Split CI Workflows by Platform and Task, Improve Formatting and Linting Reliability #108
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
|
the |
Member
Author
Yes, but we also have other tasks such as Documentation and Building Docker containers. I like it ;)
|
Contributor
|
Could you maybe also add a yml linter while we are restructuring all the yml files? |
Member
Author
I could, but this is also integrated in GitHub. If you make a mistake, it will complain. But I guess it could be good for formatting. Any suggestions on what tool to use? |
Contributor
|
Yeah, I was more thinking of formatting. With my shallow googling I found this action: https://github.com/marketplace/actions/yamllint-github-action but I have no experience |
Contributor
📝 WalkthroughSummary by CodeRabbit
WalkthroughReplaces a monolithic CI with modular GitHub Actions: adds a reusable environment selector, per‑platform CI workflows, lint/infra workflows, runner templates, and many license/header and formatting updates; removes the legacy central CI and GitLab templates. (39 words) Changes
Sequence Diagram(s)sequenceDiagram
autonumber
actor Dev as Developer
participant GH as GitHub Actions
participant Select as _select-env (reusable)
participant Runner as _runner-* (reusable)
participant Ctn as Container Image
participant Repo as Repository
participant Tests as DeeployTest
Dev->>GH: push / PR / tag / dispatch
GH->>Select: workflow_call(docker_image_deeploy)
Note right of Select: compute IMAGE from input/ref/tag/main/devel\ncompute RUNNER (self-hosted vs ubuntu)
Select-->>GH: outputs { image, runner }
loop per-platform job
GH->>Runner: workflow_call(runner=${runner}, docker-image=${image}, tests...)
Runner->>Ctn: run inside container ${image}
Runner->>Repo: checkout (submodules)
Runner->>Repo: pip install -e . (shell: bash)
Runner->>Tests: execute test scripts (params from workflow)
Tests-->>Runner: results & logs
Runner-->>GH: job status
end
GH-->>Dev: status & logs
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Possibly related PRs
Suggested reviewers
Pre-merge checks (3 passed)✅ Passed checks (3 passed)
✨ Finishing touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 39
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (15)
.github/workflows/_runner-softhier.yml (1)
18-21: Fix expressions for inputs with hyphens (use bracket notation).
${{ inputs.docker-image }}and${{ inputs.test-names }}will not parse; hyphenated keys require bracket notation.- container: - image: ${{ inputs.docker-image }} + container: + image: ${{ inputs['docker-image'] }} @@ - testNames="${{ inputs.test-names }}" + testNames="${{ inputs['test-names'] }}"Also applies to: 26-31
.github/workflows/_runner-chimera.yml (1)
21-24: Fix expressions for inputs with hyphens (use bracket notation).
${{ inputs.docker-image }}and${{ inputs.test-names }}must use brackets.- container: - image: ${{ inputs.docker-image }} + container: + image: ${{ inputs['docker-image'] }} @@ - testNames="${{ inputs.test-names }}" + testNames="${{ inputs['test-names'] }}"Also applies to: 29-31, 38-55
.github/workflows/_runner-snitch-tiled-sequential.yml (1)
38-41: Fix all hyphenated input references.These expressions won’t evaluate with dot notation:
docker-image,tests-config,num-cores,default-memory-level,memory-allocation-strategy,search-strategy. Use bracket notation.- container: - image: ${{ inputs.docker-image }} + container: + image: ${{ inputs['docker-image'] }} @@ - - name: Install jq - run: apt-get install -y jq + - name: Install jq + run: | + apt-get update + DEBIAN_FRONTEND=noninteractive apt-get install -y jq @@ - echo '${{ inputs.tests-config }}' > tests.json + echo '${{ inputs['tests-config'] }}' > tests.json @@ - python testRunner_tiled_snitch.py -t Tests/$testName --cores=${{ inputs.num-cores }} --simulator=$simulator --l1 $L1_value --defaultMemLevel=${{ inputs.default-memory-level }} --memAllocStrategy=${{ inputs.memory-allocation-strategy }} --searchStrategy=${{ inputs.search-strategy }} + python testRunner_tiled_snitch.py -t Tests/$testName --cores=${{ inputs['num-cores'] }} --simulator=$simulator --l1 $L1_value --defaultMemLevel=${{ inputs['default-memory-level'] }} --memAllocStrategy=${{ inputs['memory-allocation-strategy'] }} --searchStrategy=${{ inputs['search-strategy'] }}Also applies to: 46-51, 57-76
.github/workflows/_runner-cortexm.yml (1)
18-21: Fix expressions for inputs with hyphens (use bracket notation).- container: - image: ${{ inputs.docker-image }} + container: + image: ${{ inputs['docker-image'] }} @@ - testNames="${{ inputs.test-names }}" + testNames="${{ inputs['test-names'] }}"Also applies to: 26-29, 35-46
.github/workflows/_runner-generic.yml (1)
18-21: Fix expressions for inputs with hyphens (use bracket notation).- container: - image: ${{ inputs.docker-image }} + container: + image: ${{ inputs['docker-image'] }} @@ - testNames="${{ inputs.test-names }}" + testNames="${{ inputs['test-names'] }}"Also applies to: 26-29, 35-47
.github/workflows/_runner-siracusa-neureka-tiled.yml (1)
52-55: Fix all hyphenated input references and boolean flags.Use bracket notation for
docker-image,default-memory-level,memory-allocation-strategy,search-strategy,double-buffer,neureka-wmem; current dot notation will break expression parsing.- container: - image: ${{ inputs.docker-image }} + container: + image: ${{ inputs['docker-image'] }} @@ - - name: Run Test + - name: Run Test uses: nick-fields/retry@v3 with: timeout_minutes: 5 max_attempts: 3 retry_on: timeout command: | cd DeeployTest mkdir -p /app/.ccache export CCACHE_DIR=/app/.ccache - python testRunner_tiled_siracusa_w_neureka.py -t Tests/${{ inputs.test-name }} --cores=${{ inputs.num-cores }} --l1 ${{ matrix.L1 }} --defaultMemLevel=${{ inputs.default-memory-level }} ${{ inputs.double-buffer && '--doublebuffer' || '' }} ${{ inputs.neureka-wmem && '--neureka-wmem' || '' }} --memAllocStrategy=${{ inputs.memory-allocation-strategy }} --searchStrategy=${{ inputs.search-strategy }} + python testRunner_tiled_siracusa_w_neureka.py -t Tests/${{ inputs.test-name }} --cores=${{ inputs['num-cores'] }} --l1 ${{ matrix.L1 }} --defaultMemLevel=${{ inputs['default-memory-level'] }} ${{ inputs['double-buffer'] && '--doublebuffer' || '' }} ${{ inputs['neureka-wmem'] && '--neureka-wmem' || '' }} --memAllocStrategy=${{ inputs['memory-allocation-strategy'] }} --searchStrategy=${{ inputs['search-strategy'] }}Also applies to: 56-63, 69-79
.github/workflows/_runner-snitch.yml (1)
24-27: Fix expressions for inputs with hyphens (use bracket notation).- container: - image: ${{ inputs.docker-image }} + container: + image: ${{ inputs['docker-image'] }} @@ - testNames="${{ inputs.test-names }}" + testNames="${{ inputs['test-names'] }}" @@ - python testRunner_snitch.py -t Tests/$testName --simulator=$simulator --cores=${{ inputs.num-cores }} + python testRunner_snitch.py -t Tests/$testName --simulator=$simulator --cores=${{ inputs['num-cores'] }}Also applies to: 32-35, 41-57
.github/workflows/infra-generate-documentation.yml (2)
29-33: Pin third-party action to a commit SHA (avoid mutable branch ref).Using a PR branch ref is brittle; pin to a commit SHA or a released tag.
- - name: Prepare Multipages - uses: xeratec/gh-pages-multibranch@pr/support_tags + - name: Prepare Multipages + uses: xeratec/gh-pages-multibranch@<commit-sha>
34-40: Upgrade and pin gh-pages action.Use the maintained v4 and pin to SHA.
- - name: Deploy to GitHub Pages - uses: peaceiris/actions-gh-pages@v3 + - name: Deploy to GitHub Pages + uses: peaceiris/actions-gh-pages@<v4-commit-sha> if: ${{ github.event_name == 'push' || github.event_name == 'workflow_dispatch'}} with: publish_branch: gh-pages github_token: ${{ secrets.GITHUB_TOKEN }} publish_dir: _build/ force_orphan: true.github/workflows/_runner-siracusa-tiled.yml (2)
64-75: Invalid ‘shell’ key on a uses: step.shell is ignored/unsupported on action steps; keep it only on run steps. Remove it here.
- name: Run Test uses: nick-fields/retry@v3 with: timeout_minutes: 15 max_attempts: 3 retry_on: timeout command: | cd DeeployTest mkdir -p /app/.ccache export CCACHE_DIR=/app/.ccache python testRunner_tiled_siracusa.py -t Tests/${{ inputs.test-name }} --cores=${{ inputs.num-cores }} --l1 ${{ matrix.L1 }} --defaultMemLevel=${{ inputs.default-memory-level }} ${{ inputs.double-buffer && '--doublebuffer' || '' }} --memAllocStrategy=${{ inputs.memory-allocation-strategy }} --searchStrategy=${{ inputs.search-strategy }} - shell: bash
64-69: Pin third-party action to a commit SHA.Avoid mutable tag for retry action.
- - name: Run Test - uses: nick-fields/retry@v3 + - name: Run Test + uses: nick-fields/retry@<commit-sha>.github/workflows/ci-platform-mempool.yml (1)
56-71: Inherit secrets (models).mempool-models: needs: select-env uses: ./.github/workflows/_runner-mempool.yml + secrets: inherit.github/workflows/ci-platform-generic.yml (1)
76-94: Inherit secrets for generic-models; consider path filters to cut CI noise.generic-models: needs: select-env uses: ./.github/workflows/_runner-generic.yml + secrets: inheritOptional:
- Add on:push/pull_request paths filters to avoid running for docs-only changes.
.github/workflows/ci-platform-siracusa-tiled.yml (1)
139-168: Inherit secrets (models doublebuffer L3).siracusa-models-tiled-doublebuffer-L3: needs: select-env @@ - uses: ./.github/workflows/_runner-siracusa-tiled.yml + uses: ./.github/workflows/_runner-siracusa-tiled.yml + secrets: inherit.github/workflows/ci-platform-siracusa-neureka-tiled.yml (1)
119-143: Inherit secrets (models L3 doublebuffer wmem); add minimal token permissions once per workflow.siracusa-neureka-models-tiled-doublebuffer-L3-wmem: needs: select-env @@ - uses: ./.github/workflows/_runner-siracusa-neureka-tiled.yml + uses: ./.github/workflows/_runner-siracusa-neureka-tiled.yml + secrets: inheritAdd:
+permissions: + contents: read + packages: read
🧹 Nitpick comments (40)
.github/workflows/docker-build-deeploy.yml (7)
36-36: Nit: job title typo“Build Deploy Image” → “Build Deeploy Image” for consistency with the project name.
- name: Build Deploy Image + name: Build Deeploy Image
49-50: Verify ARM64 runner labelIs the label actually
ubuntu-22.04-arm? GitHub-hosted runners are typicallyubuntu-22.04-arm64. If you use self-hosted with a custom label, ignore.- runner: ubuntu-22.04-arm + runner: ubuntu-22.04-arm64
20-33: Use native context instead of parsing GITHUB_REF
github.ref_name/github.ref_typeare safer than string slicing.- - name: Set up environment variables + - name: Set up environment variables run: | - echo "BRANCH_NAME=${GITHUB_REF##*/}" >> $GITHUB_ENV - echo "TAG_NAME=${GITHUB_REF##*/}" >> $GITHUB_ENV - echo "IS_TAG=${GITHUB_REF_TYPE}" >> $GITHUB_ENV + echo "BRANCH_NAME=${{ github.ref_name }}" >> $GITHUB_ENV + echo "TAG_NAME=${{ github.ref_name }}" >> $GITHUB_ENV + echo "IS_TAG=${{ github.ref_type }}" >> $GITHUB_ENV
94-107: Enable provenance/SBOM on image buildsSupply-chain best practice; buildx v6 supports it.
file: Container/Dockerfile.deeploy push: true build-args: | BASE_IMAGE=${{ github.event.inputs.docker_image_toolchain }} - outputs: type=image,name=ghcr.io/${{ env.OWNER_LC }}/deeploy,annotation-index=true,name-canonical=true,push=true + outputs: type=image,name=ghcr.io/${{ env.OWNER_LC }}/deeploy,annotation-index=true,name-canonical=true,push=true + provenance: true + sbom: true
52-71: Action pinning for supply-chain safetyPin third‑party actions to commit SHAs (checkout, buildx, login, cache, free-disk-space, manifest) to avoid tag drift.
Would you like me to prepare a follow-up with current SHAs for all used actions?
3-11: Trigger scope checkCurrently workflow_dispatch-only. If images should build on branch pushes/tags, add push/release triggers; if manual-only is intended, ignore.
117-123: Optional: restrict publish to main repoAvoid accidental pushes to GHCR on forks.
- - name: GHCR Log-in + - name: GHCR Log-in + if: ${{ github.repository_owner == 'pulp-platform' }} uses: docker/login-action@v3Apply the same guard to build/push steps.
.github/workflows/_runner-softhier.yml (1)
29-41: Add timeout to prevent hung test jobs.- - name: Run Test + - name: Run Test + timeout-minutes: 30 run: | testNames="${{ inputs['test-names'] }}" export SOFTHIER_INSTALL_DIR=/app/install/softhier cd DeeployTest echo "$testNames" | while IFS= read -r testName; do if [[ -n "$testName" ]]; then echo "Running test: $testName" python testRunner_softhier.py -t Tests/$testName --toolchain=GCC fi done shell: bash.github/workflows/_runner-chimera.yml (2)
30-37: Persist ccache back to the cache.Only restoring the cache won’t update it; add a save step.
- name: Cache ccache uses: actions/cache/restore@v4 with: path: /app/.ccache key: ccache-ci @@ shell: bash + - name: Save ccache + if: always() + uses: actions/cache/save@v4 + with: + path: /app/.ccache + key: ccache-ciAlso applies to: 55-55
37-55: Bound the test duration.- - name: Run Test + - name: Run Test + timeout-minutes: 45 run: | testNames="${{ inputs['test-names'] }}" simulators="${{inputs.simulators}}" cd DeeployTest mkdir -p /app/.ccache export CCACHE_DIR=/app/.ccache export CHIMERA_SDK_HOME=/app/install/chimera-sdk.github/workflows/_runner-snitch-tiled-sequential.yml (2)
51-56: Save ccache back to cache.- name: Cache ccache uses: actions/cache/restore@v4 with: path: /app/.ccache key: ccache-ci @@ shell: bash + - name: Save ccache + if: always() + uses: actions/cache/save@v4 + with: + path: /app/.ccache + key: ccache-ciAlso applies to: 57-76
57-76: Guard against infinite test loops.- - name: Run Tests + - name: Run Tests + timeout-minutes: 60 run: | simulators="${{inputs.simulators}}".github/workflows/docker-build-toolchain.yml (3)
1-2: Scope GITHUB_TOKEN permissions explicitly.Limit token to what’s needed for GHCR pushes.
name: Docker • Build Toolchain Container on: workflow_dispatch: +permissions: + contents: read + packages: write
88-101: Enable SBOM/provenance and ensure freshest bases.- name: Build and push toolchain image id: build uses: docker/build-push-action@v6 with: platforms: linux/${{ matrix.platform }} context: . file: Container/Dockerfile.toolchain push: true outputs: type=image,name=ghcr.io/${{ env.OWNER_LC }}/deeploy-toolchain,annotation-index=true,name-canonical=true,push=true + pull: true + sbom: true + provenance: true
57-66: Pin actions to commit SHAs to reduce supply-chain risk.Use immutable SHAs for third-party actions (buildx, login, cache, manifest, free-disk-space). I can generate a patch once you confirm desired versions.
Also applies to: 82-90, 102-129
.github/workflows/_runner-cortexm.yml (1)
29-34: Restore + save ccache to keep it warm across runs.- name: Cache ccache uses: actions/cache/restore@v4 with: path: /app/.ccache key: ccache-ci - name: Run Test + timeout-minutes: 30 run: | testNames="${{ inputs['test-names'] }}" cd DeeployTest mkdir -p /app/.ccache export CCACHE_DIR=/app/.ccache echo "$testNames" | while IFS= read -r testName; do if [[ -n "$testName" ]]; then echo "Running test: $testName" python testRunner_cortexm.py -t Tests/$testName fi done shell: bash + - name: Save ccache + if: always() + uses: actions/cache/save@v4 + with: + path: /app/.ccache + key: ccache-ciAlso applies to: 35-46
.github/workflows/_runner-generic.yml (1)
29-34: Add timeout and save ccache.- name: Cache ccache uses: actions/cache/restore@v4 with: path: /app/.ccache key: ccache-ci - name: Run Test + timeout-minutes: 30 run: | testNames="${{ inputs['test-names'] }}" cd DeeployTest mkdir -p /app/.ccache export CCACHE_DIR=/app/.ccache echo "$testNames" | while IFS= read -r testName; do if [[ -n "$testName" ]]; then echo "Running test: $testName" python testRunner_generic.py -t Tests/$testName fi done shell: bash + - name: Save ccache + if: always() + uses: actions/cache/save@v4 + with: + path: /app/.ccache + key: ccache-ciAlso applies to: 35-47
.github/workflows/_runner-siracusa-neureka-tiled.yml (2)
63-68: Timeout and ccache save.- name: Cache ccache uses: actions/cache/restore@v4 with: path: /app/.ccache key: ccache-ci - - name: Run Test + - name: Run Test uses: nick-fields/retry@v3 with: timeout_minutes: 5 max_attempts: 3 retry_on: timeout command: | cd DeeployTest mkdir -p /app/.ccache export CCACHE_DIR=/app/.ccache python testRunner_tiled_siracusa_w_neureka.py ... shell: bash + - name: Save ccache + if: always() + uses: actions/cache/save@v4 + with: + path: /app/.ccache + key: ccache-ciAlso applies to: 69-79
56-63: Pin third-party action to a SHA.Consider pinning
nick-fields/retry@v3to a commit SHA..github/workflows/_runner-snitch.yml (1)
35-40: Timeout and ccache save.- name: Cache ccache uses: actions/cache/restore@v4 with: path: /app/.ccache key: ccache-ci - - name: Run Test + - name: Run Test + timeout-minutes: 45 run: | testNames="${{ inputs['test-names'] }}" simulators="${{inputs.simulators}}" cd DeeployTest mkdir -p /app/.ccache export CCACHE_DIR=/app/.ccache echo "$simulators" | while IFS= read -r simulator; do @@ shell: bash + - name: Save ccache + if: always() + uses: actions/cache/save@v4 + with: + path: /app/.ccache + key: ccache-ciAlso applies to: 41-57
.github/workflows/infra-generate-documentation.yml (2)
1-1: Exclude gh-pages to avoid unnecessary workflow triggers.Pushing to the gh-pages branch will also match '**'. Exclude it to prevent redundant runs.
on: push: branches: - - '**' + - '**' + - '!gh-pages'
20-24: Pin Python version and enable pip cache for determinism and speed.setup-python without python-version is a no-op; specify a version and enable caching.
- - uses: actions/setup-python@v5 + - uses: actions/setup-python@v5 + with: + python-version: '3.11' + cache: 'pip'.github/workflows/infra-generate-ccache.yml (2)
11-12: Cron comment/timezone mismatch.GitHub cron runs in UTC. The comment says “1AM CET”. Either fix the time or adjust the comment.
- # Runs the workflow on the default branch every day at 1AM CET to keep the cache fresh - - cron: "0 1 * * *" + # Runs daily at 01:00 UTC (note: not CET) + - cron: "0 1 * * *"
26-27: Also restore cache before build to reuse existing ccache.Speeds up warm runs.
- name: Build Deeploy shell: bash run: pip install -e . + - name: Restore CCache + uses: actions/cache/restore@v4 + with: + path: /app/.ccache + key: ccache-ci-${{ runner.os }}-${{ github.ref_name }} + restore-keys: | + ccache-ci-${{ runner.os }}-.github/workflows/ci-platform-chimera.yml (2)
5-8: Exclude gh-pages from CI triggers.Prevents CI from running on docs-only branch updates.
push: branches: - - '**' + - '**' + - '!gh-pages'
24-27: Guard optional input outside workflow_dispatch.inputs.* is undefined on push/PR. Coalesce to default to avoid passing empty values.
select-env: uses: ./.github/workflows/_select-env.yml with: - docker_image_deeploy: ${{ inputs.docker_image_deeploy }} + docker_image_deeploy: ${{ inputs.docker_image_deeploy || 'ghcr.io/pulp-platform/deeploy:devel' }}.github/workflows/ci-platform-softhier.yml (1)
3-10: Avoid duplicate runs from push and PR events.If pushing branches inside the repo, both push and PR will fire. Consider scoping or filtering to reduce CI noise.
Example options:
- Restrict push to main/release branches.
- Or add paths filters to both push/PR.
.github/workflows/ci-platform-snitch-tiled.yml (1)
28-46: Inherit secrets for runner; prefer YAML over embedded JSON for readability.snitch-kernels-tiled-singlebuffer-L2: needs: select-env uses: ./.github/workflows/_runner-snitch-tiled-sequential.yml + secrets: inherit with: @@ - tests-config: | - [ - {"name":"TestiNoNorm","L1":[5000,10000]}, - {"name":"TestAdderLarge","L1":[5000,10000]}, - {"name":"TestiSoftmaxLarge","L1":[5000,10000]}, - {"name":"testRQGEMM","L1":[2000,5000]}, - {"name":"testFloatSoftmax","L1":[2000,5000,10000]}, - {"name":"TestRQAdd","L1":[5000,10000]}, - {"name":"testFloatGEMM","L1":[2000,5000,10000]}, - {"name":"testFloatGEMMtransB","L1":[2000,5000,10000]} - ] + tests-config: | + - name: TestiNoNorm + L1: [5000, 10000] + - name: TestAdderLarge + L1: [5000, 10000] + - name: TestiSoftmaxLarge + L1: [5000, 10000] + - name: testRQGEMM + L1: [2000, 5000] + - name: testFloatSoftmax + L1: [2000, 5000, 10000] + - name: TestRQAdd + L1: [5000, 10000] + - name: testFloatGEMM + L1: [2000, 5000, 10000] + - name: testFloatGEMMtransB + L1: [2000, 5000, 10000]If the called workflow expects JSON, keep JSON; otherwise YAML is easier to diff/review.
.github/workflows/_select-env.yml (1)
5-7: Make input optional; logic already defaults IMAGE.You’re already handling empty/absent input; forcing callers to pass it adds friction for pushes/PRs.
- docker_image_deeploy: - type: string - required: true + docker_image_deeploy: + type: string + required: false.github/workflows/ci-deeploy.yml (4)
18-21: Harden token scope.Add least-privilege permissions (pulling GHCR often needs packages: read).
concurrency: group: ${{ github.workflow }}-${{ github.ref }} cancel-in-progress: true + +permissions: + contents: read + packages: read
28-41: Redundant job (unused prerequisite).build-deeploy doesn’t feed artifacts or cache and no jobs depend on it. Remove it or make test jobs depend on it and publish/cache artifacts.
42-233: Reduce duplication and add timeouts.
- Extract “Checkout + pip install -e .” into a composite/reusable workflow to DRY.
- Add timeout-minutes to long-running test jobs.
234-234: Add trailing newline.Fixes YAML lint error: “no new line character at the end of file.”
.github/workflows/ci-deeploy-testing.yml (2)
1-21: Add least-privilege permissions.Mirror the permissions block from the main CI workflow for consistency.
60-60: Add trailing newline.Ensure newline at EOF.
.github/workflows/ci-lint.yml (2)
18-21: Harden token scope.Add least-privilege permissions like other workflows.
concurrency: group: ${{ github.workflow }}-${{ github.ref }} cancel-in-progress: true + +permissions: + contents: read + packages: read
66-66: Add trailing newline.Minor YAML style fix.
.github/workflows/ci-platform-siracusa.yml (3)
18-21: Harden token scope.Add least-privilege permissions to pull images and read repo.
concurrency: group: ${{ github.workflow }}-${{ github.ref }} cancel-in-progress: true + +permissions: + contents: read + packages: read
35-67: Optional: parallelize via matrix.If _runner-siracusa runs tests serially, consider splitting test-names into a matrix for faster feedback.
86-86: Add trailing newline.Minor YAML style fix.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (34)
.github/workflows/CI.yml(0 hunks).github/workflows/GitLabCI.yml(0 hunks).github/workflows/_runner-chimera.yml(3 hunks).github/workflows/_runner-cortexm.yml(2 hunks).github/workflows/_runner-generic.yml(2 hunks).github/workflows/_runner-mempool.yml(2 hunks).github/workflows/_runner-siracusa-neureka-tiled-sequential.yml(2 hunks).github/workflows/_runner-siracusa-neureka-tiled.yml(3 hunks).github/workflows/_runner-siracusa-tiled-sequential.yml(2 hunks).github/workflows/_runner-siracusa-tiled.yml(2 hunks).github/workflows/_runner-siracusa.yml(2 hunks).github/workflows/_runner-snitch-tiled-sequential.yml(2 hunks).github/workflows/_runner-snitch.yml(3 hunks).github/workflows/_runner-softhier.yml(2 hunks).github/workflows/_select-env.yml(1 hunks).github/workflows/ci-deeploy-testing.yml(1 hunks).github/workflows/ci-deeploy.yml(1 hunks).github/workflows/ci-lint.yml(1 hunks).github/workflows/ci-platform-chimera.yml(1 hunks).github/workflows/ci-platform-cortexm.yml(1 hunks).github/workflows/ci-platform-generic.yml(1 hunks).github/workflows/ci-platform-mempool.yml(1 hunks).github/workflows/ci-platform-siracusa-neureka-tiled.yml(1 hunks).github/workflows/ci-platform-siracusa-tiled.yml(1 hunks).github/workflows/ci-platform-siracusa.yml(1 hunks).github/workflows/ci-platform-snitch-tiled.yml(1 hunks).github/workflows/ci-platform-snitch.yml(1 hunks).github/workflows/ci-platform-softhier.yml(1 hunks).github/workflows/docker-build-deeploy.yml(1 hunks).github/workflows/docker-build-toolchain.yml(1 hunks).github/workflows/infra-generate-ccache.yml(2 hunks).github/workflows/infra-generate-documentation.yml(1 hunks).gitlab/issue_templates/issue_template.md(0 hunks).gitlab/merge_request_templates/MRTemplate.md(0 hunks)
💤 Files with no reviewable changes (4)
- .gitlab/merge_request_templates/MRTemplate.md
- .gitlab/issue_templates/issue_template.md
- .github/workflows/GitLabCI.yml
- .github/workflows/CI.yml
🧰 Additional context used
🪛 YAMLlint (1.37.1)
.github/workflows/ci-deeploy.yml
[error] 234-234: no new line character at the end of file
(new-line-at-end-of-file)
.github/workflows/ci-platform-snitch.yml
[error] 47-47: no new line character at the end of file
(new-line-at-end-of-file)
🔇 Additional comments (9)
.github/workflows/docker-build-deeploy.yml (2)
1-1: Workflow name alignment LGTMThe rename improves readability and matches the new naming scheme.
134-135: Matrix outputs to merge step may be undefinedAccessing needs.build-deeploy.outputs.digest-amd64/arm64 from a matrix job isn’t guaranteed to aggregate both digests; use separate arch jobs or upload/download artifacts instead.
File: .github/workflows/docker-build-deeploy.yml Lines: 134-135
- ghcr.io/${{ env.OWNER_LC }}/deeploy@${{ needs.build-deeploy.outputs.digest-amd64 }}, - ghcr.io/${{ env.OWNER_LC }}/deeploy@${{ needs.build-deeploy.outputs.digest-arm64 }} + ghcr.io/${{ env.OWNER_LC }}/deeploy@${{ needs.build-deeploy-amd64.outputs.digest }}, + ghcr.io/${{ env.OWNER_LC }}/deeploy@${{ needs.build-deeploy-arm64.outputs.digest }}.github/workflows/docker-build-toolchain.yml (1)
40-46: Verify ARM64 runner label exists.
ubuntu-22.04-armis non-standard on GitHub-hosted runners (commonlyubuntu-22.04-arm64). Confirm you have a self-hosted runner or adjust the label..github/workflows/_runner-siracusa.yml (1)
30-31: LGTM on explicit shell for build.Specifying Bash improves consistency across runners.
.github/workflows/_runner-siracusa-tiled.yml (1)
56-57: LGTM on explicit shell for build..github/workflows/_runner-mempool.yml (1)
27-28: LGTM on explicit shell for build..github/workflows/_select-env.yml (1)
39-43: Self-hosted label specificity.If multiple self-hosted runners exist, consider emitting a more specific label set (e.g., self-hosted,linux,x64) to avoid mis-scheduling. Confirm the fleet’s labels and Docker availability.
.github/workflows/ci-lint.yml (2)
53-53: Ensure clang-format availability.Verify LLVM_INSTALL_DIR is defined in the container; otherwise install or pass a default path.
44-45: Make YAPF enforceable.Current flags only show diff; they won’t fail CI on formatting issues. Add --exit-code (or switch to in-place -ir).
- yapf -rpd -e "third_party/" -e "install/" -e "toolchain/" . + yapf -rpd --exit-code -e "third_party/" -e "install/" -e "toolchain/" .Likely an incorrect or invalid review comment.
.github/workflows/_runner-siracusa-neureka-tiled-sequential.yml
Outdated
Show resolved
Hide resolved
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (3)
CHANGELOG.md (1)
53-53: Tighten phrasing and align naming (“Lint & Licenses”).Matches workflow names listed in the PR discussion.
-- Split CI into multiple files (one per platform, one for lint & license, one general Deeploy tests, one testing infrastructure, and two Docker flows) +- Split CI into multiple workflows (one per platform, one for Lint & Licenses, one for Deeploy, one for testing infrastructure, and two Docker flows)README.md (2)
10-10: Scope the CI badge to the devel branch for accurate status.Add the branch parameter so readers see the status relevant to the default development flow.
-[](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-deeploy.yml) +[](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-deeploy.yml)
83-91: Add branch parameter to CI badges to scope them to devel branch.The CI badges currently show status for the default branch, which can be confusing when other branches have different statuses. Adding
?branch=develto each badge URL will scope them specifically to the devel branch status.-| **Generic CPU** | Your laptop CPU :) | Host | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-generic.yml) +| **Generic CPU** | Your laptop CPU :) | Host | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-generic.yml) -| **CortexM Processors** | [Documentation](https://www.arm.com/products/silicon-ip-cpu/cortex-m/cortex-m4) | [QEMU](https://www.qemu.org/) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-cortexm.yml) +| **CortexM Processors** | [Documentation](https://www.arm.com/products/silicon-ip-cpu/cortex-m/cortex-m4) | [QEMU](https://www.qemu.org/) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-cortexm.yml) -| **MemPool + ITA** | [Mempool paper](https://arxiv.org/abs/2303.17742), [ITA paper](https://arxiv.org/abs/2307.03493) | [Banshee](https://github.com/pulp-platform/banshee) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-mempool.yml) +| **MemPool + ITA** | [Mempool paper](https://arxiv.org/abs/2303.17742), [ITA paper](https://arxiv.org/abs/2307.03493) | [Banshee](https://github.com/pulp-platform/banshee) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-mempool.yml) -| **Siracusa** | [Siracusa paper](https://arxiv.org/abs/2312.14750) | [GVSoC](https://github.com/gvsoc/gvsoc) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-siracusa.yml) [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-siracusa-tiled.yml) [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-siracusa-neureka-tiled.yml) +| **Siracusa** | [Siracusa paper](https://arxiv.org/abs/2312.14750) | [GVSoC](https://github.com/gvsoc/gvsoc) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-siracusa.yml) [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-siracusa-tiled.yml) [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-siracusa-neureka-tiled.yml) -| **Snitch Cluster** | [Snitch paper](https://arxiv.org/abs/2002.10143) | [GVSoC](https://github.com/gvsoc/gvsoc) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-snitch.yml) [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-snitch-tiled.yml) +| **Snitch Cluster** | [Snitch paper](https://arxiv.org/abs/2002.10143) | [GVSoC](https://github.com/gvsoc/gvsoc) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-snitch.yml) [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-snitch-tiled.yml) -| **SoftHier** | [Repo](https://github.com/gvsoc/gvsoc/tree/soft_hier_release) | [GVSoC](https://github.com/gvsoc/gvsoc/tree/soft_hier_release) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-softhier.yml) +| **SoftHier** | [Repo](https://github.com/gvsoc/gvsoc/tree/soft_hier_release) | [GVSoC](https://github.com/gvsoc/gvsoc/tree/soft_hier_release) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-softhier.yml) -| **Chimera** | [Repo](https://github.com/pulp-platform/chimera) | [GVSoC](https://github.com/gvsoc/gvsoc) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-chimera.yml) +| **Chimera** | [Repo](https://github.com/pulp-platform/chimera) | [GVSoC](https://github.com/gvsoc/gvsoc) | [](https://github.com/pulp-platform/Deeploy/actions/workflows/ci-platform-chimera.yml)
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
CHANGELOG.md(3 hunks)README.md(2 hunks)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (160)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
- GitHub Check: deeploy-test-dmas
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L2 (microLlama/microLlama1, 60000, 10000, 5000, 8) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (2500)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Attention, 60000, 10000, 5000, 2500, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (15000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (Transformer, 60000, 30000, 15000, 8, L3) / test-runner-siracusa-tiled (30000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (CCT/CCT_2_32_32_128, 128000, 8, L3) / test-runner-siracusa-tiled (128000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (testTrainCCT/CCT1_Classifier_Training/CCT_1_16_16_128, 640... / test-runner-siracusa-tiled (64000)
Xeratec
force-pushed
the
pr/ci_cleanup
branch
from
1c50859 to
3f8b8a2
Compare
Contributor
There was a problem hiding this comment.
Actionable comments posted: 0
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
Deeploy/Targets/PULPOpen/Templates/FloatConvTemplate.py (1)
82-82: Fix typo in generated comment: “ChannelOout” → “ChannelOut”.This surfaces in generated C and looks unpolished.
Apply:
-// 2D FP Conv HWC with Im2Col and ChannelOout parallelism (Name: ${nodeName}, Op: ${nodeOp}) +// 2D FP Conv HWC with Im2Col and ChannelOut parallelism (Name: ${nodeName}, Op: ${nodeOp})Deeploy/Targets/Generic/Templates/SliceTemplate.py (1)
103-105: Incorrect loop-closing count when axes are non-contiguous — generates invalid CYou open
len(preAxes)loops pluslen(axes)loops, but you closeint(axes[-1]) + 1loops. This only matches whenaxesare contiguous from 0..axes[-1]. With axes like [0,2], this emits too many}and breaks compilation.Use the actual number of opened loops:
-% for axis in range(int(axes[-1])+1): -} -% endfor +% for _ in range(len(preAxes) + len(axes)): +} +% endforAlso, after adopting
lastAxis(see earlier comment), adjust the memcpy offset:-memcpy(ref_${data_out}, ${data_in} + ${data_out}_offset_${axis}, ${transferSize* data_out_type.referencedType.typeWidth//8}); +memcpy(ref_${data_out}, ${data_in} + ${data_out}_offset_${lastAxis}, ${(transferSize * data_out_type.referencedType.typeWidth)//8});
♻️ Duplicate comments (1)
Makefile (1)
590-623: Cover C++ sources/headers in SPDX scan.Currently only .c/.h are checked; .cc/.cpp/.cxx/.hh/.hpp/.hxx are missed. This can hide license gaps.
- echo " - Check C Files"; \ - missing_c=$$(grep -Lr "SPDX-License-Identifier: Apache-2.0" --include="*.c" \ + echo " - Check C/C++ Source Files"; \ + missing_c=$$(grep -Lr "SPDX-License-Identifier: Apache-2.0" --include="*.c" --include="*.cc" --include="*.cpp" --include="*.cxx" \ --exclude-dir="toolchain" --exclude-dir="install" --exclude-dir=".git" \ --exclude-dir="third_party" --exclude-dir="TEST_*" --exclude-dir="TestFiles" \ --exclude-dir="runtime" . || true); \ if [ -n "$$missing_c" ]; then echo "$$missing_c"; rc=1; fi; \ - echo " - Check C Header Files"; \ - missing_h=$$(grep -Lr "SPDX-License-Identifier: Apache-2.0" --include="*.h" \ + echo " - Check C/C++ Header Files"; \ + missing_h=$$(grep -Lr "SPDX-License-Identifier: Apache-2.0" --include="*.h" --include="*.hh" --include="*.hpp" --include="*.hxx" \ --exclude-dir="toolchain" --exclude-dir="install" --exclude-dir=".git" \ --exclude-dir="third_party" --exclude-dir="TEST_*" --exclude-dir="TestFiles" \ --exclude-dir="runtime" . || true); \ if [ -n "$$missing_h" ]; then echo "$$missing_h"; rc=1; fi; \
🧹 Nitpick comments (9)
Deeploy/Targets/PULPOpen/Templates/FloatConvTemplate.py (1)
3-3: Header filename casing typo.Nit: “FLoatConvTemplate.py” → “FloatConvTemplate.py” in the comment header.
-# File: FLoatConvTemplate.py +# File: FloatConvTemplate.pyMakefile (3)
569-577: Reorder and tighten format steps for cleaner diffs.Run autoflake → isort → yapf to minimize churn; current order formats before import cleanup/sort. Also make isort exclude flags consistent (use one style) and avoid mixing "./" and ".".
format: @echo "Formatting all relevant files..." - @echo " - Format Python Files" - @yapf -ipr -e "third_party/" -e "install/" -e "toolchain/" . - @echo " - Format Python Imports" - @isort --quiet --sg "**/third_party/*" --sg "install/*" --sg "toolchain/*" ./ - @autoflake -i -r --remove-all-unused-imports --ignore-init-module-imports --exclude "**/third_party/*,**/install/*,**/toolchain/*" . + @echo " - Remove unused imports" + @autoflake -i -r --remove-all-unused-imports --ignore-init-module-imports --exclude "**/third_party/*,**/install/*,**/toolchain/*" . + @echo " - Sort Python imports" + @isort --quiet --sg "**/third_party/*" --sg "install/*" --sg "toolchain/*" . + @echo " - Format Python files" + @yapf -ipr -e "third_party/" -e "install/" -e "toolchain/" .
590-623: Make the SPDX grep more robust and portable (optional).Consider matching flexible spacing and supporting BSD grep systems (macOS). Option A: use
-E 'SPDX-License-Identifier:[[:space:]]*Apache-2.0'. Option B: replacegrep --includewithfind -type f ... -name '*.ext' -print0 | xargs -0 grep -LE ...to avoid GNU-specific flags.
566-566: Add missing PHONY targets.Declare
lintandcheck-licensesas phony to avoid file name collisions.-.PHONY: docs clean-docs format +.PHONY: docs clean-docs format lint check-licenses.pre-commit-config.yaml (4)
20-27: Add a few lightweight, high-signal core hooks.Consider
end-of-file-fixer,check-merge-conflict, and optionallymixed-line-ending. They catch common issues with near-zero noise.- repo: https://github.com/pre-commit/pre-commit-hooks rev: v6.0.0 hooks: - id: check-added-large-files name: Check for added large files - id: trailing-whitespace name: Check for trailing whitespace + - id: end-of-file-fixer + name: Ensure newline at EOF + - id: check-merge-conflict + name: Check for merge conflict strings + - id: mixed-line-ending + name: Normalize line endings + args: ["--fix=lf"]
59-66: Align clang-format versions to avoid style churn.CI uses
${LLVM_INSTALL_DIR}/bin/clang-format(15.0.0 path), while pre-commit pins 15.0.7. Mismatched versions can flip formatting back and forth.Options:
- Pin pre-commit to the same major/minor as CI, or
- Make CI use the same mirror version as pre-commit.
Confirm which one you prefer and adjust accordingly.
7-17: Exclude pattern looks good; consider anchoring to repo root.Prefix with
^(already present) and add$for full-path matches if needed; otherwise OK.
27-35: Local license hook is useful; mind shell/grep availability across dev OSes.Because it shells out to
make check-licenses, portability depends on GNU tools. If macOS support is desired, document GNU grep requirement or adjust the Makefile as noted.Deeploy/Targets/Generic/Templates/SliceTemplate.py (1)
70-72: Correct choice for contiguous block; minor clarity tweak suggestedComputing
transferSizefrom the innermost iterated axis is correct. For readability and to avoid reliance on the loop variable leaking scope, consider capturinglastAxisonce and reuse it (also in the memcpy offset).-<% -transferSize = dimSteps[int(axes[-1])] -%> +<% +lastAxis = int(axes[-1]) +transferSize = dimSteps[lastAxis] +%>And later update the memcpy offset to use
lastAxis(see separate comment).
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (25)
.pre-commit-config.yaml(1 hunks)CHANGELOG.md(3 hunks)Deeploy/AbstractDataTypes.py(1 hunks)Deeploy/CommonExtensions/CodeTransformationPasses/PrintInputs.py(1 hunks)Deeploy/Targets/Generic/Templates/FloatGemmTemplate.py(1 hunks)Deeploy/Targets/Generic/Templates/FloatLayernormTemplate.py(1 hunks)Deeploy/Targets/Generic/Templates/FloatMaxPoolTemplate.py(1 hunks)Deeploy/Targets/Generic/Templates/FloatPadTemplate.py(1 hunks)Deeploy/Targets/Generic/Templates/FloatReduceSumTemplate.py(1 hunks)Deeploy/Targets/Generic/Templates/QuantTemplate.py(1 hunks)Deeploy/Targets/Generic/Templates/SliceTemplate.py(1 hunks)Deeploy/Targets/PULPOpen/Templates/FloatConvTemplate.py(2 hunks)Deeploy/Targets/PULPOpen/Templates/FloatGemmTemplate.py(1 hunks)Deeploy/Targets/PULPOpen/Templates/FloatLayernormTemplate.py(1 hunks)Deeploy/Targets/PULPOpen/Templates/FloatMatMulTemplate.py(1 hunks)Deeploy/Targets/PULPOpen/Templates/FloatMaxPoolTemplate.py(1 hunks)Deeploy/Targets/PULPOpen/Templates/SGDTemplate.py(1 hunks)Deeploy/Targets/PULPOpen/Templates/SoftmaxCrossEntropyLossTemplate.py(2 hunks)Deeploy/Targets/PULPOpen/Templates/TallGEMMTemplate.py(1 hunks)Deeploy/Targets/PULPOpen/Templates/UniformRequantShiftTemplate.py(1 hunks)Deeploy/Targets/Snitch/Templates/AddTemplate.py(1 hunks)Deeploy/Targets/Snitch/Templates/FloatSoftmaxTemplate.py(1 hunks)Deeploy/TilingExtension/CodeTransformationPasses/TilingPrototypes.py(1 hunks)Makefile(2 hunks)TargetLibraries/PULPOpen/inc/kernel/Layernorm.h(1 hunks)
✅ Files skipped from review due to trivial changes (19)
- Deeploy/Targets/Generic/Templates/FloatPadTemplate.py
- Deeploy/CommonExtensions/CodeTransformationPasses/PrintInputs.py
- Deeploy/Targets/PULPOpen/Templates/FloatLayernormTemplate.py
- Deeploy/Targets/PULPOpen/Templates/FloatGemmTemplate.py
- Deeploy/AbstractDataTypes.py
- Deeploy/Targets/PULPOpen/Templates/FloatMaxPoolTemplate.py
- Deeploy/Targets/PULPOpen/Templates/FloatMatMulTemplate.py
- Deeploy/Targets/Generic/Templates/FloatLayernormTemplate.py
- Deeploy/Targets/PULPOpen/Templates/SGDTemplate.py
- Deeploy/Targets/Generic/Templates/FloatMaxPoolTemplate.py
- Deeploy/Targets/Generic/Templates/FloatGemmTemplate.py
- Deeploy/Targets/Snitch/Templates/AddTemplate.py
- Deeploy/Targets/PULPOpen/Templates/SoftmaxCrossEntropyLossTemplate.py
- Deeploy/Targets/Generic/Templates/QuantTemplate.py
- Deeploy/TilingExtension/CodeTransformationPasses/TilingPrototypes.py
- Deeploy/Targets/Generic/Templates/FloatReduceSumTemplate.py
- Deeploy/Targets/PULPOpen/Templates/UniformRequantShiftTemplate.py
- Deeploy/Targets/PULPOpen/Templates/TallGEMMTemplate.py
- Deeploy/Targets/Snitch/Templates/FloatSoftmaxTemplate.py
🚧 Files skipped from review as they are similar to previous changes (2)
- CHANGELOG.md
- TargetLibraries/PULPOpen/inc/kernel/Layernorm.h
🧰 Additional context used
🪛 checkmake (0.2.2)
Makefile
[warning] 578-578: Target body for "lint" exceeds allowed length of 5 (10).
(maxbodylength)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (90)
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
- GitHub Check: generic-kernels / test-runner-generic
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (5000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (microLlama/microLlama1, 60000, 10000, 5000, 8, L3) / test-runner-siracusa-tiled (10000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (8000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNetv2, 60000, 16000, 12000, 8000, 8, L3) / test-runner-siracusa-tiled (12000)
- GitHub Check: siracusa-models-tiled-singlebuffer-L3 (miniMobileNet, 60000, 12000, 6000, 8, L3) / test-runner-siracusa-tiled (60000)
- GitHub Check: siracusa-kernels-tiled-singlebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: siracusa-kernels-tiled-doublebuffer-L2 / test-runner-siracusa-tiled
- GitHub Check: mempool-models / test-runner-mempool
🔇 Additional comments (5)
Deeploy/Targets/PULPOpen/Templates/FloatConvTemplate.py (2)
66-74: 2D FP Conv call formatting cleanup — OKWhitespace/argument formatting looks good; no behavior change.
87-105: Loop guard fixed (n < ${batch}) — confirm no legacy patterns remainProvided search returned no matches; verify repo-wide absence of
for (uint32_t n=0; ${batch}; ++n)by running:
rg -n -P "(?s)for\s*(\suint32_t\s+n\s=\s0\s;[^;]${?batch}?[^;];" -SMakefile (2)
463-464: LGTM on the softhier include copy.Whitespace fix is fine; path usage looks correct.
578-589: Ensure lint target fails on formatting differences.
- Add --exit-code to the yapf invocation so the lint target returns non‑zero when formatting differs:
- @yapf -rpd -e "third_party/" -e "install/" -e "toolchain/" . + @yapf -rpd --exit-code -e "third_party/" -e "install/" -e "toolchain/" .
- run_clang_format.py has no --check flag; it already returns exit code 1 (ExitStatus.DIFF) when diffs exist. The lint target calls it with -r (no -i), so C/C++ style issues will fail as-is — no change required there.
Likely an incorrect or invalid review comment.
Deeploy/Targets/Generic/Templates/SliceTemplate.py (1)
41-59: Canonicalize and validate Slice indices (starts/ends/axes/steps)Negative axes/starts/ends or zero steps can produce invalid/undefined C (e.g., uint32_t i = -1 or huge unsigned loops). Normalize & validate here unless upstream guarantees canonical, non-zero-step, sorted inputs.
File: Deeploy/Targets/Generic/Templates/SliceTemplate.py (around lines 41-59)
Proposed normalization (insert after fetching
.valuesin alignToContext):operatorRepresentation['starts'] = startsBuffer.values operatorRepresentation['ends'] = endsBuffer.values operatorRepresentation['axes'] = axesBuffer.values operatorRepresentation['steps'] = stepsBuffer.values + # Normalize axes/start/end/step + data_in_shape = operatorRepresentation['data_in_shape'] + rank = len(data_in_shape) + + axes = [int(a) for a in operatorRepresentation['axes']] + starts = [int(s) for s in operatorRepresentation['starts']] + ends = [int(e) for e in operatorRepresentation['ends']] + steps = [int(st) for st in operatorRepresentation['steps']] + + # Canonicalize negative axes and validate uniqueness/range + axes = [(a + rank) if a < 0 else a for a in axes] + if len(set(axes)) != len(axes): + raise ValueError("Slice axes must be unique") + if any(a < 0 or a >= rank for a in axes): + raise ValueError("Slice axis out of range") + if any(st == 0 for st in steps): + raise ValueError("Slice step cannot be 0") + + # Sort all triplets by axis to match template assumptions + zipped = sorted(zip(axes, starts, ends, steps), key=lambda t: t[0]) + if zipped: + axes, starts, ends, steps = map(list, zip(*zipped)) + else: + axes, starts, ends, steps = [], [], [], [] + + # Canonicalize negative starts/ends and clamp to dimension + norm_starts, norm_ends = [], [] + for a, s, e, st in zip(axes, starts, ends, steps): + dim = int(data_in_shape[a]) + s = s + dim if s < 0 else s + e = e + dim if e < 0 else e + s = max(0, min(s, dim)) + e = max(0, min(e, dim)) + norm_starts.append(s) + norm_ends.append(e) + + operatorRepresentation['axes'] = axes + operatorRepresentation['starts'] = norm_starts + operatorRepresentation['ends'] = norm_ends + operatorRepresentation['steps'] = steps + operatorRepresentation['data_in_size'] = np.prod(operatorRepresentation['data_in_shape'])If upstream already guarantees normalized, sorted, non-negative inputs and non-zero steps, confirm so this change can be skipped.
Member
Author
|
@lukamac Ready for the final review |
lukamac
approved these changes
Sep 12, 2025
Contributor
lukamac
left a comment
lukamac
left a comment
There was a problem hiding this comment.
Thank you for curing everyone's OCD. Approv'er
Contributor
|
@Xeratec actually, just one request before merging. Please update the changelog and maybe change PR title to somthing like "Linting extravaganza"? |
Xeratec
changed the title
Contributor
There was a problem hiding this comment.
Actionable comments posted: 0
♻️ Duplicate comments (1)
CHANGELOG.md (1)
7-8: Good: PR entries and capitalization look consistent.“Workflows” capitalization matches prior feedback; links render correctly.
🧹 Nitpick comments (3)
CHANGELOG.md (3)
56-58: Fix “cMake” capitalization and clarify phrasing.Use “CMake” and specify “license headers” for precision.
- Split CI into multiple workflow files: one per platform, one for lint & license, one for general Deeploy tests, one for infrastructure, and two for Docker flows, improving maintainability and status reporting - Extended CI to check license in cMake and YAML files + Split CI into multiple workflow files: one per platform, one for lint & license, one for general Deeploy tests, one for infrastructure, and two for Docker flows, improving maintainability and status reporting + Extended CI to check license headers in CMake and YAML files - Removed all trailing whitespace + Removed all trailing whitespace
18-18: Typo: “preventes” → “prevents”.User-facing doc; quick fix.
- ... This preventes registry clutter caused by unnecessary per-architecture Docker tags. + ... This prevents registry clutter caused by unnecessary per-architecture Docker tags.
19-20: Grammar: plural “DMAs” (no apostrophe).Clean up possessive/plural confusion.
- AsyncDma abstraction of DMA's + AsyncDma abstraction of DMAs - test runner per DMA and a script that tests all the DMA's + Test runner per DMA and a script that tests all the DMAs
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
CHANGELOG.md(3 hunks)
🔇 Additional comments (2)
CHANGELOG.md (2)
32-35: LGTM on CI/readability additions.Concise bullets; aligns with the PR scope (badges, YAML lint, license headers, pre-commit).
63-63: LGTM: Fix description reads clearly.Accurately reflects the jq install hardening and license check fix.
5 tasks
This was referenced Oct 17, 2025
marchioa
pushed a commit
to FondazioneChipsIT/Deeploy
that referenced
this pull request
Oct 20, 2025
…ng Reliability (pulp-platform#108) - Added CI badges to the README - Added YAML linting to CI - Added missing license headers and C header include guards - Extended the pre-commit hooks to remove trailing whitespace, check licenses, format, and lint files - Split CI into multiple workflow files: one per platform, one for lint & license, one for general Deeploy tests, one for infrastructure, and two for Docker flows, improving maintainability and status reporting - Extended CI to check license in CMake and YAML files - Removed all trailing whitespace - Fix license CI check and prevent potential issues with `jq` installation
diaconuccalin
pushed a commit
to diaconuccalin/Deeploy
that referenced
this pull request
Oct 27, 2025
…ng Reliability (pulp-platform#108) - Added CI badges to the README - Added YAML linting to CI - Added missing license headers and C header include guards - Extended the pre-commit hooks to remove trailing whitespace, check licenses, format, and lint files - Split CI into multiple workflow files: one per platform, one for lint & license, one for general Deeploy tests, one for infrastructure, and two for Docker flows, improving maintainability and status reporting - Extended CI to check license in CMake and YAML files - Removed all trailing whitespace - Fix license CI check and prevent potential issues with `jq` installation
diaconuccalin
pushed a commit
to diaconuccalin/Deeploy
that referenced
this pull request
Oct 27, 2025
…ng Reliability (pulp-platform#108) - Added CI badges to the README - Added YAML linting to CI - Added missing license headers and C header include guards - Extended the pre-commit hooks to remove trailing whitespace, check licenses, format, and lint files - Split CI into multiple workflow files: one per platform, one for lint & license, one for general Deeploy tests, one for infrastructure, and two for Docker flows, improving maintainability and status reporting - Extended CI to check license in CMake and YAML files - Removed all trailing whitespace - Fix license CI check and prevent potential issues with `jq` installation
diaconuccalin
pushed a commit
to diaconuccalin/Deeploy
that referenced
this pull request
Oct 27, 2025
…ng Reliability (pulp-platform#108) - Added CI badges to the README - Added YAML linting to CI - Added missing license headers and C header include guards - Extended the pre-commit hooks to remove trailing whitespace, check licenses, format, and lint files - Split CI into multiple workflow files: one per platform, one for lint & license, one for general Deeploy tests, one for infrastructure, and two for Docker flows, improving maintainability and status reporting - Extended CI to check license in CMake and YAML files - Removed all trailing whitespace - Fix license CI check and prevent potential issues with `jq` installation
5 tasks
This was referenced Feb 2, 2026
Merged
Merged
Victor-Jung
added a commit
that referenced
this pull request
Feb 5, 2026
This release includes improvements to the tiling and DMA code generation, new networks and operators, improved CI workflows, migration to PyTest, and support for PyPi package releases. Note: Since the release tag references the Docker container tagged with the release tag (ghcr.io/pulp-platform/deeploy:v0.2.1), the CI will initially fail. The Deeploy Docker image must be built after the release PR is merged and the CI restarted. ### List of Pull Requests - PyPi Package Deployment + Remove Banshee Dept [#154](#154) - PyTest Migration [#144](#144) - Update submodule `pulp-nn-mixed` [#145](#145) - Improve Profiling [#138](#138) - FP32 ReduceMean operator improvement [#137](#137) - Support for RMSNorm (Pow and Sqrt operators) [#136](#136) - Demo TinyViT compatibility with tiled Siracusa [#124](#124) - TinyViT on non-tiled Siracusa [#117](#117) - Support Fully Asynchronous DMAs [#114](#114) - Disallow shape inference [#128](#128) - Remove memory-aware node bindings [#123](#123) - Fix missing const's layout transformation and refactor NCHWtoNHWC passes [#122](#122) - Fix aliasing [#125](#125) - Support for 1D Autoencoder [#98](#98) - Refactor Logging for Improved Debugging [#115](#115) - Add reuse-tool as an SPDX license header linter [#113](#113) - Bug fixes, API Cleanup and Reduce Compiler Warning on PULP [#112](#112) - Fix PULP GEMM `batch` serialization [#109](#109) - Split CI Workflows by Platform and Task, Improve Formatting and Linting Reliability [#108](#108) - Refactor tiling code generation [#105](#105) - Change order of typeMatching entries [#68](#68) - Node Mangling to avoid duplication [#93](#93) - Prepare Post v0.2.0 Release [#104](#104) - Use Docker digests instead of arch-specific tags [#106](#106) - Fix `Unsqueeze` Op. when using ONNX opset 13 or higher (from attribute to input) [#119](#119) - Fix bias hoisting in generic GEMM with no bias [#126](#126)
5 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR refactors the CI files by splitting them by platform, making future maintenance easier and reducing the likelihood of merge/rebase conflicts. It also allows reporting the status for different platforms in the README.
Changed
PR Merge Checklist
develcommit and pointing todevel.CHANGELOG.mdfile has been updated.@coderabbitai summary