--- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -137,6 +137,7 @@ # #2027: defined here, before the GPU section reads it to register the native-expert tests (it was defined at the # native-experts section far below, so a first configure registered four tests fewer than the second) option(STRATA_NATIVE_EXPERTS "Native (IQ) expert support through ggml-cpu" ON) +option(STRATA_SYSTEM_GGML "Link the installed ggml libraries (sci-ml/ggml) instead of building the pinned source" OFF) if(STRATA_BUILD_CONVERSATION_TESTS) enable_testing() @@ -1836,58 +1837,71 @@ message(FATAL_ERROR "STRATA_ORCA_Q4KS_MMQ requires STRATA_NATIVE_EXPERTS=ON for GGML sources") endif() if(STRATA_NATIVE_EXPERTS) - enable_language(C) - set(BUILD_SHARED_LIBS OFF CACHE BOOL "" FORCE) - set(GGML_STATIC ON CACHE BOOL "" FORCE) - set(GGML_CUDA OFF CACHE BOOL "" FORCE) - set(GGML_OPENMP OFF CACHE BOOL "" FORCE) - set(GGML_BACKEND_DL OFF CACHE BOOL "" FORCE) - set(GGML_BUILD_TESTS OFF CACHE BOOL "" FORCE) - set(GGML_BUILD_EXAMPLES OFF CACHE BOOL "" FORCE) - if(STRATA_PORTABLE) - set(GGML_NATIVE OFF CACHE BOOL "" FORCE) - foreach(_f GGML_AVX GGML_AVX2 GGML_FMA GGML_F16C GGML_BMI2) - set(${_f} ON CACHE BOOL "" FORCE) - endforeach() - # The DeepSeek engine (STRATA_BUILD_DSV41) builds ggml-cpu with BMI2 OFF: ggml's IQ1_S / IQ1_M dots use _pdep_u64 with it, and AMD - # Zen 1/2 run PDEP in microcode (EPYC 7452, IQ1_M one-token dot 1.8 GB/s against 32 GB/s without it; IQ1_S 2.3 against 37). The - # results are bit-identical either way (ds4_gg_bench --dump, types 19 and 29, 1-16 tokens). Other engines keep BMI2 until their - # next rebuild is looked at: no Qwen package changes with this commit. - if(STRATA_BUILD_DSV41) - set(GGML_BMI2 OFF CACHE BOOL "" FORCE) + if(STRATA_SYSTEM_GGML) + # the installed ggml (sci-ml/ggml): link its libraries, and take the internal + # headers and the ggml-cuda MMQ sources compiled into the strata_mmq targets + # from its installed source tree (STRATA_GGML_DIR) + find_package(ggml REQUIRED) + add_library(ggml ALIAS ggml::ggml) + add_library(ggml-base ALIAS ggml::ggml-base) + add_library(ggml-cpu ALIAS ggml::ggml-cpu) + if(NOT STRATA_GGML_DIR OR NOT EXISTS "${STRATA_GGML_DIR}/ggml/src") + message(FATAL_ERROR "STRATA_SYSTEM_GGML=ON requires STRATA_GGML_DIR (the installed ggml source tree)") endif() - elseif(STRATA_ISA_FLOOR) - # the experimental older-CPU build (STRATA_ISA_FLOOR above): one ggml-cpu for the floor, ggml's "sandybridge" - # feature set (SSE4.2 + AVX) or SSE4.2 alone - not the build host's, which may be newer than the PC it runs on. - # From the Strata_Dirigo fork (rwkeyes). ggml's defaults turn the rest on when GGML_NATIVE is off: all off here. - set(GGML_NATIVE OFF CACHE BOOL "" FORCE) - set(GGML_CPU_ALL_VARIANTS OFF CACHE BOOL "" FORCE) - set(GGML_SSE42 ON CACHE BOOL "" FORCE) - if(STRATA_ISA_FLOOR STREQUAL "avx") - set(GGML_AVX ON CACHE BOOL "" FORCE) + else() + enable_language(C) + set(BUILD_SHARED_LIBS OFF CACHE BOOL "" FORCE) + set(GGML_STATIC ON CACHE BOOL "" FORCE) + set(GGML_CUDA OFF CACHE BOOL "" FORCE) + set(GGML_OPENMP OFF CACHE BOOL "" FORCE) + set(GGML_BACKEND_DL OFF CACHE BOOL "" FORCE) + set(GGML_BUILD_TESTS OFF CACHE BOOL "" FORCE) + set(GGML_BUILD_EXAMPLES OFF CACHE BOOL "" FORCE) + if(STRATA_PORTABLE) + set(GGML_NATIVE OFF CACHE BOOL "" FORCE) + foreach(_f GGML_AVX GGML_AVX2 GGML_FMA GGML_F16C GGML_BMI2) + set(${_f} ON CACHE BOOL "" FORCE) + endforeach() + # The DeepSeek engine (STRATA_BUILD_DSV41) builds ggml-cpu with BMI2 OFF: ggml's IQ1_S / IQ1_M dots use _pdep_u64 with it, and AMD + # Zen 1/2 run PDEP in microcode (EPYC 7452, IQ1_M one-token dot 1.8 GB/s against 32 GB/s without it; IQ1_S 2.3 against 37). The + # results are bit-identical either way (ds4_gg_bench --dump, types 19 and 29, 1-16 tokens). Other engines keep BMI2 until their + # next rebuild is looked at: no Qwen package changes with this commit. + if(STRATA_BUILD_DSV41) + set(GGML_BMI2 OFF CACHE BOOL "" FORCE) + endif() + elseif(STRATA_ISA_FLOOR) + # the experimental older-CPU build (STRATA_ISA_FLOOR above): one ggml-cpu for the floor, ggml's "sandybridge" + # feature set (SSE4.2 + AVX) or SSE4.2 alone - not the build host's, which may be newer than the PC it runs on. + # From the Strata_Dirigo fork (rwkeyes). ggml's defaults turn the rest on when GGML_NATIVE is off: all off here. + set(GGML_NATIVE OFF CACHE BOOL "" FORCE) + set(GGML_CPU_ALL_VARIANTS OFF CACHE BOOL "" FORCE) + set(GGML_SSE42 ON CACHE BOOL "" FORCE) + if(STRATA_ISA_FLOOR STREQUAL "avx") + set(GGML_AVX ON CACHE BOOL "" FORCE) + else() + set(GGML_AVX OFF CACHE BOOL "" FORCE) + endif() + foreach(_f GGML_AVX2 GGML_FMA GGML_F16C GGML_BMI2 GGML_AVX_VNNI GGML_AVX512 GGML_AVX512_VBMI GGML_AVX512_VNNI + GGML_AVX512_BF16 GGML_AMX_TILE GGML_AMX_INT8 GGML_AMX_BF16) + set(${_f} OFF CACHE BOOL "" FORCE) + endforeach() else() - set(GGML_AVX OFF CACHE BOOL "" FORCE) + set(GGML_NATIVE ON CACHE BOOL "" FORCE) endif() - foreach(_f GGML_AVX2 GGML_FMA GGML_F16C GGML_BMI2 GGML_AVX_VNNI GGML_AVX512 GGML_AVX512_VBMI GGML_AVX512_VNNI - GGML_AVX512_BF16 GGML_AMX_TILE GGML_AMX_INT8 GGML_AMX_BF16) - set(${_f} OFF CACHE BOOL "" FORCE) - endforeach() - else() - set(GGML_NATIVE ON CACHE BOOL "" FORCE) - endif() - if(STRATA_GGML_DIR) - add_subdirectory(${STRATA_GGML_DIR}/ggml ${CMAKE_BINARY_DIR}/ggml EXCLUDE_FROM_ALL) - else() - include(FetchContent) - FetchContent_Declare(strata_llamacpp - GIT_REPOSITORY https://github.com/ggml-org/llama.cpp.git - GIT_TAG 3cf03257f219afbe7334045ff7c6a06ac68c627d - GIT_SHALLOW FALSE) - FetchContent_GetProperties(strata_llamacpp) - if(NOT strata_llamacpp_POPULATED) - FetchContent_Populate(strata_llamacpp) + if(STRATA_GGML_DIR) + add_subdirectory(${STRATA_GGML_DIR}/ggml ${CMAKE_BINARY_DIR}/ggml EXCLUDE_FROM_ALL) + else() + include(FetchContent) + FetchContent_Declare(strata_llamacpp + GIT_REPOSITORY https://github.com/ggml-org/llama.cpp.git + GIT_TAG 3cf03257f219afbe7334045ff7c6a06ac68c627d + GIT_SHALLOW FALSE) + FetchContent_GetProperties(strata_llamacpp) + if(NOT strata_llamacpp_POPULATED) + FetchContent_Populate(strata_llamacpp) + endif() + add_subdirectory(${strata_llamacpp_SOURCE_DIR}/ggml ${CMAKE_BINARY_DIR}/ggml EXCLUDE_FROM_ALL) endif() - add_subdirectory(${strata_llamacpp_SOURCE_DIR}/ggml ${CMAKE_BINARY_DIR}/ggml EXCLUDE_FROM_ALL) endif() # part of strata_kernels_cpu (the pool calls these and they call its CPU probe: one library, so GNU ld's # one-pass static linking has no cycle to resolve)