From 4df931d99ba873f40597612d96a779c4d799ef26 Mon Sep 17 00:00:00 2001
From: Matt Jan <zoo868e@gmail.com>
Date: Tue, 5 Sep 2023 03:11:04 +0800
Subject: [PATCH] [CI] Pre-fetch and cache Cargo dependencies

In this commit, we optimize the population of the global Cargo registry
and cache in the CI workflow. By populating the cache before running the
jobs in the matrix, we avoid redundant downloads of crates from the
internet, improving overall workflow efficiency.
---
 .github/workflows/ci.yml | 43 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)

diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index b3517ac8e6..e350da2c08 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -31,6 +31,12 @@ env:
 jobs:
   build_test:
     runs-on: ubuntu-latest
+    # Generate and populate the global Cargo registry and cache first. Each
+    # job in the matrix runs in parallel, so without populating the cache
+    # first, most jobs would duplicate the work of downloading crates from
+    # the internet. Populating the cache first ensures that this work only
+    # happens once.
+    needs: generate_cache
 
     strategy:
       # By default, this is set to `true`, which means that a single CI job
@@ -78,6 +84,13 @@ jobs:
     steps:
     - uses: actions/checkout@3df4ab11eba7bda6032a0b82a6bb43b11571feac # v4.0.0
 
+    - uses: actions/cache@v3
+      with:
+        path: |
+          ~/.cargo/registry/index/
+          ~/.cargo/registry/cache/
+        key: ${{ runner.os }}-cargo-${{ hashFiles('**/Cargo.toml') }}
+
     # We use toolchain descriptors ("msrv", "stable", and "nightly") in the
     # matrix. This step converts the current descriptor to a particular
     # toolchain version by looking up the corresponding key in `Cargo.toml`. It
@@ -358,3 +371,33 @@ jobs:
               | tee -a $GITHUB_STEP_SUMMARY >&2
             exit 1
           fi
+
+  generate_cache:
+    runs-on: ubuntu-latest
+    name: Generate cache
+    steps:
+      - uses: actions/checkout@f43a0e5ff2bd294095638e18286ca9a3d1956744 # v3.6.0
+
+      - uses: actions/cache@v3
+        with:
+          path: |
+            ~/.cargo/registry/index/
+            ~/.cargo/registry/cache/
+          key: ${{ runner.os }}-cargo-${{ hashFiles('**/Cargo.toml') }}
+
+      - name: Populate cache
+        run: |
+          # Ensure all dependencies are downloaded - both for our crates and for tools
+          # we use in CI. We don't care about these tools succeeding for two reasons:
+          # First, this entire job is best-effort since it's just a performance optimization.
+          # Second, there may be failures due to issues other than failing to download
+          # dependencies (e.g., `cargo metadata` called with a malformed `Cargo.toml`,
+          # build failure in our own crate or in dependencies, etc). For those reasons,
+          # we discard stderr and ignore status codes.
+          #
+          # For downloading our crates' dependencies in particular, note that there is
+          # no support for doing this directly [1], so we just check all crates using --tests.
+          #
+          # [1]  https://stackoverflow.com/a/42139535/836390
+          cargo check --workspace --tests &> /dev/null || true
+          cargo metadata &> /dev/null || true