From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cooper Subject: Re: [PATCH] x86/hvm: Fix HVM guest regression introduced by c58ba78c84 Date: Wed, 4 Feb 2015 12:56:21 +0000 Message-ID: <54D216F5.1040907@citrix.com> References: <1423053312-23666-1-git-send-email-andrew.cooper3@citrix.com> <54D2227C020000780005CC1B@mail.emea.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <54D2227C020000780005CC1B@mail.emea.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Jan Beulich Cc: Xen-devel List-Id: xen-devel@lists.xenproject.org On 04/02/15 12:45, Jan Beulich wrote: >>>> On 04.02.15 at 13:35, wrote: >> --- a/xen/arch/x86/hvm/hvm.c >> +++ b/xen/arch/x86/hvm/hvm.c >> @@ -638,7 +638,7 @@ static int hvm_ioreq_server_add_vcpu(struct hvm_ioreq_server *s, >> { >> struct domain *d = s->domain; >> >> - rc = alloc_unbound_xen_event_channel(v->domain, 0, s->domid, NULL); >> + rc = alloc_unbound_xen_event_channel(v->domain, v->vcpu_id, s->domid, NULL); > But this sits in a conditional checking v->vcpu_id == 0. I.e. I don't > see what difference this makes. > > Jan > Hmm so it does. Something between d0b2caa..c58ba78 has broken HVM guests to point at which HVMloader doesn't reliably function. Back to debugging... ~Andrew