From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: high load with win7 & usb tablet Date: Fri, 10 Sep 2010 17:56:54 +0300 Message-ID: <4C8A4736.8030704@redhat.com> References: <4C80E839.50604@msgid.tls.msk.ru> <4C8A2254.7090505@redhat.com> <4C8A282A.5010506@msgid.tls.msk.ru> <4C8A29A6.60108@redhat.com> <4C8A44E6.2060008@msgid.tls.msk.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: KVM list To: Michael Tokarev Return-path: Received: from mx1.redhat.com ([209.132.183.28]:1026 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750891Ab0IJO5A (ORCPT ); Fri, 10 Sep 2010 10:57:00 -0400 In-Reply-To: <4C8A44E6.2060008@msgid.tls.msk.ru> Sender: kvm-owner@vger.kernel.org List-ID: On 09/10/2010 05:47 PM, Michael Tokarev wrote: > 10.09.2010 16:50, Avi Kivity wrote: > [] >>> During this shot, host CPU was about 30..35% in use - higher >>> than usual 25%, but I guess it's due to perfmon. >> Please build qemu with --disable-strip and use 'perf' to profile it. > Done, I tried several different options, but I can't find > anything intesting in there. Attaching a screenshot of > 3 - win7 guest (~5 minutes after startup), `perf top' > in xterm and `top' in another xterm. > > Here are random snapshots of both, in text form: > > --------------------------------------------------------------------------------------- > PerfTop: 1409 irqs/sec kernel:79.6% exact: 0.0% [1000Hz cycles], (all, 2 CPUs) > --------------------------------------------------------------------------------------- > > samples pcnt function DSO > _______ _____ _______________________________ _______________________ > > 11837.00 64.9% native_safe_halt [kernel.kallsyms] > 570.00 3.1% rb_next [kernel.kallsyms] > 206.00 1.1% hpet_next_event [kernel.kallsyms] > 149.00 0.8% do_select [kernel.kallsyms] > 149.00 0.8% svm_vcpu_run [kvm_amd] > 134.00 0.7% schedule [kernel.kallsyms] > > > Tasks: 199 total, 2 running, 196 sleeping, 0 stopped, 1 zombie > Cpu(s): 13.4%us, 10.2%sy, 0.0%ni, 76.3%id, 0.0%wa, 0.0%hi, 0.2%si, 0.0%st > Mem: 5860724k total, 4601132k used, 1259592k free, 18508k buffers > Swap: 4192928k total, 286928k used, 3906000k free, 1593908k cached > > PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND > 28492 mjt 20 0 1111m 1.0g 5456 R 27 17.7 3:50.79 qemu-system-x86 > 28763 root 20 0 105m 30m 6176 S 11 0.5 184:08.58 Xorg > 29263 mjt 20 0 637m 212m 11m S 6 3.7 92:08.82 firefox-bin > 28845 mjt 20 0 58612 5840 2372 S 2 0.1 49:27.17 compiz > > I'm still not sure where to look at.. :) > Strange. Can you also post a few lines of 'vmstat 1'? Maybe we'll see a lot of context switches in there. -- I have a truly marvellous patch that fixes the bug which this signature is too narrow to contain.