Files
llama.cpp/ggml/src/CMakeLists.txt
T
082b326fc7 ggml-et: Initial ET backend (#24179)
* ggml-et: Add performance logging

* ggml-et: Quants helpers

* ggml-et: Add MUL_MAT kernel

* ggml-et: Add ROPE kernel

* ggml-et: Add RMS_NORM kernel

* ggml-et: Add GLU kernel

* ggml-et: Add SOFT_MAX kernel

* ggml-et: Add GET_ROWS kernel

* ggml-et: Add CONT kernel

* ggml-et: Add SET_ROWS kernel

* ggml-et: Add MUL_MAT_ID kernel

* ggml-et: Build et kernels as part of ggml

* ggml-et: Embed kernels with fs fallback

* ggml-et: Build fixes

* ggml-et: Add MUL_MAT F32xF32 op

* ggml_et: Add MUL_MAT_ID op

* ggml-et: Disable offloading for debug

* ggml-et: Refactor out block ops

* ggml-et: ggml backend API changes

* ggml-et: Add RESHAPE/TRANSPOSE to supported

* ggml-et: Add CONT_F16

* ggml-et: Add supported ops doc

* gglm-et: Initial doc

* ggml-et: Remove  runtime import hacks

We can now import the runtime by a simple find_package(), so we
can cleanup the CMakeLists.txt.

* ggml-et: Fix GET_ROWS kernel

Fix lost batch dimension.

Also clean vibe-comments.

* ggml-et: Fix SET_ROWS kernel

Remove incorrect broadcasting guard.

* ggml-et: Use custom instruction for fp32->fp16

* ggml-et: Vectorize set_rows fp32->fp16

* ggml-et: Fix ROPE kernel (yarn)

ggml-et: fix et_logf

WIP: Fix ramp

WIP: fix ROPE!

* ggml-et: Better sinf

* ggml-et: Fix SOFT_MAX

Add `max_bias` and `sink` support.

* ggml-et: Fix CONT

Reorder from contiguous write to read with atomic stores.

* ggml-et: Fix elmap kernel

Remainder handlin

* ggml-et: Fix MUL_MAT MUL_MAT_ID remainders

* ggml-et: Fix ET-SOC reference

* ggml-et: Fix embed kernels scripts for old python

This allows GGML-ET to build on pre-3.8 python.

* Add sysemu support with compile time flag `-DGGML_ET_SYSEMU=ON` (#6)

* Example using ET-Soc-1 emulator configuration

Example usage:
```bash
cmake -B build -DGGML_CUDA=OFF -DGGML_ET=ON -DLLAMA_CURL=OFF -DGGML_CCACHE=ON
cmake --build build --config Release -j $(nproc)

time ./build/bin/test-backend-ops

./build/bin/llama-server \
    --model Qwen3-0.6B-Q8_0.gguf \
    --alias Qwen3-0.6B-Q8_0 \
    -fa 0 \
    --ctx-size 1024 \
    --no-warmup \
    --host 127.0.0.1 \
    --port 8080
```

* build: proper dep tracking for kernels

* support host using MOLD linker

* initial multi core GET_ROW F32 implementation

* vectorized q8 dequant

* wip: cland warning clenaups and initial logging refactor

* wip: message default message cleanup

* chore: message cleanups

* cmake cleanup

* migrate to use platform provided functions

* cmake back into subdir

* support et_print() in kernels

* fix: repair kernel building

* perf: operations run async by default

* debug: proper kernel dep tracking and error detection on kenrel launch

* fix: kernel binary dep tracking and fixing get_rows_f32 erroring

* perf: back to doing async kernel runs by default

* perf: vectorize and parallel device memset

* merge matmul work

* misc: align allocation and enable all offload

* misc: delete deadcode and respect memory limits

* fix: repair tensor debug print

* fix: loosen RMS_NORM op percision

* feat: Q4_0 GET_ROWS

* perf: FP32 MUL_MAT using TensorFMA

* update limitations

* perf: redue L1 load in compute_block_dot_product_q8_0

* feat: save kernel mapping (name to id) when profiling is enabled

* chore: memops cleanup

* perf: parallelize softmax by rows

* perf: vectorize 2nd phase of softmax

* perf: ban GET_ROWS from offloaded

* perf: vectorize and non-atomic for eltwise ops and sub support

* perf: vectorize normal rope

* perf: glu runs in parallel

* merge: manually merge saqib's work on kernel fixes

* perf: more vectorized RoPE

* perf: parallelize mul_mat_id

* perf: parallelize set_rows_f32

* perf: vectorize softmax

* feat: support kernel fusion and fuse RMS_NORM + MUL

* fix: mostly resolve test-backend-ops failure in SOFT_MAX and ROPE

* fix: bump max rope dims for gemma

* feat: GeGLU and SCALE support to fully offload Gemma

* perf: faster device memset

* feat: get_rows supporting Q4_K and avoid cont cache coherent issues

* better F32 MM

* feat: NORM for ET backend

* feat: SQR for ET backend

* feat: UNARY on ET

* feat: el_map support broadcasting for ET

* feat: SUM_ROWS in ET backend

* feat: more ops in ET backend

* feat: WKV* operators in ET backend

* perf: parallelize operators across cacheline instead of row

* perf: parallelize get_rows on cacheline

* wip: baseline FlashAttention for ET backend

* wip: enough FA and CPY f32->f16 to run llama 3.1 fully offloaded with FA on

* feat: f16 x f16 -> f32 MM using matrix engine

* wip: f16 FlashAttention using matrix engine

* wip: clean up

* feat: barriers

* perf: optimize FA_F16 in ET

* perf: vectorize pack_k_for_transpose16

* perf: prefetch next loop matrix tile

* perf: FlashAttention 2nd MM uses TensorFMA and optimizations

* cleanup: flashattention reorg

* perf: optimizations and fixes

* feat: L2SCP API and make FlashAttention support DV = 256 for gemma

* perf: parallelize norms beyond single row

* feat: GATED_DELTA_NET support and relaxed L2_NORM requirment

* feat: loosen RMS_NORM, NORM, ROPE contingous req too

* feat: repeat supports brocasting on dim 0 and loosen cont check

* feat: FILL and DIAG operator

* feat: loosen UNARY support chcek

* feat: TRI support

* feat: SOLVE_TRI support

* feat: basic SET support

* feat: loosen CONT req

* perf: fp16_to_fp32 use ASM

* feat: IMROPE support

* feat: PAD support

* feat: global barrier

* fix: view must live on the same backend as backing tensor

* feat: relax CONCAT in ET backend

* feat: dead simple CUMSUM implementation

* feat: basic SSM_CONV support

* feat: loosen CONCAT req

* feat: relax GATED_DELTA_NET and add SET support proper

* cleanup: cleanup LCM math

* feat: SWIGLU single input

* feat: SSM_SCAN support

* feat: el_map supports non aligned tensors in best effort

* feat: basic GROUP_NORM support

* feat: loosen MUL_MAT capablities slightly

* feat: loosen MUL_MAT and GET_ROWS and add IM2COL

* feat: special case for softmax 1x1x1x1

* feat: loosen SOFT_MAX req in ET backend

* fix: el_map unaligned acse fixes

* perf: optimize zero_acc_vec in flash_attn_ext_f16_me

* perf: use hart 1 for packing in MM and FA for FP16

* feat: kernel semaphore

* perf: better instruction sequence in FlashAttention

* fix: gated_delta_net with proper masking

* perf: better parallelization for GATED_DELTA_NET

* perf: parallelize SSM_CONV over nr

* perf: vectorize SSM_CONV

* perf: optimize MUL_MAT for q8

* feat: support Gemma 4

* fix: support multi-device

* feat: broader GLU support

* feat: unary ops supports view

* fix: repair fp16 MM using matrix engine

* perf: handle large N GEMV better

* perf: better q8_0 MM

* perf: better set_rows

* add back deleted files

* fix: repair after merge

* feat: POC version of uberkernel

* feat: RMS_NORM in uberkernel

* feat: add more kernels into usage

* chore: clean up uberkernel compilation

* perf: faster flash attention

* perf: opt flash attention for large seq length

* feat: loosen op bounds. clamp and mean support

* perf: vectorize ssm_scan

* perf: slightly faster FA

* perf: FlashAttention parallel MM and load

* perf: fuse Q8 MM and ADD

* feat: basic conv kernel for ET

* softMAx_test

* set_rows_f32

* get_rows and cont

* testing

* set_rows_exp

* Junk addition

* Narrowing the issue

* Update flash_attn_ext_f16_me.c

Focusing FA_ext_f16_me

* test

* Eviction updated

* Detailed cache eviction debug

* mulmat

* removeal of `BUILD_FOR_UBERKERNEL` flag

* cleaning...

* fix: balance FCC0 count

* feat: implement mul_mat and mul_mat_id for Q4_0 type

* optimize uberkernel plan upload

* add mul_mat q4 into uberkernel

* enable gating flush to just uberkernel

* update docs for ET

* update op support for ET

* et-backend: optimize Q4_0 and Q8_0 mul_mat_id row accumulations

* et-backend: specialize mul_mat_id kernels for Q4_0 and Q8_0

* et-backend: fix RoPE YaRN corr_dim formula and handle degenerate inputs

* test-backend-ops: add DeepSeek-V2-Lite RoPE test coverage

* et-backend: add Q4_0 mul_mat matrix-engine kernel using TensorFMA32

* et-backend: vectorize Q4_0 matrix-engine dequantization

* et-backend: support hybrid matrix/vector engine execution for Q4_0 mul_mat tail

* et-backend: run partial-N tiles on matrix engine for Q4_0 mul_mat

* et-backend: route Q4_0 mul_mat N < 53 to vecdot for better prefill latency

* Update uberkernel.c

* Update unary_f32.c

* gemma 4

* bisect gemma4: enable scale_f32 only

* bisect gemma4: +rms_norm_f32

* bisect gemma4: +rms_norm_mul_f32

* bisect gemma4: disable rms_norm_mul_f32 -- BREAKS OUTPUT

* bisect gemma4: +rope_f32 (skip rms_norm_mul)

* bisect gemma4: +el_map_f32

* bisect gemma4: +softmax_f32

* bisect gemma4: +get_rows_f32

* bisect gemma4: +glu_f32

* bisect gemma4: +mul_mat_f32 +mul_mat_f32_matrix_engine

* bisect gemma4: +mul_mat_f16 +mul_mat_f16_matrix_engine

* bisect gemma4: +mul_mat_Q8_0 +mul_mat_Q4_0

* bisect gemma4: +flash_attn_ext_f32 +flash_attn_ext_f16_me

* bisect gemma4: +mul_mat_id_f32

* bisect gemma4: +sum_rows_f32

* bisect gemma4: +cont_f16

* bisect gemma4: +fill_f32

* bisect gemma4: +unary_f32 (all ops re-enabled except rms_norm_mul)

* Update rms_norm_mul_f32.c

* bisect2 gemma4 n64: +scale_f32 only

* bisect2 gemma4 n64: +rms_norm_f32 +rope_f32

* bisect2 gemma4 n64: +rms_norm_mul_f32 (with ET_UBERKERNEL eviction fix)

* bisect2 gemma4 n64: +el_map +get_rows +glu +softmax (skip rms_norm_mul)

* bisect2 gemma4 n64: all ops enabled except rms_norm_mul

* bisect2 n64: test unary+cont+fill+sum_rows (no mul_mat/flash_attn)

* bisect2 n64: +mul_mat_f32 +mul_mat_f32_matrix_engine

* bisect2 n64: +mul_mat_f16 +mul_mat_f16_matrix_engine

* bisect2 n64: +mul_mat_Q8_0 +mul_mat_Q4_0

* bisect2 n64: +mul_mat_Q8_0 only (disable Q4_0)

* bisect2 n64: +mul_mat_Q4_0 only (Q8_0 breaks)

* bisect2 n64: +mul_mat_id +flash_attn_ext (skip Q8_0)

* run-3: matmul + rms_norm_mul

* run-4

* Revert "run-4"

* run5

* changes after cleanup

* cleanup before upstream

* restrict changes into ET backend

* move kernel embedding from Python to CMake

* move uberkernel gen into CMake

* apply clang format

* update CMake style

* update to match C and C++ style

* use source ggml and quant headers instead of ET's

* MROPE support

* absorb view ops into same branch as none

* fix bad rebase

* add marty1885 to codeowners

* oops

* remove redundant newline

* fix CI editor warnings

---------

Co-authored-by: Vidas <vidas@nuolat.lt>
Co-authored-by: Gianluca Guida <glguida@tlbflush.org>
Co-authored-by: Gianluca Guida <gianluca@nekko.ai>
Co-authored-by: ubergarm <leimgrub@gmail.com>
Co-authored-by: SaqibAkram-10xE <saqib.akram@10xengineers.ai>
Co-authored-by: Rehan Qasim <rehan.qasim@10xengineers.ai>
2026-07-10 12:38:34 +08:00

519 lines
20 KiB
CMake

include(CheckCXXCompilerFlag)
include("../cmake/common.cmake")
add_compile_definitions(GGML_SCHED_MAX_COPIES=${GGML_SCHED_MAX_COPIES})
# enable libstdc++ assertions for debug builds
if (CMAKE_SYSTEM_NAME MATCHES "Linux")
add_compile_definitions($<$<CONFIG:Debug>:_GLIBCXX_ASSERTIONS>)
endif()
if (NOT MSVC)
if (GGML_SANITIZE_THREAD)
add_compile_options(-fsanitize=thread)
link_libraries (-fsanitize=thread)
endif()
if (GGML_SANITIZE_ADDRESS)
add_compile_options(-fsanitize=address -fno-omit-frame-pointer)
link_libraries (-fsanitize=address)
endif()
if (GGML_SANITIZE_UNDEFINED)
add_compile_options(-fsanitize=undefined)
link_libraries (-fsanitize=undefined)
endif()
endif()
if (GGML_FATAL_WARNINGS)
if (CMAKE_CXX_COMPILER_ID MATCHES "GNU" OR CMAKE_CXX_COMPILER_ID MATCHES "Clang")
list(APPEND C_FLAGS -Werror)
list(APPEND CXX_FLAGS -Werror)
elseif (CMAKE_CXX_COMPILER_ID STREQUAL "MSVC")
add_compile_options(/WX)
endif()
endif()
if (GGML_ALL_WARNINGS)
if (NOT MSVC)
list(APPEND WARNING_FLAGS -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function)
list(APPEND C_FLAGS -Wshadow -Wstrict-prototypes -Wpointer-arith -Wmissing-prototypes
-Werror=implicit-int -Werror=implicit-function-declaration)
list(APPEND CXX_FLAGS -Wmissing-declarations -Wmissing-noreturn)
list(APPEND C_FLAGS ${WARNING_FLAGS})
list(APPEND CXX_FLAGS ${WARNING_FLAGS})
ggml_get_flags(${CMAKE_CXX_COMPILER_ID} ${CMAKE_CXX_COMPILER_VERSION})
add_compile_options("$<$<COMPILE_LANGUAGE:C>:${C_FLAGS};${GF_C_FLAGS}>"
"$<$<COMPILE_LANGUAGE:CXX>:${CXX_FLAGS};${GF_CXX_FLAGS}>")
else()
# todo : msvc
set(C_FLAGS "")
set(CXX_FLAGS "")
endif()
endif()
if (GGML_LTO)
include(CheckIPOSupported)
check_ipo_supported(RESULT result OUTPUT output)
if (result)
set(CMAKE_INTERPROCEDURAL_OPTIMIZATION TRUE)
else()
message(WARNING "IPO is not supported: ${output}")
endif()
endif()
if (GGML_CCACHE AND NOT CMAKE_C_COMPILER_LAUNCHER AND NOT CMAKE_CXX_COMPILER_LAUNCHER)
find_program(GGML_CCACHE_FOUND ccache)
find_program(GGML_SCCACHE_FOUND sccache)
if (GGML_CCACHE_FOUND OR GGML_SCCACHE_FOUND)
if(GGML_CCACHE_FOUND)
set(GGML_CCACHE_VARIANT ccache)
else()
set(GGML_CCACHE_VARIANT sccache)
endif()
# TODO: should not be set globally
if (GGML_SYCL AND GGML_CCACHE_FOUND AND WIN32)
set_property(GLOBAL PROPERTY RULE_LAUNCH_COMPILE "ccache compiler_type=icl")
else ()
set_property(GLOBAL PROPERTY RULE_LAUNCH_COMPILE "${GGML_CCACHE_VARIANT}")
endif ()
set(ENV{CCACHE_SLOPPINESS} time_macros)
message(STATUS "${GGML_CCACHE_VARIANT} found, compilation results will be cached. Disable with GGML_CCACHE=OFF.")
else()
message(STATUS "Warning: ccache not found - consider installing it for faster compilation or disable this warning with GGML_CCACHE=OFF")
endif ()
endif()
# this version of Apple ld64 is buggy
execute_process(
COMMAND ${CMAKE_C_COMPILER} ${CMAKE_EXE_LINKER_FLAGS} -Wl,-v
ERROR_VARIABLE output
OUTPUT_QUIET
)
if (output MATCHES "dyld-1015\.7")
add_compile_definitions(HAVE_BUGGY_APPLE_LINKER)
endif()
# architecture specific
# TODO: probably these flags need to be tweaked on some architectures
# feel free to update the Makefile for your architecture and send a pull request or issue
message(STATUS "CMAKE_SYSTEM_PROCESSOR: ${CMAKE_SYSTEM_PROCESSOR}")
if (MSVC)
string(TOLOWER "${CMAKE_GENERATOR_PLATFORM}" CMAKE_GENERATOR_PLATFORM_LWR)
message(STATUS "CMAKE_GENERATOR_PLATFORM: ${CMAKE_GENERATOR_PLATFORM}")
else ()
set(CMAKE_GENERATOR_PLATFORM_LWR "")
endif ()
ggml_get_system_arch()
message(STATUS "GGML_SYSTEM_ARCH: ${GGML_SYSTEM_ARCH}")
if (NOT MSVC)
if (GGML_STATIC)
if (UNIX AND NOT APPLE)
set(CMAKE_FIND_LIBRARY_SUFFIXES ".a;.so")
endif()
add_link_options(-static)
if (MINGW)
add_link_options(-static-libgcc -static-libstdc++)
endif()
endif()
if (GGML_GPROF)
add_compile_options(-pg)
endif()
endif()
#
# POSIX conformance
#
# clock_gettime came in POSIX.1b (1993)
# CLOCK_MONOTONIC came in POSIX.1-2001 / SUSv3 as optional
# posix_memalign came in POSIX.1-2001 / SUSv3
# M_PI is an XSI extension since POSIX.1-2001 / SUSv3, came in XPG1 (1985)
# Somehow in OpenBSD whenever POSIX conformance is specified
# some string functions rely on locale_t availability,
# which was introduced in POSIX.1-2008, forcing us to go higher
if (CMAKE_SYSTEM_NAME MATCHES "OpenBSD")
add_compile_definitions(_XOPEN_SOURCE=700)
elseif (CMAKE_SYSTEM_NAME MATCHES "AIX")
# Don't define _XOPEN_SOURCE. We need _ALL_SOURCE, which is the default,
# in order to define _SC_PHYS_PAGES.
else()
add_compile_definitions(_XOPEN_SOURCE=600)
endif()
# Data types, macros and functions related to controlling CPU affinity and
# some memory allocation are available on Linux through GNU extensions in libc
if (CMAKE_SYSTEM_NAME MATCHES "Linux" OR CMAKE_SYSTEM_NAME MATCHES "Android")
add_compile_definitions(_GNU_SOURCE)
endif()
# RLIMIT_MEMLOCK came in BSD, is not specified in POSIX.1,
# and on macOS its availability depends on enabling Darwin extensions
# similarly on DragonFly, enabling BSD extensions is necessary
if (
CMAKE_SYSTEM_NAME MATCHES "Darwin" OR
CMAKE_SYSTEM_NAME MATCHES "iOS" OR
CMAKE_SYSTEM_NAME MATCHES "tvOS" OR
CMAKE_SYSTEM_NAME MATCHES "DragonFly"
)
add_compile_definitions(_DARWIN_C_SOURCE)
endif()
# alloca is a non-standard interface that is not visible on BSDs when
# POSIX conformance is specified, but not all of them provide a clean way
# to enable it in such cases
if (CMAKE_SYSTEM_NAME MATCHES "FreeBSD")
add_compile_definitions(__BSD_VISIBLE)
endif()
if (CMAKE_SYSTEM_NAME MATCHES "NetBSD")
add_compile_definitions(_NETBSD_SOURCE)
endif()
if (CMAKE_SYSTEM_NAME MATCHES "OpenBSD")
add_compile_definitions(_BSD_SOURCE)
endif()
if (WIN32)
add_compile_definitions(_CRT_SECURE_NO_WARNINGS)
endif()
# ggml
if (GGML_BACKEND_DL AND NOT BUILD_SHARED_LIBS)
message(FATAL_ERROR "GGML_BACKEND_DL requires BUILD_SHARED_LIBS")
endif()
add_library(ggml-base
../include/ggml.h
../include/ggml-alloc.h
../include/ggml-backend.h
../include/ggml-cpp.h
../include/ggml-opt.h
../include/gguf.h
ggml.c
ggml.cpp
ggml-alloc.c
ggml-backend.cpp
ggml-backend-meta.cpp
ggml-opt.cpp
ggml-threading.cpp
ggml-threading.h
ggml-quants.c
ggml-quants.h
gguf.cpp)
set_target_properties(ggml-base PROPERTIES
VERSION ${GGML_VERSION}
SOVERSION ${GGML_VERSION_MAJOR}
)
target_include_directories(ggml-base PRIVATE .)
if (GGML_BACKEND_DL)
target_compile_definitions(ggml-base PUBLIC GGML_BACKEND_DL)
endif()
if (GGML_SCHED_NO_REALLOC)
target_compile_definitions(ggml-base PUBLIC GGML_SCHED_NO_REALLOC)
endif()
if (GGML_OPENMP)
find_package(OpenMP)
if (OpenMP_FOUND)
set(GGML_OPENMP_ENABLED "ON" CACHE INTERNAL "")
else()
set(GGML_OPENMP_ENABLED "OFF" CACHE INTERNAL "")
message(WARNING "OpenMP not found")
endif()
else()
set(GGML_OPENMP_ENABLED "OFF" CACHE INTERNAL "")
endif()
if (GGML_OPENMP_ENABLED)
target_compile_definitions(ggml-base PRIVATE GGML_USE_OPENMP)
target_link_libraries(ggml-base PRIVATE OpenMP::OpenMP_C OpenMP::OpenMP_CXX)
endif()
add_library(ggml
ggml-backend-dl.cpp
ggml-backend-reg.cpp)
add_library(ggml::ggml ALIAS ggml)
set_target_properties(ggml PROPERTIES
VERSION ${GGML_VERSION}
SOVERSION ${GGML_VERSION_MAJOR}
)
if (GGML_BACKEND_DIR)
if (NOT GGML_BACKEND_DL)
message(FATAL_ERROR "GGML_BACKEND_DIR requires GGML_BACKEND_DL")
endif()
target_compile_definitions(ggml PUBLIC GGML_BACKEND_DIR="${GGML_BACKEND_DIR}")
endif()
target_link_libraries(ggml PUBLIC ggml-base)
if (CMAKE_SYSTEM_NAME MATCHES "Linux")
target_link_libraries(ggml PRIVATE dl)
endif()
function(ggml_add_backend_library backend)
if (GGML_BACKEND_DL)
add_library(${backend} MODULE ${ARGN})
# write the shared library to the output directory
set_target_properties(${backend} PROPERTIES LIBRARY_OUTPUT_DIRECTORY ${CMAKE_RUNTIME_OUTPUT_DIRECTORY})
target_compile_definitions(${backend} PRIVATE GGML_BACKEND_DL)
add_dependencies(ggml ${backend})
if (GGML_BACKEND_DIR)
install(TARGETS ${backend} LIBRARY DESTINATION ${GGML_BACKEND_DIR})
else()
install(TARGETS ${backend} LIBRARY DESTINATION ${CMAKE_INSTALL_BINDIR})
endif()
else()
add_library(${backend} ${ARGN})
target_link_libraries(ggml PUBLIC ${backend})
install(TARGETS ${backend} LIBRARY)
endif()
target_link_libraries(${backend} PRIVATE ggml-base)
target_include_directories(${backend} PRIVATE ..)
if (${BUILD_SHARED_LIBS})
target_compile_definitions(${backend} PRIVATE GGML_BACKEND_BUILD)
target_compile_definitions(${backend} PUBLIC GGML_BACKEND_SHARED)
endif()
# Set versioning properties for all backend libraries
# Building a MODULE library with a version is not supported on macOS (https://gitlab.kitware.com/cmake/cmake/-/issues/20782)
if (NOT (APPLE AND GGML_BACKEND_DL))
set_target_properties(${backend} PROPERTIES
VERSION ${GGML_VERSION}
SOVERSION ${GGML_VERSION_MAJOR}
)
endif()
if(NOT GGML_AVAILABLE_BACKENDS)
set(GGML_AVAILABLE_BACKENDS "${backend}"
CACHE INTERNAL "List of backends for cmake package")
else()
list(FIND GGML_AVAILABLE_BACKENDS "${backend}" has_backend)
if(has_backend EQUAL -1)
set(GGML_AVAILABLE_BACKENDS "${GGML_AVAILABLE_BACKENDS};${backend}"
CACHE INTERNAL "List of backends for cmake package")
endif()
endif()
endfunction()
function(ggml_add_backend backend)
string(TOUPPER "GGML_${backend}" backend_id)
if (${backend_id})
string(TOLOWER "ggml-${backend}" backend_target)
add_subdirectory(${backend_target})
message(STATUS "Including ${backend} backend")
if (NOT GGML_BACKEND_DL)
string(TOUPPER "GGML_USE_${backend}" backend_use)
target_compile_definitions(ggml PUBLIC ${backend_use})
endif()
endif()
endfunction()
function(ggml_add_cpu_backend_variant tag_name)
set(GGML_CPU_TAG_NAME ${tag_name})
# other: OPENMP LLAMAFILE CPU_HBM
if (GGML_SYSTEM_ARCH STREQUAL "x86")
foreach (feat NATIVE
SSE42
AVX AVX2 BMI2 AVX_VNNI FMA F16C
AVX512 AVX512_VBMI AVX512_VNNI AVX512_BF16
AMX_TILE AMX_INT8 AMX_BF16)
set(GGML_${feat} OFF)
endforeach()
foreach (feat ${ARGN})
set(GGML_${feat} ON)
endforeach()
elseif (GGML_SYSTEM_ARCH STREQUAL "ARM")
foreach (feat ${ARGN})
set(GGML_INTERNAL_${feat} ON)
endforeach()
elseif (GGML_SYSTEM_ARCH STREQUAL "PowerPC")
foreach (feat ${ARGN})
set(GGML_INTERNAL_${feat} ON)
endforeach()
elseif (GGML_SYSTEM_ARCH STREQUAL "s390x")
foreach (feat VXE2 NNPA)
set(GGML_INTERNAL_${feat} OFF)
endforeach()
foreach (feat ${ARGN})
set(GGML_INTERNAL_${feat} ON)
endforeach()
elseif (GGML_SYSTEM_ARCH STREQUAL "riscv64")
foreach (feat RVV)
set(GGML_INTERNAL_${feat} OFF)
endforeach()
foreach (feat ${ARGN})
set(GGML_INTERNAL_${feat} ON)
endforeach()
endif()
ggml_add_cpu_backend_variant_impl(${tag_name})
endfunction()
ggml_add_backend(CPU)
if (GGML_CPU_ALL_VARIANTS)
if (NOT GGML_BACKEND_DL)
message(FATAL_ERROR "GGML_CPU_ALL_VARIANTS requires GGML_BACKEND_DL")
elseif (GGML_CPU_ARM_ARCH)
message(FATAL_ERROR "Cannot use both GGML_CPU_ARM_ARCH and GGML_CPU_ALL_VARIANTS")
endif()
if (GGML_SYSTEM_ARCH STREQUAL "x86")
ggml_add_cpu_backend_variant(x64)
ggml_add_cpu_backend_variant(sse42 SSE42)
ggml_add_cpu_backend_variant(sandybridge SSE42 AVX)
if (NOT MSVC)
# __FMA__ and __F16C__ are not defined in MSVC, however they are implied with AVX2/AVX512
ggml_add_cpu_backend_variant(ivybridge SSE42 AVX F16C)
ggml_add_cpu_backend_variant(piledriver SSE42 AVX F16C FMA)
endif()
ggml_add_cpu_backend_variant(haswell SSE42 AVX F16C FMA AVX2 BMI2)
ggml_add_cpu_backend_variant(skylakex SSE42 AVX F16C FMA AVX2 BMI2 AVX512)
ggml_add_cpu_backend_variant(cannonlake SSE42 AVX F16C FMA AVX2 BMI2 AVX512 AVX512_VBMI)
ggml_add_cpu_backend_variant(cascadelake SSE42 AVX F16C FMA AVX2 BMI2 AVX512 AVX512_VNNI)
ggml_add_cpu_backend_variant(icelake SSE42 AVX F16C FMA AVX2 BMI2 AVX512 AVX512_VBMI AVX512_VNNI)
if (NOT MSVC)
# MSVC 2022 doesn't support BF16 intrinsics without `/arch:AVX10.1` ?!
# https://learn.microsoft.com/en-us/cpp/intrinsics/x64-amd64-intrinsics-list?view=msvc-170
# https://learn.microsoft.com/en-us/cpp/build/reference/arch-x64?view=msvc-170
ggml_add_cpu_backend_variant(cooperlake SSE42 AVX F16C FMA AVX2 BMI2 AVX512 AVX512_VNNI AVX512_BF16)
ggml_add_cpu_backend_variant(zen4 SSE42 AVX F16C FMA AVX2 BMI2 AVX512 AVX512_VBMI AVX512_VNNI AVX512_BF16)
endif()
ggml_add_cpu_backend_variant(alderlake SSE42 AVX F16C FMA AVX2 BMI2 AVX_VNNI)
if (NOT MSVC)
# MSVC doesn't support AMX
ggml_add_cpu_backend_variant(sapphirerapids SSE42 AVX F16C FMA AVX2 BMI2 AVX512 AVX512_VBMI AVX512_VNNI AVX512_BF16 AMX_TILE AMX_INT8)
endif()
elseif(GGML_SYSTEM_ARCH STREQUAL "ARM")
if (CMAKE_SYSTEM_NAME MATCHES "Linux")
# Many of these features are optional so we build versions with popular
# combinations and name the backends based on the version they were
# first released with
ggml_add_cpu_backend_variant(armv8.0_1)
ggml_add_cpu_backend_variant(armv8.2_1 DOTPROD)
ggml_add_cpu_backend_variant(armv8.2_2 DOTPROD FP16_VECTOR_ARITHMETIC)
ggml_add_cpu_backend_variant(armv8.2_3 DOTPROD FP16_VECTOR_ARITHMETIC SVE)
ggml_add_cpu_backend_variant(armv8.6_1 DOTPROD FP16_VECTOR_ARITHMETIC SVE MATMUL_INT8)
ggml_add_cpu_backend_variant(armv8.6_2 DOTPROD FP16_VECTOR_ARITHMETIC SVE MATMUL_INT8 SVE2)
ggml_add_cpu_backend_variant(armv9.2_1 DOTPROD FP16_VECTOR_ARITHMETIC SVE MATMUL_INT8 SME)
ggml_add_cpu_backend_variant(armv9.2_2 DOTPROD FP16_VECTOR_ARITHMETIC SVE MATMUL_INT8 SVE2 SME)
elseif (CMAKE_SYSTEM_NAME MATCHES "Android")
# Android-specific backends with SoC-compatible feature sets
ggml_add_cpu_backend_variant(android_armv8.0_1)
ggml_add_cpu_backend_variant(android_armv8.2_1 DOTPROD)
ggml_add_cpu_backend_variant(android_armv8.2_2 DOTPROD FP16_VECTOR_ARITHMETIC)
ggml_add_cpu_backend_variant(android_armv8.6_1 DOTPROD FP16_VECTOR_ARITHMETIC MATMUL_INT8)
ggml_add_cpu_backend_variant(android_armv9.0_1 DOTPROD MATMUL_INT8 FP16_VECTOR_ARITHMETIC SVE2)
ggml_add_cpu_backend_variant(android_armv9.2_1 DOTPROD MATMUL_INT8 FP16_VECTOR_ARITHMETIC SVE SME)
ggml_add_cpu_backend_variant(android_armv9.2_2 DOTPROD MATMUL_INT8 FP16_VECTOR_ARITHMETIC SVE SVE2 SME)
elseif (APPLE)
ggml_add_cpu_backend_variant(apple_m1 DOTPROD)
ggml_add_cpu_backend_variant(apple_m2_m3 DOTPROD MATMUL_INT8)
ggml_add_cpu_backend_variant(apple_m4 DOTPROD MATMUL_INT8 NOSVE SME)
else()
message(FATAL_ERROR "Unsupported ARM target OS: ${CMAKE_SYSTEM_NAME}")
endif()
elseif (GGML_SYSTEM_ARCH STREQUAL "PowerPC")
if (CMAKE_SYSTEM_NAME MATCHES "Linux")
ggml_add_cpu_backend_variant(power0)
ggml_add_cpu_backend_variant(power7_1 POWER7)
ggml_add_cpu_backend_variant(power7_2 POWER7 VSX)
ggml_add_cpu_backend_variant(power8_1 POWER8)
ggml_add_cpu_backend_variant(power8_2 POWER8 VSX)
ggml_add_cpu_backend_variant(power9 POWER9 VSX)
ggml_add_cpu_backend_variant(power10 POWER10 VSX)
# POWER11 backend: only if compiler supports -mcpu=power11
check_cxx_compiler_flag("-mcpu=power11" GGML_CXX_SUPPORTS_POWER11)
if (GGML_CXX_SUPPORTS_POWER11)
message(STATUS "Compiler supports -mcpu=power11, enabling POWER11 backend")
ggml_add_cpu_backend_variant(power11 POWER11 VSX)
else()
message(STATUS "Skipping POWER11 backend: compiler does not support -mcpu=power11")
endif()
else()
message(FATAL_ERROR "Unsupported PowerPC target OS: ${CMAKE_SYSTEM_NAME}")
endif()
elseif (GGML_SYSTEM_ARCH STREQUAL "s390x")
if (CMAKE_SYSTEM_NAME MATCHES "Linux")
ggml_add_cpu_backend_variant(z15 Z15 VXE2)
ggml_add_cpu_backend_variant(z16 Z16 VXE2 NNPA)
else()
message(FATAL_ERROR "Unsupported s390x target OS: ${CMAKE_SYSTEM_NAME}")
endif()
elseif (GGML_SYSTEM_ARCH STREQUAL "riscv64")
if (CMAKE_SYSTEM_NAME MATCHES "Linux")
ggml_add_cpu_backend_variant(riscv64_0)
ggml_add_cpu_backend_variant(riscv64_v RVV)
else()
message(FATAL_ERROR "Unsupported RISC-V target OS: ${CMAKE_SYSTEM_NAME}")
endif()
else()
message(FATAL_ERROR "GGML_CPU_ALL_VARIANTS not yet supported with ${GGML_SYSTEM_ARCH} on ${CMAKE_SYSTEM_NAME}")
endif()
elseif (GGML_CPU)
ggml_add_cpu_backend_variant_impl("")
endif()
ggml_add_backend(BLAS)
ggml_add_backend(CANN)
ggml_add_backend(CUDA)
ggml_add_backend(ET)
ggml_add_backend(HIP)
ggml_add_backend(METAL)
ggml_add_backend(MUSA)
ggml_add_backend(RPC)
ggml_add_backend(VirtGPU)
ggml_add_backend(SYCL)
ggml_add_backend(Vulkan)
ggml_add_backend(WebGPU)
ggml_add_backend(zDNN)
ggml_add_backend(OpenCL)
ggml_add_backend(Hexagon)
ggml_add_backend(ZenDNN)
ggml_add_backend(OPENVINO)
foreach (target ggml-base ggml)
target_include_directories(${target} PUBLIC $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/../include> $<INSTALL_INTERFACE:include>)
target_compile_features (${target} PRIVATE c_std_11 cxx_std_17) # don't bump
endforeach()
target_link_libraries(ggml-base PRIVATE Threads::Threads)
if (DEFINED MATH_LIBRARY)
target_link_libraries(ggml-base PRIVATE ${MATH_LIBRARY})
elseif (NOT WIN32 AND NOT DEFINED ENV{ONEAPI_ROOT})
target_link_libraries(ggml-base PRIVATE m)
endif()
if (CMAKE_SYSTEM_NAME MATCHES "Android")
target_link_libraries(ggml-base PRIVATE dl)
endif()
if(CMAKE_SYSTEM_NAME MATCHES "visionOS")
target_compile_definitions(ggml-base PUBLIC _DARWIN_C_SOURCE)
endif()
if (BUILD_SHARED_LIBS)
foreach (target ggml-base ggml)
set_target_properties(${target} PROPERTIES POSITION_INDEPENDENT_CODE ON)
target_compile_definitions(${target} PRIVATE GGML_BUILD)
target_compile_definitions(${target} PUBLIC GGML_SHARED)
endforeach()
endif()