spectrometerHBH opened a new pull request, #20226:
URL: https://github.com/apache/tvm/pull/20226

   This change allows `tirx.required_block_size` to be combined with CUDA 
launch-bounds attributes while preserving the existing incompatibility with 
`tirx.max_registers`.
   
   The combination is needed for kernels that require exact block/cluster 
dimensions and also need a minimum blocks-per-SM launch bound. The CUDA code 
generator now emits both attributes, and SplitHostDevice retains the supported 
combination.
   
   Tests:
   - `python -m pytest -q 
tests/python/tirx-transform/test_tir_transform_split_host_device.py 
tests/python/tirx/codegen/test_codegen_cuda.py` (225 passed)
   - `pre-commit run --files src/backend/cuda/codegen/codegen_cuda.cc 
src/tirx/transform/split_host_device.cc 
tests/python/tirx-transform/test_tir_transform_split_host_device.py 
tests/python/tirx/codegen/test_codegen_cuda.py`


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to