-
Notifications
You must be signed in to change notification settings - Fork 18k
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: trybot windows/amd64 failure in gc shade #18867
Comments
There are two crashes here. The first one crashes accessing the heap bitmap for a goroutine pointer, which should be mapped. The second is in cmd/dist when it makes a cgo call into the DLL to exit(1) (because the first crash happened). During that DLL call, the DLL itself faults on address -1. It seems more likely that the machine is sick in some way than that both the net/http test and cmd/dist simultaneously hit memory faults due to problems in Go (unless there was a coordinated external event like a leap second!). |
The windows builders are all fresh VMs per build on GCE. No leap seconds. I don't think this is bad hardware. |
The leap second thing was a joke, but there's no useful information here for further debugging. Still looks like a hardware or OS problem to me. If this happens again we can look harder. Dunno if that state is open or closed. |
I don't know if this is interesting, but Sunday night I started a trybot run and I noticed that there was something wrong with the windows-386-gce and windows-amd64-gce bots. They were super slow (like 20 minutes, all the other bots were done) and then they failed with a gce error message (something like "not enough resources on this zone"). |
@ALTree, that's not interesting. That's just the build coordinator not having enough quota to run N VMs at once and giving up waiting. The trybots keep waiting for quota, though, so you'd only see that message if you click through to one of them from its web status page. |
@bradfitz ah, thanks for the explanation. Sorry for the noise. |
@aclements, can you grep your collected failure logs for this? |
I only found one that looks similar, but it's actually from
|
Not high enough frequency to worry for Go 1.8. |
This has never been observed again on the dashboard. Closing. |
This happened in the second trybot run (but not the first, same tree) on my recent CL to merge master into release-branch.go1.8:
https://storage.googleapis.com/go-build-log/d8d2f036/windows-amd64-gce_dc014ee3.log
/cc @aclements @RLH
The text was updated successfully, but these errors were encountered: