leavelet/cutlass
1
0
Fork 0
Code Issues Pull Requests Actions 25 Packages Projects Releases Wiki Activity

2.9 fixes for nvrtc (#480)

Browse Source 
* Use platform::is_same instead of std::is_same

* Don't hide cuComplex include from nvrtc

* Typo fixed

* Remove comment rename
This commit is contained in:
Stepan Tezyunichev
2022-04-29 09:06:52 -04:00
committed by GitHub
parent 21c1fa3849
commit 86ce09aed1
13 changed files with 36 additions and 35 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/cutlass/epilogue/threadblock/default_epilogue_tensor_op.h
+4 -4
View File
@@ -215,10 +215,10 @@ struct DefaultIteratorsTensorOp<
InstructionShape,
ThreadMap> {
static_assert(cutlass::platform::is_same<ElementOutput, cutlass::int4b_t>::value ||
cutlass::platform::is_same<ElementOutput, cutlass::uint4b_t>::value ||
cutlass::platform::is_same<ElementOutput, int8_t>::value ||
cutlass::platform::is_same<ElementOutput, uint8_t>::value,
static_assert(platform::is_same<ElementOutput, cutlass::int4b_t>::value ||
platform::is_same<ElementOutput, cutlass::uint4b_t>::value ||
platform::is_same<ElementOutput, int8_t>::value ||
platform::is_same<ElementOutput, uint8_t>::value,
"ElementOutput needs to be 4 or 8 bit (unsigned) int.");
static_assert((ElementsPerAccess == 16 || ElementsPerAccess == 8),