ggml : move AMX to the CPU backend

.clang-tidy

@@ -17,8 +17,10 @@ Checks: >
     -clang-analyzer-security.insecureAPI.DeprecatedOrUnsafeBufferHandling,
     performance-*,
     portability-*,
+    -portability-simd-intrinsics,
     misc-*,
     -misc-const-correctness,
     -misc-non-private-member-variables-in-classes,
     -misc-no-recursion,
+    -misc-use-anonymous-namespace,
 FormatStyle: none

.github/workflows/build.yml

@@ -1121,6 +1121,11 @@ jobs:
         run: |
           & 'C:\Program Files\AMD\ROCm\*\bin\clang.exe' --version
 
+      - name: Install ccache
+        uses: hendrikmuhs/ccache-action@v1.2
+        with:
+          key: ${{ github.job }}
+
       - name: Build
         id: cmake_build
         run: |

Makefile

@@ -254,8 +254,8 @@ endif
 # keep standard at C11 and C++11
 MK_CPPFLAGS  = -Iggml/include -Iggml/src -Iinclude -Isrc -Icommon -DGGML_USE_CPU
 MK_CFLAGS    = -std=c11   -fPIC
-MK_CXXFLAGS  = -std=c++11 -fPIC
-MK_NVCCFLAGS = -std=c++11
+MK_CXXFLAGS  = -std=c++17 -fPIC
+MK_NVCCFLAGS = -std=c++17
 
 ifdef LLAMA_NO_CCACHE
 GGML_NO_CCACHE := 1
@@ -575,9 +575,12 @@ endif
 
 ifndef GGML_NO_AMX
 	MK_CPPFLAGS += -DGGML_USE_AMX
-	OBJ_GGML_EXT += ggml/src/ggml-amx/ggml-amx.o ggml/src/ggml-amx/mmq.o
+	OBJ_GGML_EXT += ggml/src/ggml-cpu/amx/amx.o ggml/src/ggml-cpu/amx/mmq.o
 endif
 
+# only necessary for the CPU backend files
+MK_CPPFLAGS += -Iggml/src/ggml-cpu
+
 ifdef GGML_RPC
 	MK_CPPFLAGS  += -DGGML_USE_RPC
 	OBJ_GGML_EXT += ggml/src/ggml-rpc.o

Package.swift

@@ -28,13 +28,16 @@ var cSettings: [CSetting] =  [
     .unsafeFlags(["-Wno-shorten-64-to-32", "-O3", "-DNDEBUG"]),
     .unsafeFlags(["-fno-objc-arc"]),
     .headerSearchPath("ggml/src"),
+    .headerSearchPath("ggml/src/ggml-cpu"),
     // NOTE: NEW_LAPACK will required iOS version 16.4+
     // We should consider add this in the future when we drop support for iOS 14
     // (ref: ref: https://developer.apple.com/documentation/accelerate/1513264-cblas_sgemm?language=objc)
     // .define("ACCELERATE_NEW_LAPACK"),
     // .define("ACCELERATE_LAPACK_ILP64")
+    .define("GGML_USE_CPU"),
 ]
 
+
 #if canImport(Darwin)
 sources.append("ggml/src/ggml-common.h")
 sources.append("ggml/src/ggml-metal/ggml-metal.m")
@@ -44,7 +47,6 @@ cSettings.append(
     contentsOf: [
         .define("GGML_USE_ACCELERATE"),
         .define("GGML_USE_METAL"),
-        .define("GGML_USE_CPU")
     ]
 )
 #endif

common/CMakeLists.txt

@@ -88,5 +88,5 @@ if (LLAMA_CURL)
 endif ()
 
 target_include_directories(${TARGET} PUBLIC .)
-target_compile_features   (${TARGET} PUBLIC cxx_std_11)
+target_compile_features   (${TARGET} PUBLIC cxx_std_17)
 target_link_libraries     (${TARGET} PRIVATE ${LLAMA_COMMON_EXTRA_LIBS} PUBLIC llama Threads::Threads)

common/common.cpp

@@ -652,7 +652,17 @@ bool fs_validate_filename(const std::string & filename) {
 
     std::u32string filename_utf32;
     try {
+#if defined(__clang__)
+        // disable C++17 deprecation warning for std::codecvt_utf8
+#    pragma clang diagnostic push
+#    pragma clang diagnostic ignored "-Wdeprecated-declarations"
+#endif
         std::wstring_convert<std::codecvt_utf8<char32_t>, char32_t> converter;
+
+#if defined(__clang__)
+#    pragma clang diagnostic pop
+#endif
+
         filename_utf32 = converter.from_bytes(filename);
 
         // If the reverse conversion mismatches, it means overlong UTF-8 sequences were used,

examples/batched-bench/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-batched-bench)
 add_executable(${TARGET} batched-bench.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/batched/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-batched)
 add_executable(${TARGET} batched.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/convert-llama2c-to-ggml/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-convert-llama2c-to-ggml)
 add_executable(${TARGET} convert-llama2c-to-ggml.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/cvector-generator/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-cvector-generator)
 add_executable(${TARGET} cvector-generator.cpp pca.hpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/embedding/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-embedding)
 add_executable(${TARGET} embedding.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/eval-callback/CMakeLists.txt

@@ -2,7 +2,7 @@ set(TARGET llama-eval-callback)
 add_executable(${TARGET} eval-callback.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)
 
 set(TEST_TARGET test-eval-callback)
 add_test(NAME ${TEST_TARGET}

examples/export-lora/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-export-lora)
 add_executable(${TARGET} export-lora.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/gbnf-validator/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-gbnf-validator)
 add_executable(${TARGET} gbnf-validator.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/gen-docs/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-gen-docs)
 add_executable(${TARGET} gen-docs.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/gguf-hash/CMakeLists.txt

@@ -19,4 +19,4 @@ add_library(sha256 OBJECT deps/sha256/sha256.c deps/sha256/sha256.h)
 target_link_libraries(${TARGET} PRIVATE sha256)
 
 target_link_libraries(${TARGET} PRIVATE ggml ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/gguf-split/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-gguf-split)
 add_executable(${TARGET} gguf-split.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/gguf/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-gguf)
 add_executable(${TARGET} gguf.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE ggml ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/gritlm/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-gritlm)
 add_executable(${TARGET} gritlm.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/imatrix/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-imatrix)
 add_executable(${TARGET} imatrix.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/infill/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-infill)
 add_executable(${TARGET} infill.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/llama-bench/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-bench)
 add_executable(${TARGET} llama-bench.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/llava/CMakeLists.txt

@@ -11,7 +11,7 @@ target_include_directories(llava PUBLIC .)
 target_include_directories(llava PUBLIC ../..)
 target_include_directories(llava PUBLIC ../../common)
 
-target_compile_features(llava PRIVATE cxx_std_11)
+target_compile_features(llava PRIVATE cxx_std_17)
 
 add_library(llava_static STATIC $<TARGET_OBJECTS:llava>)
 if (BUILD_SHARED_LIBS)
@@ -35,11 +35,11 @@ add_executable(${TARGET} llava-cli.cpp)
 set_target_properties(${TARGET} PROPERTIES OUTPUT_NAME llama-llava-cli)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llava ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)
 
 set(TARGET llama-minicpmv-cli)
 add_executable(${TARGET} minicpmv-cli.cpp)
 set_target_properties(${TARGET} PROPERTIES OUTPUT_NAME llama-minicpmv-cli)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llava ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/lookahead/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-lookahead)
 add_executable(${TARGET} lookahead.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/lookup/CMakeLists.txt

@@ -2,22 +2,22 @@ set(TARGET llama-lookup)
 add_executable(${TARGET} lookup.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)
 
 set(TARGET llama-lookup-create)
 add_executable(${TARGET} lookup-create.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)
 
 set(TARGET llama-lookup-merge)
 add_executable(${TARGET} lookup-merge.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)
 
 set(TARGET llama-lookup-stats)
 add_executable(${TARGET} lookup-stats.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/main-cmake-pkg/CMakeLists.txt

@@ -29,4 +29,4 @@ add_executable(${TARGET} ${CMAKE_CURRENT_LIST_DIR}/../main/main.cpp)
 target_include_directories(${TARGET} PRIVATE ${_common_path})
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/main/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-cli)
 add_executable(${TARGET} main.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/parallel/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-parallel)
 add_executable(${TARGET} parallel.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/passkey/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-passkey)
 add_executable(${TARGET} passkey.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/perplexity/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-perplexity)
 add_executable(${TARGET} perplexity.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/quantize-stats/CMakeLists.txt

@@ -3,4 +3,4 @@ add_executable(${TARGET} quantize-stats.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE llama build_info ${CMAKE_THREAD_LIBS_INIT})
 target_include_directories(${TARGET} PRIVATE ../../common)
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/quantize/CMakeLists.txt

@@ -3,4 +3,4 @@ add_executable(${TARGET} quantize.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
 target_include_directories(${TARGET} PRIVATE ../../common)
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/retrieval/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-retrieval)
 add_executable(${TARGET} retrieval.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/run/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-run)
 add_executable(${TARGET} run.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/save-load-state/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-save-load-state)
 add_executable(${TARGET} save-load-state.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/server/CMakeLists.txt

@@ -50,4 +50,4 @@ if (WIN32)
     TARGET_LINK_LIBRARIES(${TARGET} PRIVATE ws2_32)
 endif()
 
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/simple-chat/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-simple-chat)
 add_executable(${TARGET} simple-chat.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/simple/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-simple)
 add_executable(${TARGET} simple.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/speculative-simple/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-speculative-simple)
 add_executable(${TARGET} speculative-simple.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/speculative/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-speculative)
 add_executable(${TARGET} speculative.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

examples/tokenize/CMakeLists.txt

@@ -2,4 +2,4 @@ set(TARGET llama-tokenize)
 add_executable(${TARGET} tokenize.cpp)
 install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
-target_compile_features(${TARGET} PRIVATE cxx_std_11)
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

ggml/CMakeLists.txt

@@ -161,7 +161,6 @@ set   (GGML_METAL_MACOSX_VERSION_MIN "" CACHE STRING
 set   (GGML_METAL_STD "" CACHE STRING       "ggml: metal standard version (-std flag)")
 option(GGML_OPENMP                          "ggml: use OpenMP"                                ON)
 option(GGML_RPC                             "ggml: use RPC"                                   OFF)
-option(GGML_AMX                             "ggml: use AMX"                                   OFF)
 option(GGML_SYCL                            "ggml: use SYCL"                                  OFF)
 option(GGML_SYCL_F16                        "ggml: use 16 bit floats for sycl calculations"   OFF)
 set   (GGML_SYCL_TARGET "INTEL" CACHE STRING