Error Building llama.cpp on MacBook Pro #6806

Spider-netizen · 2024-04-21T11:44:48Z

Spider-netizen
Apr 21, 2024

Hi,

I am trying to build llama.cpp on my MacBook Pro using "make".

The install fails with following output:

% make
I ccache not found. Consider installing it for faster compilation.
I llama.cpp build info: 
I UNAME_S:   Darwin
I UNAME_P:   i386
I UNAME_M:   x86_64
I CFLAGS:    -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -std=c11   -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wshadow -Wstrict-prototypes -Wpointer-arith -Wmissing-prototypes -Werror=implicit-int -Werror=implicit-function-declaration -pthread -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wdouble-promotion 
I CXXFLAGS:  -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL 
I NVCCFLAGS: -std=c++11 -O3 
I LDFLAGS:   -framework Accelerate -framework Foundation -framework Metal -framework MetalKit 
I CC:        Apple clang version 12.0.0 (clang-1200.0.32.29)
I CXX:       Apple clang version 12.0.0 (clang-1200.0.32.29)

cc  -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -std=c11   -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wshadow -Wstrict-prototypes -Wpointer-arith -Wmissing-prototypes -Werror=implicit-int -Werror=implicit-function-declaration -pthread -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wdouble-promotion    -c ggml.c -o ggml.o
c++ -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -c llama.cpp -o llama.o
c++ -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -c common/common.cpp -o common.o
c++ -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -c common/sampling.cpp -o sampling.o
c++ -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -c common/grammar-parser.cpp -o grammar-parser.o
c++ -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -c common/build-info.cpp -o build-info.o
c++ -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -c common/json-schema-to-grammar.cpp -o json-schema-to-grammar.o
c++ -std=c++11 -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wmissing-declarations -Wmissing-noreturn -pthread  -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wmissing-prototypes -Wextra-semi -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -c common/console.cpp -o console.o
cc -I. -Icommon -D_XOPEN_SOURCE=600 -D_DARWIN_C_SOURCE -DNDEBUG -DGGML_USE_ACCELERATE -DACCELERATE_NEW_LAPACK -DACCELERATE_LAPACK_ILP64 -DGGML_USE_METAL  -std=c11   -fPIC -O3 -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wshadow -Wstrict-prototypes -Wpointer-arith -Wmissing-prototypes -Werror=implicit-int -Werror=implicit-function-declaration -pthread -march=native -mtune=native -Wunreachable-code-break -Wunreachable-code-return -Wdouble-promotion  -c ggml-metal.m -o ggml-metal.o
ggml-metal.m:414:69: error: use of undeclared identifier 'MTLGPUFamilyApple7'
    ctx->support_simdgroup_reduction  = [ctx->device supportsFamily:MTLGPUFamilyApple7];
                                                                    ^
ggml-metal.m:417:61: error: use of undeclared identifier 'MTLGPUFamilyApple7'
    ctx->support_simdgroup_mm = [ctx->device supportsFamily:MTLGPUFamilyApple7];
                                                            ^
ggml-metal.m:797:5: error: use of undeclared identifier 'MTLComputePassDescriptor'
    MTLComputePassDescriptor * edesc = MTLComputePassDescriptor.computePassDescriptor;
    ^
ggml-metal.m:797:32: error: use of undeclared identifier 'edesc'
    MTLComputePassDescriptor * edesc = MTLComputePassDescriptor.computePassDescriptor;
                               ^
ggml-metal.m:797:40: error: use of undeclared identifier 'MTLComputePassDescriptor'
    MTLComputePassDescriptor * edesc = MTLComputePassDescriptor.computePassDescriptor;
                                       ^
ggml-metal.m:798:5: error: use of undeclared identifier 'edesc'
    edesc.dispatchType = MTLDispatchTypeSerial;
    ^
ggml-metal.m:841:64: warning: instance method '-computeCommandEncoderWithDescriptor:' not found (return type defaults to 'id') [-Wobjc-method-access]
        id<MTLComputeCommandEncoder> encoder = [command_buffer computeCommandEncoderWithDescriptor: edesc];
                                                               ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
ggml-metal.m:841:101: error: use of undeclared identifier 'edesc'
        id<MTLComputeCommandEncoder> encoder = [command_buffer computeCommandEncoderWithDescriptor: edesc];
                                                                                                    ^
ggml-metal.m:942:39: warning: unused variable 'nb' [-Wunused-variable]
                        const int64_t nb = ne00;
                                      ^
ggml-metal.m:944:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_CONCAT].pipeline;
                                                    ^
ggml-metal.m:976:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN(1024, ne0);
                                  ^
ggml-metal.m:1048:43: warning: unused variable 'n' [-Wunused-variable]
                            const int64_t n = ggml_nelements(dst)/4;
                                          ^
ggml-metal.m:1052:39: warning: unused variable 'nth' [-Wunused-variable]
                            const int nth = MIN((int) pipeline.maxTotalThreadsPerThreadgroup, ne0);
                                      ^
ggml-metal.m:984:38: warning: unused variable 'offs' [-Wunused-variable]
                        const size_t offs = 0;
                                     ^
ggml-metal.m:1100:39: warning: unused variable 'nth' [-Wunused-variable]
                            const int nth = MIN((int) pipeline.maxTotalThreadsPerThreadgroup, ne00);
                                      ^
ggml-metal.m:1066:38: warning: unused variable 'pnb1' [-Wunused-variable]
                        const size_t pnb1 = ((int32_t *) dst->op_params)[0];
                                     ^
ggml-metal.m:1067:38: warning: unused variable 'pnb2' [-Wunused-variable]
                        const size_t pnb2 = ((int32_t *) dst->op_params)[1];
                                     ^
ggml-metal.m:1068:38: warning: unused variable 'pnb3' [-Wunused-variable]
                        const size_t pnb3 = ((int32_t *) dst->op_params)[2];
                                     ^
ggml-metal.m:1069:38: warning: unused variable 'offs' [-Wunused-variable]
                        const size_t offs = ((int32_t *) dst->op_params)[3];
                                     ^
ggml-metal.m:1137:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN((int) pipeline.maxTotalThreadsPerThreadgroup, ne00);
                                  ^
ggml-metal.m:1168:49: warning: unused variable 'pipeline' [-Wunused-variable]
                    id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_CLAMP].pipeline;
                                                ^
ggml-metal.m:1181:35: warning: unused variable 'n' [-Wunused-variable]
                    const int64_t n = ggml_nelements(dst);
                                  ^
ggml-metal.m:1192:61: warning: unused variable 'pipeline' [-Wunused-variable]
                                id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_TANH].pipeline;
                                                            ^
ggml-metal.m:1198:47: warning: unused variable 'n' [-Wunused-variable]
                                const int64_t n = ggml_nelements(dst);
                                              ^
ggml-metal.m:1204:61: warning: unused variable 'pipeline' [-Wunused-variable]
                                id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_RELU].pipeline;
                                                            ^
ggml-metal.m:1210:47: warning: unused variable 'n' [-Wunused-variable]
                                const int64_t n = ggml_nelements(dst);
                                              ^
ggml-metal.m:1281:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_SQR].pipeline;
                                                    ^
ggml-metal.m:1287:39: warning: unused variable 'n' [-Wunused-variable]
                        const int64_t n = ggml_nelements(dst);
                                      ^
ggml-metal.m:1295:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_SUM_ROWS].pipeline;
                                                    ^
ggml-metal.m:1357:37: warning: unused variable 'm0' [-Wunused-variable]
                        const float m0 = powf(2.0f, -(max_bias       ) / n_head_log2);
                                    ^
ggml-metal.m:1358:37: warning: unused variable 'm1' [-Wunused-variable]
                        const float m1 = powf(2.0f, -(max_bias / 2.0f) / n_head_log2);
                                    ^
ggml-metal.m:1387:35: warning: unused variable 'n_past' [-Wunused-variable]
                        const int n_past = ((int32_t *)(dst->op_params))[0];
                                  ^
ggml-metal.m:1450:57: error: use of undeclared identifier 'MTLGPUFamilyApple7'
                        if ([ctx->device supportsFamily:MTLGPUFamilyApple7] &&
                                                        ^
ggml-metal.m:1698:43: warning: unused variable 'mem_size' [-Wunused-variable]
                                const int mem_size = src0t == GGML_TYPE_IQ2_XXS ? 256*8+128 : 512*8+128;
                                          ^
ggml-metal.m:1703:43: warning: unused variable 'mem_size' [-Wunused-variable]
                                const int mem_size = src0t == GGML_TYPE_IQ3_XXS ? 256*4+128 : 512*4;
                                          ^
ggml-metal.m:1708:43: warning: unused variable 'mem_size' [-Wunused-variable]
                                const int mem_size = 32*sizeof(float);
                                          ^
ggml-metal.m:1728:47: warning: unused variable 'ny' [-Wunused-variable]
                                const int64_t ny = (ne11 + nrows - 1)/nrows;
                                              ^
ggml-metal.m:1773:57: error: use of undeclared identifier 'MTLGPUFamilyApple7'
                        if ([ctx->device supportsFamily:MTLGPUFamilyApple7] &&
                                                        ^
ggml-metal.m:2023:43: warning: unused variable 'mem_size' [-Wunused-variable]
                                const int mem_size = src0t == GGML_TYPE_IQ2_XXS ? 256*8+128 : 512*8+128;
                                          ^
ggml-metal.m:2028:43: warning: unused variable 'mem_size' [-Wunused-variable]
                                const int mem_size = src0t == GGML_TYPE_IQ3_XXS ? 256*4+128 : 512*4;
                                          ^
ggml-metal.m:2033:43: warning: unused variable 'mem_size' [-Wunused-variable]
                                const int mem_size = 32*sizeof(float);
                                          ^
ggml-metal.m:2053:47: warning: unused variable 'ny' [-Wunused-variable]
                                const int64_t ny = (_ne1 + nrows - 1)/nrows; // = _ne1
                                              ^
ggml-metal.m:2015:39: warning: unused variable 'tgz' [-Wunused-variable]
                            const int tgz = dst_rows;
                                      ^
ggml-metal.m:2116:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_RMS_NORM].pipeline;
                                                    ^
ggml-metal.m:2126:39: warning: unused variable 'nrows' [-Wunused-variable]
                        const int64_t nrows = ggml_nrows(src0);
                                      ^
ggml-metal.m:2137:37: warning: unused variable 'eps' [-Wunused-variable]
                        const float eps = 1e-6f; // TODO: temporarily hardcoded
                                    ^
ggml-metal.m:2139:39: warning: unused variable 'n_groups' [-Wunused-variable]
                        const int32_t n_groups = ((int32_t *) dst->op_params)[0];
                                      ^
ggml-metal.m:2141:29: warning: unused variable 'nth' [-Wunused-variable]
                        int nth = 32; // SIMD width
                            ^
ggml-metal.m:2147:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_GROUP_NORM].pipeline;
                                                    ^
ggml-metal.m:2169:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN(256, ne00);
                                  ^
ggml-metal.m:2171:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_NORM].pipeline;
                                                    ^
ggml-metal.m:2181:39: warning: unused variable 'nrows' [-Wunused-variable]
                        const int64_t nrows = ggml_nrows(src0);
                                      ^
ggml-metal.m:2189:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN(1024, ne00);
                                  ^
ggml-metal.m:2198:37: warning: unused variable 'm0' [-Wunused-variable]
                        const float m0 = powf(2.0f, -(max_bias) / n_heads_log2_floor);
                                    ^
ggml-metal.m:2199:37: warning: unused variable 'm1' [-Wunused-variable]
                        const float m1 = powf(2.0f, -(max_bias / 2.0f) / n_heads_log2_floor);
                                    ^
ggml-metal.m:2201:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_ALIBI_F32].pipeline;
                                                    ^
ggml-metal.m:2232:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN(1024, ne00);
                                  ^
ggml-metal.m:2234:35: warning: unused variable 'n_past' [-Wunused-variable]
                        const int n_past     = ((int32_t *) dst->op_params)[0];
                                  ^
ggml-metal.m:2235:35: warning: unused variable 'n_dims' [-Wunused-variable]
                        const int n_dims     = ((int32_t *) dst->op_params)[1];
                                  ^
ggml-metal.m:2236:35: warning: unused variable 'mode' [-Wunused-variable]
                        const int mode       = ((int32_t *) dst->op_params)[2];
                                  ^
ggml-metal.m:2238:35: warning: unused variable 'n_orig_ctx' [-Wunused-variable]
                        const int n_orig_ctx = ((int32_t *) dst->op_params)[4];
                                  ^
ggml-metal.m:2295:39: warning: unused variable 's0' [-Wunused-variable]
                        const int32_t s0 = ((const int32_t *)(dst->op_params))[0];
                                      ^
ggml-metal.m:2296:39: warning: unused variable 's1' [-Wunused-variable]
                        const int32_t s1 = ((const int32_t *)(dst->op_params))[1];
                                      ^
ggml-metal.m:2297:39: warning: unused variable 'p0' [-Wunused-variable]
                        const int32_t p0 = ((const int32_t *)(dst->op_params))[2];
                                      ^
ggml-metal.m:2298:39: warning: unused variable 'p1' [-Wunused-variable]
                        const int32_t p1 = ((const int32_t *)(dst->op_params))[3];
                                      ^
ggml-metal.m:2299:39: warning: unused variable 'd0' [-Wunused-variable]
                        const int32_t d0 = ((const int32_t *)(dst->op_params))[4];
                                      ^
ggml-metal.m:2300:39: warning: unused variable 'd1' [-Wunused-variable]
                        const int32_t d1 = ((const int32_t *)(dst->op_params))[5];
                                      ^
ggml-metal.m:2304:39: warning: unused variable 'N' [-Wunused-variable]
                        const int32_t N  = src1->ne[is_2D ? 3 : 2];
                                      ^
ggml-metal.m:2306:39: warning: unused variable 'IH' [-Wunused-variable]
                        const int32_t IH = is_2D ? src1->ne[1] : 1;
                                      ^
ggml-metal.m:2307:39: warning: unused variable 'IW' [-Wunused-variable]
                        const int32_t IW =         src1->ne[0];
                                      ^
ggml-metal.m:2312:39: warning: unused variable 'OH' [-Wunused-variable]
                        const int32_t OH = is_2D ? dst->ne[2] : 1;
                                      ^
ggml-metal.m:2313:39: warning: unused variable 'OW' [-Wunused-variable]
                        const int32_t OW =         dst->ne[1];
                                      ^
ggml-metal.m:2315:39: warning: unused variable 'CHW' [-Wunused-variable]
                        const int32_t CHW = IC * KH * KW;
                                      ^
ggml-metal.m:2317:39: warning: unused variable 'ofs0' [-Wunused-variable]
                        const int32_t ofs0 = src1->nb[is_2D ? 3 : 2] / 4;
                                      ^
ggml-metal.m:2318:39: warning: unused variable 'ofs1' [-Wunused-variable]
                        const int32_t ofs1 = src1->nb[is_2D ? 2 : 1] / 4;
                                      ^
ggml-metal.m:2349:35: warning: unused variable 'sf' [-Wunused-variable]
                        const int sf = dst->op_params[0];
                                  ^
ggml-metal.m:2374:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN((int) pipeline.maxTotalThreadsPerThreadgroup, ne0);
                                  ^
ggml-metal.m:2382:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_PAD_F32].pipeline;
                                                    ^
ggml-metal.m:2404:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN(1024, ne0);
                                  ^
ggml-metal.m:2418:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_ARANGE_F32].pipeline;
                                                    ^
ggml-metal.m:2426:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN(1024, ne0);
                                  ^
ggml-metal.m:2435:35: warning: unused variable 'max_period' [-Wunused-variable]
                        const int max_period = dst->op_params[1];
                                  ^
ggml-metal.m:2439:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_TIMESTEP_EMBEDDING_F32].pipeline;
                                                    ^
ggml-metal.m:2448:35: warning: unused variable 'nth' [-Wunused-variable]
                        const int nth = MIN(1024, half);
                                  ^
ggml-metal.m:2457:35: warning: unused variable 'nrows' [-Wunused-variable]
                        const int nrows = ggml_nrows(src0);
                                  ^
ggml-metal.m:2469:35: warning: unused variable 'mem_size' [-Wunused-variable]
                        const int mem_size = GGML_PAD(ne00_padded*sizeof(int32_t), 16);
                                  ^
ggml-metal.m:2495:53: warning: unused variable 'pipeline' [-Wunused-variable]
                        id<MTLComputePipelineState> pipeline = ctx->kernels[GGML_METAL_KERNEL_TYPE_LEAKY_RELU_F32].pipeline;
                                                    ^
ggml-metal.m:2502:39: warning: unused variable 'n' [-Wunused-variable]
                        const int64_t n = ggml_nelements(dst);
                                      ^
ggml-metal.m:2512:29: warning: unused variable 'nth' [-Wunused-variable]
                        int nth = MIN(1024, ne00/ggml_blck_size(src0->type));
                            ^
80 warnings and 9 errors generated.
make: *** [ggml-metal.o] Error 1

Does anyone know what I'm missing?

ggerganov · 2024-04-21T12:27:44Z

ggerganov
Apr 21, 2024
Maintainer

Try LLAMA_NO_METAL=1 make

5 replies

Spider-netizen Apr 21, 2024
Author

@ggerganov
Thanks for the prompt response. Building using LLAMA_NO_METAL=1 make seems to have worked, but why can't I build with Metal? I'd like to make use of it for faster generation...

arjansiddhpura Apr 21, 2024

@Spider-netizen
Do you have Xcode installed?

Spider-netizen Apr 21, 2024
Author

Thanks for the response @arjansiddhpura.
I do have Xcode installed. I'm also running macOS 10.15.7, if it matters.

Spider-netizen Apr 22, 2024
Author

@ggerganov
Any chance you can direct me how I can find a solution to this?

Thank you for the great work BTW.

oldgithubman Jul 9, 2024

@ggerganov Any chance you can direct me how I can find a solution to this?

...?

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Error Building llama.cpp on MacBook Pro #6806

Uh oh!

{{title}}

Uh oh!

Replies: 1 comment 5 replies

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

Error Building llama.cpp on MacBook Pro #6806

Uh oh!

Spider-netizen Apr 21, 2024

Replies: 1 comment · 5 replies

Uh oh!

ggerganov Apr 21, 2024 Maintainer

Uh oh!

Uh oh!

Spider-netizen Apr 21, 2024 Author

Uh oh!

arjansiddhpura Apr 21, 2024

Uh oh!

Spider-netizen Apr 21, 2024 Author

Uh oh!

Spider-netizen Apr 22, 2024 Author

Uh oh!

oldgithubman Jul 9, 2024

Spider-netizen
Apr 21, 2024

Replies: 1 comment 5 replies

ggerganov
Apr 21, 2024
Maintainer

Spider-netizen Apr 21, 2024
Author

Spider-netizen Apr 21, 2024
Author

Spider-netizen Apr 22, 2024
Author