[User] Error building server; possibly faulty libllama.so build; possible fix

[alk0]

Really sorry for not following the whole template, I just believe it will save some time.

I believe the current Makefile is missing ggml-quants-k.o in libllama.so: and server: targets. I couldn't build server before adding it, and, while libllama.so was built successfully, it seemed to be unusable. Adding ggml-quants-k.o to both target solved these two problems for me.

Environment: Linux, Arch, OpenBLAS, no-AVX CPU (although I don't think any of that matters).

[alk0]

Relevant output:

$ make -j2 LLAMA_OPENBLAS=1 LLAMA_BUILD_SERVER=1
I llama.cpp build info: 
I UNAME_S:  Linux
I UNAME_P:  unknown
I UNAME_M:  x86_64
I CFLAGS:   -I.              -O3 -std=c11   -fPIC -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wdouble-promotion -Wshadow -Wstrict-prototypes -Wpointer-arith -pthread -march=native -mtune=native -DGGML_USE_OPENBLAS -I/usr/local/include/openblas -I/usr/include/openblas
I CXXFLAGS: -I. -I./examples -O3 -std=c++11 -fPIC -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -pthread -march=native -mtune=native
I LDFLAGS:  -lopenblas -lcblas
I CC:       cc (GCC) 13.1.1 20230429
I CXX:      g++ (GCC) 13.1.1 20230429

g++ -I. -I./examples -O3 -std=c++11 -fPIC -DNDEBUG -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function -Wno-multichar -pthread -march=native -mtune=native -Iexamples/server examples/server/server.cpp ggml.o llama.o common.o -o server -lopenblas -lcblas
/usr/bin/ld: ggml.o: in function `ggml_quantize_chunk':
ggml.c:(.text+0x343b9): undefined reference to `ggml_quantize_q2_k'
/usr/bin/ld: ggml.c:(.text+0x343f0): undefined reference to `ggml_quantize_q3_k'
/usr/bin/ld: ggml.c:(.text+0x34427): undefined reference to `ggml_quantize_q4_k'
/usr/bin/ld: ggml.c:(.text+0x3445e): undefined reference to `ggml_quantize_q5_k'
/usr/bin/ld: ggml.c:(.text+0x34492): undefined reference to `ggml_quantize_q6_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x1e0): undefined reference to `dequantize_row_q2_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x1e8): undefined reference to `quantize_row_q2_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x1f0): undefined reference to `quantize_row_q2_k_reference'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x1f8): undefined reference to `quantize_row_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x200): undefined reference to `ggml_vec_dot_q2_k_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x210): undefined reference to `dequantize_row_q3_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x218): undefined reference to `quantize_row_q3_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x220): undefined reference to `quantize_row_q3_k_reference'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x228): undefined reference to `quantize_row_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x230): undefined reference to `ggml_vec_dot_q3_k_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x240): undefined reference to `dequantize_row_q4_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x248): undefined reference to `quantize_row_q4_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x250): undefined reference to `quantize_row_q4_k_reference'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x258): undefined reference to `quantize_row_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x260): undefined reference to `ggml_vec_dot_q4_k_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x270): undefined reference to `dequantize_row_q5_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x278): undefined reference to `quantize_row_q5_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x280): undefined reference to `quantize_row_q5_k_reference'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x288): undefined reference to `quantize_row_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x290): undefined reference to `ggml_vec_dot_q5_k_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x2a0): undefined reference to `dequantize_row_q6_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x2a8): undefined reference to `quantize_row_q6_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x2b0): undefined reference to `quantize_row_q6_k_reference'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x2b8): undefined reference to `quantize_row_q8_k'
/usr/bin/ld: ggml.o:(.data.rel.ro+0x2c0): undefined reference to `ggml_vec_dot_q6_k_q8_k'
collect2: error: ld returned 1 exit status
make: *** [Makefile:278: server] Error 1