gemini-code-assist[bot] commented on code in PR #19896:
URL: https://github.com/apache/tvm/pull/19896#discussion_r3486328539
##########
python/tvm/backend/cuda/operator/tile_primitive/permute_layout/warp_xor_swizzle.py:
##########
@@ -305,28 +305,70 @@ def _project(iter_idx, st_list):
tid_x = sctx.launch_params["threadIdx.x"]
dtype = src_buf.dtype
+ # Shared 32/64b: base ptr + stride offset avoids buf[] flatten IMAD path.
+ direct = (
+ dtype_bytes in (4, 8)
Review Comment:
This is a great optimization to use direct `ld.shared`/`st.shared` for
shared memory copies. It seems this could be extended to support 16-bit types
as well, since `ld.shared.b16` is a valid PTX instruction.
```suggestion
dtype_bytes in (2, 4, 8)
```
--
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]
