From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH] KVM: x86: fix vcpu->mmio_fragments overflow Date: Mon, 22 Oct 2012 14:58:23 +0200 Message-ID: <508542EF.5050401@redhat.com> References: <5081033C.4060503@linux.vnet.ibm.com> <20121022091615.GG29310@redhat.com> <50852972.305@linux.vnet.ibm.com> <20121022112314.GO29310@redhat.com> <50852F9C.9020808@siemens.com> <20121022114311.GQ29310@redhat.com> <508531E1.2030307@siemens.com> <508539A8.40404@redhat.com> <50853FF1.8010809@siemens.com> <20121022125301.GS29310@redhat.com> <50854232.8090309@siemens.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Gleb Natapov , Xiao Guangrong , Marcelo Tosatti , LKML , KVM To: Jan Kiszka Return-path: In-Reply-To: <50854232.8090309@siemens.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org On 10/22/2012 02:55 PM, Jan Kiszka wrote: >> Since the userspace change is needed the idea is dead, but if we could >> implement it I do not see how it can hurt the latency if it would be the >> only mechanism to use coalesced mmio buffer. Checking that the ring buffer >> is empty is cheap and if it is not empty it means that kernel just saved >> you a lot of 8 bytes exists so even after iterating over all the entries there >> you still saved a lot of time. > > When taking an exit for A, I'm not interesting in flushing stuff for B > unless I have a dependency. Thus, buffers would have to be per device > before extending their use. Any mmio exit has to flush everything. For example a DMA caused by an e1000 write has to see any writes to the framebuffer, in case the guest is transmitting its framebuffer to the outside world. -- error compiling committee.c: too many arguments to function