tqchen commented on code in PR #18681:
URL: https://github.com/apache/tvm/pull/18681#discussion_r2721230246
##########
python/tvm/contrib/nvcc.py:
##########
@@ -410,10 +532,88 @@ def _compile_cuda_nvrtc(code, target_format=None,
arch=None, options=None):
nvrtc.nvrtcDestroyProgram(prog)
raise RuntimeError(f"Failed to get PTX:
{nvrtc.nvrtcGetErrorString(result)}")
- # Clean up
+ # Clean up NVRTC program
nvrtc.nvrtcDestroyProgram(prog)
- return bytearray(binary_buf)
+ # link stage for NVSHMEM
+ if use_nvshmem:
+ import ctypes # pylint: disable=import-outside-toplevel
Review Comment:
conside move into a separate function
--
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]
