Re: [PATCH 0/2] thread-pool: fix performance regression

[Lukáš Doktor]

Hello Paolo, folks, I gave it a try (on top of the f9fc8932) and it's better 
than the f9fc8932, better than the previous patch by Stefan, but still I'm not 
reaching the performance of d7482ffe97 (before the f9fc8932 commit):

f9f    |  0.0 | -2.8 |  0.6
stefan | -3.1 | -1.2 | -2.2
paolo  |  5.3 |  5.4 |  7.1
d74    |  7.2 |  9.1 |  8.2

Anyway it's definitely closer to the previous baseline (~-2%). Note I have not 
tried other scenarios, just the 4K nbd writes on rotational disk. I'll try 
running more throughout the night.

Regards,
Lukáš

Dne 06. 05. 22 v 13:47 Paolo Bonzini napsal(a):
> Together, these two patches fix the performance regression induced by
> QemuSemaphore; individually they don't though.
> 
> 6.2:
>    iops        : min=58051, max=62260, avg=60282.57, stdev=1081.18, samples=30
>     clat percentiles (usec):   1.00th=[  490],   99.99th=[  775]
>    iops        : min=59401, max=61290, avg=60651.27, stdev=468.24, samples=30
>     clat percentiles (usec):   1.00th=[  490],   99.99th=[  717]
>    iops        : min=59583, max=60816, avg=60353.43, stdev=282.69, samples=30
>     clat percentiles (usec):   1.00th=[  490],   99.99th=[  701]
>    iops        : min=58099, max=60713, avg=59739.53, stdev=755.49, samples=30
>     clat percentiles (usec):   1.00th=[  494],   99.99th=[  717]
> 
> 
> patched:
>    iops        : min=60616, max=62522, avg=61654.37, stdev=555.67, samples=30
>     clat percentiles (usec):   1.00th=[  474],   99.99th=[ 1303]
>    iops        : min=61841, max=63600, avg=62878.47, stdev=442.40, samples=30
>     clat percentiles (usec):   1.00th=[  465],   99.99th=[  685]
>    iops        : min=62976, max=63910, avg=63531.60, stdev=261.05, samples=30
>     clat percentiles (usec):   1.00th=[  461],   99.99th=[  693]
>    iops        : min=60803, max=63623, avg=62653.37, stdev=808.76, samples=30
>     clat percentiles (usec):   1.00th=[  465],   99.99th=[  685]
> 
> Paolo
> 
> Supersedes: <20220505131346.823941-1-pbonzini@redhat.com>
> 
> Paolo Bonzini (2):
>   thread-pool: optimize scheduling of completion bottom half
>   thread-pool: replace semaphore with condition variable
> 
>  util/thread-pool.c | 34 +++++++++++++---------------------
>  1 file changed, 13 insertions(+), 21 deletions(-)
>

OpenPGP_0x26B362E47FCF22C1.asc
Description: OpenPGP public key

OpenPGP_signature
Description: OpenPGP digital signature