CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-18 11:46:58 +00:00
Code Issues Packages Projects Releases Wiki Activity

ggml-cpu : optimize RVV kernels (#15720)

Browse Source 
* ggml-cpu : optimize rvv ggml_vec_dot_f32

* ggml-cpu : optimize 128-bit rvv ggml_vec_dot_q4_K_q8_K

* ggml-cpu : fix riscv arch flags

* ggml-cpu : add more rvv ops

* ggml-cpu : optimize rvv ggml_vec_dot_q4_K_q8_K

* ggml-cpu : optimize rvv ggml_vec_dot_q6_K_q8_K

* ggml-cpu : minor rvv adjustments

* ggml-cpu : fix riscv include
This commit is contained in:
xctan
2025-09-03 16:16:21 +08:00
committed by GitHub
parent 8c3fdf44ec
commit 05c0380f2a
6 changed files with 289 additions and 114 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
ggml/src/ggml-cpu/CMakeLists.txt
View File

@@ -433,15 +433,22 @@ function(ggml_add_cpu_backend_variant_impl tag_name)
ggml-cpu/arch/riscv/quants.c
ggml-cpu/arch/riscv/repack.cpp
)
if (GGML_RVV)
if (GGML_XTHEADVECTOR)
list(APPEND ARCH_FLAGS -march=rv64gc_zfhmin_xtheadvector -mabi=lp64d)
elseif (GGML_RV_ZFH)
list(APPEND ARCH_FLAGS -march=rv64gcv_zfhmin -mabi=lp64d)
else()
list(APPEND ARCH_FLAGS -march=rv64gcv -mabi=lp64d)
set(MARCH_STR "rv64gc")
if (GGML_RV_ZFH)
string(APPEND MARCH_STR "_zfh")
endif()
if (GGML_XTHEADVECTOR)
string(APPEND MARCH_STR "_xtheadvector")
elseif (GGML_RVV)
string(APPEND MARCH_STR "_v")
if (GGML_RV_ZVFH)
string(APPEND MARCH_STR "_zvfh")
endif()
endif()
if (GGML_RV_ZICBOP)
string(APPEND MARCH_STR "_zicbop")
endif()
list(APPEND ARCH_FLAGS "-march=${MARCH_STR}" -mabi=lp64d)
elseif (GGML_SYSTEM_ARCH STREQUAL "s390x")
message(STATUS "s390x detected")
list(APPEND GGML_CPU_SOURCES ggml-cpu/arch/s390/quants.c)