Re: [PATCH 2/2] KVM: selftests: Use getcpu() instead of sched_getcpu() in rseq_test

[Gavin Shan]

On 8/10/22 6:53 AM, Sean Christopherson wrote:
> On Tue, Aug 09, 2022, Gavin Shan wrote:
> > On 8/9/22 5:17 PM, Florian Weimer wrote:
> > > * Florian Weimer:
> > >
> > > > * Gavin Shan:
> > > >
> > > > > sched_getcpu() is glibc dependent and it can simply return the CPU
> > > > > ID from the registered rseq information, as Florian Weimer pointed.
> > > > > In this case, it's pointless to compare the return value from
> > > > > sched_getcpu() and that fetched from the registered rseq information.
> > > > >
> > > > > Fix the issue by replacing sched_getcpu() with getcpu(), as Florian
> > > > > suggested. The comments are modified accordingly.
> > > >
> > > > Note that getcpu was added in glibc 2.29, so perhaps you need to perform
> > > > a direct system call?
> > >
> > > One more thing: syscall(__NR_getcpu) also has the advantage that it
> > > wouldn't have to be changed again if node IDs become available via rseq
> > > and getcpu is implemented using that.
> > >
> > > Thanks,
> > > Florian
> >
> > Thanks, Florian. It makes sense to me to use syscall(__NR_getcpu) in
> > next revision. Thanks for your quick review :)
>
> +1, and definitely add a comment to prevent future "cleanup".

Yep, I will have something like below in next revision:

    /*
     * We have to perform direct system call for getcpu() because it's not
     * available until glic 2.29.
     */

Thanks,
Gavin