On Fri, Apr 02, 2010 at 11:54:59AM -0400, Laine Stump wrote:
> On 04/01/2010 12:22 PM, Daniel Veillard wrote:
> >@@ -5928,6 +6096,29 @@ static int qemudDomainStart(virDomainPtr dom) {
> > goto endjob;
> > }
> >
> >+ /*
> >+ * If there is a managed saved state restore it instead of starting
> >+ * from scratch. In any case the old state is removed.
> >+ */
> >+ managed_save = qemuDomainManagedSavePath(driver, vm);
> >+ if ((managed_save)&& (virFileExists(managed_save))) {
> >+ virDomainObjUnlock(vm);
> >+ qemuDriverUnlock(driver);
> >+ ret = qemudDomainRestore(dom->conn, managed_save);
>
> There is problem here.
>
> qemudDomainRestore will call qemuDomainObjBeginJobWithDriver, which
> was already called just above this bit of code in qemudDomainStart.
> The result will be a temporary deadlock, while the semaphore wait
> times out, followed by failure.
>
> Either qemudDomainStart needs to call that function later, or else
> it needs to call qemuDomainObjEndJob before calling
> qemudDomainRestore.
>
> (Unfortunately, I haven't yet had the time to do any more of a
> review other than build and try running the code.)
Right, spot on !
With the fix and the restore patch from Chris the cycle seems to work
just fine for me now :-)
I will integrate the feedback and update the patches,
thanks !
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
daniel@xxxxxxxxxxxx | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library http://libvirt.org/
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list