Skip to content

Commit 77f4b99

Browse files
author
mike dupont
committed
working
Naming the unnamed ggml structures Here we add names for the nested structures of ggml adding print statements to main. This inserts the print probes at key points adding include for refl working now server has it
1 parent 04814e7 commit 77f4b99

20 files changed

+2976
-537
lines changed

‎.gitignore

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -99,3 +99,6 @@ tests/test-tokenizer-0-llama
9999
tests/test-tokenizer-0-falcon
100100
tests/test-tokenizer-1-llama
101101
tests/test-tokenizer-1-bpe
102+
/#llama.cpp#
103+
#*
104+
\\#*

‎CMakeLists.txt

Lines changed: 22 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -104,7 +104,7 @@ option(LLAMA_BUILD_SERVER "llama: build server example"
104104
# Compile flags
105105
#
106106

107-
set(CMAKE_CXX_STANDARD 11)
107+
set(CMAKE_CXX_STANDARD 20)
108108
set(CMAKE_CXX_STANDARD_REQUIRED true)
109109
set(CMAKE_C_STANDARD 11)
110110
set(CMAKE_C_STANDARD_REQUIRED true)
@@ -230,7 +230,12 @@ if (LLAMA_BLAS)
230230

231231
message(STATUS "BLAS found, Includes: ${BLAS_INCLUDE_DIRS}")
232232
add_compile_options(${BLAS_LINKER_FLAGS})
233-
add_compile_definitions(GGML_USE_OPENBLAS)
233+
234+
# from https://github.com/NVIDIA/cutlass
235+
make_directory("${PROJECT_BINARY_DIR}/nvcc_tmp")
236+
set(cuda_flags --keep "SHELL:--keep-dir ${PROJECT_BINARY_DIR}/nvcc_tmp" ${cuda_flags})
237+
238+
# add_compile_definitions(GGML_USE_OPENBLAS)
234239
if (${BLAS_INCLUDE_DIRS} MATCHES "mkl" AND (${LLAMA_BLAS_VENDOR} MATCHES "Generic" OR ${LLAMA_BLAS_VENDOR} MATCHES "Intel"))
235240
add_compile_definitions(GGML_BLAS_USE_MKL)
236241
endif()
@@ -312,7 +317,7 @@ if (LLAMA_MPI)
312317
if (MPI_C_FOUND)
313318
message(STATUS "MPI found")
314319
set(GGML_HEADERS_MPI ggml-mpi.h)
315-
set(GGML_SOURCES_MPI ggml-mpi.c ggml-mpi.h)
320+
set(GGML_SOURCES_MPI ggml-mpi.cpp ggml-mpi.h)
316321
add_compile_definitions(GGML_USE_MPI)
317322
add_compile_definitions(${MPI_C_COMPILE_DEFINITIONS})
318323
if (NOT MSVC)
@@ -438,6 +443,9 @@ if (NOT cuda_host_flags STREQUAL "")
438443
set(cuda_flags ${cuda_flags} -Xcompiler ${cuda_host_flags})
439444
endif()
440445

446+
#
447+
set(cuda_flags --verbose -G ${cuda_flags})
448+
441449
add_compile_options("$<$<COMPILE_LANGUAGE:CUDA>:${cuda_flags}>")
442450

443451
if (WIN32)
@@ -485,8 +493,10 @@ if (NOT MSVC)
485493
add_link_options(-static-libgcc -static-libstdc++)
486494
endif()
487495
endif()
496+
add_link_options("-Wl,-Map=${TARGET}.map")
497+
488498
if (LLAMA_GPROF)
489-
add_compile_options(-pg)
499+
add_compile_options(-pg)
490500
endif()
491501
endif()
492502

@@ -645,13 +655,16 @@ if (GGML_USE_CPU_HBM)
645655
endif()
646656

647657
add_library(ggml OBJECT
648-
ggml.c
658+
ggml.cpp
649659
ggml.h
650-
ggml-alloc.c
660+
print.hpp
661+
ggml-internal.hpp
662+
llama-internal.hpp
663+
ggml-alloc.cpp
651664
ggml-alloc.h
652-
ggml-backend.c
665+
ggml-backend.cpp
653666
ggml-backend.h
654-
ggml-quants.c
667+
ggml-quants.cpp
655668
ggml-quants.h
656669
${GGML_SOURCES_CUDA} ${GGML_HEADERS_CUDA}
657670
${GGML_SOURCES_OPENCL} ${GGML_HEADERS_OPENCL}
@@ -683,7 +696,7 @@ add_library(llama
683696
)
684697

685698
target_include_directories(llama PUBLIC .)
686-
target_compile_features(llama PUBLIC cxx_std_11) # don't bump
699+
target_compile_features(llama PUBLIC cxx_std_20) # don't bump
687700
target_link_libraries(llama PRIVATE
688701
ggml
689702
${LLAMA_EXTRA_LIBS}

‎Makefile

Lines changed: 12 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -116,7 +116,7 @@ endif
116116
# keep standard at C11 and C++11
117117
MK_CPPFLAGS = -I. -Icommon
118118
MK_CFLAGS = -std=c11 -fPIC
119-
MK_CXXFLAGS = -std=c++11 -fPIC
119+
MK_CXXFLAGS = -std=c++20 -fPIC -fpermissive -DCPP_ONLY
120120

121121
# -Ofast tends to produce faster code, but may not be available for some compilers.
122122
ifdef LLAMA_FAST
@@ -502,7 +502,7 @@ ggml-metal.o: ggml-metal.m ggml-metal.h
502502
endif # LLAMA_METAL
503503

504504
ifdef LLAMA_MPI
505-
ggml-mpi.o: ggml-mpi.c ggml-mpi.h
505+
ggml-mpi.o: ggml-mpi.cpp ggml-mpi.h
506506
$(CC) $(CFLAGS) -c $< -o $@
507507
endif # LLAMA_MPI
508508

@@ -537,17 +537,17 @@ $(info )
537537
# Build library
538538
#
539539

540-
ggml.o: ggml.c ggml.h ggml-cuda.h
541-
$(CC) $(CFLAGS) -c $< -o $@
540+
ggml.o: ggml.cpp ggml.h ggml-cuda.h
541+
$(CXX) $(CXXFLAGS) -c $< -o $@
542542

543-
ggml-alloc.o: ggml-alloc.c ggml.h ggml-alloc.h
544-
$(CC) $(CFLAGS) -c $< -o $@
543+
ggml-alloc.o: ggml-alloc.cpp ggml.h ggml-alloc.h
544+
$(CXX) $(CXXFLAGS) -c $< -o $@
545545

546-
ggml-backend.o: ggml-backend.c ggml.h ggml-backend.h
547-
$(CC) $(CFLAGS) -c $< -o $@
546+
ggml-backend.o: ggml-backend.cpp ggml.h ggml-backend.h
547+
$(CXX) $(CXXFLAGS) -c $< -o $@
548548

549-
ggml-quants.o: ggml-quants.c ggml.h ggml-quants.h
550-
$(CC) $(CFLAGS) -c $< -o $@
549+
ggml-quants.o: ggml-quants.cpp ggml.h ggml-quants.h
550+
$(CXX) $(CXXFLAGS) -c $< -o $@
551551

552552
OBJS += ggml-alloc.o ggml-backend.o ggml-quants.o
553553

@@ -734,5 +734,5 @@ tests/test-tokenizer-1-bpe: tests/test-tokenizer-1-bpe.cpp ggml.o llama.o $(COMM
734734
tests/test-tokenizer-1-llama: tests/test-tokenizer-1-llama.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
735735
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
736736

737-
tests/test-c.o: tests/test-c.c llama.h
738-
$(CC) $(CFLAGS) -c $(filter-out %.h,$^) -o $@
737+
tests/test-c.o: tests/test-c.cpp llama.h
738+
$(CXX) $(CXXFLAGS) -c $(filter-out %.h,$^) -o $@

0 commit comments

Comments
 (0)