You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
$ gotip version
go version devel +220e0e0f73 Thu Apr 27 04:42:21 2017 +0000 linux/amd64
The 2-files go program at the bottom of this issue reliably triggers a race report after a few seconds when built in a loop on a -race enabled toolchain on my 4-cpu system with
Do we care about short reproducers for data races or the race report is good enough? As you can imagine this stuff is hard to reduce because a little single-package single-function program won't trigger the race.
ALTree
changed the title
cmd/compile: compiler data race on dowidth( ) with -c=4
cmd/compile: compiler data races on dowidth( ) with -c=4
Apr 27, 2017
The 2-files go program at the bottom of this issue reliably triggers a race report after a few seconds when built in a loop on a
-race
enabled toolchain on my 4-cpu system withThe race report:
The issue is reproducible on a non-race-enabled toolchain, where the program sometimes kills the compiler with a SIGSEGV:
The program that triggers the problem:
a/a.go
main/main.go
The text was updated successfully, but these errors were encountered: