-
Notifications
You must be signed in to change notification settings - Fork 147
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
Github CI Anomaly Zoo #1394
Comments
To shed a bit more light on thisIf you go to the PR at mit-plv/bedrock2#280, we see |
I'm thinking this might actually be an OOM kill, especially given that we're running the checker while replacing |
Plausibly related: actions/runner-images#736 (comment) (but this is only about mac) Maybe worth asking about at https://github.com/orgs/community/discussions/ |
https://github.com/mit-plv/fiat-crypto/actions/runs/3245312644/jobs/5322679386
|
This is new
I've seen a bunch of errors about dynlink failing, but not this one |
We should include snippets of the logs in the comments; GH deletes logs after a couple months |
How do we debug this? Maybe we shouldn't suppress stderr in bytedump? |
validate failed on bytedump after previous steps succeeded: |
In fact bytedump is failing very frequently now, can we do something about this? |
This suspiciously started happening when I started not trying to rebuild on Mac OS |
Can anyone reproduce the (seemingly consistent?) issue that |
https://github.com/mit-plv/fiat-crypto/actions/runs/3281473086/jobs/5403532417#step:17:1812 |
https://github.com/mit-plv/bedrock2/actions/runs/3292916290/jobs/5428839023 self-cancelled on start |
Sometimes the version of Coq changes between one step and the next in the docker action This results in a bytedump failure because the bedrock2 files are compiled with an outdated version of Coq and so we get inconsistent assumptions |
https://github.com/mit-plv/bedrock2/actions/runs/3305594557/jobs/5455825957 self-cancelled on submit |
https://github.com/mit-plv/bedrock2/actions/runs/3337004947/jobs/5522821413 self-cancelled on submit |
https://github.com/mit-plv/bedrock2/actions/runs/3352143850/jobs/5554084845 self-cancelled on submit |
https://github.com/mit-plv/bedrock2/actions/runs/3390987808/jobs/5635703710 self-cancelled on submit |
|
That's probably just me messing up how to get the version of
The actual failure here is
which presumably is #1394 (comment) |
https://github.com/mit-plv/fiat-crypto/actions/runs/3446654504/jobs/5751826046 Windows: Retrieve the Cygwin cache |
https://github.com/mit-plv/fiat-crypto/actions/runs/3457159795/jobs/5770405148#step:6:116
|
https://github.com/mit-plv/fiat-crypto/actions/runs/3470364042/jobs/5798475008 |
This is because I have the
|
validate times out in https://github.com/mit-plv/fiat-crypto/actions/runs/3470357503/jobs/5798461079 and https://github.com/mit-plv/fiat-crypto/actions/runs/3470364029/jobs/5798476434 because changed Coq version between steps results in project rebuild. We should really have a better way of dealing with Coq changing versions under us... (no progress on coq-community/docker-coq-action#80 yet...) |
boringssl test failure at https://github.com/mit-plv/fiat-crypto/actions/runs/3524875493/jobs/5910863477#step:7:111 on a PR changing
|
Package g++-7 is not available, but is referred to by another package. E: Package 'g++-7' has no installation candidate https://github.com/mit-plv/fiat-crypto/actions/runs/3524875493/jobs/5936926857#step:3:67 |
https://github.com/mit-plv/bedrock2/actions/runs/3653687403/jobs/6176913824 timed out for no reason I can see |
Windows does this sometimes, I think: it fails to quit terminal after all scripts complete |
https://github.com/mit-plv/fiat-crypto/actions/runs/7700820211/job/20985462543?pr=1811
|
https://github.com/mit-plv/rewriter/actions/runs/7705023867/job/20998315736?pr=148 |
|
https://github.com/mit-plv/fiat-crypto/actions/runs/9056914029/job/24884198262#step:7:29
|
Maybe we need to bump |
This should hopefully deal with #1394 (comment)
I've reported https://gitlab.haskell.org/ghc/ghc/-/issues/24927 and https://gitlab.haskell.org/ghc/ghc/-/issues/20683 about the memory issues |
This should hopefully deal with #1394 (comment)
This should hopefully deal with #1394 (comment)
https://github.com/mit-plv/fiat-crypto/actions/runs/10189997626/job/28191187326#step:8:3829 |
The Debian CI has started dying with Error: Process completed with exit code 137. And no other info
|
On Windows:
Probably related to upgrading setup-ocaml? Edit: Reported upstream as ocaml/setup-ocaml#866 |
https://github.com/mit-plv/fiat-crypto/actions/runs/11099505411/job/30836037467#step:3:24 |
The text was updated successfully, but these errors were encountered: