From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoffer Dall Subject: Re: [PATCH 0/2] arm/arm64: KVM: Fix arthictected timer issues Date: Fri, 4 Sep 2015 17:53:14 +0200 Message-ID: <20150904155314.GA20138@cbox> References: <1441376679-8341-1-git-send-email-christoffer.dall@linaro.org> <55E9BA38.7030305@arm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: kvmarm@lists.cs.columbia.edu, linux-arm-kernel@lists.infradead.org, kvm@vger.kernel.org To: Marc Zyngier Return-path: Received: from mail-lb0-f182.google.com ([209.85.217.182]:32999 "EHLO mail-lb0-f182.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759024AbbIDPvr (ORCPT ); Fri, 4 Sep 2015 11:51:47 -0400 Received: by lbcjc2 with SMTP id jc2so13659599lbc.0 for ; Fri, 04 Sep 2015 08:51:45 -0700 (PDT) Content-Disposition: inline In-Reply-To: <55E9BA38.7030305@arm.com> Sender: kvm-owner@vger.kernel.org List-ID: On Fri, Sep 04, 2015 at 04:35:20PM +0100, Marc Zyngier wrote: > Hi Christoffer, > > On 04/09/15 15:24, Christoffer Dall wrote: > > These two patches fix two separate issues with the architected timer and > > the corresponding interrupt injection to VMs on KVM/ARM. > > > > The first patch fixes an issue introduced with the active timer state > > switching series recently merged for v4.3, which could cause a guest to > > loop without progress if another VCPU is run on the same physical CPU > > and preempts the original VCPU while the guest is running the ISR for > > the timer interrupt. > > > > The second patch resets the architected timer's control register to zero > > on system reset, ensuring that interrupts are not injected when a system > > resets. This fixes a long-standing issue with UEFI, where soft reset > > initiated from within UEFI prevented the system from booting again. > > Thanks for respinning these patches. I've queued them in our -next > queue. I'll send this to Paolo some time next week, once they get some > hammering. > Awesome, fyi I left these patches running in a loop with reboots and running hackbench on 2 simultaneous VMs with 4 vcpus each on X-gene for 100 iterations, and left the guest and host run cyclictest for ~20 minutes without issues. -Christoffer