Re: [PATCH v5 1/2] powerpc/vcpu: Assume dedicated processors as non-preempt

Srikar Dronamraju [off-list ref] writes:

* Michael Ellerman [off-list ref] [2019-12-13 14:50:35]:

quoted

Waiman Long suggested using static_keys.

Fixes: 247f2f6f3c70 ("sched/core: Don't schedule threads on pre-empted vCPUs")
Cc: stable@vger.kernel.org # v4.18+
Reported-by: Parth Shah <redacted>
Reported-by: Ihor Pasichnyk <redacted>
Tested-by: Juri Lelli <juri.lelli@redhat.com>
Acked-by: Waiman Long <longman@redhat.com>
Reviewed-by: Gautham R. Shenoy <redacted>
Signed-off-by: Srikar Dronamraju <redacted>
Acked-by: Phil Auld <redacted>
Reviewed-by: Vaidyanathan Srinivasan <redacted>
Tested-by: Parth Shah <redacted>
[mpe: Move the key and setting of the key to pseries/setup.c]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 arch/powerpc/include/asm/spinlock.h    | 4 +++-
 arch/powerpc/platforms/pseries/setup.c | 7 +++++++
 2 files changed, 10 insertions(+), 1 deletion(-)

Tested with your version of the patch and its working as expected

Thanks, I took the values below and put them in the changelog for the
patched case.

cheers

Latency percentiles (usec)
        50.0th: 45
        75.0th: 63
        90.0th: 74
        95.0th: 78
        *99.0th: 82
        99.5th: 83
        99.9th: 86
        min=0, max=96
Latency percentiles (usec)
        50.0th: 46
        75.0th: 64
        90.0th: 75
        95.0th: 79
        *99.0th: 83
        99.5th: 85
        99.9th: 91
        min=0, max=117
Latency percentiles (usec)
        50.0th: 46
        75.0th: 64
        90.0th: 75
        95.0th: 79
        *99.0th: 83
        99.5th: 84
        99.9th: 90
        min=0, max=95
Latency percentiles (usec)
        50.0th: 47
        75.0th: 65
        90.0th: 75
        95.0th: 79
        *99.0th: 84
        99.5th: 85
        99.9th: 90
        min=0, max=117
Latency percentiles (usec)
        50.0th: 45
        75.0th: 64
        90.0th: 75
        95.0th: 79
        *99.0th: 82
        99.5th: 83
        99.9th: 93
        min=0, max=111


-- 
Thanks and Regards
Srikar Dronamraju