]> www.infradead.org Git - users/jedix/linux-maple.git/commit
selftests/resctrl: Use cache size to determine "fill_buf" buffer size
authorReinette Chatre <reinette.chatre@intel.com>
Thu, 24 Oct 2024 21:18:49 +0000 (14:18 -0700)
committerShuah Khan <skhan@linuxfoundation.org>
Tue, 5 Nov 2024 00:02:03 +0000 (17:02 -0700)
commitf77b9672536e581c945b2623b521a284fdbf75ff
tree7c9055082837b5ede6a1ba98b2ecaabfe66610e7
parent3cb3f0b8755919648281064d44c06319743db343
selftests/resctrl: Use cache size to determine "fill_buf" buffer size

By default the MBM and MBA tests use the "fill_buf" benchmark to
read from a buffer with the goal to measure the memory bandwidth
generated by this buffer access.

Care should be taken when sizing the buffer used by the "fill_buf"
benchmark. If the buffer is small enough to fit in the cache then
it cannot be expected that the benchmark will generate much memory
bandwidth. For example, on a system with 320MB L3 cache the existing
hardcoded default of 250MB is insufficient.

Use the measured cache size to determine a buffer size that can be
expected to trigger memory access while keeping the existing default
as minimum, now renamed to MINIMUM_SPAN, that has been appropriate for
testing so far.

Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
tools/testing/selftests/resctrl/fill_buf.c
tools/testing/selftests/resctrl/mba_test.c
tools/testing/selftests/resctrl/mbm_test.c
tools/testing/selftests/resctrl/resctrl.h
tools/testing/selftests/resctrl/resctrl_tests.c