Branch: refs/heads/main
Revision: 1f9780c
Author: bradcray
Link: Remove -O flags for ra correctness testing by bradcray · Pull Request #19451 · chapel-lang/chapel · GitHub
Log Message:
Merge pull request #19451 from bradcray/no-dash-O-for-ra
Remove -O flags for ra correctness testing
[trivial PR, discussed with @ronawho]
This removes the -O flags that we throw when doing ra correctness
testing, in part because it doesn't seem necessary or valuable (in
that, on my Mac at least, it makes testing take longer due to the
increased compile-times), in part because this is more consistent with
what we do for most other tests (don't throw -O for correctness
testing), and in part because it causes warnings to crop up in our
testing of --incremental that this seemed like the easiest way to
quiet (where other options would be to throw --no-warnings or come up
with some sort of prediff or incremental-specific .good file). I'm fairly
confident that the reason we started throwing -O was that we used to
run a much larger problem size for correctness testing than we do now,
and it was necessary for making things complete in a timely manner.
Modified Files:
R test/release/examples/benchmarks/hpcc/ra-atomics.compopts
R test/release/examples/benchmarks/hpcc/ra.compopts
R test/release/examples/benchmarks/hpcc/variants/ra-cleanloop.compopts
Compare: https://github.com/chapel-lang/chapel/compare/8ba1f585a758...1f9780c042e3