Issue building the connector on Amazon EC2 instance running Amazon Linux 2

[tbolis-at-mulesoft]

Hi,
I'm trying to build the project on my Amazon EC2 instance running Amazon Linux 2, but I'm getting the following error.

Thank you.

[root@ip-172-31-38-20 whisper.cpp]# cmake -B build

-- The C compiler identification is GNU 7.3.1
-- The CXX compiler identification is GNU 7.3.1
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: /bin/git (found version "2.47.1")
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Check if compiler accepts -pthread
-- Check if compiler accepts -pthread - yes
-- Found Threads: TRUE
-- ccache found, compilation results will be cached. Disable with GGML_CCACHE=OFF.
-- CMAKE_SYSTEM_PROCESSOR: x86_64
-- Including CPU backend
-- Found OpenMP_C: -fopenmp (found version "4.5")
-- Found OpenMP_CXX: -fopenmp (found version "4.5")
-- Found OpenMP: TRUE (found version "4.5")
-- x86 detected
-- Adding CPU backend variant ggml-cpu: -march=native
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Configuring done (1.2s)
-- Generating done (0.0s)
-- Build files have been written to: /home/ec2-user/whisper.cpp/build

[root@ip-172-31-38-20 whisper.cpp]# cmake --build build --config Release

[  2%] Building C object ggml/src/CMakeFiles/ggml-base.dir/ggml.c.o
[  4%] Building C object ggml/src/CMakeFiles/ggml-base.dir/ggml-alloc.c.o
[  7%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/ggml-backend.cpp.o
[  9%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/ggml-opt.cpp.o
[ 11%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/ggml-threading.cpp.o
[ 14%] Building C object ggml/src/CMakeFiles/ggml-base.dir/ggml-quants.c.o
[ 16%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/gguf.cpp.o
[ 19%] Linking CXX shared library libggml-base.so
[ 19%] Built target ggml-base
[ 21%] Building C object ggml/src/CMakeFiles/ggml-cpu.dir/ggml-cpu/ggml-cpu.c.o
[ 23%] Building CXX object ggml/src/CMakeFiles/ggml-cpu.dir/ggml-cpu/ggml-cpu.cpp.o
[ 26%] Building CXX object ggml/src/CMakeFiles/ggml-cpu.dir/ggml-cpu/ggml-cpu-aarch64.cpp.o
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp: In function ‘void ggml_gemm_q4_0_8x8_q8_0(int, float*, size_t, const void*, const void*, int, int)’:
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:105:54: error: ‘_mm256_set_m128i’ was not declared in this scope
 #define GGML_F32Cx8x2_LOAD(x, y)     _mm512_cvtph_ps(_mm256_set_m128i(_mm_loadu_si128((const __m128i *)(y)), _mm_loadu_si128((const __m128i *)(x))))
                                                      ^
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:3183:50: note: in expansion of macro ‘GGML_F32Cx8x2_LOAD’
                     const __m512 col_scale_f32 = GGML_F32Cx8x2_LOAD(b_ptr_0[b].d, b_ptr_1[b].d);
                                                  ^~~~~~~~~~~~~~~~~~
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:105:54: note: suggested alternative: ‘_mm256_set_epi8’
 #define GGML_F32Cx8x2_LOAD(x, y)     _mm512_cvtph_ps(_mm256_set_m128i(_mm_loadu_si128((const __m128i *)(y)), _mm_loadu_si128((const __m128i *)(x))))
                                                      ^
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:3183:50: note: in expansion of macro ‘GGML_F32Cx8x2_LOAD’
                     const __m512 col_scale_f32 = GGML_F32Cx8x2_LOAD(b_ptr_0[b].d, b_ptr_1[b].d);
                                                  ^~~~~~~~~~~~~~~~~~
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:105:54: error: ‘_mm256_set_m128i’ was not declared in this scope
 #define GGML_F32Cx8x2_LOAD(x, y)     _mm512_cvtph_ps(_mm256_set_m128i(_mm_loadu_si128((const __m128i *)(y)), _mm_loadu_si128((const __m128i *)(x))))
                                                      ^
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:3377:50: note: in expansion of macro ‘GGML_F32Cx8x2_LOAD’
                     const __m512 col_scale_f32 = GGML_F32Cx8x2_LOAD(b_ptr_0[b].d, b_ptr_1[b].d);
                                                  ^~~~~~~~~~~~~~~~~~
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:105:54: note: suggested alternative: ‘_mm256_set_epi8’
 #define GGML_F32Cx8x2_LOAD(x, y)     _mm512_cvtph_ps(_mm256_set_m128i(_mm_loadu_si128((const __m128i *)(y)), _mm_loadu_si128((const __m128i *)(x))))
                                                      ^
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:3377:50: note: in expansion of macro ‘GGML_F32Cx8x2_LOAD’
                     const __m512 col_scale_f32 = GGML_F32Cx8x2_LOAD(b_ptr_0[b].d, b_ptr_1[b].d);
                                                  ^~~~~~~~~~~~~~~~~~
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp: In function ‘void ggml_gemm_q4_K_8x8_q8_K(int, float*, size_t, const void*, const void*, int, int)’:
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:4405:54: error: ‘_mm256_set_m128’ was not declared in this scope
                         const __m256 row_scale_f32 = _mm256_set_m128(row_scale_f32_sse, row_scale_f32_sse);//GGML_F32Cx8_REPEAT_LOAD(a_ptrs[rp][b].d, loadMask);
                                                      ^~~~~~~~~~~~~~~
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:4405:54: note: suggested alternative: ‘_mm256_set_epi8’
                         const __m256 row_scale_f32 = _mm256_set_m128(row_scale_f32_sse, row_scale_f32_sse);//GGML_F32Cx8_REPEAT_LOAD(a_ptrs[rp][b].d, loadMask);
                                                      ^~~~~~~~~~~~~~~
                                                      _mm256_set_epi8
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:4735:50: error: ‘_mm256_set_m128’ was not declared in this scope
                     const __m256 row_scale_f32 = _mm256_set_m128(row_scale_f32_sse, row_scale_f32_sse); //GGML_F32Cx8_REPEAT_LOAD(a_ptrs[rp][b].d, loadMask);
                                                  ^~~~~~~~~~~~~~~
/home/ec2-user/whisper.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp:4735:50: note: suggested alternative: ‘_mm256_set_epi8’
                     const __m256 row_scale_f32 = _mm256_set_m128(row_scale_f32_sse, row_scale_f32_sse); //GGML_F32Cx8_REPEAT_LOAD(a_ptrs[rp][b].d, loadMask);
                                                  ^~~~~~~~~~~~~~~
                                                  _mm256_set_epi8
gmake[2]: *** [ggml/src/CMakeFiles/ggml-cpu.dir/ggml-cpu/ggml-cpu-aarch64.cpp.o] Error 1
gmake[1]: *** [ggml/src/CMakeFiles/ggml-cpu.dir/all] Error 2
gmake: *** [all] Error 2

[fujimotos]

@tbolis-at-mulesoft The probability is that your gcc/g++ version is too old.

error: ‘_mm256_set_m128i’ was not declared in this scope

This particular instruction _mm256_set_m128 was supported in GCC 8.
See the following ticket for more details:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80582

-- The C compiler identification is GNU 7.3.1
-- The CXX compiler identification is GNU 7.3.1

As shown in the build log, you are using GCC v7.3.1. I estimate that
you will need to upgrade your base OS to Amazon Linux 2023.

FWIW I could successfully build whisper.cpp on aarch64 AWS in #89 (comment).