Greeting, FYI, we noticed a -62.3% improvement of kernel-selftests.dma.dma_map_benchmark.average_unmap_latency due to commit: commit: a44f4a480620b3f88e3b16a9407903d06d7704bc ("mm: slub: Move flush_cpu_slab() invocations __free_slab() invocations out of IRQ context") https://git.kernel.org/cgit/linux/kernel/git/vbabka/linux.git slub-local-lock-v2r3 in testcase: kernel-selftests on test machine: 4 threads Intel(R) Core(TM) i5-6500 CPU @ 3.20GHz with 32G memory with following parameters: group: group-01 ucode: 0xe2 test-description: The kernel contains a set of "self tests" under the tools/testing/selftests/ directory. These are intended to be small unit tests to exercise individual code paths in the kernel. test-url: https://www.kernel.org/doc/Documentation/kselftest.txt Details are as below: --------------------------------------------------------------------------------------------------> To reproduce: git clone https://github.com/intel/lkp-tests.git cd lkp-tests bin/lkp install job.yaml # job file is attached in this email bin/lkp split-job --compatible job.yaml # generate the yaml file for lkp run bin/lkp run generated-yaml-file ========================================================================================= compiler/group/kconfig/rootfs/tbox_group/testcase/ucode: gcc-9/group-01/x86_64-rhel-8.3-kselftests/debian-10.4-x86_64-20200603.cgz/lkp-skl-d05/kernel-selftests/0xe2 commit: 5c6ac9a0a6 ("mm, slab: make flush_slab() possible to call with irqs enabled") a44f4a4806 ("mm: slub: Move flush_cpu_slab() invocations __free_slab() invocations out of IRQ context") 5c6ac9a0a606fc17 a44f4a480620b3f88e3b16a9407 ---------------- --------------------------- fail:runs %reproduction fail:runs | | | :13 100% 13:13 dmesg.WARNING:possible_circular_locking_dependency_detected :13 15% 2:13 kernel-selftests.kmod.kmod.sh.pass %stddev %change %stddev \ | \ 0.70 -57.1% 0.30 kernel-selftests.dma.dma_map_benchmark.average_map_latency 1.11 ± 7% -35.4% 0.72 ± 5% kernel-selftests.dma.dma_map_benchmark.average_map_latency_stddev 0.53 ± 8% -62.3% 0.20 kernel-selftests.dma.dma_map_benchmark.average_unmap_latency 0.99 ± 6% -38.8% 0.61 ± 4% kernel-selftests.dma.dma_map_benchmark.average_unmap_latency_stddev 629.91 -10.1% 566.35 kernel-selftests.time.elapsed_time 629.91 -10.1% 566.35 kernel-selftests.time.elapsed_time.max 2163591 +4.8% 2266526 kernel-selftests.time.minor_page_faults 41.62 -16.6% 34.69 kernel-selftests.time.percent_of_cpu_this_job_got 253.43 -26.3% 186.65 kernel-selftests.time.system_time 64020 +18.5% 75863 ± 16% kernel-selftests.time.voluntary_context_switches 287345 ± 6% +24.5% 357646 ± 6% meminfo.VmallocUsed kernel-selftests.time.system_time 300 +---------------------------------------------------------------------+ | | 250 |+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++| | | | | 200 |OOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOO | | O | 150 |-+ OO | | | 100 |-+ | | | | | 50 |-+ | | | 0 +---------------------------------------------------------------------+ kernel-selftests.time.percent_of_cpu_this_job_got 110 +---------------------------------------------------------------------+ | | 100 |-+ | 90 |-+ | | | 80 |-+ | | | 70 |-+ | | | 60 |-+ | 50 |-+ | | | 40 |+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++| |OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOO | 30 +---------------------------------------------------------------------+ kernel-selftests.time.elapsed_time 700 +---------------------------------------------------------------------+ |++++++++++++++++++++++ ++++++++++++ +++++++++++++++++++++++++++++++++| 600 |-O O O O + + + | |O OOO O OOOO OOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOO | 500 |-+ | | OO | 400 |-+ | | | 300 |-+ | | | 200 |-+ | | | 100 |-+ | | | 0 +---------------------------------------------------------------------+ kernel-selftests.time.elapsed_time.max 700 +---------------------------------------------------------------------+ |++++++++++++++++++++++ ++++++++++++ +++++++++++++++++++++++++++++++++| 600 |-O O O O + + + | |O OOO O OOOO OOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOO | 500 |-+ | | OO | 400 |-+ | | | 300 |-+ | | | 200 |-+ | | | 100 |-+ | | | 0 +---------------------------------------------------------------------+ kernel-selftests.time.minor_page_faults 9e+06 +-------------------------------------------------------------------+ | | 8e+06 |-+ | 7e+06 |-+ | | | 6e+06 |-+ | 5e+06 |-+ | | | 4e+06 |-+ | 3e+06 |-+ | | OO OOO OOOO OOOOO OO OO OOOOOO OOOOOOO OOOOO OO OOOOOOOOOO | 2e+06 |+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++| 1e+06 |-+ | | | 0 +-------------------------------------------------------------------+ kernel-selftests.dma.dma_map_benchmark.average_map_latency 0.7 +---------------------------------------------------------------------+ | | 0.6 |-+ | | | 0.5 |-+ | | | 0.4 |O+OOOOO OOOOO O | | | 0.3 |-O O O OOOOOOOOOOOOO OOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOO | | | 0.2 |-+ | | | 0.1 |-+ | | | 0 +---------------------------------------------------------------------+ kernel-selftests.dma.dma_map_benchmark.average_unmap_latency 0.6 +---------------------------------------------------------------------+ |: :::: :::::::::: : : : :: : ::: :::: : : | 0.5 |-++ + + + + ++ ++ +++++++++++++++++ ++++++++++++++ + +++ + +++++ | | | | | 0.4 |-+ | | | 0.3 |-+ | | | 0.2 |OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOO | | | | | 0.1 |-+ | | | 0 +---------------------------------------------------------------------+ kernel-selftests.dma.dma_map_benchmark.average_unmap_latency_stddev 1.2 +---------------------------------------------------------------------+ |+ + + ++ + + ++ ++ + + | 1 |-+++ ++++ + + +++++++ : +++ + : ++: + ++ + : + : ++++ + ++ ++ + +| | :: ::::::: : :: :: : :: :: :: : :: : :: :: ::: : | | + + + +: + + + ++ + + + ++ ++++ + + + ++ | 0.8 |-+ + | | O O O | 0.6 |OOOOOOO OOO O OOOO OOO OO OOOO OOO OOOO OOOOOOOO OOOOOOOO OOOOO | | O OO O O OO OOOO O O | 0.4 |-+ | | | | | 0.2 |-+ | | | 0 +---------------------------------------------------------------------+ [*] bisect-good sample [O] bisect-bad sample Disclaimer: Results have been estimated based on internal Intel analysis and are provided for informational purposes only. Any difference in system hardware or software design or configuration may affect actual performance. --- 0DAY/LKP+ Test Infrastructure Open Source Technology Center https://lists.01.org/hyperkitty/list/lkp@lists.01.org Intel Corporation Thanks, Rong Chen