Make better use of the GHA cache (#4489)

* Make better use of the GHA cache
IntersectMBO · Jul 22, 2024 · 8972703 · 8972703
1 parent 3c1b576
commit 8972703
Show file tree

Hide file tree

Showing 2 changed files with 76 additions and 20 deletions.
diff --git a/.github/workflows/haskell.yml b/.github/workflows/haskell.yml
@@ -80,28 +80,60 @@ jobs:
     - name: Cabal update
       run: cabal update
 
-    - name: Cabal configure
-      run: cabal configure --enable-tests --enable-benchmarks --write-ghc-environment-files=always
-
-    - uses: actions/cache@v4
-      if: matrix.os != 'macos-latest'
-      name: Cache cabal store
+    # We create a dependencies.txt file that can be used to index the cabal
+    # store cache.
+    #
+    # We do not use plan.json directly because adding a dependency to our
+    # Cabal files which was already present somewhere else would result in a
+    # diferent plan, even though the set of dependencies is the same.
+    - name: Record dependencies to be used as cache keys
+      id: record-deps
+      run: |
+        cabal build all --enable-tests --dry-run --minimize-conflict-set
+        cat dist-newstyle/cache/plan.json \
+        | jq '.["install-plan"][].id' \
+        | sort \
+        | uniq \
+        | tee dependencies.txt
+
+    - name: Restore cache
+      uses: actions/cache/restore@v4
+      id: restore-cabal-cache
+      env:
+        cache-name: cache-cabal-build
       with:
         path: |
           ${{ steps.install-haskell.outputs.cabal-store }}
-        # cache is invalidated upon a change to cabal.project (and/or cabal.project.local), a bump to
-        # CABAL_CACHE_VERSION or after a week of inactivity
-        key: cache-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-${{ hashFiles('cabal.project*') }}
-        # Restoring attempts are from current branch then master
+        # A new cache is created upon a change to the cabal build plan,
+        # cabal.project (and/or cabal.project.local), or a bump to
+        # CABAL_CACHE_VERSION.
+        key: ${{ env.cache-name }}-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-${{ hashFiles('dependencies.txt') }}-${{ hashFiles('cabal.project*') }}
+        # Restoring attempts are from current branch then master. The key above
+        # is by default already a restore-key.
         restore-keys: |
-          cache-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-${{ hashFiles('cabal.project*') }}
+          ${{ env.cache-name }}-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-${{ hashFiles('dependencies.txt') }}
+          ${{ env.cache-name }}-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-
 
     - name: Check workflow test matrix
       run: cabal build all --dry-run && scripts/check-workflow-test-matrix.hs
 
     - name: Build dependencies
+      id: build-dependencies
       run: cabal build all --only-dependencies
 
+    # Save the cache of built dependencies early, so that it is available for
+    # the next GHA run if the Build step fails.
+    - name: Save cache
+      uses: actions/cache/save@v4
+      id: save-cabal-cache
+      # Note: cache-hit will be set to true only when cache hit occurs for the
+      # exact key match. For a partial key match via restore-keys or a cache
+      # miss, it will be set to false.
+      if: steps.build-dependencies.outcome == 'success' && steps.restore-cabal-cache.outputs.cache-hit != 'true'
+      with:
+        path: ${{ steps.install-haskell.outputs.cabal-store }}
+        key:  ${{ steps.restore-cabal-cache.outputs.cache-primary-key }}
+
     - name: Build
       run: cabal build all
 
@@ -228,18 +260,26 @@ jobs:
     - name: Unarchive working directory
       run: tar -xf state.tar && rm state.tar
 
-    - uses: actions/cache@v4
-      if: matrix.os != 'macos-latest'
-      name: Cache cabal store
+    # A dependencies.txt file should have been created by the build job, so we
+    # check if it exists and is not empty.
+    - name: Check dependencies to be used as cache keys
+      id: check-deps
+      run: |
+        ./scripts/file-not-null.sh dependencies.txt
+
+    - name: Restore cache
+      uses: actions/cache/restore@v4
+      id: restore-cabal-cache
+      env:
+        cache-name: cache-cabal-build
       with:
         path: |
           ${{ steps.install-haskell.outputs.cabal-store }}
-        # cache is invalidated upon a change to cabal.project (and/or cabal.project.local), a bump to
-        # CABAL_CACHE_VERSION or after a week of inactivity
-        key: cache-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-${{ hashFiles('cabal.project*') }}
-        # Restoring attempts are from current branch then master
-        restore-keys: |
-          cache-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-${{ hashFiles('cabal.project*') }}
+        key: ${{ env.cache-name }}-${{ runner.os }}-${{ matrix.ghc }}-${{ env.CABAL_CACHE_VERSION }}-${{ hashFiles('dependencies.txt') }}-${{ hashFiles('cabal.project*') }}
+        # The cache with this specific key should have been created by the build
+        # job. If the cache is missing, something is terribly wrong! We fail the
+        # test job, or otherwise we would start rebuilding all the dependencies.
+        fail-on-cache-miss: true
 
     - name: Cabal update
       run: cabal update

diff --git a/scripts/file-not-null.sh b/scripts/file-not-null.sh
@@ -0,0 +1,16 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+if [ ! -f "$1" ]; then
+  echo "Error: $1 does not exist!"
+  exit 1
+fi
+
+if [ ! -s "$1" ]; then
+  echo "Error: $1 is empty!"
+  exit 1
+fi
+
+echo "$1 exists and is non-empty. Printing..."
+cat "$1"