RE: [Xen-devel] DomU crash during migration when suspending source domain

["Graham, Simon" <Simon.Graham@xxxxxxxxxxx>]

> A broader set of fixes might get accepted upstream because
> cache_add_dev()
> can fail for other reasons too (at least out-of-memory) and any such
> failure
> will cause cache_remove_dev() to barf. But it's not such a simple
thing
> to
> fix and it does not solve the general problem for us.

OK -- after some work today, I have a patch that would perhaps be
acceptable upstream -- I started off looking into ways of ensuring the
information on the cache got reset properly after migration but quickly
decided that this was not going to be at all easy (plus it wouldn't
address the larger issues of processor compatibility), so instead I
concentrated on handling all the cases where pointers might be NULL
because of an allocation failure (which also happens to fix the crash we
see here).

Attached is a patch to intel_cacheinfo-xen.c in unstable -- I basically
found everywhere that you might have a NULL due to allocation failure
and tested for it; I've run this a little and confirmed that it solves
the original crash migrating back and forth between the systems with
different processor families. Will run more extensive regressions on it
tonight and tomorrow, but I thought I'd send it now for review; will
resend with signed-off-by and comment if you think it is OK.

Simon

Attachment: xen-unstable-intel-cachinfo.patch
Description: xen-unstable-intel-cachinfo.patch

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel