Flush mutator buffers in destroy_mutator

[k-sareen]

No description provided.

[wks]

LGTM.

I think a deeper problem is that the semantics of flush_mutator is unclear, and it should be removed from the API for the VM binding.

See: #1047

And this PR LGTM because it makes the semantics of destroy_mutator clear. We should remove flush_mutator and fix the OpenJDK binding in another PR.

[qinsoon]

The Ruby tests failed, but it could be unrelated with the PR: https://github.com/mmtk/mmtk-core/actions/runs/7192554303/job/19589222684. @wks Can you check this before merging?

[wks]

The Ruby tests failed, but it could be unrelated with the PR: https://github.com/mmtk/mmtk-core/actions/runs/7192554303/job/19589222684. @wks Can you check this before merging?

Very strange. It is a segmentation fault without any further message.

Actually I am about to make an update to the MMTk-ruby binding. I'll re-run that after updating the Ruby binding, first.

[qinsoon]

The Ruby tests failed, but it could be unrelated with the PR: https://github.com/mmtk/mmtk-core/actions/runs/7192554303/job/19589222684. @wks Can you check this before merging?

Very strange. It is a segmentation fault without any further message.

Actually I am about to make an update to the MMTk-ruby binding. I'll re-run that after updating the Ruby binding, first.

There was an error message in the first attemp:

  ERROR: An MMTk GC thread panicked.  This is a bug.
  panicked at 'Object 0x200fff78718 does not have HAS_MOVED_GIVTBL flag or original givtbl', src/abi.rs:111:17
  Backtrace is disabled.
  run with `RUST_BACKTRACE=1` environment variable to display a backtrace

[wks]

The Ruby tests failed, but it could be unrelated with the PR: https://github.com/mmtk/mmtk-core/actions/runs/7192554303/job/19589222684. @wks Can you check this before merging?

Very strange. It is a segmentation fault without any further message.
Actually I am about to make an update to the MMTk-ruby binding. I'll re-run that after updating the Ruby binding, first.

There was an error message in the first attemp:

  ERROR: An MMTk GC thread panicked.  This is a bug.
  panicked at 'Object 0x200fff78718 does not have HAS_MOVED_GIVTBL flag or original givtbl', src/abi.rs:111:17
  Backtrace is disabled.
  run with `RUST_BACKTRACE=1` environment variable to display a backtrace

This is reproducible locally. I think there is a bug.

Another bug is that the CI script in mmtk-core does not look at Cargo.toml in mmtk-ruby for the chosen revision of repo ruby, so it automatically checked out the latest revision of the ruby repo which I am just about to run CI scripts against (which failed, too). I have made a force push to the ruby repo to roll back to the previous head. I'll fix the CI script for mmtk-core later.

[wks]

OK. Now all tests passed. @k-sareen You can hit the merge button if you don't have further changes.

[k-sareen]

You should also set RUST_BACKTRACE=1 in the CI if you haven't already.