[Profiler] Prefer TSC to wall clock when available #73855
Closed
Conversation
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
CI Flow Status⚛️ CI FlowRuleset - Version:
|
Contributor
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 2cb6dd4 (more details on the Dr. CI page):
🕵️ 1 new failure recognized by patternsThe following CI failures do not appear to be due to upstream breakages:
|
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D34231071 |
robieta
force-pushed
the
export-D34231071
branch
from
00ea484 to
d864e09
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D34231071 |
aaronenyeshi
approved these changes
Mar 11, 2022
Member
aaronenyeshi
left a comment
aaronenyeshi
left a comment
There was a problem hiding this comment.
LGTM! See comments internal!
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D34231071 |
robieta
force-pushed
the
export-D34231071
branch
from
d864e09 to
0b2be5e
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D34231071 |
robieta
force-pushed
the
export-D34231071
branch
from
0b2be5e to
f99f2ca
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D34231071 |
robieta
force-pushed
the
export-D34231071
branch
from
f99f2ca to
458ede4
Compare
Summary: Pull Request resolved: pytorch#73855 Calling the clock is one of the most expensive parts of profiling. We can reduce the profiling overhead by using `rdtsc` instead. The tradeoff is that we have to measure and convert. (shift and scale) Test Plan: I added a cpp unit test with *very* aggressive anti-flake measures. I also ran the overhead benchmark (9 replicates) with `--stressTestKineto` (0.94 -> 0.89 us) and `--stressTestKineto --kinetoProfileMemory` (1.27 -> 1.17 us) Reviewed By: chaekit Differential Revision: D34231071 fbshipit-source-id: a7c2e9a05d5f1328444231dc439570afe8b82fc8
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D34231071 |
robieta
force-pushed
the
export-D34231071
branch
from
458ede4 to
2cb6dd4
Compare
facebook-github-bot
pushed a commit
that referenced
this pull request
Mar 13, 2022
Summary: Pull Request resolved: #73855 Calling the clock is one of the most expensive parts of profiling. We can reduce the profiling overhead by using `rdtsc` instead. The tradeoff is that we have to measure and convert. (shift and scale) Test Plan: I added a cpp unit test with *very* aggressive anti-flake measures. I also ran the overhead benchmark (9 replicates) with `--stressTestKineto` (0.94 -> 0.89 us) and `--stressTestKineto --kinetoProfileMemory` (1.27 -> 1.17 us) Reviewed By: chaekit Differential Revision: D34231071 fbshipit-source-id: e3b3dd7580d93bcc783e87c7f2fc726cb74f4df8
Contributor
|
Hey @robieta. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Summary: Calling the clock is one of the most expensive parts of profiling. We can reduce the profiling overhead by using
rdtscinstead. The tradeoff is that we have to measure and convert. (shift and scale)Test Plan: I added a cpp unit test with very aggressive anti-flake measures. I also ran the overhead benchmark (9 replicates) with
--stressTestKineto(0.94 -> 0.89 us) and--stressTestKineto --kinetoProfileMemory(1.27 -> 1.17 us)Differential Revision: D34231071