Rafi Shamim
Rafi Shamim
i'll merge this to resolve the failure before the release triage meeting bors r+
> That comment has to do with how options are represented in the CRDB code base. As far as I can tell, the options you pass from the client get...
you might need to do `rm -rf vendor; git checkout vendor; git submodule update --init --recursive`
tftr! failures are flakes from https://github.com/cockroachdb/cockroach/issues/85944 bors r=ajwerner
``` FAIL! Expected sql doesn't match generated sql! Deleting expectedSql file will test that new sql works correctly and will auto-generate a new version if it passes. EXPECTED SQL: CREATE...
closing since this is an empty diff
I thought we had a linter for not returning errors ([returnerrcheck](https://github.com/cockroachdb/cockroach/blob/master/pkg/testutils/lint/passes/returnerrcheck/returnerrcheck.go))? But it looks like this problem (i.e. not returning an error that could have caused the txn to abort)...
> The returnerrcheck linter doesn't do anything if the function doesn't itself return an error. The optimizer doesn't really return errors, it panics them, so the linter does not trigger....
@mgartner regarding your earlier comment: > I want to take some time to audit all these casts - any cast that has to read some mutable state should not be...
the many failures look like this: https://stackoverflow.com/questions/72297870/minitest-how-to-solve-mocked-method-call-expects-0-arguments-got-1 a test library issue, so not a blocker