From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:35151) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1YijfY-0002yo-LO for qemu-devel@nongnu.org; Thu, 16 Apr 2015 09:17:25 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1YijfX-0000tE-QB for qemu-devel@nongnu.org; Thu, 16 Apr 2015 09:17:24 -0400 Message-ID: <552FB652.4040400@redhat.com> Date: Thu, 16 Apr 2015 15:17:06 +0200 From: Paolo Bonzini MIME-Version: 1.0 References: <1429186730-3866-1-git-send-email-pl@kamp.de> <1429186730-3866-6-git-send-email-pl@kamp.de> <552FAE53.7050102@redhat.com> <552FB2DD.9010902@kamp.de> In-Reply-To: <552FB2DD.9010902@kamp.de> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH for-2.4 05/10] block/iscsi: optimize WRITE10/16 if cache.writeback is not set List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Peter Lieven , qemu-devel@nongnu.org Cc: ronniesahlberg@gmail.com, qemu-block@nongnu.org On 16/04/2015 15:02, Peter Lieven wrote: >>> >> Also, I think it is iscsi_co_generic_cb that should set >> force_next_flush, so that it is only set on failure. Not really for the >> optimization value, but because it's clearer. > > I don't get what you mean with it should only "set on failure". > My approach would be to add a flag to the iTask that the command > requires to set force_flush after successful completion. Is it that > what you mean? Set on success. Lack of sleep. Paolo