leavelet/cutlass
1
0
Fork 0
Code Issues Pull Requests Actions 26 Packages Projects Releases Wiki Activity

Replace CUDA driver API with runtime API (#2928)

Browse Source 
Co-authored-by: dePaul Miller <23461061+depaulmillz@users.noreply.github.com>
This commit is contained in:
dePaul Miller
2026-01-05 13:50:44 -05:00
committed by GitHub
co-authored by dePaul Miller
parent 2aee73922c
commit 7127592069
2 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/cute/atom/copy_traits_sm90_im2col.hpp
+2 -2
View File
@@ -472,8 +472,8 @@ make_im2col_tma_copy_desc(
tma_oob_fill);
int driver_version = 0;
CUresult driver_version_result = cuDriverGetVersion(&driver_version);
assert(driver_version_result == CUDA_SUCCESS);
cudaError_t driver_version_err = cudaDriverGetVersion(&driver_version);
assert(driver_version_err == cudaSuccess);
if (driver_version <= 13010) {
if (cute::bits_to_bytes(
cute::cosize(tensor_cwhdn.layout()) *
include/cute/atom/copy_traits_sm90_tma.hpp
+2 -2
View File
@@ -1053,8 +1053,8 @@ make_tma_copy_desc(Tensor<GEngine,GLayout> const& gtensor, // The origin
tma_oobFill);
int driver_version = 0;
CUresult driver_version_result = cuDriverGetVersion(&driver_version);
assert(driver_version_result == CUDA_SUCCESS);
cudaError_t driver_version_err = cudaDriverGetVersion(&driver_version);
assert(driver_version_err == cudaSuccess);
if (driver_version <= 13010) {
if (cute::bits_to_bytes(
cute::cosize(gtensor.layout()) *