Closed matjin closed 2 years ago
General question: do you have a unit test to test what was fixed? Does it mitigate any false positives?
And this PR seems to have some duplicate changes from this PR. Could you keep either this one or that one and abandon the other?
General question: do you have a unit test to test what was fixed? Does it mitigate any false positives?
Not really an easy way to test this -- it was observed within an X++ netmodule, so we would have to make a whole project for that. Not a false positive so much as literally not being able to translate these instructions quite frequently -- a substantial and unnecessary reduction in coverage
And this PR seems to have some duplicate changes from this PR. Could you keep either this one or that one and abandon the other?
It is actually intentional. 105 derives from 103. By checking in 103 before 105, we get a cleaner commit history that shows the incremental diffs. We want to check in 103, then we check in 105.
This patch introduces a bug fix for the scenario where the instruction operand (ldarg/stloc/ldloc) is not an integer, but rather a Cecil object.