Closed
Description
https://build.golang.org/log/16f4dca098af2d81699f154757b5f6b77ada1d3f
Occurred at CL 40095, which definitely touched a lot of the compiler. Frustratingly, there is no indication in the logs whatsoever what failed, and I don't see anything obviously racy in that CL. Running a bunch of race-enabled builds locally has yet to generate a failure. I'll keep at it.
Metadata
Metadata
Assignees
Type
Projects
Relationships
Development
No branches or pull requests
Activity
[-]cmd/compile: racecompile buildure failure on dashboard[/-][+]cmd/compile: racecompile build failure on dashboard[/+]bradfitz commentedon May 19, 2017
That looks more like the builder's (my) fault than yours.
Maybe the new RPC just added for #20222 failed but something up the chain didn't handle the temporary error nicely.
In any case, those logs suck. I had assumed an RPC failure would result in a retry, but perhaps that's only for trybots. It should write more debug info the build log on failures, just in case it's not retried.
I wouldn't worry about this.
josharian commentedon May 19, 2017
OK. I've gotten no failures locally either, so I'm going to switch this to be a builders bug about the logs.
[-]cmd/compile: racecompile build failure on dashboard[/-][+]x/build: more logs when something goes wrong[/+]josharian commentedon May 19, 2017
This happened again today, at
495f55d mvdan 19 May 18:11
: https://build.golang.org/log/5ca06376c703c8190614928d381ffa7c67346b9ebradfitz commentedon May 19, 2017
https://go-review.googlesource.com/c/43774/ today might've fixed it.
josharian commentedon May 19, 2017
Oh I see—those were both very old commits, pre-dash-c. Thanks!
bradfitz commentedon May 30, 2019
I think this is done.