On 2/15/22 15:51, Mel Gorman wrote:
> free_pcppages_bulk() has taken two passes through the pcp lists since
> commit 0a5f4e5b4562 ("mm/free_pcppages_bulk: do not hold lock when picking
> pages to free") due to deferring the cost of selecting PCP lists until
> the zone lock is held. Now that list selection is simplier, the main
> cost during selection is bulkfree_pcp_prepare() which in the normal case
> is a simple check and prefetching. As the list manipulations have cost
> in itself, go back to freeing pages in a single pass.
>
> The series up to this point was evaulated using a trunc microbenchmark
> that is truncating sparse files stored in page cache (mmtests config
> config-io-trunc). Sparse files were used to limit filesystem interaction.
>
> The results versus a revert of storing high-order pages in the PCP lists is
>
> 1-socket Skylake
> 5.17.0-rc3 5.17.0-rc3 5.17.0-rc3
> vanilla mm-reverthighpcp-v1r1 mm-highpcpopt-v1
> Min elapsed 540.00 ( 0.00%) 530.00 ( 1.85%) 530.00 ( 1.85%)
> Amean elapsed 543.00 ( 0.00%) 530.00 * 2.39%* 530.00 * 2.39%*
> Stddev elapsed 4.83 ( 0.00%) 0.00 ( 100.00%) 0.00 ( 100.00%)
> CoeffVar elapsed 0.89 ( 0.00%) 0.00 ( 100.00%) 0.00 ( 100.00%)
> Max elapsed 550.00 ( 0.00%) 530.00 ( 3.64%) 530.00 ( 3.64%)
> BAmean-50 elapsed 540.00 ( 0.00%) 530.00 ( 1.85%) 530.00 ( 1.85%)
> BAmean-95 elapsed 542.22 ( 0.00%) 530.00 ( 2.25%) 530.00 ( 2.25%)
> BAmean-99 elapsed 542.22 ( 0.00%) 530.00 ( 2.25%) 530.00 ( 2.25%)
>
> 2-socket CascadeLake
> 5.17.0-rc3 5.17.0-rc3 5.17.0-rc3
> vanilla mm-reverthighpcp-v1 mm-highpcpopt-v1
> Min elapsed 510.00 ( 0.00%) 500.00 ( 1.96%) 500.00 ( 1.96%)
> Amean elapsed 529.00 ( 0.00%) 521.00 ( 1.51%) 516.00 * 2.46%*
> Stddev elapsed 16.63 ( 0.00%) 12.87 ( 22.64%) 9.66 ( 41.92%)
> CoeffVar elapsed 3.14 ( 0.00%) 2.47 ( 21.46%) 1.87 ( 40.45%)
> Max elapsed 550.00 ( 0.00%) 540.00 ( 1.82%) 530.00 ( 3.64%)
> BAmean-50 elapsed 516.00 ( 0.00%) 512.00 ( 0.78%) 510.00 ( 1.16%)
> BAmean-95 elapsed 526.67 ( 0.00%) 518.89 ( 1.48%) 514.44 ( 2.32%)
> BAmean-99 elapsed 526.67 ( 0.00%) 518.89 ( 1.48%) 514.44 ( 2.32%)
>
> The original motivation for multi-passes was will-it-scale page_fault1
> using $nr_cpu processes.
>
> 2-socket CascadeLake (40 cores, 80 CPUs HT enabled)
> 5.17.0-rc3 5.17.0-rc3
> vanilla mm-highpcpopt-v1r4
> Hmean page_fault1-processes-2 2694662.26 ( 0.00%) 2696801.07 ( 0.08%)
> Hmean page_fault1-processes-5 6425819.34 ( 0.00%) 6426573.21 ( 0.01%)
> Hmean page_fault1-processes-8 9642169.10 ( 0.00%) 9647444.94 ( 0.05%)
> Hmean page_fault1-processes-12 12167502.10 ( 0.00%) 12073323.10 * -0.77%*
> Hmean page_fault1-processes-21 15636859.03 ( 0.00%) 15587449.50 * -0.32%*
> Hmean page_fault1-processes-30 25157348.61 ( 0.00%) 25111707.15 * -0.18%*
> Hmean page_fault1-processes-48 27694013.85 ( 0.00%) 27728568.63 ( 0.12%)
> Hmean page_fault1-processes-79 25928742.64 ( 0.00%) 25920933.41 ( -0.03%) <---
> Hmean page_fault1-processes-110 25730869.75 ( 0.00%) 25695727.57 * -0.14%*
> Hmean page_fault1-processes-141 25626992.42 ( 0.00%) 25675346.68 * 0.19%*
> Hmean page_fault1-processes-172 25611651.35 ( 0.00%) 25650940.14 * 0.15%*
> Hmean page_fault1-processes-203 25577298.75 ( 0.00%) 25584848.65 ( 0.03%)
> Hmean page_fault1-processes-234 25580686.07 ( 0.00%) 25601794.52 * 0.08%*
> Hmean page_fault1-processes-265 25570215.47 ( 0.00%) 25553191.25 ( -0.07%)
> Hmean page_fault1-processes-296 25549488.62 ( 0.00%) 25530311.58 ( -0.08%)
> Hmean page_fault1-processes-320 25555149.05 ( 0.00%) 25585059.83 ( 0.12%)
>
> The differences are mostly within the noise and the difference close to
> $nr_cpus is negligible.
>
> Signed-off-by: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Reviewed-by: Vlastimil Babka <vbabka@xxxxxxx>
