From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chris J Arges Subject: Re: [PATCH] kvm-unit-tests: x86: pmu: call measure for every counter in check_counters_many Date: Mon, 25 Aug 2014 14:38:03 -0500 Message-ID: <53FB909B.4030507@canonical.com> References: <1408049924-18848-1-git-send-email-chris.j.arges@canonical.com> <53FB6832.6080800@redhat.com> <53FB76AB.80206@canonical.com> <1883019063.35938857.1408995147601.JavaMail.zimbra@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org To: Paolo Bonzini Return-path: Received: from youngberry.canonical.com ([91.189.89.112]:60577 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751127AbaHYTiH (ORCPT ); Mon, 25 Aug 2014 15:38:07 -0400 In-Reply-To: <1883019063.35938857.1408995147601.JavaMail.zimbra@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 08/25/2014 02:32 PM, Paolo Bonzini wrote: >> Ok I see now where this patch doesn't make sense. >> With the latest kvm tree I get: >> >> sudo ./x86-run x86/pmu.flat -smp 1 -cpu host | grep -v PASS >> >> >> qemu-system-x86_64 -enable-kvm -device pc-testdev -device >> isa-debug-exit,iobase=0xf4,iosize=0x4 -display none -serial stdio >> -device pci-testdev -kernel x86/pmu.flat -smp 1 -cpu host >> enabling apic >> paging enabled >> cr0 = 80010011 >> cr3 = 7fff000 >> cr4 = 20 >> PMU version: 2 >> GP counters: 4 >> GP counter width: 48 >> Mask length: 7 >> Fixed counters: 3 >> Fixed counter width: 48 >> FAIL: all counters >> >> SUMMARY: 67 tests, 1 unexpected failures >> Return value from qemu: 3 >> >> I've tested this on a few Intel platforms (sandybridge/haswell), I'll >> look into the code more then. > > > Are you using the NMI watchdog in the host? It eats one PMU counter > and makes this test fail. > > Paolo > Ah, I didn't know that. Yes disabling NMI watchdog via: echo 0 | sudo tee /proc/sys/kernel/nmi_watchdog Allows this test to pass. Would it make sense to have a check if nmi_watchdog is enabled in this test case, and skip the all counters test? --chris j arges