Re: [PATCH v1 06/11] configure: disable MTTCG for MIPS guests

[Aleksandar Markovic]

20:17 Pon, 23.03.2020. Richard Henderson <address@hidden> је написао/ла:
>
> On 3/23/20 9:15 AM, Alex Bennée wrote:
> > While debugging check-acceptance failures I found an instability in
> > the mips64el test case. Briefly the test case:
> >
> >   retry.py -n 100 -c -- ./mips64el-softmmu/qemu-system-mips64el \
> >     -display none -vga none -serial mon:stdio \
> >     -machine malta -kernel ./vmlinux-4.7.0-rc1.I6400 \
> >     -cpu I6400 -smp 8 -vga std \
> >     -append "printk.time=0 clocksource=GIC console=tty0 console=ttyS0 panic=-1" \
> >     --no-reboot
> >
> > Reports about a 9% failure rate:
> >
> >   Results summary:
> >   0: 91 times (91.00%), avg time 5.547 (0.45 varience/0.67 deviation)
> >   -6: 9 times (9.00%), avg time 3.394 (0.02 varience/0.13 deviation)
> >   Ran command 100 times, 91 passes
> >
> > When re-run with "--accel tcg,thread=single" the instability goes
> > away.
> >
> >   Results summary:
> >   0: 100 times (100.00%), avg time 17.318 (249.76 varience/15.80 deviation)
> >   Ran command 100 times, 100 passes
> >
> > Which seems to indicate there is some aspect of the MIPS MTTCG fixes
> > that has been missed. Ideally we would fix that but I'm afraid I don't
> > have time to investigate and am not super familiar with the
> > architecture anyway. In lieu of someone tracking down the failure lets
> > disable it for now.
> >
> > Signed-off-by: Alex Bennée <address@hidden>
> > Acked-by: Philippe Mathieu-Daudé <address@hidden>
> > Cc: Aleksandar Markovic <address@hidden>
> > Cc: Aurelien Jarno <address@hidden>
> > Cc: Aleksandar Rikalo <address@hidden>
>
> Acked-by: Richard Henderson <address@hidden>
>

Reviewed-by: Aleksandar Markovic <aleksandar.qem address@hidden>

But, Alex, I expect this patch will go through your queue, not MIPS queue (unless you told me otherwise).

Thanks,
Aleksandar

>
> r~