Add test showing rdpmc of hardware counters occasionally jumping by 2^48#11
Open
nkurz wants to merge 1 commit into
Open
Add test showing rdpmc of hardware counters occasionally jumping by 2^48#11nkurz wants to merge 1 commit into
nkurz wants to merge 1 commit into
Conversation
I've been having problems using rdpmc to read hardware counters that are set with perf_event_open(). I think the issue is that 'offset' field in the mmap buffer periodically periodically has its high bit set. Here's what I see when I run this test program on Skylake with Linux 4.4.4: nate@skylake:~/git/perf_event_tests/tests/rdpmc$ rdpmc_jump Testing if hardware counters exhibit unexpected jumps with rdpmc. Trying up to 100000 iterations (10000000 allowed per iteration) Got spurious count of 281474977720397 in iteration 230 Max count for previous iterations was 1010185 Testing if hardware counters exhibit unexpected jumps with rdpmc FAILED nate@skylake:~/git/perf_event_tests/tests/rdpmc$ rdpmc_jump Testing if hardware counters exhibit unexpected jumps with rdpmc. Trying up to 100000 iterations (10000000 allowed per iteration) Got spurious count of 281474977714973 in iteration 7295 Max count for previous iterations was 1007879 Testing if hardware counters exhibit unexpected jumps with rdpmc FAILED
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I've been having problems using rdpmc to read hardware counters that are
set with perf_event_open(). I think the issue is that 'offset' field in
the mmap buffer periodically periodically has its high bit set.
Here's what I see when I run this test program on Skylake with Linux 4.4.4:
nate@skylake:~/git/perf_event_tests/tests/rdpmc$ rdpmc_jump
Testing if hardware counters exhibit unexpected jumps with rdpmc.
Trying up to 100000 iterations (10000000 allowed per iteration)
Got spurious count of 281474977720397 in iteration 230
Max count for previous iterations was 1010185
Testing if hardware counters exhibit unexpected jumps with rdpmc FAILED
nate@skylake:~/git/perf_event_tests/tests/rdpmc$ rdpmc_jump
Testing if hardware counters exhibit unexpected jumps with rdpmc.
Trying up to 100000 iterations (10000000 allowed per iteration)
Got spurious count of 281474977714973 in iteration 7295
Max count for previous iterations was 1007879
Testing if hardware counters exhibit unexpected jumps with rdpmc FAILED