EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

[AMD] Support Hierarchical Caching on AMD GPUs (#8236)

Browse Source
This commit is contained in:
Hubert Lu
2025-08-28 15:27:07 -07:00
committed by GitHub
parent 5343058875
commit 711390a971
10 changed files with 105 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
sgl-kernel/include/pytorch_extension_utils_rocm.h Normal file
View File

@@ -0,0 +1,20 @@
#include <torch/library.h>
#define CHECK_CUDA(x) TORCH_CHECK(x.is_cuda(), #x " must be a CUDA tensor")
#define CHECK_CONTIGUOUS(x) TORCH_CHECK(x.is_contiguous(), #x " must be contiguous")
#define CHECK_LAST_DIM_CONTIGUOUS(x) \
TORCH_CHECK(x.strides()[x.strides().size() - 1] == 1, #x "must be contiguous at last dimension")
#define CHECK_INPUT(x) \
CHECK_CUDA(x); \
CHECK_CONTIGUOUS(x)
#define CHECK_LAST_DIM_CONTIGUOUS_INPUT(x) \
CHECK_CUDA(x); \
CHECK_LAST_DIM_CONTIGUOUS(x)
#define CHECK_DIM(d, x) TORCH_CHECK(x.dim() == d, #x " must be a " #d "D tensor")
#define CHECK_EQ(a, b) TORCH_CHECK((a) == (b), "CHECK_EQ(" #a ", " #b ") failed. ", a, " vs ", b)
#define CHECK_GE(a, b) TORCH_CHECK((a) >= (b), "CHECK_GE(" #a ", " #b ") failed. ", a, " vs ", b)