xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-21 16:27:19 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix] Fix CUDA version check for mma warning suppression (#5642)

Browse Source
This commit is contained in:
Tyler Michael Smith 2024-06-18 19:48:49 -04:00 committed by GitHub
parent 2bd231a7b7
commit b23ce92032
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
csrc/quantization/marlin/sparse/common/mma.h
View File

@ -17,6 +17,7 @@
#pragma once #pragma once
#include "base.h" #include "base.h"
#include <cudaTypedefs.h>
namespace marlin_24 { namespace marlin_24 {
@ -26,7 +27,7 @@ namespace marlin_24 {
// | Advisory: Modifier ‘.sp::ordered_metadata’ should be used on instruction // | Advisory: Modifier ‘.sp::ordered_metadata’ should be used on instruction
// | ‘mma’ instead of modifier ‘.sp’ as it is expected to have substantially // | ‘mma’ instead of modifier ‘.sp’ as it is expected to have substantially
// | reduced performance on some future architectures // | reduced performance on some future architectures
#if defined CUDA_VERSION && CUDA_VERSION >= 12500 #if defined CUDA_VERSION && CUDA_VERSION >= 12050
#define MMA_SP_INST \ #define MMA_SP_INST \
"mma.sp::ordered_metadata.sync.aligned.m16n8k32.row.col.f32.f16.f16.f32 " "mma.sp::ordered_metadata.sync.aligned.m16n8k32.row.col.f32.f16.f16.f32 "
#else #else