Re: [PATCH 1/2] sched/topology: Allow archs to override cpu_smt_mask

From: Srikar Dronamraju
Date: Thu Aug 06 2020 - 12:45:08 EST

* peterz@xxxxxxxxxxxxx <peterz@xxxxxxxxxxxxx> [2020-08-06 10:54:29]:

> On Thu, Aug 06, 2020 at 03:32:25PM +1000, Michael Ellerman wrote:
> > That brings with it a bunch of problems, such as existing software that
> > has been developed/configured for Power8 and expects to see SMT8.
> >
> > We also allow LPARs to be live migrated from Power8 to Power9 (and back), so
> > maintaining the illusion of SMT8 is considered a requirement to make that work.
> So how does that work if the kernel booted on P9 and demuxed the SMT8
> into 2xSMT4? If you migrate that state onto a P8 with actual SMT8 you're
> toast again.

To add to what Michael already said, the reason we don't expose the demux of
SMT8 into 2xSMT4 to userspace, is to make the userspace believe they are on
a SMT8. When the kernel is live migrated from P8 to P9, till the time of reboot
they would only have the older P8 topology. After reboot the kernel topology
would change, but the userspace is made to believe that they are running on
SMT8 core by way of keeping the sibling_cpumask at SMT8 core level.

Thanks and Regards
Srikar Dronamraju