From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:57910) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aAyAg-0002Fz-0U for qemu-devel@nongnu.org; Mon, 21 Dec 2015 05:58:30 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1aAyAb-00043C-VT for qemu-devel@nongnu.org; Mon, 21 Dec 2015 05:58:29 -0500 Received: from mx0.arrikto.com ([212.71.252.59]:33198) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aAyAb-000422-Hp for qemu-devel@nongnu.org; Mon, 21 Dec 2015 05:58:25 -0500 References: <1450284917-10508-1-git-send-email-apyrgio@arrikto.com> <5671AA78.8050603@redhat.com> <5672768E.70101@arrikto.com> <56728F08.7070702@redhat.com> <5672B451.1050302@arrikto.com> <5672B4E6.8000504@redhat.com> From: Alex Pyrgiotis Message-ID: <5677DB4F.6020308@arrikto.com> Date: Mon, 21 Dec 2015 12:58:23 +0200 MIME-Version: 1.0 In-Reply-To: <5672B4E6.8000504@redhat.com> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH 0/9] Add full scatter-gather support for SCSI generic devices List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini Cc: qemu-devel@nongnu.org Hi Paolo, On 12/17/2015 03:13 PM, Paolo Bonzini wrote: > > > On 17/12/2015 14:10, Alex Pyrgiotis wrote: >>>>>> Which commands have large payloads and are on the data path, for >>>>>> scsi-block? Or is the use case just scsi-generic (e.g. tape devices?)? >> >> If I understand correctly, what you're saying is that if "scsi-block" is >> started with "cache=writeback" and internally uses ioctl()s to bypass >> the page cache, why not set "cache=none" beforehand and use >> readv()/writev()? >> >> This is a valid suggestion, but this patch does not target only the >> "scsi-block" device type. Its purpose is to allow faster read/writes via >> ioctl()s, either to a "scsi-block" device or to a "scsi-generic" device. >> Note that the latter device type can only use ioctl()s, so it cannot >> benefit from the readv()/writev() DMA interface and currently has to use >> a bounce buffer. > > Okay, so that answers my questions; there is still a valid use case for > e.g. tape devices, and of course for when someone forgets to use scsi-block. Nice. I'm looking forward to your comments and I'll reply promptly. Thanks, Alex >>> We can improve the code to print a warning if you don't. (It needs some >>> care: iscsi never caches, independent of the cache= argument, so we >>> don't want to warn for it. But it can be done). >> >> I wasn't particularly concerned about that issue. I'd may prefer if this >> was explicitly addressed in the QEMU doc, under the "cache=" section, >> but that's a different discussion. >> >> Thanks, >> Alex >>