Re: [PATCH-for-4.17] xen/sched: migrate timers to correct cpus after suspend

[Juergen Gross <jgross@xxxxxxxx>]

On 28.10.22 12:22, Marek Marczykowski-Górecki wrote:
> On Fri, Oct 28, 2022 at 10:12:36AM +0000, Andrew Cooper wrote:
> > On 28/10/2022 11:08, Juergen Gross wrote:
> > > On 27.10.22 20:13, Marek Marczykowski-Górecki wrote:
> > >
> > > This is very weird. The data suggests that the scheduling resource
> > > pointer
> > > for cpu 1 was NULL, but I can't see how this can be the case without
> > > causing
> > > similar crashes without this patch.
> > >
> > > Are there any additional patches related to cpu on/offlining or
> > > suspend/resume
> > > in the hypervisor?
> No such patches, it was this:
> https://github.com/marmarek/xen/commits/master-credit-timers
>
> > QubesOS runs with smt=0 by default.  Siblings ought to be parked at this
> > point.
> Yes, indeed this test was with smt=off.
Ah, this is subtle. The CPU_ONLINE notifiers are called for the siblings,
too, resulting in above error.

Preparing V2 of the patch.


Juergen

Attachment: OpenPGP_0xB0DE9DD628BF132F.asc
Description: OpenPGP public key

Attachment: OpenPGP_signature
Description: OpenPGP digital signature