allispaul

**Reference issues/PRs** Fixes #680. **Types of changes** - [ ] Bug fix (non-breaking change which fixes an issue) - [ ] New feature (non-breaking change which adds functionality) - [x]...

**Reference issues/PRs** Fixes #682. Fixes #683. **Types of changes** - [x] Bug fix (non-breaking change which fixes an issue) - [ ] New feature (non-breaking change which adds functionality) -...

I'm noticing two weird things when font-lock-mode and cuda-mode are both enabled, both shown in the attached screenshot: 1. Identifiers only get `font-lock-variable-name-face` starting at the *letter before the first...

[BUG] Logic issue in nondeterministic reduction mode of Stream-K tile scheduler.

4

**Describe the bug** The nondeterministic reduction mode of the Stream-K tile scheduler, as described [here](https://github.com/NVIDIA/cutlass/blob/main/include/cutlass/gemm/kernel/tile_scheduler_params.h#L374), is supposed to have all CTAs collaborating on a tile wait for the first one...