On 26/04/2019 17:22, Julien Grall wrote: > Hi Dario, > > On 26/04/2019 16:13, Dario Faggioli wrote: >> On Tue, 2019-04-23 at 12:13 +0300, Andrii Anisov wrote: >>> Hello Dario, >>> >>> On 20.04.19 18:24, Dario Faggioli wrote: >>>> In schedule(), if we pick, as the next vcpu to run (next) the same >>>> one >>>> that is running already (prev), we never get to call >>>> context_switch(). >>> >>> And what about `if ( prev != current )` in >>> arm/domain.c:schedule_tail() ? >>> >> You're suggesting that's redundant too, aren't you? >> >> From a quick look, it seems to me as well that it would be. I'm ok >> taking care of it, not sure if in this patch, or in another one. > > I thought Andrew already queued this patch?
I did. > > Anyway, I think you are right. We can drop the check in > schedule_tail() as well. That should really be a separate patch, because it is ARM-specific scheduling behaviour. The same is not true on x86. ~Andrew _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
