From mboxrd@z Thu Jan 1 00:00:00 1970 From: Julien Grall Subject: Re: [PATCH v6 1/5] xen/arm: observe itargets setting in vgic_enable_irqs and vgic_disable_irqs Date: Tue, 24 Jun 2014 13:07:43 +0100 Message-ID: <53A96A0F.3020006@linaro.org> References: <1403541463-23734-1-git-send-email-stefano.stabellini@eu.citrix.com> <53A866CC.5030201@linaro.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Stefano Stabellini Cc: julien.grall@citrix.com, xen-devel@lists.xensource.com, Ian.Campbell@citrix.com List-Id: xen-devel@lists.xenproject.org On 24/06/14 12:38, Stefano Stabellini wrote: >> Sorry for not having catch this earlier. I don't really like the idea to >> extend the rank lock to vgic_{enable/disable}_irqs. The 2 functions >> can be long to execute as it may touch the GIC distributor. >> >> In another way, the rank lock is only taken in the distributor emulation. >> >> Assuming we consider that distributor access may be slow: > > We could end up enabling or disabling the wrong set of interrupts > without this change. I think it is necessary. AFAIU, this lock only protects the rank when we retrieve the target VCPU, the other part of the function will still work without it. What I meant is to call vgic_get_target_vcpu, so the lock will protect only what is necessary and not more. -- Julien Grall