[PATCH v2] qemuDomainSaveInternal: fix memoryleak of virDomainDef

Michal Privoznik mprivozn at redhat.com
Thu Jul 30 08:06:33 UTC 2020

On 7/29/20 9:51 PM, Laine Stump wrote:

> I thought about not pushing this until after the freeze is over and 
> 6.7.0 is released (since it is a bug, but a bug that's been in the code 
> for 9 years), but in the end decided to push it, because 1) it's a 
> bonafide leak in the  *non*-error path of a libvirt public API, it's 
> very simple, and 3) I would probably forget to push it if I waited until 
> after the freeze is over.

(my apologies for hijacking this thread)

Actually, I just thought about this yesterday. We have these freezes 
which in most cases mean "I will postpone pushing a patch for a week". I 
mean, does anybody do any testing of release candidates and how 
extensive is that? Because what I do is merely some very basic testing 
"virsh start/hotplug/destroy/shutdown". Hopefully, we will have some 
integration testing set up soon at which point, the master will always 
be in release ready state and we won't need much (if any) additional 
testing before making a release.


More information about the libvir-list mailing list