fix(profiling) upscaling of I/O profiles#3146
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #3146 +/- ##
============================================
- Coverage 56.98% 52.26% -4.72%
Complexity 2883 2883
============================================
Files 141 114 -27
Lines 15844 11438 -4406
Branches 1094 0 -1094
============================================
- Hits 9028 5978 -3050
+ Misses 6243 5460 -783
+ Partials 573 0 -573
Flags with carried forward coverage won't be shown. Click here to find out more. Continue to review full report in Codecov by Sentry.
🚀 New features to boost your workflow:
|
Benchmarks [ profiler ]Benchmark execution time: 2025-03-28 16:47:23 Comparing candidate commit d1c5443 in PR branch Found 0 performance improvements and 0 performance regressions! Performance is the same for 29 metrics, 7 unstable metrics. |
87f5c70 to
76e3aed
Compare
bdddf26 to
8f4fd59
Compare
|
Why are we hooking into readv()? At least a quick source search for |
2f74ba4 to
a16ae44
Compare
bwoebi
left a comment
There was a problem hiding this comment.
Thanks for the readability refactor :-)
Co-authored-by: Bob Weinand <[email protected]>
Description
I thought I was clever with the upscaling, but turns out that I can't bend the rules as far as I did and still have correctly upscaled numbers. So back to how poisson upscaling actually works and tada, numbers are looking good again.
Also this adds hooking into
recvfrom()libc function calls.PROF-11288
Reviewer checklist