On Mon, Nov 2, 2020 at 1:24 PM Peter Xu peterx@redhat.com wrote:
On Tue, Oct 27, 2020 at 04:37:29PM -0700, Ben Gardon wrote:
Much of the code in demand_paging_test can be reused by other, similar multi-vCPU-memory-touching-perfromance-tests. Factor that common code out for reuse.
No functional change expected.
Is there explicit reason to put the common code in a header rather than perf_test_util.c? No strong opinion on this especially this is test code, just curious. Since iiuc .c file is still preferred for things like this.
I don't have a compelling reason not to put the common code in a .c file, and I agree that would probably be a better place for it. I'll do that in a v2.
This series was tested by running the following invocations on an Intel Skylake machine: dirty_log_perf_test -b 20m -i 100 -v 64 dirty_log_perf_test -b 20g -i 5 -v 4 dirty_log_perf_test -b 4g -i 5 -v 32 demand_paging_test -b 20m -v 64 demand_paging_test -b 20g -v 4 demand_paging_test -b 4g -v 32 All behaved as expected.
May move this chunk to the cover letter to avoid keeping it in every commit (btw, you mentioned "this series" but I feel like you meant "you verified that after applying each of the commits").
I can move the testing description to the cover letter. I can only definitively say I ran those tests at the last commit in the series, so I'll amend the description to specify that the commits were tested all together and not one by one.
Thanks,
-- Peter Xu