Port-xen archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Xen DOM0 pirq panic with lots of cores



On 13.04 22:15, Manuel Bouyer wrote:
On Mon, Apr 13, 2026 at 03:03:01PM -0400, Greg Troxel wrote:
Staffan Thomen <staffan%shangtai.net@localhost> writes:

[...]

The system is not yet in production so I can do tests for a brief time if necessary.

I (who is unlikely to fix this) would be interested in

what is the highest number of cpus that is stable? even roughly from binary search, might be interesting.

Greg:

40 an 20 were too many 10 booted. But this might be highly variable from
how many pseudo-interrupts are needed on the cpus, and these seem to use
a lot. The standard msgbuf is too small to fit all the boot messages
for all the stuff going on.

what is the value of NR_PIRQS in GENERIC and XEN3_DOM0? Is this as simple as a default being different?

NR_PIRQS doens't exists in GENERIC. It's an array mapping physical IRQs to Xen events. An array doesn't scale well on hardware with lots of core; something else is needed here.

No need to rush to fix this on my account, although I suppose this could
be made to dynamically reallocate itself either in chunks or singly as
needed when a new one is discovered? Presumably it's a lookup table that
is generated on boot and then never changes. Then again I know next to
nothing about kernel programming and even less of xen, only a suspicion
that it might be a more restricted environment.

Staffan

Attachment: OpenPGP_signature.asc
Description: OpenPGP digital signature



Home | Main Index | Thread Index | Old Index