From mboxrd@z Thu Jan 1 00:00:00 1970 From: George Dunlap Subject: Re: Early ACPI events prevent subsequent ACPI functionality on xen 4.3 + HVM domU Date: Wed, 22 May 2013 12:50:50 +0100 Message-ID: <519CB11A.5030907@eu.citrix.com> References: <752956577.8634488.1368537190555.JavaMail.root@zimbra002> <86B77960F7BEC777A3C5C339@Ximines.local> <519B947D.8030607@eu.citrix.com> <519B9801.3080901@eu.citrix.com> <20130521164531.GE3669@phenom.dumpdata.com> <6D7B8289F0F8A65C8C2A25C7@Ximines.local> <1809930721.71216.1369219520131.JavaMail.root@flexiant.com> <519CA407.7000609@eu.citrix.com> <8AFD571E8A65CC3529A5604A@Ximines.local> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <8AFD571E8A65CC3529A5604A@Ximines.local> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Alex Bligh Cc: Ian Campbell , Konrad Rzeszutek Wilk , xen-devel@lists.xen.org, Paul Durrant , Stefano Stabellini , Anthony Perard , Diana Crisan , Dave Scott List-Id: xen-devel@lists.xenproject.org On 22/05/13 12:16, Alex Bligh wrote: > George, > > --On 22 May 2013 11:55:03 +0100 George Dunlap > wrote: > >>> I have tested both cases with Ubuntu. >>> Sending the trigger is reliable if you wait for boot to fully complete. >>> However, if I issue it during boot it does not get executed. Any >>> subsequent triggers do not get executed as well until one is sent when >>> the vm has fully booted. >> >> Right -- so what I hear you saying is, "ACPI commands issued before the >> OS is paying attention are ignored." I think that's expected behavior. >> >> I can see that "Shut this vm down as soon as possible" is a useful thing >> to have. The problem at the moment guest OSes can ignore signals sent >> too >> early, it doesn't really seem within the scope of libxl to work around >> that. > > I (now) think xl trigger power is working as well as could be expected. > xl shutdown has a problem (in that an early use of this > breaks later use), but we can avoid that by using xl trigger power > (or rather the libxl equivalent). So we have a workaround for this one. OK -- this will still be on our list of bugs to track, but as it's probably a linux issue, it may take some time to filter back into distributions. > FWIW our workaround actually will be send xl trigger power every > second until the machine disappears, as there is in general no IP > connectivity between hypervisor and guest in our environment. (Just > in case anyone is reading the archive and wants a hint). Ah right -- that makes sense. Poking it until it shuts off it a bit blunt, but should be effective. :-) -George