342C.4070101@redhat.com> In-Reply-To: <4AC2342C.4070101@redhat.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: [Qemu-devel] Re: [PATCH 0/2] fix virtio_blk serial pci config breakage List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Avi Kivity Cc: john cooper , "Michael S. Tsirkin" , john cooper , Rusty Russell , qemu-devel@nongnu.org, Vadim Rozenfeld , jens.axboe@oracle.com Avi Kivity wrote: > On 09/29/2009 03:51 PM, Anthony Liguori wrote: >>> The change this patch implements is passing of the >>> identify data through a mapping established by >>> PCI BAR #5 rather than the PCI config area, the >>> latter of which resulted in the above breakage. >> >> >> This is a massive layering violation. The virtio-blk ABI cannot make >> demands of the transport. > > True. > >> >> The better solution would be to move the entire virtio-pci config >> space to a separate BAR that's an MMIO region. Then there is no >> practical limit on the size of the config area. > > Don't some fast-paths accesses go through the config space? Using > mmio will slow them down. Not that I know of. Regards, Anthony Liguori