sleepy/llama.cpp
1
0
Fork 0
Code Issues 11 Pull Requests Actions 109 Packages Projects Releases Wiki Activity

ci : re-enable mac workflows (#21894)

Browse Source 
* ci : re-enable mac workflows

* vulkan : fix compile warning
This commit is contained in:
Georgi Gerganov
2026-04-14 15:58:09 +03:00
committed by GitHub
parent aa0f1897b7
commit f4b5bf2f32
3 changed files with 93 additions and 94 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/build-self-hosted.yml
+53 -55
View File
@@ -141,61 +141,59 @@ jobs:
# amd-smi static # amd-smi static
# GG_BUILD_ROCM=1 GG_BUILD_AMDGPU_TARGETS="gfx1101" bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp # GG_BUILD_ROCM=1 GG_BUILD_AMDGPU_TARGETS="gfx1101" bash ./ci/run.sh ~/results/llama.cpp /mnt/llama.cpp
# TODO: sandbox Mac runners ggml-ci-mac-metal:
# ggml-ci-mac-metal: runs-on: [self-hosted, macOS, ARM64]
# runs-on: [self-hosted, macOS, ARM64]
# steps:
# steps: - name: Clone
# - name: Clone id: checkout
# id: checkout uses: actions/checkout@v6
# uses: actions/checkout@v6
# - name: Test
# - name: Test id: ggml-ci
# id: ggml-ci run: |
# run: | GG_BUILD_METAL=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
# GG_BUILD_METAL=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
# ggml-ci-mac-webgpu:
# ggml-ci-mac-webgpu: runs-on: [self-hosted, macOS, ARM64]
# runs-on: [self-hosted, macOS, ARM64]
# steps:
# steps: - name: Clone
# - name: Clone id: checkout
# id: checkout uses: actions/checkout@v6
# uses: actions/checkout@v6
# - name: Dawn Dependency
# - name: Dawn Dependency id: dawn-depends
# id: dawn-depends run: |
# run: | DAWN_VERSION="v20260317.182325"
# DAWN_VERSION="v2.0.0" DAWN_OWNER="google"
# DAWN_OWNER="reeselevine" DAWN_REPO="dawn"
# DAWN_REPO="dawn" DAWN_ASSET_NAME="Dawn-18eb229ef5f707c1464cc581252e7603c73a3ef0-macos-latest-Release"
# DAWN_ASSET_NAME="Dawn-5e9a4865b1635796ccc77dd30057f2b4002a1355-macos-latest-Release" echo "Fetching release asset from https://github.com/google/dawn/releases/download/${DAWN_VERSION}/${DAWN_ASSET_NAME}.tar.gz"
# echo "Fetching release asset from https://github.com/${DAWN_OWNER}/${DAWN_REPO}/releases/download/${DAWN_VERSION}/${DAWN_ASSET_NAME}.zip" curl -L -o artifact.tar.gz \
# curl -L -o artifact.zip \ "https://github.com/google/dawn/releases/download/${DAWN_VERSION}/${DAWN_ASSET_NAME}.tar.gz"
# "https://github.com/${DAWN_OWNER}/${DAWN_REPO}/releases/download/${DAWN_VERSION}/${DAWN_ASSET_NAME}.zip" mkdir dawn
# mkdir dawn tar -xvf artifact.tar.gz -C dawn --strip-components=1
# unzip artifact.zip
# tar -xvf ${DAWN_ASSET_NAME}.tar.gz -C dawn --strip-components=1 - name: Test
# id: ggml-ci
# - name: Test run: |
# id: ggml-ci GG_BUILD_WEBGPU=1 GG_BUILD_WEBGPU_DAWN_PREFIX="$GITHUB_WORKSPACE/dawn" \
# run: | bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
# GG_BUILD_WEBGPU=1 GG_BUILD_WEBGPU_DAWN_PREFIX="$GITHUB_WORKSPACE/dawn" \
# bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp ggml-ci-mac-vulkan:
# runs-on: [self-hosted, macOS, ARM64]
# ggml-ci-mac-vulkan:
# runs-on: [self-hosted, macOS, ARM64] steps:
# - name: Clone
# steps: id: checkout
# - name: Clone uses: actions/checkout@v6
# id: checkout
# uses: actions/checkout@v6 - name: Test
# id: ggml-ci
# - name: Test run: |
# id: ggml-ci vulkaninfo --summary
# run: | GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
# vulkaninfo --summary
# GG_BUILD_VULKAN=1 bash ./ci/run.sh ~/results/llama.cpp ~/mnt/llama.cpp
ggml-ci-linux-intel-vulkan: ggml-ci-linux-intel-vulkan:
runs-on: [self-hosted, Linux, Intel] runs-on: [self-hosted, Linux, Intel]
.github/workflows/server-self-hosted.yml
+39 -38
View File
@@ -84,41 +84,42 @@ jobs:
export ${{ matrix.extra_args }} export ${{ matrix.extra_args }}
pytest -v -x -m "not slow" pytest -v -x -m "not slow"
server-cuda: # TODO: provision CUDA runner
runs-on: [self-hosted, llama-server, Linux, NVIDIA] # server-cuda:
# runs-on: [self-hosted, llama-server, Linux, NVIDIA]
name: server-cuda (${{ matrix.wf_name }}) #
strategy: # name: server-cuda (${{ matrix.wf_name }})
matrix: # strategy:
build_type: [Release] # matrix:
wf_name: ["GPUx1"] # build_type: [Release]
include: # wf_name: ["GPUx1"]
- build_type: Release # include:
extra_args: "LLAMA_ARG_BACKEND_SAMPLING=1" # - build_type: Release
wf_name: "GPUx1, backend-sampling" # extra_args: "LLAMA_ARG_BACKEND_SAMPLING=1"
fail-fast: false # wf_name: "GPUx1, backend-sampling"
# fail-fast: false
steps: #
- name: Clone # steps:
id: checkout # - name: Clone
uses: actions/checkout@v6 # id: checkout
with: # uses: actions/checkout@v6
fetch-depth: 0 # with:
ref: ${{ github.event.inputs.sha || github.event.pull_request.head.sha || github.sha || github.head_ref || github.ref_name }} # fetch-depth: 0
# ref: ${{ github.event.inputs.sha || github.event.pull_request.head.sha || github.sha || github.head_ref || github.ref_name }}
- name: Build #
id: cmake_build # - name: Build
run: | # id: cmake_build
cmake -B build -DGGML_SCHED_NO_REALLOC=ON # run: |
cmake --build build --config ${{ matrix.build_type }} -j $(sysctl -n hw.logicalcpu) --target llama-server # cmake -B build -DGGML_SCHED_NO_REALLOC=ON
# cmake --build build --config ${{ matrix.build_type }} -j $(sysctl -n hw.logicalcpu) --target llama-server
- name: Tests #
id: server_integration_tests # - name: Tests
if: ${{ (!matrix.disabled_on_pr || !github.event.pull_request) }} # id: server_integration_tests
run: | # if: ${{ (!matrix.disabled_on_pr || !github.event.pull_request) }}
cd tools/server/tests # run: |
python3 -m venv venv # cd tools/server/tests
source venv/bin/activate # python3 -m venv venv
pip install -r requirements.txt # source venv/bin/activate
export ${{ matrix.extra_args }} # pip install -r requirements.txt
pytest -v -x -m "not slow" # export ${{ matrix.extra_args }}
# pytest -v -x -m "not slow"
ggml/src/ggml-webgpu/ggml-webgpu.cpp
+1 -1
View File
@@ -3485,7 +3485,7 @@ static bool create_webgpu_device(ggml_backend_webgpu_reg_context * ctx) {
dev_desc.requiredFeatureCount = required_features.size(); dev_desc.requiredFeatureCount = required_features.size();
dev_desc.SetDeviceLostCallback( dev_desc.SetDeviceLostCallback(
wgpu::CallbackMode::AllowSpontaneous, wgpu::CallbackMode::AllowSpontaneous,
[ctx](const wgpu::Device & device, wgpu::DeviceLostReason reason, wgpu::StringView message) { [](const wgpu::Device & device, wgpu::DeviceLostReason reason, wgpu::StringView message) {
if (reason == wgpu::DeviceLostReason::Destroyed) { if (reason == wgpu::DeviceLostReason::Destroyed) {
return; return;
} }