github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-15 12:48:05 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #4573] Update llama.cpp to b2938 or newer to fix Vulkan build #2868

New Issue
Closed
opened 2026-04-12 13:13:09 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 13:13:09 -05:00
Owner
Copy Link

Originally created by @dreirund on GitHub (May 22, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4573

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

Can it be that you use some older commit of llama.cpp?

Building with Vulkan (and testing options LLAMA_VULKAN_CHECK_RESULTS=ON and LLAMA_VULKAN_RUN_TESTS=ON), I get the error ggml-vulkan.cpp:6880:80: error: cannot convert ‘ggml_tensor*’ to ‘float’. ↗ llama.cpp upstream says that it is already fixed there, but the fix does not seem to have arrived in ollama.

Details:

I am on Artix GNU/Linux (rolling release), GCC 14.1.1, and I build ollama-vulkan which pulls in and uses llama.cpp from it's git repository.

When building, I get the error
ggml-vulkan.cpp:6880:80: error: cannot convert ‘ggml_tensor*’ to ‘float’:

[...]
+ init_vars
+ case "${GOARCH}" in
+ ARCH=x86_64
+ LLAMACPP_DIR=../llama.cpp
+ CMAKE_DEFS=
+ CMAKE_TARGETS='--target ollama_llama_server'
+ echo ''
+ grep -- -g
+ CMAKE_DEFS='-DCMAKE_BUILD_TYPE=Release -DLLAMA_SERVER_VERBOSE=off '
+ case $(uname -s) in
++ uname -s
+ LIB_EXT=so
+ WHOLE_ARCHIVE=-Wl,--whole-archive
+ NO_WHOLE_ARCHIVE=-Wl,--no-whole-archive
+ GCC_ARCH=
+ '[' -z '50;52;61;70;75;80' ']'
+ echo 'OLLAMA_CUSTOM_CPU_DEFS="
  -DBUILD_TESTING=ON
  -DCMAKE_BUILD_TYPE=Release
  -DCMAKE_INSTALL_PREFIX=/usr
  -DLLAMA_ACCELERATE=ON
  -DLLAMA_ALL_WARNINGS=OFF
  -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF
  -DLLAMA_FATAL_WARNINGS=OFF
  -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON
  -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON
  -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF
  -DLLAMA_LTO=OFF
  -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF 
  -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON
 -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF"'
OLLAMA_CUSTOM_CPU_DEFS="
  -DBUILD_TESTING=ON
  -DCMAKE_BUILD_TYPE=Release
  -DCMAKE_INSTALL_PREFIX=/usr
  -DLLAMA_ACCELERATE=ON
  -DLLAMA_ALL_WARNINGS=OFF
  -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF
  -DLLAMA_FATAL_WARNINGS=OFF
  -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON
  -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON
  -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF
  -DLLAMA_LTO=OFF
  -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF 
  -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON
 -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF"
+ CMAKE_DEFS='
  -DBUILD_TESTING=ON
  -DCMAKE_BUILD_TYPE=Release
  -DCMAKE_INSTALL_PREFIX=/usr
  -DLLAMA_ACCELERATE=ON
  -DLLAMA_ALL_WARNINGS=OFF
  -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF
  -DLLAMA_FATAL_WARNINGS=OFF
  -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON
  -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON
  -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF
  -DLLAMA_LTO=OFF
  -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF 
  -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON
 -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF -DCMAKE_POSITION_INDEPENDENT_CODE=on -DCMAKE_BUILD_TYPE=Release -DLLAMA_SERVER_VERBOSE=off '
+ BUILD_DIR=../build/linux/x86_64/cpu
+ echo 'Building custom CPU'
Building custom CPU
+ build
+ cmake -S ../llama.cpp -B ../build/linux/x86_64/cpu -DBUILD_TESTING=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr -DLLAMA_ACCELERATE=ON -DLLAMA_ALL_WARNINGS=OFF -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF -DLLAMA_FATAL_WARNINGS=OFF -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF -DLLAMA_LTO=OFF -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF -DCMAKE_POSITION_INDEPENDENT_CODE=on -DCMAKE_BUILD_TYPE=Release -DLLAMA_SERVER_VERBOSE=off
-- The C compiler identification is GNU 14.1.1
-- The CXX compiler identification is GNU 14.1.1
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/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: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: /usr/bin/git (found version "2.45.1")
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE
-- Found Vulkan: /lib/libvulkan.so (found version "1.3.285") found components: glslc glslangValidator
-- Vulkan found
-- ccache found, compilation results will be cached. Disable with LLAMA_CCACHE=OFF.
-- CMAKE_SYSTEM_PROCESSOR: x86_64
-- x86 detected
-- Found OpenSSL: /usr/lib/libcrypto.so (found version "3.3.0")
-- Configuring done (0.6s)
-- Generating done (0.1s)
-- Build files have been written to: /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/build/linux/x86_64/cpu
+ cmake --build ../build/linux/x86_64/cpu --target ollama_llama_server -j8
[  6%] Generating build details from Git
[ 20%] Building C object CMakeFiles/ggml.dir/ggml-alloc.c.o
[ 20%] Building C object CMakeFiles/ggml.dir/ggml.c.o
[ 20%] Building C object CMakeFiles/ggml.dir/ggml-backend.c.o
[ 26%] Building C object CMakeFiles/ggml.dir/ggml-quants.c.o
[ 26%] Building CXX object CMakeFiles/ggml.dir/sgemm.cpp.o
[ 33%] Building CXX object CMakeFiles/ggml.dir/ggml-vulkan.cpp.o
-- Found Git: /usr/bin/git (found version "2.45.1")
[ 33%] Building CXX object common/CMakeFiles/build_info.dir/build-info.cpp.o
[ 33%] Built target build_info
/var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp: In function ‘void ggml_vk_soft_max(ggml_backend_vk_context*, vk_context*, const ggml_tensor*, const ggml_tensor*, const ggml_tensor*, ggml_tensor*)’:
/var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:4288:119: note: ‘#pragma message: TODO: src2 is no longer used in soft_max - should be removed and ALiBi calculation should be updated’
 4288 | #pragma message("TODO: src2 is no longer used in soft_max - should be removed and ALiBi calculation should be updated")
      |                                                                                                                       ^
/var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:4289:73: note: ‘#pragma message: ref:  https://github.com/ggerganov/llama.cpp/pull/7192’
 4289 | #pragma message("ref:  https://github.com/ggerganov/llama.cpp/pull/7192")
      |                                                                         ^
/var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp: In function ‘void ggml_vk_check_results_0(ggml_backend_vk_context*, ggml_compute_params*, ggml_tensor*)’:
/var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:6880:80: error: cannot convert ‘ggml_tensor*’ to ‘float’
 6880 |             tensor_clone = ggml_soft_max_ext(ggml_ctx, src0_clone, src1_clone, src2_clone, ((float *)tensor->op_params)[0], ((float *)tensor->op_params)[1]);
      |                                                                                ^~~~~~~~~~
      |                                                                                |
      |                                                                                ggml_tensor*
In file included from /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.h:3,
                 from /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:1:
/var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml.h:1446:35: note:   initializing argument 4 of ‘ggml_tensor* ggml_soft_max_ext(ggml_context*, ggml_tensor*, ggml_tensor*, float, float)’
 1446 |             float                 scale,
      |             ~~~~~~~~~~~~~~~~~~~~~~^~~~~
make[3]: *** [CMakeFiles/ggml.dir/build.make:132: CMakeFiles/ggml.dir/ggml-vulkan.cpp.o] Error 1
make[2]: *** [CMakeFiles/Makefile2:838: CMakeFiles/ggml.dir/all] Error 2
make[1]: *** [CMakeFiles/Makefile2:3322: ext_server/CMakeFiles/ollama_llama_server.dir/rule] Error 2
make: *** [Makefile:1336: ollama_llama_server] Error 2
llm/generate/generate_linux.go:3: running "bash": exit status 2

Regards!

OS

Artix GNU/Linux

GPU

AMD (gfx1103)

CPU

AMD (7840U)

Ollama version

  • ollama --version: ollama version is 0.0.0,
  • git --describe tags: v0.1.39-rc1-4-g955c317c,
  • git checkout from 2024-05-22, git commit hash 955c317c.
Originally created by @dreirund on GitHub (May 22, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4573 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? Can it be that you use some older commit of `llama.cpp`? Building with Vulkan (and testing options `LLAMA_VULKAN_CHECK_RESULTS=ON` and `LLAMA_VULKAN_RUN_TESTS=ON`), I get the error `ggml-vulkan.cpp:6880:80: error: cannot convert ‘ggml_tensor*’ to ‘float’`. [↗ `llama.cpp` upstream says that it is already fixed there](https://github.com/ggerganov/llama.cpp/issues/7446#issuecomment-2124212932), but the fix does not seem to have arrived in `ollama`. --- Details: I am on [Artix GNU/Linux](http://artixlinux.org/) (rolling release), GCC 14.1.1, and I build [`ollama-vulkan`](https://aur.archlinux.org/pkgbase/ollama-nogpu-git) which pulls in and uses [`llama.cpp` from it's git repository](https://github.com/ggerganov/llama.cpp). When building, I get the error `ggml-vulkan.cpp:6880:80: error: cannot convert ‘ggml_tensor*’ to ‘float’`: ``` [...] + init_vars + case "${GOARCH}" in + ARCH=x86_64 + LLAMACPP_DIR=../llama.cpp + CMAKE_DEFS= + CMAKE_TARGETS='--target ollama_llama_server' + echo '' + grep -- -g + CMAKE_DEFS='-DCMAKE_BUILD_TYPE=Release -DLLAMA_SERVER_VERBOSE=off ' + case $(uname -s) in ++ uname -s + LIB_EXT=so + WHOLE_ARCHIVE=-Wl,--whole-archive + NO_WHOLE_ARCHIVE=-Wl,--no-whole-archive + GCC_ARCH= + '[' -z '50;52;61;70;75;80' ']' + echo 'OLLAMA_CUSTOM_CPU_DEFS=" -DBUILD_TESTING=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr -DLLAMA_ACCELERATE=ON -DLLAMA_ALL_WARNINGS=OFF -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF -DLLAMA_FATAL_WARNINGS=OFF -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF -DLLAMA_LTO=OFF -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF"' OLLAMA_CUSTOM_CPU_DEFS=" -DBUILD_TESTING=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr -DLLAMA_ACCELERATE=ON -DLLAMA_ALL_WARNINGS=OFF -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF -DLLAMA_FATAL_WARNINGS=OFF -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF -DLLAMA_LTO=OFF -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF" + CMAKE_DEFS=' -DBUILD_TESTING=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr -DLLAMA_ACCELERATE=ON -DLLAMA_ALL_WARNINGS=OFF -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF -DLLAMA_FATAL_WARNINGS=OFF -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF -DLLAMA_LTO=OFF -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF -DCMAKE_POSITION_INDEPENDENT_CODE=on -DCMAKE_BUILD_TYPE=Release -DLLAMA_SERVER_VERBOSE=off ' + BUILD_DIR=../build/linux/x86_64/cpu + echo 'Building custom CPU' Building custom CPU + build + cmake -S ../llama.cpp -B ../build/linux/x86_64/cpu -DBUILD_TESTING=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr -DLLAMA_ACCELERATE=ON -DLLAMA_ALL_WARNINGS=OFF -DLLAMA_ALL_WARNINGS_3RD_PARTY=OFF -DLLAMA_FATAL_WARNINGS=OFF -DLLAMA_AVX=ON -DLLAMA_AVX2=ON -DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON -DLLAMA_F16C=ON -DLLAMA_FMA=ON -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=ON -DLLAMA_BUILD_TESTS=ON -DLLAMA_CPU_HBM=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_CUDA=OFF -DLLAMA_HIPBLAS=OFF -DLLAMA_HIP_UMA=OFF -DLLAMA_METAL=OFF -DLLAMA_SYCL=OFF -DLLAMA_KOMPUTE=OFF -DLLAMA_LTO=OFF -DLLAMA_GPROF=OFF -DLLAMA_PERF=OFF -DLLAMA_SANITIZE_ADDRESS=OFF -DLLAMA_SANITIZE_THREAD=OFF -DLLAMA_SANITIZE_UNDEFINED=OFF -DLLAMA_SERVER_SSL=ON -DLLAMA_SERVER_VERBOSE=ON -DLLAMA_VULKAN=ON -DLLAMA_VULKAN_CHECK_RESULTS=ON -DLLAMA_VULKAN_DEBUG=OFF -DLLAMA_VULKAN_RUN_TESTS=ON -DLLAMA_VULKAN_VALIDATE=OFF -DCMAKE_POSITION_INDEPENDENT_CODE=on -DCMAKE_BUILD_TYPE=Release -DLLAMA_SERVER_VERBOSE=off -- The C compiler identification is GNU 14.1.1 -- The CXX compiler identification is GNU 14.1.1 -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working C compiler: /usr/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: /usr/bin/c++ - skipped -- Detecting CXX compile features -- Detecting CXX compile features - done -- Found Git: /usr/bin/git (found version "2.45.1") -- Performing Test CMAKE_HAVE_LIBC_PTHREAD -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success -- Found Threads: TRUE -- Found Vulkan: /lib/libvulkan.so (found version "1.3.285") found components: glslc glslangValidator -- Vulkan found -- ccache found, compilation results will be cached. Disable with LLAMA_CCACHE=OFF. -- CMAKE_SYSTEM_PROCESSOR: x86_64 -- x86 detected -- Found OpenSSL: /usr/lib/libcrypto.so (found version "3.3.0") -- Configuring done (0.6s) -- Generating done (0.1s) -- Build files have been written to: /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/build/linux/x86_64/cpu + cmake --build ../build/linux/x86_64/cpu --target ollama_llama_server -j8 [ 6%] Generating build details from Git [ 20%] Building C object CMakeFiles/ggml.dir/ggml-alloc.c.o [ 20%] Building C object CMakeFiles/ggml.dir/ggml.c.o [ 20%] Building C object CMakeFiles/ggml.dir/ggml-backend.c.o [ 26%] Building C object CMakeFiles/ggml.dir/ggml-quants.c.o [ 26%] Building CXX object CMakeFiles/ggml.dir/sgemm.cpp.o [ 33%] Building CXX object CMakeFiles/ggml.dir/ggml-vulkan.cpp.o -- Found Git: /usr/bin/git (found version "2.45.1") [ 33%] Building CXX object common/CMakeFiles/build_info.dir/build-info.cpp.o [ 33%] Built target build_info /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp: In function ‘void ggml_vk_soft_max(ggml_backend_vk_context*, vk_context*, const ggml_tensor*, const ggml_tensor*, const ggml_tensor*, ggml_tensor*)’: /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:4288:119: note: ‘#pragma message: TODO: src2 is no longer used in soft_max - should be removed and ALiBi calculation should be updated’ 4288 | #pragma message("TODO: src2 is no longer used in soft_max - should be removed and ALiBi calculation should be updated") | ^ /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:4289:73: note: ‘#pragma message: ref: https://github.com/ggerganov/llama.cpp/pull/7192’ 4289 | #pragma message("ref: https://github.com/ggerganov/llama.cpp/pull/7192") | ^ /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp: In function ‘void ggml_vk_check_results_0(ggml_backend_vk_context*, ggml_compute_params*, ggml_tensor*)’: /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:6880:80: error: cannot convert ‘ggml_tensor*’ to ‘float’ 6880 | tensor_clone = ggml_soft_max_ext(ggml_ctx, src0_clone, src1_clone, src2_clone, ((float *)tensor->op_params)[0], ((float *)tensor->op_params)[1]); | ^~~~~~~~~~ | | | ggml_tensor* In file included from /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.h:3, from /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml-vulkan.cpp:1: /var/cache/makepkg/build/ollama-nogpu-git/src/ollama-vulkan/llm/llama.cpp/ggml.h:1446:35: note: initializing argument 4 of ‘ggml_tensor* ggml_soft_max_ext(ggml_context*, ggml_tensor*, ggml_tensor*, float, float)’ 1446 | float scale, | ~~~~~~~~~~~~~~~~~~~~~~^~~~~ make[3]: *** [CMakeFiles/ggml.dir/build.make:132: CMakeFiles/ggml.dir/ggml-vulkan.cpp.o] Error 1 make[2]: *** [CMakeFiles/Makefile2:838: CMakeFiles/ggml.dir/all] Error 2 make[1]: *** [CMakeFiles/Makefile2:3322: ext_server/CMakeFiles/ollama_llama_server.dir/rule] Error 2 make: *** [Makefile:1336: ollama_llama_server] Error 2 llm/generate/generate_linux.go:3: running "bash": exit status 2 ``` Regards! ### OS Artix GNU/Linux ### GPU AMD (gfx1103) ### CPU AMD (7840U) ### Ollama version * `ollama --version`: `ollama version is 0.0.0`, * `git --describe tags`: `v0.1.39-rc1-4-g955c317c`, * git checkout from 2024-05-22, git commit hash 955c317c.
GiteaMirror added the bug label 2026-04-12 13:13:09 -05:00
GiteaMirror commented 2026-04-12 13:13:10 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (May 22, 2024):

We're currently at b2871. We'll be updating soon...

<!-- gh-comment-id:2125657716 --> @dhiltgen commented on GitHub (May 22, 2024): We're currently at b2871. We'll be updating soon...
GiteaMirror commented 2026-04-12 13:13:11 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Jun 21, 2024):

Forgot to close this a while back after we updated llama.cpp

<!-- gh-comment-id:2183579609 --> @dhiltgen commented on GitHub (Jun 21, 2024): Forgot to close this a while back after we updated llama.cpp
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#2868
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?