Skip to content

vulkan: linux builds + small subgroup size fixes #11767

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Feb 14, 2025
Merged

vulkan: linux builds + small subgroup size fixes #11767

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9feb8e3
mm subgroup size
netrunnereve Feb 8, 2025
3cbdbe8
upload vulkan x86 builds
netrunnereve Feb 8, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 28 additions & 0 deletions .github/workflows/build.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -403,6 +403,34 @@ jobs:
# This is using llvmpipe and runs slower than other backends
ctest -L main --verbose --timeout 1800

- name: Determine tag name
id: tag
shell: bash
run: |
BUILD_NUMBER="$(git rev-list --count HEAD)"
SHORT_HASH="$(git rev-parse --short=7 HEAD)"
if [[ "${{ env.BRANCH_NAME }}" == "master" ]]; then
echo "name=b${BUILD_NUMBER}" >> $GITHUB_OUTPUT
else
SAFE_NAME=$(echo "${{ env.BRANCH_NAME }}" | tr '/' '-')
echo "name=${SAFE_NAME}-b${BUILD_NUMBER}-${SHORT_HASH}" >> $GITHUB_OUTPUT
fi

- name: Pack artifacts
id: pack_artifacts
if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
run: |
cp LICENSE ./build/bin/
cp examples/run/linenoise.cpp/LICENSE ./build/bin/LICENSE.linenoise.cpp
zip -r llama-${{ steps.tag.outputs.name }}-bin-ubuntu-vulkan-x64.zip ./build/bin/*

- name: Upload artifacts
if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
uses: actions/upload-artifact@v4
with:
path: llama-${{ steps.tag.outputs.name }}-bin-ubuntu-vulkan-x64.zip
name: llama-bin-ubuntu-vulkan-x64.zip

ubuntu-22-cmake-hip:
runs-on: ubuntu-22.04
container: rocm/dev-ubuntu-22.04:6.0.2
Expand Down
13 changes: 7 additions & 6 deletions ggml/src/ggml-vulkan/ggml-vulkan.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1395,6 +1395,7 @@ static void ggml_vk_load_shaders(vk_device& device) {
VK_LOG_DEBUG("ggml_vk_load_shaders(" << device->name << ")");

// some shaders have a minimum subgroup size
const uint32_t subgroup_size_8 = std::max(device->subgroup_size, 8u);
const uint32_t subgroup_size_16 = std::max(device->subgroup_size, 16u);
const uint32_t subgroup_size_32 = std::max(device->subgroup_size, 32u);

Expand Down Expand Up @@ -1457,13 +1458,13 @@ static void ggml_vk_load_shaders(vk_device& device) {
const uint32_t tk_m = device->coopmat_support ? device->coopmat_k : 1;
const uint32_t tk_s = device->coopmat_support ? device->coopmat_k : 1;

l_warptile = { 128, 128, 128, 16, device->subgroup_size * 2, 64, 2, tm_l, tn_l, tk_l, device->subgroup_size };
m_warptile = { 128, 64, 64, 16, device->subgroup_size, 32, 2, tm_m, tn_m, tk_m, device->subgroup_size };
s_warptile = { subgroup_size_16, 32, 32, 16, 32, 32, 2, tm_s, tn_s, tk_s, device->subgroup_size };
l_warptile = { 128, 128, 128, 16, subgroup_size_8 * 2, 64, 2, tm_l, tn_l, tk_l, subgroup_size_8 };
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I imagine the coopmat path doesn't handle faking the subgroup size, maybe add an assert to that effect? Coopmat implementations probably have at least 8 invocations per subgroup, so this seems fine.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Does our coopmat shader even work with a subgroup size of 8? We should probably find the actual limit and set up the assert based on that.

Honestly I don't know exactly why the regular mul_mat shaders break down with a subgroup size less than 8, but with the Vulkan backend becoming more and more popular I'd rather have it run slowly than fail mysteriously on those devices.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The warptile parameters are not independent. There is probably a minimum there, coming from the hardcoded values.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm going to merge this as I've confirmed that 8 is the minimum requirement on several different systems, including an ARM Llvmpipe machine which has a subgroup size of 4. If we manage to figure out why this is happening and fix the shaders we can then remove the subgroup_size_8.

m_warptile = { 128, 64, 64, 16, subgroup_size_8, 32, 2, tm_m, tn_m, tk_m, subgroup_size_8 };
s_warptile = { subgroup_size_16, 32, 32, 16, 32, 32, 2, tm_s, tn_s, tk_s, subgroup_size_8 };

l_warptile_mmq = { 128, 128, 128, 32, device->subgroup_size * 2, 64, 2, tm_l, tn_l, tk_l, device->subgroup_size };
m_warptile_mmq = { 128, 64, 64, 32, device->subgroup_size, 32, 2, tm_m, tn_m, tk_m, device->subgroup_size };
s_warptile_mmq = { subgroup_size_32, 32, 32, 32, 32, 32, 2, tm_s, tn_s, tk_s, device->subgroup_size };
l_warptile_mmq = { 128, 128, 128, 32, subgroup_size_8 * 2, 64, 2, tm_l, tn_l, tk_l, subgroup_size_8 };
m_warptile_mmq = { 128, 64, 64, 32, subgroup_size_8, 32, 2, tm_m, tn_m, tk_m, subgroup_size_8 };
s_warptile_mmq = { subgroup_size_32, 32, 32, 32, 32, 32, 2, tm_s, tn_s, tk_s, subgroup_size_8 };

l_mmq_wg_denoms = l_wg_denoms = {128, 128, 1 };
m_mmq_wg_denoms = m_wg_denoms = { 64, 64, 1 };
Expand Down
Loading