hm. i don’t think that the rust tests are running ...
# development
w
hm. i don’t think that the rust tests are running on main. just saw a test compilation failure during a cherry-pick to 2.13.
1
will investigate in a bit.
…false alarm. it looks like the rust tests were just incorrectly skipped for the top commit in https://github.com/pantsbuild/pants/pull/16093
👍 1
h
our skipping logic is def faulty
w
in this case, i’m sure it was due to my rebases or something. but yea, the CI portion of the skips should almost certainly check all of the commits in a branch, rather than the top one
h
it checks based on the merge base iirc, so your most recent push
instead it should probably be based on diff w/
upstream/main
, but then you can get issues with unrelated PRs causing your CI to not be skipped when it should, if your branch is not based on latest
upstream/main
w
should use the merge-base with main… that’s exactly the PR’s commits
h
We ~know how to skip with in-CI logic now (since we're doing it for docs-only changes)
So could expand that instead of relying on commit message hackery
b
Hmm did this lead to this failure? https://github.com/pantsbuild/pants/runs/7307810181?check_suite_focus=true error[E0425]: cannot find function
update_env
in module
local
--> process_execution/src/local_tests.rs35710 | 357 | local::update_env(&work_dir.path(), &mut req);
Copy code
|          ^^^^^^^^^^ not found in `local`
w
yes
i … might skip CI to fix CI.
😲 1
done.