EngineX-Ascend/enginex-ascend-910-llama.cpp
10
0
Fork 0
Code Issues Pull Requests Actions 4 Projects Releases Wiki Activity

HIP: remove the use of __HIP_PLATFORM_AMD__, explicitly support only AMD targets (#14945)

Browse Source
This commit is contained in:
uvos
2025-07-29 20:23:04 +02:00
committed by GitHub
parent b77d11179d
commit aa79524c51
10 changed files with 67 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-cuda/fattn-tile-f16.cu
View File

@@ -5,9 +5,9 @@
#define FATTN_KQ_STRIDE_TILE_F16 64
template<int D, int ncols, int nwarps, bool use_logit_softcap> // D == head size
#if !(defined(GGML_USE_HIP) && defined(__HIP_PLATFORM_AMD__))
#if !defined(GGML_USE_HIP)
__launch_bounds__(nwarps*WARP_SIZE, 2)
#endif // !(defined(GGML_USE_HIP) && defined(__HIP_PLATFORM_AMD__))
#endif // !defined(GGML_USE_HIP)
static __global__ void flash_attn_tile_ext_f16(
const char * __restrict__ Q,
const char * __restrict__ K,