New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
runtime/race: false negative at tip #13097
Comments
CL https://golang.org/cl/16490 mentions this issue. |
Kill off that field and simplify the code. Also update some docs. Thanks to Totoro W for the diagnosis in https://golang.org/cl/16465 Sadly this race was never detected due to golang/go#13097 Fixes golang/go#13089 Change-Id: Ia5400ea6733c7824e872affc3c6720bd436911fe Reviewed-on: https://go-review.googlesource.com/16490 Reviewed-by: Blake Mizerany <blake.mizerany@gmail.com>
What is the revision of golang.org/x/net? |
See first sentence of bug report. |
The race is still detectable on tip, it's just detection rate significantly decreased. On 1.5 it is detected in 45 runs out of 100; on tip - in 3 runs out of 100. @aclements did something wrt scheduling changed since 1.5? |
@dvyukov, GC scheduling behavior has changed a bit, but as far as I'm aware general goroutine scheduling hasn't changed. Would this be difficult to bisect? |
Run bisect using
Two times it pointed to:
So, Brad, it was you :) I guess it is just some parasitic synchronization and we need to bear with it. |
When running the race detector against x/net/http2 at rev 2fd7f15, Go 1.5.1 on OS X detects this race, but Go tip (rev ab7e82e) does not:
And the missed detection:
Even with a large
--count=
value, in case it was scheduling or something.The text was updated successfully, but these errors were encountered: