A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 00:45 UTC
Revert "[Build/CI] Fix libcuda.so linkage" (#12552)
Tyler Michael Smith committed
Jan 29, 2025 at 16:12 UTC
73aa6cfdf789ddc67a3d2924ef52fd791554fe2a
1 file changed
-4
CMakeLists.txt
-4
@@ -446,9 +446,6 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
446
endif()
447
448
message(STATUS "Enabling C extension.")
449
-if(VLLM_GPU_LANG STREQUAL "CUDA")
450
- list(APPEND VLLM_C_LIBS cuda)
451
-endif()
449
define_gpu_extension_target(
450
_C
451
DESTINATION vllm
@@ -457,7 +454,6 @@ define_gpu_extension_target(
454
COMPILE_FLAGS ${VLLM_GPU_FLAGS}
455
ARCHITECTURES ${VLLM_GPU_ARCHES}
456
INCLUDE_DIRECTORIES ${CUTLASS_INCLUDE_DIR};${CUTLASS_TOOLS_UTIL_INCLUDE_DIR}
460
- LIBRARIES ${VLLM_C_LIBS}
457
USE_SABI 3
458
WITH_SOABI)
459