)]}'
{
  "log": [
    {
      "commit": "46761c66a7f475a1621e638538a74e5622e28c60",
      "tree": "a6ea3096e73b6356bc54134a5bb0d3862408b787",
      "parents": [
        "72f2223d3a317a3e796b1b3c205884ec140bc45d"
      ],
      "author": {
        "name": "Jerry Belton",
        "email": "jcecil@google.com",
        "time": "Thu Nov 19 14:17:15 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Nov 19 14:17:15 2026 -0800"
      },
      "message": "[project] Support materialize-in-source\u003d\"false\"\n\nAllow manifest \u003cpackage\u003e elements to set materialize-in-source\u003d\"false\"\nso build-only CIPD packages can be resolved in jiri.lock without\nunpacking them into $FUCHSIA_DIR/prebuilt/ during checkout sync:\n\n- Add MaterializeInSource (*bool), ShouldMaterializeInSource(), and\n  Packages.MaterializedInSource() in project/manifest.go, omitting the\n  default true value when serializing snapshots while preserving\n  materialize-in-source\u003d\"false\" in .jiri_root/update_history/latest.\n- Update LoadWorktreePackageIndex, FetchPackages, and\n  RunUnifiedPackageCachePipeline to exclude unmaterialized packages from\n  workspace destinations while pre-filtering cached instances against\n  instances/ (and .cipd/pkgs for unpinned tags) so jiri update retains\n  current on-demand instances without resolving uncached packages, and\n  prunes stale rolled cache instances when the active index is empty.\n- Update tryLinkPackagesFromParentCache to skip unmaterialized packages\n  when linking from the parent checkout while keeping them in\n  pkgsWAccess for the worktree snapshot.\n- Update garbageCollectObsoletePrebuilt to exclude unmaterialized\n  packages from neededRelPaths so stamped leftover directories are\n  cleaned up and empty parent directories under prebuilt/ are pruned.\n\nBug: 578684789\nBug: 566304584\nTest: go test -race ./project\nChange-Id: I5b5f3abd64e6a05a8ba9e3a1e8b99bd211304fd3\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1967194\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Chase Latta \u003cchaselatta@google.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nCommit-Queue: Jerry Belton \u003cjcecil@google.com\u003e\n"
    },
    {
      "commit": "72f2223d3a317a3e796b1b3c205884ec140bc45d",
      "tree": "f099c43f1ced3f6fb157222cad7991d4f768b4b5",
      "parents": [
        "8dae45656b6c676959af50cc7431d7946da6bb39"
      ],
      "author": {
        "name": "Jerry Belton",
        "email": "jcecil@google.com",
        "time": "Thu Nov 19 12:48:35 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Nov 19 12:48:35 2026 -0800"
      },
      "message": "[jiri] Simplify and optimize fetch-package\n\nSimplify and optimize the on-demand package fetching pipeline:\n\n- Avoid duplicate `collectDestsWithRequiredPaths` and\n  `areAllDestsUpToDate` passes in `FetchPackagePaths` when no uncached\n  internal ACL check is required, and unify up-to-date/post-fetch path\n  verification while preserving `\u0027cipd auth-login\u0027` diagnostics.\n- Short-circuit `Package.ResolvePathForPlatform` when `p.Path` has no\n  `{{` template directives, and fast-path `cipd.ResolvePlatforms` in\n  `matchPackagesForPaths` for non-templated package names.\n- Avoid falling through to process cwd in `resolveCheckoutRoot` when an\n  explicit `WorkingDir` is provided and not inside a checkout.\n- Deduplicate subset cache format-version stamping into\n  `PackageCache.syncMissingInstances`, drop the unused `includeSelf`\n  parameter on `containsNestedCacheStamp`, and share lock-file opening\n  and unlock helpers in `cache_lock.go`.\n\nBug: 566304584\nTest: go test -race ./cmd/jiri/subcommands ./project\nChange-Id: Ia25674d24987af22d77d6992fd82d06322c1ebda\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1976294\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nCommit-Queue: Jerry Belton \u003cjcecil@google.com\u003e\nReviewed-by: Chase Latta \u003cchaselatta@google.com\u003e\n"
    },
    {
      "commit": "8dae45656b6c676959af50cc7431d7946da6bb39",
      "tree": "26f380996cd781704e071c5ccf9f5d36167c8444",
      "parents": [
        "4cab204bac1218f006f0a58e18ce5a8732b13aa2"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Mon Nov 16 14:15:44 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Nov 16 14:15:44 2026 -0800"
      },
      "message": "[worktree] List subcommands in `jiri worktree -h`\n\nRunning `jiri worktree` with no arguments prints a list of its\nsubcommands, but `jiri worktree -h` and `jiri help worktree` only\nprinted \"Manage Jiri worktrees.\", which isn\u0027t useful for discovering\nthe subcommands. Now all three print the same subcommand list.\n\nTAG\u003dagy\nCONV\u003d6dbd0003-dfd0-4663-92bc-d45dc46435f0\n\nChange-Id: I1b0dc96e889175ec65b98534553f1348aad375aa\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1983697\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "4cab204bac1218f006f0a58e18ce5a8732b13aa2",
      "tree": "603bdacf5e10419235e02148f82e8946cd3ae7dd",
      "parents": [
        "af3df068220ccac574d5f4966022bf5328d0cb7a"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Fri Nov 13 12:59:11 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Nov 13 12:59:11 2026 -0800"
      },
      "message": "[project] Respect -local-manifest-project in package GC\n\n`jiri update -local-manifest-project\u003dfoo` fetched packages using foo\u0027s\nlocal manifest, but the package GC that runs at the end of the update\nreloaded the manifest with only the default local manifest projects\n(the top-level imports in .jiri_manifest). If foo was imported\nindirectly, GC read foo\u0027s manifest from JIRI_HEAD and deleted any\npackage that only existed in the local manifest changes, right after\nthe update had installed it.\n\nNow GC uses the same local manifest projects as the update for the\ncurrent workspace. Other worktrees still use their defaults.\n\nTAG\u003dagy\nCONV\u003dcb13f62a-2b39-439c-8e7e-8bf371eaee9b\n\nChange-Id: I9b9ad48efb07a0a37186ce008583c438ce136c00\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1979114\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nReviewed-by: David Turner \u003cdigit@google.com\u003e\n"
    },
    {
      "commit": "af3df068220ccac574d5f4966022bf5328d0cb7a",
      "tree": "3e6ee1d56db6effe40137cbaf544747ec13d1350",
      "parents": [
        "37a83faf8d5e69a03315dadbe7a86fad9c052dbd"
      ],
      "author": {
        "name": "Chase Latta",
        "email": "chaselatta@google.com",
        "time": "Fri Nov 13 09:05:20 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Nov 13 09:05:20 2026 -0800"
      },
      "message": "[project] Support LUCI_AUTH_CREDENTIAL_HELPER for RBE auth\n\nDeveloper and agent GCE VMs without local Application Default Credentials\npreviously fell back to the GCE metadata server, whose default service\naccount token lacked RBE scopes and failed requests with permission errors.\n\nCheck LUCI_AUTH_CREDENTIAL_HELPER in LiveRBEClient after ADC files and\nstrictly before GCE metadata, using the Reclient credential helper JSON\nprotocol with caching, invalidation, and clean fallback on helper failures.\n\nTAG\u003dagy\nCONV\u003daa2cbb7f-078f-4471-b287-ee62c9c27387\nChange-Id: I8131fb26546b13a708b4ca3810cfb86929f7977c\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1975400\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nReviewed-by: Jiaming Li \u003clijiaming@google.com\u003e\nCommit-Queue: Chase Latta \u003cchaselatta@google.com\u003e\nReviewed-by: Darren Chan \u003cchandarren@google.com\u003e\n"
    },
    {
      "commit": "37a83faf8d5e69a03315dadbe7a86fad9c052dbd",
      "tree": "d5a6f64aae8e0f5dfe68ceb878552fb6c67a3628",
      "parents": [
        "1b4a7dc8c78204a02a06b68486697ec04e4e2e7d"
      ],
      "author": {
        "name": "Jerry Belton",
        "email": "jcecil@google.com",
        "time": "Thu Nov 12 16:17:42 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Nov 12 16:17:42 2026 -0800"
      },
      "message": "[cmd/jiri] Implement on-demand jiri fetch-package (4/4)\n\nWire `jiri fetch-package` in `cmd/jiri/subcommands/fetch_pkg.go` to\n`project.FetchPackagePaths` so `fx` commands and GN/Bazel build actions\ncan fetch missing `prebuilt/` paths on demand:\n\n- Resolve the checkout root across symlinked checkouts, symlinked build\n  directories, and relative `-root` flags via `resolveCheckoutRoot`.\n- Normalize relative, `//prebuilt/...`, and absolute arguments under\n  `prebuilt/` via `normalizePrebuiltPath`.\n- Fast-return without initializing `jiri.NewContext` or parsing\n  manifests when `JIRI_ON_DEMAND_PACKAGES` is disabled.\n- Otherwise initialize `jiri.NewContext` and materialize the enclosing\n  package destinations via `project.FetchPackagePaths`.\n\nBug: 571564570\nBug: 566304584\nTest: go test -race ./cmd/jiri/subcommands ./project\nChange-Id: I535f05fb0d058862fd948108046c17539b79dfa7\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1957834\nReviewed-by: Chase Latta \u003cchaselatta@google.com\u003e\nCommit-Queue: Jerry Belton \u003cjcecil@google.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "1b4a7dc8c78204a02a06b68486697ec04e4e2e7d",
      "tree": "b90dd0d3b7cc64bb6216462433b1d943b4d77d43",
      "parents": [
        "a483c768081210ea487597d20f168a4e7a5e2f4c"
      ],
      "author": {
        "name": "Jerry Belton",
        "email": "jcecil@google.com",
        "time": "Thu Nov 12 02:38:57 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Nov 12 02:38:57 2026 -0800"
      },
      "message": "[project] Add FetchPackagePaths for on-demand prebuilts (3/4)\n\nAdd `project/fetch_package.go` and `project/fetch_package_test.go` to\nresolve checkout-relative `prebuilt/` paths against the checkout\u0027s\nmanifest snapshot or `.jiri_manifest` and materialize them on demand:\n\n- Resolve requested `prebuilt/` paths against\n  `.jiri_root/update_history/latest` (or `WorktreeParentRoot`\u0027s latest\n  snapshot when the worktree has no snapshot of its own), falling back\n  to `.jiri_manifest` when no snapshot exists.\n- Filter optional package attributes via\n  `FilterOptionalProjectsPackages` and internal packages via\n  `readCachedInternalAccess` / `FilterACL`.\n- Match each requested path to the most-specific enclosing package\n  destination or descendant destinations under an umbrella directory,\n  attach `RequiredRelPaths`, and materialize out-of-date destinations\n  via `NewPrebuiltProvider`.\n\nBug: 571564570\nBug: 566304584\nTest: go test -race ./project\nChange-Id: I1b9d2dd777fb7c4cc3f2f061082999dc7a09cf2a\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1964714\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nCommit-Queue: Jerry Belton \u003cjcecil@google.com\u003e\nReviewed-by: Chase Latta \u003cchaselatta@google.com\u003e\n"
    },
    {
      "commit": "a483c768081210ea487597d20f168a4e7a5e2f4c",
      "tree": "886abb85f3bec2cf5e969425a0c56c4252b327aa",
      "parents": [
        "60a6ea2354e08af0aa2b47c966280e0d4fae94b7"
      ],
      "author": {
        "name": "Jerry Belton",
        "email": "jcecil@google.com",
        "time": "Wed Nov 11 15:10:34 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Nov 11 15:10:34 2026 -0800"
      },
      "message": "[project] Add non-pruning subset sync to PackageCache (2/4)\n\nSupport on-demand subset package synchronization in `PackageCache` and\n`CIPDPrebuiltProvider`:\n\n- Add `PackageCache.LockWithContext`, `InstalledInstances`, and\n  `SyncSubset` in `project/package_cache.go` so subset fetches preserve\n  already-cached instances in `.cipd/pkgs/*/description.json` instead of\n  pruning unreferenced instances during `cipd ensure`.\n- Update `CIPDPrebuiltProvider.EnsureDestinations` in subset mode\n  (`worktreeIndex \u003d\u003d nil \u0026\u0026 !fullCheckoutSync`) to return immediately\n  before acquiring `.cache.lock` when all destinations are up to date,\n  wait on `LockWithContext` with `fetchTimeout`, and synchronize missing\n  instances via `SyncSubset` (or `opts.InstanceFetcher` in tests).\n\nBug: 571564570\nBug: 566304584\nTest: go test -race ./project\nChange-Id: I5728b8f92d6744d9b51b837ddc2a8ed723987a00\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1964713\nCommit-Queue: Jerry Belton \u003cjcecil@google.com\u003e\nReviewed-by: Chase Latta \u003cchaselatta@google.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "60a6ea2354e08af0aa2b47c966280e0d4fae94b7",
      "tree": "0c0f979a93bbff5a0ccf5fe1736fe1ffda05d95d",
      "parents": [
        "10a0d07c4d3b9f53117de3463048821492b9fea2"
      ],
      "author": {
        "name": "Jerry Belton",
        "email": "jcecil@google.com",
        "time": "Wed Nov 11 15:02:15 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Nov 11 15:02:15 2026 -0800"
      },
      "message": "[project] Add RequiredRelPaths and blocking cache lock (1/4)\n\nAdd context-aware cache lock acquisition and per-destination required\nfile verification to support on-demand subset prebuilt materialization:\n\n- Add `AcquireCacheLockWithContext` in `project/cache_lock.go` so\n  concurrent subset fetches (`jiri fetch-package`) wait for `.cache.lock`\n  instead of failing fast when parallel build actions materialize\n  packages simultaneously.\n- Add `RequiredRelPaths` to `WorkspaceDestination` and check\n  `hasRequiredPaths` in `IsUpToDate`, `isRBECasfsDestinationUpToDate`,\n  and `EnsureDestinationViaRBE` so a stamped destination directory whose\n  requested file was deleted is detected as out of date and re-linked.\n- Update `WorkspaceDestination.Link` and `fixSymlinksAfterRestoration`\n  to preserve nested stamped child destinations via\n  `cleanDestinationPreservingNestedDestinations` when re-linking an\n  enclosing parent destination.\n\nBug: 571564570\nBug: 566304584\nTest: go test -race ./project\nChange-Id: I19500cf2f74454cc691f73698e908761cb42d546\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1964712\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nReviewed-by: Chase Latta \u003cchaselatta@google.com\u003e\nCommit-Queue: Jerry Belton \u003cjcecil@google.com\u003e\n"
    },
    {
      "commit": "10a0d07c4d3b9f53117de3463048821492b9fea2",
      "tree": "34d443a8faa95c968049952343affeeb14a3daae",
      "parents": [
        "2c9db754b8e8a76a65ff004ddeceab499df54071"
      ],
      "author": {
        "name": "Otto FDW (Agent)",
        "email": "groupagent-fuchsia-developer-workflows-team@google.com",
        "time": "Tue Nov 10 10:57:45 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Nov 10 10:57:45 2026 -0800"
      },
      "message": "[project] Normalize rpc:// remotes for project keys\n\nRewrite rpc://\u003chost\u003e/\u003cpath\u003e remotes to https://\u003chost\u003e.googlesource.com/\nbefore stripping the URL scheme in rewriteAndNormalizeRemote, matching\nthe existing sso:// normalization.\n\nThis ensures that sso://, rpc://, and https:// remotes for the same\nrepository produce identical project keys in MakeProjectKey, allowing a\nlocal project or import checked out with one scheme to match a manifest\nusing another.\n\nTest: go test -count\u003d1 -run TestProjectKeysMatch ./project\nTAG\u003dagy\nCONV\u003d0775ca48-0548-4e6c-a933-86aec9e100e3\nChange-Id: I9a4cbdc0cae78eaa423c9d3e49c4500711f83bcc\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1967374\nFuchsia-Auto-Submit: Darren Chan \u003cchandarren@google.com\u003e\nReviewed-by: Jiaming Li \u003clijiaming@google.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nCommit-Queue: Darren Chan \u003cchandarren@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "2c9db754b8e8a76a65ff004ddeceab499df54071",
      "tree": "50c7ba954280b3d2087d1e28f22d94951e3f7c2f",
      "parents": [
        "8450522c47cd49ac4903fa4e5341b01c72584fb6"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Fri Nov 06 11:56:30 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Nov 06 11:56:30 2026 -0800"
      },
      "message": "[project] Add revisionfrom attribute to \u003cimport\u003e\n\nLet an \u003cimport\u003e take its revision from the \u003cimport\u003e with the same name\nand remote in another manifest in the same repository. integration.git\nis splitting fuchsia/stem (Copybara-mirrored to public integration) from\ninternal/stem, and both pin the same fuchsia revision. Today every\nrecipe that rolls, cherry-picks, or snaps fuchsia would need config\ntelling it to update both files in lockstep. With revisionfrom there\u0027s\nonly one pin to update.\n\nThe source import must have a literal revision and the same remote, so\nreferences can\u0027t chain or point at a different repository. jiri edit\nfollows the reference and edits the source manifest, and jiri manifest\nand snapshots report the resolved revision.\n\nNothing about this is specific to \u003cimport\u003e. For consistency it would\nmake sense to support revisionfrom on \u003cproject\u003e too, and versionfrom on\n\u003cpackage\u003e. We only need it on \u003cimport\u003e for now, so this only adds it\nthere to keep the change simple.\n\nTAG\u003dagy\nCONV\u003d5754636e-4dc3-4aeb-9555-b4c93bcc67a9\n\nBug: 580927649\nChange-Id: I344383c221220f4cac97f94591c694ad34101e70\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1957392\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nLint: android-build-ayeaye@system.gserviceaccount.com \u003candroid-build-ayeaye@system.gserviceaccount.com\u003e\nReviewed-by: Hunter Freyer \u003chjfreyer@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "8450522c47cd49ac4903fa4e5341b01c72584fb6",
      "tree": "e8d4e82c6707a78b97fd313f58612cf2a822bb6c",
      "parents": [
        "930be6a4e924c1c0cc506f7fc85b0374556365d4"
      ],
      "author": {
        "name": "Hunter Freyer",
        "email": "hjfreyer@google.com",
        "time": "Thu Nov 05 06:41:34 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Nov 05 06:41:34 2026 -0800"
      },
      "message": "[project] Include import topology in jiri project -list-remote-projects\n\nExpose imported_by and import_manifest in \u0027jiri project -list-remote-projects\n-json-output\u0027 so callers can resolve the full repository pin graph directly\nfrom Jiri without re-parsing XML manifests.\n\nBug: 580924899\nChange-Id: I1253d619d2322dce3a518f28012a8a8b6a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1942512\nCommit-Queue: Hunter Freyer \u003chjfreyer@google.com\u003e\nFuchsia-Auto-Submit: Hunter Freyer \u003chjfreyer@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "930be6a4e924c1c0cc506f7fc85b0374556365d4",
      "tree": "6ef017abaeb7487f9415f05eb18240e1bdbae689",
      "parents": [
        "3e0b6b770b7f154c5824bbe0d96b9ff661e6b89f"
      ],
      "author": {
        "name": "Sen Jiang",
        "email": "senj@google.com",
        "time": "Wed Nov 04 16:03:13 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Nov 04 16:03:13 2026 -0800"
      },
      "message": "[worktree] Sync to parent JIRI_HEAD\n\nWhen jiri update runs while a parent project is checked out on a local\nbranch or local commit, Jiri updates the project\u0027s JIRI_HEAD and\nrefs/remotes/jiri/head to the manifest revision, leaves HEAD on the\nlocal commit, and records that local HEAD in\n.jiri_root/update_history/latest.\n\nSyncing a worktree previously read target revisions solely from\nupdate_history/latest, causing worktrees to check out the parent\u0027s local\nbranch commit and overwrite the shared refs/remotes/jiri/head ref with\nthat commit. Prefer each parent project\u0027s JIRI_HEAD when\nupdate_history/latest is present so worktrees stay aligned with the\nparent\u0027s Jiri sync point, and warn when a project is skipped due to\nshared local-config ignore or no-update settings.\n\nBug: None\nTest: go test -race -count\u003d1 -run Worktree ./project ./integrationtests\nTest: go test -short -count\u003d1 ./...\nChange-Id: Ic7e9435104f383c53e94670cbb9f1e9160bac336\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1942635\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nCommit-Queue: Sen Jiang \u003csenj@google.com\u003e\n"
    },
    {
      "commit": "3e0b6b770b7f154c5824bbe0d96b9ff661e6b89f",
      "tree": "855b428901c6f667563eb502f485a2baf520b7f0",
      "parents": [
        "9df99485589d9a9cfe54e791510112386e788be0"
      ],
      "author": {
        "name": "global-integration-roller",
        "email": "global-integration-roller@fuchsia-infra.iam.gserviceaccount.com",
        "time": "Tue Nov 03 09:34:55 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Nov 03 09:34:55 2026 -0800"
      },
      "message": "[cipd] Update Jiri\u0027s version of the cipd CLI\n\nThis is an automated change that updates Jiri\u0027s pinned version of the\n\u0027cipd\u0027 command-line tool. It\u0027s important to use an up-to-date version\nthe cipd CLI in case because the CIPD backend doesn\u0027t guarantee support\nfor arbitrarily old versions of the cipd CLI.\n\nThis change does not require review, just a Code-Review+2 and\nCommit-Queue+2.\n\nHowever, this change should only be submitted by a human during normal\nworking hours. Due to Jiri\u0027s self-updating functionality Jiri changes\nroll out automatically to users the next time they run \u0027jiri update\u0027,\nso it\u0027s important to remain online shortly after submitting this\nchange to respond in case users report issues.\n\nGenerated by \u0027scripts/update_cipd.sh\u0027.\n\nRoller-URL: https://cr-buildbucket.appspot.com/build/8666136511725705857\nCQ-Do-Not-Cancel-Tryjobs: true\nChange-Id: I2c5dc4ee221f356294c448fd9056796b88a9857c\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1944457\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Advik Sachdeva \u003cadviks@google.com\u003e\nCommit-Queue: Advik Sachdeva \u003cadviks@google.com\u003e\n"
    },
    {
      "commit": "9df99485589d9a9cfe54e791510112386e788be0",
      "tree": "42f3cad8ce0054a9c6eafa224566cc3ed3edfd5d",
      "parents": [
        "5abf9c4d533e7a24fa1dbbfaa5b1905be8ba12c8"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Nov 03 09:29:58 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Nov 03 09:29:58 2026 -0800"
      },
      "message": "[project] Hide prebuilt GC logs\n\nPrebuilt GC runs on every package fetch and logs one line per removed\ndirectory or symlink. These lines aren\u0027t actionable, and they show up\nin large bursts the first time stale prebuilts get cleaned up. Failed\nremovals are still logged as warnings.\n\nAlso fix the package cache doc, which said this cleanup only runs with\njiri update -gc.\n\nTAG\u003dagy\nCONV\u003d4439b753-c0fc-405e-81c4-8ac665d43f15\n\nChange-Id: I156f78cf4e1ef56f98ac564dff72b1e2e7c3f626\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1942513\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "5abf9c4d533e7a24fa1dbbfaa5b1905be8ba12c8",
      "tree": "ed1ea3245f63c51cb1baf841291d753b31cb96d4",
      "parents": [
        "802eccb9cbc9abb9552ec5373f09cf862869734d"
      ],
      "author": {
        "name": "Chase Latta",
        "email": "chaselatta@google.com",
        "time": "Mon Nov 02 09:55:30 2026 -0800"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Nov 02 09:55:30 2026 -0800"
      },
      "message": "[jiri] Support generic delegated subcommands\n\nSupport delegating subcommands configured via .jiri_root/config to external tools (such as fuchsia-workspace sync for update). This prevents non-standard workspace checkouts from running uncoordinated operations.\n\nAdds DelegatedCommands configuration to Config, introduces --no-delegate-subcommands and JIRI_NO_DELEGATE_SUBCOMMANDS bypass controls, and intercepts configured subcommands in NewCommander to forward execution and propagate exit codes.\n\nBug: None\nTest: go test -v -run TestDelegated ./cmd/jiri/subcommands\nTest: go test -v -run TestDelegatedCommandsConfig .\nChange-Id: I38e7006bb17d2db72e8dde47e66437c5faddf4d4\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1825449\nReviewed-by: Jiaming Li \u003clijiaming@google.com\u003e\nCommit-Queue: Chase Latta \u003cchaselatta@google.com\u003e\n"
    },
    {
      "commit": "802eccb9cbc9abb9552ec5373f09cf862869734d",
      "tree": "b481d2ba6ed64d972be3258b9d00f0f5e8974a8b",
      "parents": [
        "d53f7d5fa5bfdfcb6fe644b3b0d1d64237d7c567"
      ],
      "author": {
        "name": "Suraj Malhotra",
        "email": "surajmalhotra@google.com",
        "time": "Wed Oct 28 14:06:48 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 28 14:06:48 2026 -0700"
      },
      "message": "[project] Apply git hooks concurrently\n\napplyGitHooks, which runs at the end of updateProjects (e.g. in\n`jiri update`), iterated serially over every operation, spawning git\ncommands (CurrentGitHooksPath and TopLevel) and writing hook files one\nproject at a time.\n\nRun applyProjectGitHooks concurrently across operations (bounded by\n-j), since each project writes to its own hooks directory. Collect\nper-index errors and return the first failing operation\u0027s error in\norder.\n\nApplying hooks to the 176 projects of a worktree (timed with a\ntemporary harness that loads its local projects and calls\napplyGitHooks, with -j 50) went from ~0.57s serially to ~0.07s.\n\nTest: go test -race ./...; new TestApplyGitHooks; benchmarked\napplyGitHooks serially and concurrently on a 176-project checkout.\n\nChange-Id: Ia1a10ecc5a899c39ba9b34f765d0493c6a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1908826\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Suraj Malhotra \u003csurajmalhotra@google.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\n"
    },
    {
      "commit": "d53f7d5fa5bfdfcb6fe644b3b0d1d64237d7c567",
      "tree": "06689ff4bb771b737d3850febc90b55ba0de7f0d",
      "parents": [
        "d9672bb61270ad769ba91b7791022ffb2e6fd026"
      ],
      "author": {
        "name": "Suraj Malhotra",
        "email": "surajmalhotra@google.com",
        "time": "Wed Oct 28 13:49:03 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 28 13:49:03 2026 -0700"
      },
      "message": "[project] Match snapshot projects concurrently\n\nIn LocalProjects(FastScan), projectsExistLocally checked each snapshot\nproject serially (running git commands via IsLocalProject), and then\nthe caller read each project\u0027s LocalConfig serially (running\nCommonGitDir again). Across 176 projects this took ~1.35s per FastScan\ncall (~0.9s in \"match manifest\" plus ~0.45s reading local configs).\nFastScan runs for most commands that look at local projects, e.g.\n`jiri status` and `jiri fetch-packages`, and once per worktree when\nbuilding the worktree package index.\n\nReplace projectsExistLocally with matchLocalProjects, which checks\nprojects and reads their LocalConfig concurrently (up to -j at once),\nreusing the common git directory from localProjectGitDir instead of\nquerying git twice per project. Sort results by project path and\ninspect all results before returning so error precedence and ordering\nstay deterministic.\n\nOn a checkout with 176 projects and 8 worktrees (warm runs):\n- `jiri -time status`: \"local projects\" went from ~1.36s to ~0.13s\n  (\"match manifest\" ~0.88s to ~0.08s), and the total from ~5.4s to\n  ~4.2s.\n- `jiri -time fetch-packages`: \"local projects\" went from ~1.36s to\n  ~0.13s, \"load worktree manifests\" from ~2.0s to ~0.75s, and the\n  total from ~9.3s to ~6.9s.\n\nTest: go test -race ./...; new TestLocalProjectsFastScanManyProjects;\nbenchmarked `jiri -time status` and `jiri -time fetch-packages`.\n\nChange-Id: If8475360586903abcbdfae5f3a855bb06a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1908824\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nCommit-Queue: Suraj Malhotra \u003csurajmalhotra@google.com\u003e\n"
    },
    {
      "commit": "d9672bb61270ad769ba91b7791022ffb2e6fd026",
      "tree": "383419689c09ce5d289dcab5750b2ca3b1bc0aef",
      "parents": [
        "ac5b478eb83996760840211001dfa7db633fd23d"
      ],
      "author": {
        "name": "Suraj Malhotra",
        "email": "surajmalhotra@google.com",
        "time": "Wed Oct 28 13:33:08 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 28 13:33:08 2026 -0700"
      },
      "message": "[project] Load worktree manifests concurrently\n\nFetching packages (`jiri fetch-packages`, `jiri update`, and\n`jiri worktree sync` when it can\u0027t link them from the main checkout\u0027s\npackage cache) spends most of \"fetch cipd packages\" in\nLoadWorktreePackageIndex, which serially scans local projects and\nloads the manifest of every registered worktree, and of the main\ncheckout when run from a worktree. The active checkout is already\nskipped, since its packages are passed in by the caller, but each\nother workspace still takes ~1.6s: with 8 worktrees, building the\nindex took ~13s of a ~20.3s `jiri fetch-packages`.\n\nLoad the other workspaces\u0027 manifests concurrently, timed as a whole as\n\"load worktree manifests\". Load failures are returned rather than\nlogged, and everything else stays serial in workspace order: warning\nabout failures, filtering package ACLs, resolving destinations and\nmerging instances. So warnings are logged in the same order as before,\nnothing is logged for workspaces after one that fails to resolve, each\npackage\u0027s ACL is still checked at most once per command, and the index\nis deterministic.\n\nBound the concurrent loads by min(-j, 16). Each load scans local\nprojects with up to -j concurrent git commands, so bounding by -j\nalone could run -j*-j git processes. With 9 checkouts, a bound of 8\ntakes ~3.4s vs ~1.9s unbounded; 16 doesn\u0027t bind at 9 checkouts and\nleaves headroom while capping that fan-out.\n\nOn a checkout with 8 worktrees, warm `jiri -time fetch-packages` went\nfrom ~20.3s to ~9.3s, \"fetch cipd packages\" from ~18.5s to ~7.5s, and\nloading the worktree manifests from ~13s to ~2.0s (~0.75s with\nfxr/1908824, which checks each workspace\u0027s projects concurrently).\n\nTest: go test -race ./...; new TestWorktreePackageIndex_ManyWorktrees\nand TestWorktreePackageIndex_ReturnsFirstResolveError; benchmarked\n`jiri -time fetch-packages` before and after.\n\nChange-Id: I72371d5733a7c65b2f8b1be4d61719f06a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1909486\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nCommit-Queue: Suraj Malhotra \u003csurajmalhotra@google.com\u003e\n"
    },
    {
      "commit": "ac5b478eb83996760840211001dfa7db633fd23d",
      "tree": "aca7ead1bb7901dc62b2a45b30d1676974ea9da1",
      "parents": [
        "600bbfe3358f7030504917346b1b773b4caaa251"
      ],
      "author": {
        "name": "Suraj Malhotra",
        "email": "surajmalhotra@google.com",
        "time": "Fri Oct 23 16:06:50 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 23 16:06:50 2026 -0700"
      },
      "message": "[project] Make LoadManifestFile concurrency-safe\n\nPrepare for loading the manifests of several worktrees concurrently\nwhen building the worktree package index:\n\n- Add jiri.WithoutTimer, which returns a context whose TimerPush and\n  TimerPop are no-ops. The command\u0027s timing.Timer keeps a single stack\n  of nested operations, so code run on concurrent goroutines must not\n  use it.\n- Make MarkOverridesWarned atomically report whether it was the first\n  call, and drop OverridesWarned, so that concurrent manifest loads\n  can\u0027t both see that the user wasn\u0027t warned about overrides and both\n  warn.\n- Serialize updateOrCreateCache per git cache directory within the\n  process, replacing updateCache\u0027s per-call mutexes. LoadManifestFile\n  calls it via cloneManifestRepo when a manifest project is missing\n  locally; concurrent calls for one cache race to clone and configure\n  it, and a failed `git config` is treated as corruption that wipes\n  the cache.\n\nTest: go test -race ./...; new TestWithoutTimer and\nTestUpdateOrCreateCacheConcurrent (fails without the lock), updated\nTestRunStateSharedByDerivedContexts.\n\nChange-Id: I120fc82ef813ea03ac7aa1685c3902556a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1909485\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nCommit-Queue: Suraj Malhotra \u003csurajmalhotra@google.com\u003e\n"
    },
    {
      "commit": "600bbfe3358f7030504917346b1b773b4caaa251",
      "tree": "e7b772f2954283f665915a6dbdfd867c83a16f7e",
      "parents": [
        "a7d916b153ae1b66cf740ec117a6b35eb81f2931"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 23 14:13:25 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 23 14:13:25 2026 -0700"
      },
      "message": "[worktree] Project CIPD prebuilts from parent\n\nProject CIPD prebuilt hardlink forests into worktrees directly from\nparent .jiri_cache_hash stamps and shared .jiri_root/packages/instances\nwhile holding cache.Lock(), without spawning cipd (auth-info,\nensure-file-resolve, or ensure) or indexing every registered worktree.\n\nFall back to FetchPackages when a non-CIPD prebuilt provider (such as\nrbe-casfs) is configured, when parent stamps are virtualized, missing,\nor out of date, or when checkout attribute sets do not match.\n\nTest: go test -count\u003d1 ./project ./integrationtests\nTest: scripts/shac check --all\nBug: None\nTAG: agy\nCONV: ee18f0c0-0e92-49a7-baad-27717dcce9ca\nChange-Id: I5043b5b88796fac3d03e6f77b74535d7125144db\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1916730\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\n"
    },
    {
      "commit": "a7d916b153ae1b66cf740ec117a6b35eb81f2931",
      "tree": "72c557f6816a3eede28553b534fcd450fc9542bb",
      "parents": [
        "bf520127d04be75d04cc99898fa0f6f5c85c52f4"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 23 14:09:21 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 23 14:09:21 2026 -0700"
      },
      "message": "[worktree] Sync projects and hooks from parent\n\njiri worktree sync previously ran full tree scans and delegated to\nupdateProjects, so even an up-to-date Fuchsia worktree spent ~18s on\ngit queries and hooks. It also left .jiri_manifest and checkout config\nunsynced and cloned newly added manifest projects as standalone\nrepositories instead of git worktrees or colocated jj workspaces.\n\nRefactor WorktreeSync to synchronize projects, manifests, snapshots,\nand hooks directly from the parent checkout:\n- Load the parent manifest offline under parentCtx (disallowClone) and\n  load parent and worktree projects from update_history/latest and .git\n  metadata without walking the tree or spawning git when on detached\n  HEAD with current JIRI_HEAD and refs/remotes/jiri/head.\n- Provision newly added manifest projects from the parent as git\n  worktrees or colocated jj workspaces, validate repository identity on\n  unrepresented paths, and update JIRI_HEAD and refs/remotes/jiri/head\n  before checking out so secondary jj workspaces auto-import target\n  revisions.\n- Sync .jiri_manifest atomically (rewriting external or parent-only\n  relative symlinks), write .jiri_root/update_history/latest from\n  in-memory project state, propagate parent config settings, and skip\n  RunHooks on no-op syncs.\n- Remove unreachable UpdateUniverseParams.SkipStatus, worktree guards\n  in updateProjects, and the isJJWorkspace branch in computeOp.\n\nTest: go test -count\u003d1 ./...\nTest: scripts/shac check --all\nBug: None\nTAG: agy\nCONV: ee18f0c0-0e92-49a7-baad-27717dcce9ca\nChange-Id: I5436d2ca20493ebe6723502e1dda1463f698bf68\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1910222\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "bf520127d04be75d04cc99898fa0f6f5c85c52f4",
      "tree": "28d05dfb2b9974cd271a5635a36f2c9afb4b80c1",
      "parents": [
        "57024c4071851a23405e75c2ac0e536851f40ac2"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 23 13:05:56 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 23 13:05:56 2026 -0700"
      },
      "message": "[project] Fix ProjectKey.Less and active-root scan\n\nFix `ProjectKey.Less` so it only falls back to comparing `remote` when\n`name` is equal. Previously, `ProjectKey.Less` returned true for both\n`a.Less(b)` and `b.Less(a)` when one key had a smaller `name` and the\nother had a smaller `remote`, violating strict weak ordering. Add\n`sortedProjectKeys` and use it in `convertMismatchedWorktreeProjects`\nfor deterministic map iteration.\n\nAlso reuse the caller\u0027s `checkout` and `activePkgs` for `checkout.Root`\nin `LoadWorktreePackageIndex` instead of re-running `FastScan`,\n`LoadManifestFile`, and `FilterACL` on the active workspace.\n\nTest: go test -count\u003d1 ./project\nTest: scripts/shac check --all\nBug: None\nTAG: agy\nCONV: ee18f0c0-0e92-49a7-baad-27717dcce9ca\nChange-Id: I149c6d515033669c8a818fa90354c238b4b93aa1\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1916729\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "57024c4071851a23405e75c2ac0e536851f40ac2",
      "tree": "c530301da260fa17bdb45ec1d91318113f5f267a",
      "parents": [
        "f818a3f4d7e5edf47aede8ba1d5f4e8a3297e05b"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 23 13:01:54 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 23 13:01:54 2026 -0700"
      },
      "message": "[jiri] Centralize checkout config in ApplyConfig\n\nExtract `Checkout.ApplyConfig` and `Config.InheritParentPrebuilt` so\n`newCheckout` and `getWorkspaceContext` share a single implementation\nfor applying `.jiri_root/config` settings onto a `Checkout`.\n\nPreviously, `getWorkspaceContext` shallow-copied the caller\u0027s `Checkout`\nand only updated prebuilt/RBE fields, leaking settings such as\n`KeepGitHooks`, `RewriteSsoToHttps`, `FetchingAttrs`, `ExcludeDirs`, and\ncustom `PackageCachePath` when switching between workspaces or when a\ntarget workspace had no `.jiri_root/config` file.\n\nTest: go test -count\u003d1 . ./project\nTest: scripts/shac check --all\nBug: None\nTAG: agy\nCONV: ee18f0c0-0e92-49a7-baad-27717dcce9ca\nChange-Id: I4101e4a948bc4dd8cb77429de64685bcc8bc06c4\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1916728\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "f818a3f4d7e5edf47aede8ba1d5f4e8a3297e05b",
      "tree": "b91620e248f4d5a0c5ab21f19a0286defa78f90f",
      "parents": [
        "cd76783873c5c928b755dd65102253ec127a18ab"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Fri Oct 23 10:48:28 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 23 10:48:28 2026 -0700"
      },
      "message": "[project] Avoid redundant work when cloning manifest repos\n\nOn cold checkouts, manifest loading clones imported manifest repos into\na temporary directory just to read their manifest and lock files before\nupdateProjects creates the actual checkout. Previously this:\n- Cloned from the remote URL over the network even after populating the\n  local git cache.\n- Passed --dissociate when configured, repacking multi-GB packfiles into\n  a temporary directory that gets deleted at the end of the command\n  (taking 92.7s for fuchsia.git and 45.6s for integration.git on build\n  8667207747838549457).\n- Checked out the full working tree (12.3s for fuchsia.git) even when\n  reading manifests via `git show`.\n- Mutated the caller\u0027s localProjects map with the temporary directory,\n  causing non-root manifest repos to be moved across filesystems and\n  requiring the checkout.Root workaround from commit\n  8e4cda3268fbbeed1100ed7aeb5c7586c0d7a5aa.\n\nAlso update `jiri init` so re-running without -dissociate clears\ndissociate in .jiri_root/config on existing warm checkouts.\n\nTAG\u003dagy\nCONV\u003dc353bdc6-7034-4219-8410-e174eb5db03e\n\nBug: 576801109\nChange-Id: I90ceef3314afcb4cfee9de0fb7056b9bacb8242f\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1914486\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: David Iserovich \u003cdiserovich@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "cd76783873c5c928b755dd65102253ec127a18ab",
      "tree": "d7c22f5b134a95ff0ca6fabde967f42bd66985a8",
      "parents": [
        "58cc39c7600572514ab47feff453817dbddf6435"
      ],
      "author": {
        "name": "Chase Latta",
        "email": "chaselatta@google.com",
        "time": "Wed Oct 21 17:38:41 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 21 17:38:41 2026 -0700"
      },
      "message": "[cmd/jiri] Wire rbe-casfs prebuilt-provider in jiri init and fetch-packages\n\nAdd -prebuilt-provider, -prebuilt-ttl-refresh, -rbe-instance,\n-rbe-service, and -abfs-cache-dir flags to `jiri init` and route\n`jiri fetch-packages` and `RunUnifiedPackageCachePipeline` through\n`NewPrebuiltProvider` when `prebuilt-provider` is set to `rbe-casfs`.\n\nWhen `prebuilt-provider\u003drbe-casfs`:\n- `jiri fetch-packages --local-manifest` uses the fast\n  `update_history/latest` project snapshot and local manifest project\n  resolution, and includes manifest projects when filtering optional\n  packages.\n- `FetchPackages` short-circuits before `FilterACL` when all pinned\n  package destinations already have fresh virtualized `.jiri_cache_hash`\n  stamps on disk.\n\nChange-Id: I31e8f20a477e8fa9f35a0ef967dd932a6d6fbedc\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1893756\nReviewed-by: Jerry Belton \u003cjcecil@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Chase Latta \u003cchaselatta@google.com\u003e\n"
    },
    {
      "commit": "58cc39c7600572514ab47feff453817dbddf6435",
      "tree": "035d8e35c0cb1adab5368f91aefe6ae33303a394",
      "parents": [
        "ec8392d2394300f30aec6ab7bbcae35cf7a69105"
      ],
      "author": {
        "name": "Chase Latta",
        "email": "chaselatta@google.com",
        "time": "Wed Oct 21 17:35:33 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 21 17:35:33 2026 -0700"
      },
      "message": "[project] Add PrebuiltProvider, EnsureDestinationViaRBE, and ABFSRBEPrebuiltProvider\n\nAdd PrebuiltProvider (CIPDPrebuiltProvider and ABFSRBEPrebuiltProvider)\nand RBE CAS + casfs.ari_size destination materialization:\n\n- Add DefaultPrebuiltTTLRefresh, PrebuiltProvider, PrebuiltTTLRefresh,\n  RBEInstance, RBEService, and ABFSCacheDir to jiri.Config and\n  jiri.Checkout, and implement NewRBEClientForCheckout with cacheman\n  /read-config auto-discovery.\n- Introduce PrebuiltProvider, CIPDPrebuiltProvider (supporting both\n  full-tree sync and subset destination sync), and\n  ABFSRBEPrebuiltProvider executing independent destinations\n  concurrently in 16-worker ancestry-ordered waves.\n- Implement EnsureDestinationViaRBE: resolves cached RBETree manifests\n  via abfs-maps, performs cold CIPD unpacking into ephemeral temp dirs\n  under \u003cabfsCacheDir\u003e/tmp (with \u0027$OverrideInstallMode copy\u0027 and .cipd/\n  symlink dereferencing), seeds local \u003cabfsCacheDir\u003e/aris/ blobs via\n  atomic rename, uploads missing blobs to RBE CAS, refreshes stale (\u003e7d)\n  stamps, and materializes 0-byte casfs.ari_size files.\n\nBug: 566304584, 567114481\nChange-Id: Ic1f16c8cc5ae4c1535ef0f53a6e61d93da896829\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1903306\nReviewed-by: Jerry Belton \u003cjcecil@google.com\u003e\nCommit-Queue: Chase Latta \u003cchaselatta@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "ec8392d2394300f30aec6ab7bbcae35cf7a69105",
      "tree": "b4cfa9853f11ec87a6ea072fd0303766522e99d5",
      "parents": [
        "7e09eade7fcf6f17451b36da97c9c525eede4add"
      ],
      "author": {
        "name": "Chase Latta",
        "email": "chaselatta@google.com",
        "time": "Wed Oct 21 17:34:21 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 21 17:34:21 2026 -0700"
      },
      "message": "[project] Add RBETree, casfs.ari_size virtualizer, and LiveRBEClient\n\nAdd the core RBE CAS, cacheman UDS/HTTP, RBETree, and casfs.ari_size\nvirtualization primitives for prebuilt package materialization:\n\n- Extend CacheStampData with \u0027virtualized\u0027 and \u0027casfs_aris\u0027 fields,\n  MatchesDestinationMode, WriteWithMode, NeedsTTLRefresh, and Touch.\n- Add RBEDigest, RBETree, BuildTreeFromInstanceDir,\n  BuildTreeAndFilePathsFromInstanceDir, and MaterializeRBETreeVirtual\n  (creating 0-byte placeholder inodes with casfs.ari_size extended\n  attributes).\n- Add ABFSMapClient and RBEClient interfaces, SynthesizeABFSMapKey,\n  EncodeRBEResultGitBlob, and DecodeRBEResultGitBlob.\n- Implement LiveRBEClient backed by local ABFS cacheman Unix domain\n  sockets (\u003cabfsCacheDir\u003e/socket and \u003cabfsCacheDir\u003e/http.socket) and\n  HTTP/2 gRPC to ContentAddressableStorage / ByteStream.\n\nBug: 566304584, 567114481\nChange-Id: I160a90c6888393b05a819b8532e030814595ebcb\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1903305\nReviewed-by: Jerry Belton \u003cjcecil@google.com\u003e\nCommit-Queue: Chase Latta \u003cchaselatta@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "7e09eade7fcf6f17451b36da97c9c525eede4add",
      "tree": "09a4d76ec39ec75aacf2a70b7c1badd1671d0d99",
      "parents": [
        "b153055b1cf04767da268e17f252c7d3240820fb"
      ],
      "author": {
        "name": "Curtis Galloway",
        "email": "curtisgalloway@google.com",
        "time": "Mon Oct 19 14:30:53 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 19 14:30:53 2026 -0700"
      },
      "message": "[cipd] Cache package ACL checks per command invocation\n\nWhen syncing the shared package cache across N registered worktrees,\nFetchPackages and LoadWorktreePackageIndex invoke Packages.FilterACL\nonce for the active checkout and once per sibling worktree, spawning\nduplicate `cipd acl-check` subprocesses for the same internal\u003d\"true\"\npackage prefixes while holding .cache.lock.\n\nCache CIPD read-access results on jiri.runState (via CIPDReadAccess and\nRecordCIPDReadAccess) so CheckPackageACL checks each unique package path\nat most once per command invocation, skipping CIPD bootstrap and\ntemporary directory creation when all requested packages are already\ncached and avoiding caching false negatives when the context is\ncanceled.\n\nAlso fix Packages.FilterACL to look up strings.TrimRight(pkg.Name, \"/\")\nin the second loop without mutating Package.Name so inaccessible\ninternal packages with a trailing slash are filtered out, and make\nTestWorktreePackageIndex_FiltersInaccessibleInternalPackages hermetic.\n\nBug: 573678825\nTest: go test -race . ./cipd ./project -run \u0027TestRunState|TestWithRunState|TestCheckPackageACL|TestWorktreePackageIndex\u0027\nChange-Id: Ic7e0e3ccce1843de656e554d65e543663a0d2b88\nTAG\u003dagy\nCONV\u003db4d6a874-4abb-42f9-86e7-69888e1cc22c\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1893755\nCommit-Queue: Curtis Galloway \u003ccurtisgalloway@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "b153055b1cf04767da268e17f252c7d3240820fb",
      "tree": "f3f64017d05337013cdea31f5e383e2c4bd71a25",
      "parents": [
        "4a5d26cbf21e2b484434207a4a3b8a28bfe7ad05"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Oct 19 08:12:23 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 19 08:12:23 2026 -0700"
      },
      "message": "[project] Delete Package.Flag and WritePackageFlags\n\nWith the removal of the flag attribute from the brcmfmac package\nentry in manifests/prebuilts (where build/cipd.gni is now checked into\nfuchsia.git and inspects //.jiri_root/prebuilt.json directly),\nPackage.Flag has no remaining users across any manifest.\n\nDelete Package.Flag, WritePackageFlags, TestWritePackageFlags, and\nassociated documentation in manifest.md.\n\nTest: go test ./project/...\nTest: scripts/shac check\nBug: 571336644\nDepends-on: I77efb8a603e8a094d13466912b7693b8bb239d16\nChange-Id: I646e25d2ef23e1b754fdc9584c2b2c329b9d47e5\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1894640\nDependencies-Satisfied: cl-deps-checker@fuchsia-internal-service-accts.iam.gserviceaccount.com \u003ccl-deps-checker@fuchsia-internal-service-accts.iam.gserviceaccount.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "4a5d26cbf21e2b484434207a4a3b8a28bfe7ad05",
      "tree": "062cdf18fcbfe9e51df706487da9f82cabb9e7fc",
      "parents": [
        "0628b5705aad8edd2a9c2dbd0a01c313c7e16ef1"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 16 11:59:59 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 16 11:59:59 2026 -0700"
      },
      "message": "[project] Remove remotebranch attribute\n\nRemove the obsolete remotebranch XML attribute from Project and Import\nstructs in project/project.go and project/manifest.go.\n\nAn audit showed that remotebranch was unused across all Fuchsia, vendor,\nand integration manifests, and the only two infra manifest entries\nredundantly specified remotebranch\u003d\"main\", which is already Jiri\u0027s\ndefault.\n\nDrop RemoteBranch from Project and Import, remove remoteBranch\npropagation in loader.go, remove the -remote-branch flag in jiri import,\nand update documentation and tests.\n\nTest: go test -count\u003d1 ./... -skip TestWorktreeJJLifecycle\nTest: scripts/shac check\nBug: 571336841\nChange-Id: Ide428ff5fe898ac89c98509f15216a5b81430b7a\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1896563\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "0628b5705aad8edd2a9c2dbd0a01c313c7e16ef1",
      "tree": "4efb19e1c61d72b40695ceda29f566850746af73",
      "parents": [
        "a6caaf5c61ed16c3e9bdda2926ad7724097528e2"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Fri Oct 16 09:12:51 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 16 09:12:51 2026 -0700"
      },
      "message": "[gitutil] Don\u0027t lazy-fetch in IsRevAvailable\n\nIn a partial clone, IsRevAvailable\u0027s `git cat-file` on a missing\ncommit makes git fetch it from the promisor remote. That fetch sends\nno \"have\" lines, so the server sends the commit\u0027s whole commit and tree\nhistory. In fuchsia-roller builds, this check for a new integration.git\ncommit in the blob:none cache took 87 of the 115 seconds of `jiri\nupdate`, before jiri ran the real incremental fetch (11s). The result\nof that lookup was always \"not available\" anyway, since the cache\u0027s\nHEAD can\u0027t be a commit that wasn\u0027t there yet.\n\nSet GIT_NO_LAZY_FETCH\u003d1 for the check, so a missing revision is\nreported as unavailable (logged at debug level, not as an error) and\njiri goes on to fetch it normally. It also skips resolving the revision\nwhen it\u0027s already the checked-out commit. Older gits without\nGIT_NO_LAZY_FETCH (\u003c 2.44) ignore the variable.\n\nTAG\u003dagy\nCONV\u003dcbf04936-7608-4c03-ace7-b6a76416846c\n\nChange-Id: Iaba318ad7df4fea5b79e9efadcec9c84a0fbe46c\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1896718\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "a6caaf5c61ed16c3e9bdda2926ad7724097528e2",
      "tree": "66fe98d2d88ab1992fc8640f4a1ad581c1244358",
      "parents": [
        "1f3cec517faaff0fa2b67f1c17ee9b9b93879de2"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 16 08:52:28 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 16 08:52:28 2026 -0700"
      },
      "message": "[gerrit] Drop dead commitMsgFetcher and SSO client\n\nSince commit 5a573a80 (\"[project] Embed commit-msg hook\"),\napplyGitHooks embeds commitMsgHookBytes rather than fetching\n/tools/hooks/commit-msg over HTTP/SSO.\n\nAs a result, commitMsgFetcher, fuchsiaGerritHost, gerrit/sso.go,\nConfig.SsoCookiePath, Checkout.SsoCookiePath, and the -sso-cookie-path\nflag in \u0027jiri init\u0027 are unreachable dead code.\n\nDelete the unused fetcher, SSO cookie client, CLI flag, and unused\ngolang.org/x/net dependency.\n\nTest: go test -count\u003d1 ./...\nTest: scripts/shac check\nBug: 571337128\nChange-Id: I69deab862bbb8d1f5ed9c7ee082d8adbf0ae57a5\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1894637\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "1f3cec517faaff0fa2b67f1c17ee9b9b93879de2",
      "tree": "083625b10f2c8d37c0b43231cc58af5d7d3b1bf5",
      "parents": [
        "addf9c7dac15cfc2d68d0fd7f789760a58f3e240"
      ],
      "author": {
        "name": "Travis Geiselbrecht",
        "email": "travisg@google.com",
        "time": "Thu Oct 15 15:49:02 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Oct 15 15:49:02 2026 -0700"
      },
      "message": "[cache] Filter package ACLs for sibling worktrees\n\nWhen building the shared package index in LoadWorktreePackageIndex, only\nthe active checkout root received activePkgs (which FetchPackages had\nalready filtered via FilterACL). Packages loaded from sibling worktree\nmanifests only ran through FilterOptionalProjectsPackages, allowing\ninaccessible internal\u003d\"true\" packages from other worktrees to be added\nto the shared CIPD ensure file and causing `jiri update` to fail for\ncallers without internal CIPD access.\n\nUpdate LoadWorktreePackageIndex to also filter sibling worktree packages\nwith FilterACL before collecting workspace destinations, and add a unit\ntest verifying that inaccessible internal packages from a registered\nworktree are excluded from the package index and ensure file.\n\nBug: 557237866\nTest: go test -v -count\u003d1 -run \"TestWorktreePackageIndex\" ./project\nChange-Id: I8963fdb9c85f9c04252ba54daee29701e00156a1\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1893193\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Travis Geiselbrecht \u003ctravisg@google.com\u003e\n"
    },
    {
      "commit": "addf9c7dac15cfc2d68d0fd7f789760a58f3e240",
      "tree": "09705c48fc32b4b708e8020e9a569f8ae942880e",
      "parents": [
        "80c40580b1a66bd96e5b1d44f54eb35698f57048"
      ],
      "author": {
        "name": "Travis Geiselbrecht",
        "email": "travisg@google.com",
        "time": "Thu Oct 15 15:19:52 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Oct 15 15:19:52 2026 -0700"
      },
      "message": "[metadata] Update expected base64 zlib outputs for Go 1.27\n\nGo 1.27 updated compress/flate\u0027s default DEFLATE encoding (emitting\nstored uncompressed blocks for tiny inputs when smaller than Huffman\nheaders, and slightly different match distances on larger inputs). As a\nresult, ToBase64() produces valid zlib streams with different compressed\nbytes than Go 1.26, causing TestToBase64 and TestLDFlag to fail after\nthe CI Go toolchain rolled to 1.27.1.\n\nAdd the Go 1.27 base64-encoded zlib strings to the allTests allowlist in\nmetadata_test.go alongside the existing Go \u003c\u003d 1.26 encodings so both\nTestToBase64 and TestFromBase64 pass across Go versions.\n\nBug: None\nTest: go test -v ./metadata/...\nChange-Id: Ib3a26f01570d3ae6393e4aa3dcf1f9611605637b\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1893333\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Travis Geiselbrecht \u003ctravisg@google.com\u003e\n"
    },
    {
      "commit": "80c40580b1a66bd96e5b1d44f54eb35698f57048",
      "tree": "7d28c632378ea66a3bf5420252d07597730b2db5",
      "parents": [
        "d50425f4ea8f1b97d970bd7db92802f96face9cf"
      ],
      "author": {
        "name": "Adam Snaider",
        "email": "asnaider@google.com",
        "time": "Thu Oct 15 11:39:18 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Oct 15 11:39:18 2026 -0700"
      },
      "message": "[worktree] Add configurable worktrees directory\n\nTools that manage pools of jiri worktrees, such as `fx worktree`, keep\nthem in .jiri_root/worktrees. Some users want them elsewhere, e.g. next\nto the checkout, where file watchers such as Watchman treat each\nworktree as its own project instead of ignoring it with .jiri_root.\n\nAdd `jiri init -worktrees-dir\u003d\u003cpath\u003e`, stored as \u003cworktrees_dir\u003e in\n.jiri_root/config. A relative path is relative to the checkout root and\nis stored as given, so `../fuchsia-worktrees` keeps working if the\ncheckout moves. Checkout.WorktreesDir() resolves it, defaulting to\n.jiri_root/worktrees; in a worktree it reads the main checkout\u0027s\ncurrent setting. `jiri worktree list` uses it for its fallback when\nthere is no worktrees registry. `jiri worktree add` still takes an\nexplicit path.\n\nBug: 570715187\nTest: go test . ./cmd/jiri/subcommands/ -run \u0027TestWorktreesDir|TestInitWorktreesDir\u0027\nTest: go test ./integrationtests/ -run TestWorktreeListLegacyFallbackUsesWorktreesDir\nChange-Id: I8996c4983bc87cb2c4a72efdc406643b6a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1880651\nCommit-Queue: Adam Snaider \u003casnaider@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "d50425f4ea8f1b97d970bd7db92802f96face9cf",
      "tree": "8dda7bf450928dc5d7e567610dd0d3c5b701b91d",
      "parents": [
        "b0772badff57772060a1c69655afe0c7dbd4e71c"
      ],
      "author": {
        "name": "Adam Snaider",
        "email": "asnaider@google.com",
        "time": "Thu Oct 15 08:40:02 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Oct 15 08:40:02 2026 -0700"
      },
      "message": "[worktree] Support Jujutsu workspaces\n\nGive projects whose repository in the main checkout is a Jujutsu (jj)\nrepo colocated jj workspaces in worktrees, instead of plain git\nworktrees. A colocated workspace (`jj workspace add --colocate`) is\nalso a git worktree whose HEAD jj keeps at the workspace\u0027s parent\ncommit, so git commands, and the rest of jiri, work in it unchanged.\nProjects that don\u0027t use jj still get git worktrees, so a checkout can\nmix both.\n\nOnly the operations that would otherwise fight jj over the working\ncopy use jj commands:\n\n- WorktreeAdd creates the workspace, named after the worktree.\n- WorktreeSync moves it with `jj new \u003crev\u003e` (after updating a stale\n  working copy), leaving uncommitted changes in the previous\n  working-copy commit. computeOp leaves jj workspaces alone.\n- WorktreeRemove forgets the workspace, which also removes its git\n  worktree, and deletes it.\n- `jiri clean` starts a fresh working-copy commit with `jj new @-`.\n- Relocating a moved project also repairs the workspace\u0027s .jj/repo\n  pointer, which jj may store as a relative path.\n\nAdd `jiri worktree sync -force` to convert worktree projects in place\nwhen the main checkout starts or stops using jj. Without -force, such a\nmismatch is an error. `jj workspace add` refuses non-empty directories,\nso git-to-jj conversion creates the workspace in a temporary directory\nand moves its .git and .jj into place, keeping the project\u0027s files,\nincluding uncommitted changes.\n\nBug: 570715187\nTest: go test ./integrationtests/ -run TestWorktreeJJ\nChange-Id: I65942970bb3e173bbffdbf66cf6e1ded6a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870249\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Adam Snaider \u003casnaider@google.com\u003e\n"
    },
    {
      "commit": "b0772badff57772060a1c69655afe0c7dbd4e71c",
      "tree": "288965427ad3d5ff3d196eaff8eab98b9e3cbca5",
      "parents": [
        "88ec68a5e3d3a73a032a043c6d4a29ea5ef27b0b"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Oct 15 08:28:29 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Oct 15 08:28:29 2026 -0700"
      },
      "message": "[cache] Remove package cache migration logic\n\nNow that all checkouts and infrastructure builders use the\ncontent-addressable package cache, remove bespoke in-place migration\nroutines (MigrateLegacyGroupCache, MigrateStandaloneToPackageCache,\nadoptCIPDPackages), in-place destination attachment (canAttachInPlace,\nCacheStamp.Remove), and v1-to-v2 cache instance contamination scanning\n(sanitizeContaminatedInstances).\n\nHandle any remaining legacy or pre-v2 checkouts via clean\nre-initialization: when a pre-v2 .jiri_cache_format_version, standalone\n\u003croot\u003e/.cipd directory, legacy symlink destination, or unexpected\n.jiri_root/packages entry is detected, reset the entire package cache\nand redownload clean CIPD instances before relinking destination\nhardlink trees.\n\nTest: go test -count\u003d1 ./...\nBug: 570540155\nTAG: agy\nCONV: b61c1b8b-31b5-475e-9656-5bc6c13b0794\nChange-Id: I9c562106bb570528159e8f8d114a278e617c5e2c\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1889113\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "88ec68a5e3d3a73a032a043c6d4a29ea5ef27b0b",
      "tree": "d1a23636e0707ee1b8ef6c17e15924c07633e104",
      "parents": [
        "749c5089cc2b865cf94b55a3fbe63f5d1a6aa895"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Wed Oct 14 11:29:11 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 14 11:29:11 2026 -0700"
      },
      "message": "[project] Delete GitAttributes and project Flag\n\nRemove unused Project.GitAttributes, the importTree DAG in\nproject/loader.go (which only existed to populate .gitattributes in\nthe deleted generate-gitmodules subcommand), and Project.Flag /\nWriteProjectFlags (unused across all Fuchsia, vendor, and infra\nmanifests). Package.Flag and WritePackageFlags remain in place for\n\u003cpackage\u003e elements.\n\nTest: go test ./...\nBug: 571336260\nChange-Id: If9fd766072f7ff4e7c5353c2d8613a151dcaa356\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1886819\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "749c5089cc2b865cf94b55a3fbe63f5d1a6aa895",
      "tree": "626d46420707aae62b5dbbe6284497c6c1bfe255",
      "parents": [
        "569d4564499186f7d487af4b2fa945b001625a48"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Tue Oct 13 12:25:53 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 13 12:25:53 2026 -0700"
      },
      "message": "[gitutil] Drop ctx from New and support bare ctx\n\nEvery method on *gitutil.Git already accepts a context.Context and\ndelegates to runGit(ctx, ...), making the ctx parameter on gitutil.New\nredundant. In addition, requiring a jiri.Checkout, cmdline.Env, and\nlog.Logger at gitutil.New construction time forced tests in\nintegrationtests, gitutil, and selfupdate to call xtest.NewContext(t)\n(creating a temporary .jiri_root and linking a cached cipd binary) just\nto run local git commands.\n\nRemove ctx from gitutil.New(opts ...gitOpt), resolve the checkout,\nenvironment, and logger from ctx at command execution time, and allow\njiri.CheckoutFromContext, jiri.EnvVars, and log.FromContext to tolerate\na bare context.Context such as t.Context().\n\nTest: scripts/shac check\nTest: go test -count\u003d1 ./...\nBug: 572839458\nChange-Id: I108c503a8e22b21c14116046446648bb51502d15\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1885628\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "569d4564499186f7d487af4b2fa945b001625a48",
      "tree": "c3d93eebf250699c886f1b7f8811150af12183f2",
      "parents": [
        "76c89f42c3907c09457e17a33d3f043717877425"
      ],
      "author": {
        "name": "Adam Snaider",
        "email": "asnaider@google.com",
        "time": "Tue Oct 13 10:48:33 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 13 10:48:33 2026 -0700"
      },
      "message": "[worktree] Repair gitlinks after relocating projects\n\nrelocateMovedWorktreeProjects repaired the bidirectional gitlinks before\nmoving a worktree project\u0027s directory, so a failed move left the\nworktree\u0027s .git file and the parent\u0027s gitdir pointing at the new path\nwhile the directory stayed at the old one. Repair them only after the\nmove succeeds, and simplify repairWorktreeGitLink to take the\nproject\u0027s (new) path.\n\nExtend TestWorktreeSyncSubdirPathChange to move the project past a\nstray directory at the destination and to check that the parent lists\nthe worktree at its new path.\n\nTest: go test ./integrationtests/ -run TestWorktreeSyncSubdirPathChange\nChange-Id: I515cc77d933f36f8d16dfcd4dc7d7c046a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1884363\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Adam Snaider \u003casnaider@google.com\u003e\n"
    },
    {
      "commit": "76c89f42c3907c09457e17a33d3f043717877425",
      "tree": "b2be4d14a5c090a822bd6789686499c2b443147d",
      "parents": [
        "7140d63954ca11d1200e22d41b4a1cb02eb4c47d"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Mon Oct 12 20:56:42 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 12 20:56:42 2026 -0700"
      },
      "message": "[project] Retry fetch if pinned revision missing\n\nWhen fetching a project or manifest import pinned to a specific commit,\n`git fetch` can succeed with exit code 0 against a stale Git-on-Borg\nreplica that does not yet have the target commit. Consolidate the\nrevision-fetching fallback logic into `fetchRevision` and\n`ensureRevisionFetched` so `fetchAll`, `updateOrCreateCache`, and\n`checkoutHeadRevision` verify that the pinned revision is present after\nfetching, falling back to a direct revision fetch and tag fetch, and\nretrying with backoff if the revision is still unavailable.\n\nAlso include `git show` stderr when `loadManifestFile` fails and check\nwhether the target revision itself is missing from the local repository\nso the error does not misleadingly claim the manifest file is missing.\n\nTAG\u003dagy\nCONV\u003d5a1829fb-6f87-4642-89b9-2a63a670e703\n\nChange-Id: Id84adab7d18e9f4b4efd27aebb8519ae5a389c7e\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1879877\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "7140d63954ca11d1200e22d41b4a1cb02eb4c47d",
      "tree": "f74ce96440be571f91c9c051577f59966863044b",
      "parents": [
        "c92e7404d727efb841636cca29a740677713d7a6"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Mon Oct 12 14:27:02 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 12 14:27:02 2026 -0700"
      },
      "message": "[shac] Add scripts/shac wrapper\n\nThe shac presubmit builder runs shac from a bare checkout without Go\ninstalled, so add a script that installs pinned versions of Go and shac\nfrom CIPD and runs shac with Go on $PATH. The builder will use it via\nthe shac recipe\u0027s `shac_path` property, and it also gives developers a\nway to run the checks with the same versions as presubmit.\n\nTAG\u003dagy\nCONV\u003d513dc628-3cd1-4e8c-9e75-f69cdda4a123\n\nBug: 572867624\nTest: scripts/shac check --all \u0026\u0026 scripts/shac test\nChange-Id: Ied9d645d3a5e8ebe07e5c56264ab6fe204fed1e0\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1882125\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "c92e7404d727efb841636cca29a740677713d7a6",
      "tree": "897c89f19ff13c40324b1c93b7f8f25f0b330db7",
      "parents": [
        "83583b7bc2b5d8658f6c3aff178214fabc82221e"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Mon Oct 12 14:23:52 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 12 14:23:52 2026 -0700"
      },
      "message": "[shac] Add gofmt, go vet and go fix checks\n\nAdd a shac config that runs gofmt (as a `shac fmt` formatter), go vet\nand go fix, so these are enforced consistently instead of relying on\nreviewers and AGENTS.md guidance.\n\ngo fix\u0027s suggested fixes often make several edits far apart in a file\n(e.g. adding an import and rewriting a loop), so they can\u0027t be attached\nto each diagnostic\u0027s span. Instead, the gofix check reports each\ndiagnostic without a replacement, plus one finding per file whose\nreplacement is the whole file with all of the file\u0027s fixes applied and\ngofmt run on the result. So `shac fix` gives the same result as\n`go fix ./...`.\n\nAlso point AGENTS.md at shac and drop the guidance it now enforces.\n\nTAG\u003dagy\nCONV\u003d513dc628-3cd1-4e8c-9e75-f69cdda4a123\n\nBug: 572867624\nTest: shac test \u0026\u0026 shac check --all\nChange-Id: I7e6775376b7d964c8ff8eeb9580fa5dd67e26fda\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1882124\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "83583b7bc2b5d8658f6c3aff178214fabc82221e",
      "tree": "6329e3ccadb4622b36d34fe6c86a3363a765401f",
      "parents": [
        "9961ac1c7dc01e34ab51de2dc9815987a701f0db"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Mon Oct 12 14:17:59 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 12 14:17:59 2026 -0700"
      },
      "message": "[jiri] Fix gofmt and go fix findings\n\nApply `go fix ./...` (Go 1.27) and `gofmt -s -w .` across the repo so\nthat the upcoming shac checks start out clean. Most of the changes\nreplace manual loops with range-over-int, slices.Contains,\nstrings.SplitSeq, maps.Copy, slices.Backward and sync.WaitGroup.Go,\nand remove obsolete `// +build` lines.\n\nTAG\u003dagy\nCONV\u003d513dc628-3cd1-4e8c-9e75-f69cdda4a123\n\nBug: 572867624\nTest: go test ./...\nChange-Id: I4ff3294b91e4239813d60125402529754d742d79\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1882123\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "9961ac1c7dc01e34ab51de2dc9815987a701f0db",
      "tree": "355ee6afbfee722728879221efe748fbfe09c3a0",
      "parents": [
        "06476e035662cbffa1d04d2dc3a451be7ac07b11"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Oct 12 13:00:49 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 12 13:00:49 2026 -0700"
      },
      "message": "[cache] Drop package-cache\u003dfalse support\n\nNow that the content-addressable package cache is always enabled across\nall checkouts and infrastructure builders, remove package-cache\u003dfalse\nsupport, MigratePackageCacheToStandalone, and fetchPackagesWithoutPackageCache.\n\nMark the -package-cache flag in jiri init as a deprecated no-op and\nremove PackageCacheEnabled (\u003cpackage_cache\u003e\u003cenabled\u003e) from jiri.Config\nso rewriting .jiri_root/config automatically purges the legacy XML field\nwhile preserving \u003cpackage_cache\u003e\u003cpath\u003e.\n\nTest: go test -count\u003d1 ./...\nBug: 570539173\nTAG: agy\nCONV: 5dbfcc4d-3c12-4fa5-b9da-068b05ba6706\nChange-Id: Iffc7ae3ddc45dd5c8416ad7a483f4290075b43fc\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1884579\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "06476e035662cbffa1d04d2dc3a451be7ac07b11",
      "tree": "0b8fdc406416b4367d3fde53d64ac99703b61645",
      "parents": [
        "a777d356c72af3b7e8e9ca3219f0694417923b2e"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Oct 12 13:00:28 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 12 13:00:28 2026 -0700"
      },
      "message": "[docs] Add AGENTS.md with repo architecture and Go guidelines\n\nDocument Jiri repository architecture, context.Context checkout\nplumbing, package layout, unit vs. integration testing conventions, and\nGo style and concurrency guidelines.\n\nTest: None, documentation change.\nChange-Id: I8cda4f2e0c4d7bb22c872c2453f7acce643031e6\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1879876\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "a777d356c72af3b7e8e9ca3219f0694417923b2e",
      "tree": "f8439217784645d9f8befaed032960f3439d07ae",
      "parents": [
        "bd59831323447484996072b8f24f77c63a09c573"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Oct 12 12:10:34 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 12 12:10:34 2026 -0700"
      },
      "message": "[tests] Consolidate FakeJiriRoot tests\n\nTests using jiritest.NewFakeJiriRoot paid the full I/O and subprocess\ncost of multi-repo Git integration tests while bypassing jiri init,\nCLI flag parsing, config loading, and .jiri_root discovery, and\nduplicating coverage across project, subcommands, and integrationtests.\n\nConsolidate multi-repo and CLI subcommand tests into integrationtests/\nwith t.Parallel() enabled across all tests, retain only fast unit tests\nin package-level test files, and remove jiritest.NewFakeJiriRoot. Add\ncmdline.Env.WorkingDir and shared CIPD binary pre-linking in\nintegrationtests so parallel CLI tests can execute in arbitrary working\ndirectories without t.Chdir or redundant network downloads.\n\nBug: 572041515\nTest: go test -count\u003d1 ./...\nChange-Id: I56efe7cc68905e5d7a9573a788fbf0e86d0b699b\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1881551\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "bd59831323447484996072b8f24f77c63a09c573",
      "tree": "0df719ef0b11b2353982d7404a07245ff2cbb335",
      "parents": [
        "c5f77021faabf7a7a0671621367bc69cf5574b15"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 09 15:21:45 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 09 15:21:45 2026 -0700"
      },
      "message": "[worktree] Unwrap parent root in worktree commands\n\nWhen jiri worktree commands run with cwd inside an existing worktree\nwithout -root, FindRoot resolves checkout.Root to the worktree and\nrecords the main checkout in checkout.WorktreeParentRoot. Previously,\nWorktreeAdd, WorktreeRemove, and WorktreePrune operated directly on\nthe worktree checkout, configuring newly added worktrees with the\nsibling worktree as their upstream parent.\n\nResolve relative wtRoot paths against the caller\u0027s cwd and unwrap\nWorktreeParentRoot via getMainCheckoutContext so worktree lifecycle\noperations always target the main checkout.\n\nTest: go test ./integrationtests/... ./project/...\nBug: 572022489\nTAG: agy\nCONV: 0333355a-59f2-4d92-afe1-b22bed7e864a\nChange-Id: Idc5aaab813589487a7f7566c96b751fc0563d54e\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1878781\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Chris Masone \u003ccmasone@google.com\u003e\n"
    },
    {
      "commit": "c5f77021faabf7a7a0671621367bc69cf5574b15",
      "tree": "8c66b00247bc2388740f134d8b9762961f3af580",
      "parents": [
        "710003d26dcc33861fc99981384fb11db521761f"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Oct 09 13:46:28 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 09 13:46:28 2026 -0700"
      },
      "message": "[worktree] Use renameDir when relocating moved projects\n\nWhen a project path in the manifest moves into a subdirectory of its\nprevious path (e.g., third_party/markupsafe -\u003e\nthird_party/markupsafe/markupsafe) or back up to its parent directory,\ncalling os.Rename directly in relocateMovedWorktreeProjects fails with\nEINVAL or destination-already-exists.\n\nUse renameDir (which moves via .jiri_root/swap) and allow moving into a\nsubdirectory of the source or into a parent directory whose only child\nis the source, matching moveOperation in project/operations.go.\n\nTest: go test ./integrationtests -run TestWorktreeSyncSubdirPathChange\nBug: 571639170\nTAG: agy\nCONV: 86b00259-2a66-400c-b468-de19013c9fe8\nChange-Id: I0fc8f67ffeb7aac7f90bdd8773e9c3ced2e56352\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1877816\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "710003d26dcc33861fc99981384fb11db521761f",
      "tree": "6b59da0ce35c342bd918c7464dc917165c930de8",
      "parents": [
        "c1a9c1668a06bc487b0390c39a433362570ff957"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Oct 08 15:02:18 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Oct 08 15:02:18 2026 -0700"
      },
      "message": "[jiri] Speed up test suite execution\n\nReduce \u0027go test -count\u003d1 ./...\u0027 wall-clock time from ~27.5s to ~4.1s (and\ncached \u0027go test ./...\u0027 from ~16.5s to ~2.4s) by:\n- Enabling t.Parallel() across unit and integration tests in project,\n  cmd/jiri/subcommands, cipd, gitutil, and integrationtests.\n- Caching the bootstrapped CIPD client binary across tests and hardlinking\n  it into each test checkout instead of writing a 22MB binary per test.\n- Caching exec.LookPath lookups for \u0027git\u0027 and \u0027git-remote-sso\u0027 via\n  sync.OnceValue to avoid logging ~700k redundant PATH stat entries into\n  Go\u0027s testlog cache on every test run.\n- Setting a shared CIPD_CACHE_DIR and default GIT_AUTHOR_*/GIT_COMMITTER_*\n  variables in IsolateGitEnv(), and removing redundant git config calls in\n  CreateRemoteProject.\n- Batching remote manifest updates in test setupUniverse/createProjects\n  helpers into a single commit instead of committing once per project.\n- Reusing the read-only 6-project fixture across subtests in\n  TestStatusFlags and using xtest.NewContext(t) instead of\n  jiritest.NewFakeJiriRoot(t) where a git fixture is unnecessary.\n- Fixing jiriInit in integrationtests to pass finalArgs (including\n  -attempts 1) to NewCommander.\n\nTest: go test -count\u003d1 ./...\nChange-Id: I8434cd1e872c4accc895870ef9a10d0cb9d56a9b\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1878286\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "c1a9c1668a06bc487b0390c39a433362570ff957",
      "tree": "882761f91a5b0350f7266e12cf6eea54c1828b93",
      "parents": [
        "24357c75d1afdb29125d21d8b0f2d046fca8c788"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Oct 08 10:04:55 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Oct 08 10:04:55 2026 -0700"
      },
      "message": "[cache] Fix multi-package destination poisoning\n\nWhen migrating legacy standalone checkouts, adopting destinations\nshared by multiple packages (such as prebuilt toolchain host and\ntarget runtimes) hardlinks the merged destination into target\npackage instances. This contaminates target packages with foreign\nbinaries, clobbering updated host tools and masking failures with\na matching .jiri_cache_hash.\n\nExample failure scenario (Rust prebuilt):\nIn Fuchsia, `prebuilt/third_party/rust/linux-x64` is shared by 8\npackages: 1 host compiler (host/linux-amd64) and 7 target runtimes\n(e.g. target/wasm32-unknown-unknown). In standalone checkouts, all\npackages are deployed to the same merged destination.\n\n1. During standalone cache migration, `adoptCIPDPackages` iterated\n   over packages and hardlinked `actualSourceDir` into each package\u0027s\n   cache directory (`instances/\u003cid\u003e`). For target runtimes, this\n   copied the old compiler (`bin/rustc` 1.70.0) into the target\n   runtime\u0027s instance cache (`instances/\u003cwasm32_id\u003e/bin/rustc`).\n2. Later, when Rust rolled to 1.80.0, the new host compiler was\n   downloaded to `instances/\u003cnew_host_id\u003e`. Because the target\n   runtime did not roll, its poisoned instance was left untouched.\n3. During `WorkspaceDestination.Link`, instances were overlaid in\n   manifest order. First, `instances/\u003cnew_host_id\u003e` linked the new\n   `bin/rustc` (1.80.0). Then, `instances/\u003cwasm32_id\u003e` linked its\n   target libraries AND its contaminated `bin/rustc`, overwriting\n   `bin/rustc` back to 1.70.0.\n4. Jiri then wrote `.jiri_cache_hash` containing the new instance\n   IDs.\n\nThis left the repository in a poisoned state:\n- `jiri package` and `.jiri_cache_hash` reported rustc 1.80.0.\n- `bin/rustc` on disk ran the old rustc 1.70.0.\n- Because `.jiri_cache_hash` matched the new instance IDs, Jiri\n  treated the directory as up-to-date and never repaired it.\n\nFix this issue across the migration, central cache, and workspaces:\n1. Skip local adoption from actualSourceDir in adoptCIPDPackages\n   when a destination is shared across multiple packages, allowing\n   CIPD to download clean, isolated instances.\n2. In PackageCache.Sync, inspect existing instances against their\n   CIPD package manifest.json and purge any contaminated instance\n   directories and their .cipd/pkgs metadata slots.\n3. Bump CurrentCacheStampFormatVersion to 2. Stale v1 stamps fail\n   MatchesDestination, allowing canAttachInPlace to detect mismatched\n   binaries, purge and relink contaminated trees, and upgrade clean\n   destinations in-place without touching inodes.\n\nTest: go test -v -count\u003d1 ./project/...\nChange-Id: I32288d2be601adb35073f26f2756d41326426a17\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1874366\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "24357c75d1afdb29125d21d8b0f2d046fca8c788",
      "tree": "11785757c88514ffcf86ebdd7c097732a26278dd",
      "parents": [
        "57e6b78a685fe271e1da0344223bda43bd464a2b"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Wed Oct 07 08:39:21 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 07 08:39:21 2026 -0700"
      },
      "message": "[jiri] Replace simplemr in runp with sync primitives\n\nReplace the obsolete simplemr package in the runp subcommand with\nstandard Go concurrency (sync.WaitGroup, bounded worker pool, and\nchannels) and delete the simplemr package.\n\nPreserves context cancellation, exit-on-error, interactive mode,\nprefixes, and collated output.\n\nBug: 567111720\nTest: go test ./...\nChange-Id: I0aafb416f070c0a737861ca8831272159b261797\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870053\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "57e6b78a685fe271e1da0344223bda43bd464a2b",
      "tree": "3e74235b678728afb6b53512159461df1ba64133",
      "parents": [
        "83b612cba9711749217b46cc90c7b35dac370db8"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Wed Oct 07 08:13:59 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 07 08:13:59 2026 -0700"
      },
      "message": "[jiri] Delete obsolete Windows OS support and platform shims\n\nFuchsia and Jiri do not support or build for Windows. Remove Windows\nspecific implementations:\n- project/cache_lock_windows.go (using golang.org/x/sys/windows)\n- osutil/exec_windows.go (calling kernel32.dll / GetModuleFileNameW)\n- Consolidate project/cache_lock_unix.go into project/cache_lock.go\n- Remove runtime.GOOS \u003d\u003d \"windows\" handling in osutil/command.go\n\nBug: 570526895\nBug: 567111720\nTest: go test ./...\nChange-Id: I00f3db082701c86086cbd3d12ab8978c8514753f\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870052\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "83b612cba9711749217b46cc90c7b35dac370db8",
      "tree": "cbd29f5c25176bc5e594945fb4107d87b6db0fbb",
      "parents": [
        "4ca9cce9b15d6682f103f6f9592fa6bc379efd96"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Wed Oct 07 08:12:57 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 07 08:12:57 2026 -0700"
      },
      "message": "[jiri] Delete obsolete Google Universal Analytics telemetry\n\nUniversal Analytics has been shut down and all telemetry collection\ncalls to UA-101128147-1 fail. Remove analytics_util, analytics session\nlifecycle hooks, update command timing, and the interactive opt-in\nprompt on checkout initialization. Retain -analytics-opt and\n-show-analytics-data flags as deprecated no-ops for backward\ncompatibility with existing scripts and tests.\n\nBug: 567111720\nTest: go test ./...\nChange-Id: Ibd5c1910477c360d9073ac5ac96b5172a55a8a72\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870051\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "4ca9cce9b15d6682f103f6f9592fa6bc379efd96",
      "tree": "21b0b003919a5733635009bad81878a9e2c6ba0e",
      "parents": [
        "261be9e4177c67f23c09e528b719210f50944c00"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Wed Oct 07 08:11:55 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 07 08:11:55 2026 -0700"
      },
      "message": "[jiri] Delete unused branch subcommand\n\nThe branch subcommand listed or deleted branches across multiple\nrepositories, including parsing submitted Gerrit Change-IDs.\nIt is unused across Fuchsia and infra.\n\nBug: 567111720\nTest: go test ./...\nChange-Id: I9443c0d695338453110c1346d3f0396a8aa14b43\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870050\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "261be9e4177c67f23c09e528b719210f50944c00",
      "tree": "298900b6385f98c938cacdfe7f2d134de4792a78",
      "parents": [
        "f98d205cfab5cc7e33a8e39aec3d967f9f6baa75"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Wed Oct 07 08:06:42 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Oct 07 08:06:42 2026 -0700"
      },
      "message": "[jiri] Delete unused grep subcommand\n\nThe grep subcommand ran git grep across multiple repositories. In\npractice, ripgrep (rg) across the source tree is standard and far more\nefficient, making jiri grep obsolete and unused.\n\nBug: 567111720\nTest: go test ./...\nChange-Id: Iad141d11bd93505885dddc3775420817dabb5d10\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870049\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "f98d205cfab5cc7e33a8e39aec3d967f9f6baa75",
      "tree": "d0bfe6656a38e0a177675bb68d1d2d73e85786a8",
      "parents": [
        "e7c50f007a45d5001659068009a7950b29cd545a"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Tue Oct 06 09:02:01 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 09:02:01 2026 -0700"
      },
      "message": "[jiri] Delete unused generate-gitmodules subcommand\n\nThe generate-gitmodules subcommand generated .gitmodules and\n.gitattributes files from a Jiri project manifest. It was unused\nacross Fuchsia and infra. Removing it also allows removing the\nassociated submodule prefix-tree pruning logic in project/project.go.\n\nBug: 567111720\nTest: go test ./...\nChange-Id: I208b68b849fa0f9e916622f1b6cd8e7dea5171ab\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870048\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "e7c50f007a45d5001659068009a7950b29cd545a",
      "tree": "0279925d1ac500356158d07d96caac67e699a7d5",
      "parents": [
        "c12b2803422f21e49caa0609c7116a0175e5b427"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Tue Oct 06 08:59:28 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 08:59:28 2026 -0700"
      },
      "message": "[jiri] Delete unused diff subcommand\n\nThe diff subcommand computed differences between two snapshot files\nand queried Gerrit for changes. It is unused across Fuchsia and infra.\n\nBug: 567111720\nTest: go test ./...\nChange-Id: I75d34cce5ba64d52b9c27d5b0fbab8c1c4fbb582\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1870047\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "c12b2803422f21e49caa0609c7116a0175e5b427",
      "tree": "516ce26c560c247ddddc0b8ae1c8a9f91c907787",
      "parents": [
        "6f163947cfa7bea91f6dcbdc942bb66b5b1caf08"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 08:37:15 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 08:37:15 2026 -0700"
      },
      "message": "[cmdline] Cancel the root ctx on SIGINT and SIGTERM\n\nAn interrupt used to kill jiri without running its cleanups, and a\nSIGTERM sent only to jiri left its git children running. Now the\nfirst SIGINT or SIGTERM cancels the root ctx, which interrupts\nsubprocesses and stops the project loops from starting new work.\nAfter cleanups run, jiri re-sends the signal to itself so that shells\nsee it was killed by the signal. It doesn\u0027t print the command\u0027s\nerror in that case, since that\u0027s mostly canceled subprocesses. A\nsecond signal kills jiri immediately.\n\nCleanup git calls (restoring branches and remotes, aborting a rebase\nor cherry-pick) use context.WithoutCancel. Code that falls back to\nsomething else when git fails, like wiping a corrupt cache or\nretrying update with a full scan, returns the error instead when ctx\nis canceled.\n\nrunp interrupts its commands through ctx instead of SIGKILL, and\n-exit-on-error cancels the others the same way. simplemr waits for\nrunning mappers when canceled, and Cancel works before Run.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: Iac4560c786c37144bf8ff8743317b391e6667b5d\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867433\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "6f163947cfa7bea91f6dcbdc942bb66b5b1caf08",
      "tree": "594026b288d97a48496aa59543c0d3e9ec361a10",
      "parents": [
        "3adfcf07ec9b53e922e2ceff638781c958ad5575"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:49:49 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:49:49 2026 -0700"
      },
      "message": "[gitutil,gerrit,cipd] Interrupt subprocesses on cancel\n\nEvery gitutil.Git method that runs git takes a ctx, and git and cipd\nrun through a new osutil.CommandContext that sends SIGINT when ctx is\ndone instead of SIGKILL. git cleans up lock files like\n.git/index.lock when interrupted, but leaves them behind when killed.\n\nIt doesn\u0027t set WaitDelay by default, because that makes Wait fail\nwhen a grandchild such as an ssh control master holds the output\npipes open. Hooks and cipd commands that run under a timeout do set\nit, so they\u0027re still killed if they ignore the interrupt. If an\ninterrupted command exits 0, it\u0027s still treated as failed.\n\nThe root ctx is still never canceled, so the only change in behavior\nis that timed-out hooks and cipd commands get 10s to exit before\nbeing killed.\n\nGit now copies only OffloadPackfiles from the Checkout. Committer and\nthe interactive Git methods had no callers and are deleted.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: I365d8c97e75148810f9db96eb79604ef245c3de7\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867432\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "3adfcf07ec9b53e922e2ceff638781c958ad5575",
      "tree": "ae67d65b3caae0dc2122a8aa226f6be5f74d7885",
      "parents": [
        "2de791f0c1994fd02bd939248d4fceb4fa191c44"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:48:44 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:48:44 2026 -0700"
      },
      "message": "[jiri] Rename X to Checkout\n\nWhat\u0027s left of X describes the checkout a command operates on, so\nrename it to Checkout (and x.go to checkout.go, WithX to\nWithCheckout, FromContext to CheckoutFromContext, jirix to checkout).\n\nAlso stop changing Checkout fields in the middle of a command.\nCheckoutSnapshot sets UsingSnapshot on a copy attached to its own\nctx, and UsingImportOverride becomes a field of the manifest loader,\nits only reader.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: Ib547f477baf0908c90468377d50d4b3f45716033\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867431\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "2de791f0c1994fd02bd939248d4fceb4fa191c44",
      "tree": "ce722136d12cc82b0d59a30bb9b710ea8857efd1",
      "parents": [
        "dae32862762e7f3d2b6de7952480903c709738a9"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:47:26 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:47:26 2026 -0700"
      },
      "message": "[jiri] Move stdio, env, timer and run state to ctx\n\nX copied stdio, env vars, the timer and the usage func out of the\ncmdline.Env that\u0027s already in every command\u0027s ctx. Read them from the\nEnv instead, via jiri.Stdout(ctx), jiri.EnvVars(ctx), etc.\njiri.WithStdout/WithStderr replace the remaining uses of X.Clone,\nwhich is deleted.\n\nFailures, cleanups, the override warning flag and the analytics\nsession belong to a command invocation, not a checkout, so they move\nto a run state attached to ctx by WithRunState.\n\nSome tests relied on X\u0027s fmt.Errorf fallback for usage errors.\ncmdline.Env.UsageErrorf prints to stderr and returns ErrUsage, which\nis what production already did, so those tests now check stderr.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: I9c5e9529cdd39eb0a843df7806e2d6d42684bc56\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867430\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "dae32862762e7f3d2b6de7952480903c709738a9",
      "tree": "2f447227cf0377dc177d04f99b72ef5e84aac444",
      "parents": [
        "f31b0166649817d579dac69e87cf540af6b52c57"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:37:07 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:37:07 2026 -0700"
      },
      "message": "[jiri] Read the logger and color from ctx\n\nAdd log.WithLogger/FromContext, color.WithColor/FromContext and\nleveled helpers like log.Debugf(ctx, ...), and remove X.Logger and\nX.Color. jiri.NewContext replaces NewXFromContext and returns a ctx\nwith the logger, color and X attached.\n\ndiff.go silenced logging by swapping the shared X\u0027s logger and\nrestoring it afterwards. It now uses a derived ctx with a quiet\nlogger instead.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: I7ec578be848dee183803a51b3816945a46a2b4be\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867429\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "f31b0166649817d579dac69e87cf540af6b52c57",
      "tree": "0cd340add96994b17f8b9669fb11bed0d5953503",
      "parents": [
        "49541620e7e3ba5e407e04b75fa18af5f026b033"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:31:45 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:31:45 2026 -0700"
      },
      "message": "[jiri] Clean up after the ctx conversion\n\nGerrit and PackageCache methods take ctx instead of holding a\n*jiri.X, and calls that re-attached the X already in ctx pass ctx\nthrough. GetProjectStates and runp stop cloning the X per goroutine,\nwhich dates from when tool.Context held a non-thread-safe runner.\n\nCommand timeouts and HTTP requests derive from the caller\u0027s ctx\nrather than context.Background(), and retry stops early when ctx is\ndone. This doesn\u0027t change behavior yet since the root ctx is never\ncanceled.\n\nFakeJiriRoot gets a Context method to replace hundreds of copies of\njiri.WithX(t.Context(), fake.X) in tests.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: I86e630f775f60dd6e829a67ab25a796b5d123857\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867428\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "49541620e7e3ba5e407e04b75fa18af5f026b033",
      "tree": "433a44621df9cf6bd4ba1f3cbe7f1a22fc9b7e66",
      "parents": [
        "2c170c3018de66c3d64ce63415ebd83712d77aa0"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:26:24 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:26:24 2026 -0700"
      },
      "message": "[project,cmd] Take ctx instead of *jiri.X\n\nDo the same for the project package, the subcommands and the test\nhelpers. Subcommands use the ctx from google/subcommands, with the X\nattached by executeWrapper.\n\nCode that switches to a different X (e.g. the parent checkout in\n`jiri update` from a worktree) attaches it to ctx with jiri.WithX.\nrelocateMovedWorktreeProjects still takes the parent X explicitly\nsince it needs both.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: I3c549e04b18ef746fd6180b11135bc793f571cbf\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867427\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "2c170c3018de66c3d64ce63415ebd83712d77aa0",
      "tree": "1aa9df644d46df2050bd232f829663762b06c5cb",
      "parents": [
        "42fdf6428b337f5981424b854cd28ac607adc559"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:21:23 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:21:23 2026 -0700"
      },
      "message": "[gitutil,cipd,gerrit] Take ctx instead of *jiri.X\n\nAdd jiri.WithX and jiri.FromContext, and change retry, gitutil, cipd,\ngerrit and selfupdate to take a ctx carrying the X instead of a\n*jiri.X. Callers that don\u0027t have a ctx yet pass\njiri.WithX(context.TODO(), jirix) for now.\n\nMost of this was generated with a go/types-based rewriter. The local\ntimeout contexts in cipd and hook running are renamed to cmdCtx so\nthey don\u0027t shadow ctx.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: Iffa768c7b1511f128c9d6308567a8121bfbc3c6d\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867426\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "42fdf6428b337f5981424b854cd28ac607adc559",
      "tree": "0895c06806b68c300ea3968fb35da2c1b9969876",
      "parents": [
        "6dad33a9bb3751f592b9a7bb9d246ef671adf7c0"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Tue Oct 06 07:04:43 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Oct 06 07:04:43 2026 -0700"
      },
      "message": "[jiri] Delete tool.Context\n\ntool.Context is a custom context type left over from Vanadium, and it\ngets in the way of passing a standard context.Context through jiri.\nMove its fields (env, stdio, timer) onto jiri.X, keeping accessor\nmethods with the same names so callers don\u0027t change.\n\nX.Clone copied fields by hand and had fallen out of date, silently\nzeroing about 20 of them (config, OffloadPackfiles, LockfileEnabled,\netc.). Make it a shallow copy, and keep failures and cleanups in a\nshared runState so they aren\u0027t lost on copies of X. Previously a\ncleanup registered on the parent-checkout copy in `jiri update` from\na worktree never ran.\n\nTAG\u003dagy\nCONV\u003d170938af-04f9-4e19-98ca-89bea64ceebb\n\nBug: 570045154\nChange-Id: I7ded221e0a2837a7bad057462f79f3ffdfa74e8b\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1867425\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "6dad33a9bb3751f592b9a7bb9d246ef671adf7c0",
      "tree": "fde8c0e5689e3310aba3d810697f565592040568",
      "parents": [
        "c5a5e3024e199bb95da23b63e926f301c84eb8e5"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Oct 05 09:53:35 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Oct 05 09:53:35 2026 -0700"
      },
      "message": "[cache] Default -package-cache to true\n\nUpdate jiri so that the package cache is enabled by default. If users\nhave not explicitly configured package_cache in their jiri config,\nNewX defaults PackageCacheEnabled to true.\n\nOnly when users explicitly configure -package-cache\u003dfalse is it\ndisabled. In jiri.Config, PackageCacheEnabled is changed from bool to\n*bool with omitempty to distinguish unset (nil) from explicitly disabled\n(\u0026false).\n\nTest: go test -v ./...\nTest: go test -v -run TestInitPackageCache ./cmd/jiri/subcommands\nTest: go test -v -run TestPackageCacheToggle ./integrationtests\nChange-Id: I184d7e7acb5d822da0e019f9828c67d5b580acd9\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1866805\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "c5a5e3024e199bb95da23b63e926f301c84eb8e5",
      "tree": "94101f46754e5980c06aa5d63cfeb57628e32926",
      "parents": [
        "889b278392a8c18766fa2d29c4576a8b34ad306f"
      ],
      "author": {
        "name": "eakammer",
        "email": "eakammer@google.com",
        "time": "Fri Oct 02 09:45:04 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Oct 02 09:45:04 2026 -0700"
      },
      "message": "[selfupdate] Improve HTTP error reporting and body truncation\n\nCapture the response status code and truncated error body when\nCIPD binary downloads fail, improving diagnosability.\n\nBug: 567589427\nChange-Id: Iaf5e4e5a1159c1b50f6be222a6d80d096a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1857018\nCommit-Queue: Liz Kammer \u003ceakammer@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "889b278392a8c18766fa2d29c4576a8b34ad306f",
      "tree": "7050e8cb02e8d8caadebb17408e858f3229311f0",
      "parents": [
        "09054ad1e3e7078dc8f25f28206c9eac32ebbd4a"
      ],
      "author": {
        "name": "eakammer",
        "email": "eakammer@google.com",
        "time": "Wed Sep 30 12:31:47 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Sep 30 12:31:47 2026 -0700"
      },
      "message": "[selfupdate] Use git ls-remote to check latest commit\n\nSwitch from unauthenticated HTTP Gitiles queries to git ls-remote via\ngitutil.\n\nBug: 567589427\nChange-Id: I960baea6db78e2ed51dd1cb47c40da116a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1857017\nCommit-Queue: Liz Kammer \u003ceakammer@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "09054ad1e3e7078dc8f25f28206c9eac32ebbd4a",
      "tree": "32711b5acde6562738e6ee1fd35e73aeb60f0fd2",
      "parents": [
        "ae0ef7147de023f14d5b07d663087e3866223ce8"
      ],
      "author": {
        "name": "eakammer",
        "email": "eakammer@google.com",
        "time": "Wed Sep 30 12:12:33 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Sep 30 12:12:33 2026 -0700"
      },
      "message": "[selfupdate] Move self-update logic to selfupdate package\n\nThis is to support moving to rely on git ls-remote to use authenticated\nchecks for the latest commit.\n\nBug: 567589427\nChange-Id: I1cf2b606d956bee59064de6d960e1db06a6a6964\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1857016\nCommit-Queue: Liz Kammer \u003ceakammer@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "ae0ef7147de023f14d5b07d663087e3866223ce8",
      "tree": "87bb03dc1cd4dcc1308d8f18f294d01e0f6bb025",
      "parents": [
        "dc75b05842abae0eea8d380fe0ea27c27ffb7f89"
      ],
      "author": {
        "name": "Jerry Belton",
        "email": "jcecil@google.com",
        "time": "Wed Sep 30 09:01:46 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Sep 30 09:01:46 2026 -0700"
      },
      "message": "[cmd] Add `jiri fetch-package` no-op subcommand\n\nAdd the `jiri fetch-package \u003cpath\u003e...` subcommand as the initial entry\npoint for on-demand prebuilt fetching.\n\nWhen on-demand packages are disabled (the current default), the command\nvalidates that at least one path is provided and that all paths resolve\nunder `prebuilt/` (accepting relative `prebuilt/...`, `//prebuilt/...`,\nand absolute `\u003cjiri-root\u003e/prebuilt/...` paths), then exits 0 without\nloading manifests or lockfiles.\n\n`Execute` invokes `errToExitStatus` directly rather than\n`executeWrapper` so that hot-path invocations from `fx` and build\nactions avoid `jiri.NewX` analytics initialization (which would\notherwise emit stdout warnings on unconfigured checkouts or block on\nanalytics HTTP requests).\n\nTest: go test -v -run TestFetchPackage ./cmd/jiri/subcommands\nBug: 567114481\nTAG: agy\nChange-Id: I2336280b703d09847ecd760a10bdaf3e26e0c8c3\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1854398\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Anni Zhang \u003cannizhang@google.com\u003e\nCommit-Queue: Jerry Belton \u003cjcecil@google.com\u003e\n"
    },
    {
      "commit": "dc75b05842abae0eea8d380fe0ea27c27ffb7f89",
      "tree": "dff0bc334905d595f7d839e18fb4e52c246c1efc",
      "parents": [
        "8139b88144642a8a72e75fbf5a053b64eb712e13"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Tue Sep 29 15:33:38 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Sep 29 15:33:38 2026 -0700"
      },
      "message": "[cache] Fix multi-instance symlink adjustment\n\nWhen multiple CIPD package instances share a workspace destination (such\nas prebuilt/.build-id), WorkspaceDestination.Link and\nMigratePackageCacheToStandalone previously called\nfixSymlinksAfterRestoration in a loop for every instance.\n\nBecause fixSymlinksAfterRestoration resolves external relative symlinks\n(such as CIPD symlinks pointing into .jiri_root/packages/.cipd/pkgs)\nrelative to the cache instance directory rather than the destination\ndirectory, calling it repeatedly on the same destination prepended an\nextra .jiri_root/packages/ segment on each pass. Furthermore, returning\nerrBrokenSymlink inside the filepath.WalkDir callback aborted the walk\non the first broken symlink.\n\nCall fixSymlinksAfterRestoration once after linking all instances for a\ndestination, pass the PackageCache directly so the instance depth\ninvariant is handled in one place, and finish walking the destination\ndirectory before returning errBrokenSymlink.\n\nTest: go test -v -run \\\n    \"TestWorkspaceDestination_MultiInstanceCIPDSymlinks\" ./project\nBug: 567346876\nTAG: agy\nCONV: 2f7d84a7-177f-451b-a255-a7c51a271b46\nChange-Id: Icad856fb7f0411552c004d0f0a2cb0db61c898f9\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1855435\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "8139b88144642a8a72e75fbf5a053b64eb712e13",
      "tree": "9dba869efb9e714be2284f61b142dc85c8e4d415",
      "parents": [
        "ba95af9d10bd5cd47df6049ef83fc2ed76a26235"
      ],
      "author": {
        "name": "Chase Latta",
        "email": "chaselatta@google.com",
        "time": "Tue Sep 29 10:35:49 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Sep 29 10:35:49 2026 -0700"
      },
      "message": "[cache] Make PackageCacheDir configurable per checkout\n\nThis adds a `PackageCachePath` configuration parameter to\n`.jiri_root/config`. It enables users of external mounts (like CartFS)\nto configure the package cache path to reside in the same filesystem,\nbut outside the local checkout directory. This preserves native hardlink\nprojection (avoiding cross-device EXDEV penalties) without polluting the\nlocalized checkout state.\n\nIt also exposes a `-package-cache-path` flag into `jiri init` to\nspecify this during initial turn-up. Git worktrees continue\nreferencing their parent root by default.\n\nChange-Id: Ia580819b0fb2554c5cffdcbe42434788ac8fc976\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1846087\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Chase Latta \u003cchaselatta@google.com\u003e\n"
    },
    {
      "commit": "ba95af9d10bd5cd47df6049ef83fc2ed76a26235",
      "tree": "85a91cba66d2ea0ace841a6a8223c8af0da58ed6",
      "parents": [
        "f3533eb9a9abf653c6056a6cca4f2446014d34e2"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Sep 25 09:34:05 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Sep 25 09:34:05 2026 -0700"
      },
      "message": "[cache] Canonicalize relative symlinks in canAttachInPlace\n\nCIPD packages frequently contain relative symlinks (such as clang++\npointing to clang or shared libraries referencing relative paths).\nfixSymlinksAfterRestoration rewrites internal relative symlinks to their\ncanonical form (e.g. \"./clang\" becomes \"clang\") in destDir, so the\ndestination never matches the raw cache copy for such links.\n\nIf raw symlink text differs syntactically between the destination and\ncache instance, a direct string comparison fails, causing\ncanAttachInPlace to return false and triggering an unnecessary directory\nwipe and relink.\n\nWhen raw symlink targets differ, resolve both targets relative to their\nrespective tree roots (destDir and instDir) to verify whether they point\nto equivalent relative paths.\n\nTest: go test -v -run \"TestWorkspaceDestination_AttachInPlace_SymlinkCanonicalization\" ./project\nBug: 557237866\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: Ie75f58d2da622d1d1c2dbf3eda9791bf7d1e962e\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1843813\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "f3533eb9a9abf653c6056a6cca4f2446014d34e2",
      "tree": "ce9d72ea08d556cdea07043888f1b90a23e36513",
      "parents": [
        "f8e102fe2e0319e7f6d6f8b466e63e52591d5245"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Sep 25 09:32:45 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Sep 25 09:32:45 2026 -0700"
      },
      "message": "[cache] Ignore runtime artifacts and skip nested destinations\n\nInvoking prebuilt compilers and build tools generates runtime artifacts\n(__pycache__, *.pyc bytecode, and *.lock files like Bazel\u0027s\ninstall_base.lock) within prebuilt trees in the workspace checkout.\nFurthermore, certain packages install into subdirectories of other packages.\n\nWhen evaluating destinations in canAttachInPlace, encountering these\nruntime files or nested package destinations caused file count and entry\nmismatches, triggering full directory wipes and invalidating downstream\nbuild systems.\n\nVerify all pristine package entries present in the cache instances,\nwhile tolerating runtime-generated artifacts in the workspace destination\n(such as __pycache__, unbundled .pyc files, and install_base.lock) that\nare not part of the package itself. Skip subdirectories that belong to\nnested workspace destinations.\n\nTest: go test -v -run \"TestWorkspaceDestination_AttachInPlace.*\" ./project\nBug: 557237866\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: Ib434e930b65c6d232894992b3b9338793727a9f1\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1843812\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "f8e102fe2e0319e7f6d6f8b466e63e52591d5245",
      "tree": "4a3bb64476a4d2511aaa94de7bfd548d690afede",
      "parents": [
        "97fd03f6426a1076079dfe025d56327e5dd88657"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Fri Sep 25 09:30:21 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Sep 25 09:30:21 2026 -0700"
      },
      "message": "[cache] Attach matching package destinations in-place\n\nWhen a workspace destination directory lacks an active cache stamp,\nWorkspaceDestination.Link previously removed the entire directory via\nos.RemoveAll and relinked all files from scratch.\n\nEven if every file already matched the target cache instances, wiping\nthe directory created new directory and file inodes and updated the\ndirectory mtime. Downstream build tools (Ninja, Bazel) detect these\ninode and timestamp changes and trigger redundant rebuilds across the\nbuild graph.\n\nBefore wiping the destination, check whether the directory already\ncontains an exact match of target instance files and symlinks via\ncanAttachInPlace. If so, attach it in-place by writing the verification\nstamp, preserving existing timestamps and inodes.\n\nTest: go test -v -run \"TestWorkspaceDestination_AttachInPlace.*\" ./project\nBug: 557237866\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: I26fd2a463d2010d62e2db101d410caf745ecb31e\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1843811\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "97fd03f6426a1076079dfe025d56327e5dd88657",
      "tree": "ff37ab041ae4939143720893535e4077b1c5ff42",
      "parents": [
        "9891682befa98428599d22fbaa5ad8500ad58760"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Sep 24 11:00:05 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Sep 24 11:00:05 2026 -0700"
      },
      "message": "[cache] Sort workspace destinations deterministically\n\nIn collectWorkspaceDestinations, destination groups were iterated\ndirectly over Go\u0027s groups map. Map iteration in Go has randomized\norder.\n\nFor nested package destinations (where one package installs inside a\nsubdirectory of another package, e.g. foo/subtools inside foo),\nprocessing the child before the parent could cause parent linking\nor wipe operations to clobber the child directory.\n\nSort destination paths lexicographically with slices.Sort, guaranteeing\nthat parent destinations are always evaluated and linked before nested\nchild destinations.\n\nTest: go test -v -run \"TestCollectWorkspaceDestinations_SortedPaths\" ./project\nBug: 557237866\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: I9303bd80ab7e431b747552fc3e750ceaca192b3e\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1843810\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "9891682befa98428599d22fbaa5ad8500ad58760",
      "tree": "73da6117e9f71730bb6efea2ccfb50d8febb735a",
      "parents": [
        "cea73d12beacdb03616f50c39f31a3b1b571b438"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Sep 24 10:58:36 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Sep 24 10:58:36 2026 -0700"
      },
      "message": "[cache] Make cache stamp instance matching order-independent\n\nWhen multiple CIPD package instances share a workspace destination,\ntheir order in memory depends on manifest inclusion order or grouping.\n\nPreviously, CacheStamp.MatchesDestination compared instances strictly\nby slice index (data.Instances[i].InstanceID !\u003d inst.InstanceID). If\nthe slice ordering differed between the on-disk stamp and the memory\ndestination, Jiri concluded the directory was out-of-date, triggering\nneedless directory wipes and rebuilds.\n\nCompare instances using multiset frequency counts rather than\npositional indices, verifying exact cardinality without ordering\ndependencies.\n\nTest: go test -v -run \"TestCacheStamp_MatchesDestination_OrderIndependent\" ./project\nBug: 557237866\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: I2e00f020997d73f65f70a2e345d361daaba5beb4\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1843809\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "cea73d12beacdb03616f50c39f31a3b1b571b438",
      "tree": "cb6825e2be85d96170143a52c75860e2e6fe3030",
      "parents": [
        "a6c9452273b0e9ce7e7ea19ce2e6aa579b37cc05"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Sep 24 10:57:26 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Sep 24 10:57:26 2026 -0700"
      },
      "message": "[cache] Preserve directory mtime on cache stamp write and remove\n\nAdding or deleting the .jiri_cache_hash stamp file inside a package\ndestination directory modifies the destination directory modification\ntime (mtime).\n\nDownstream build systems (such as GN and Ninja) frequently declare\nprebuilt directories as direct inputs (e.g. prebuilt_python_dir).\nModifying the directory mtime causes Ninja to treat dependent generator\nrules and targets as dirty, triggering unnecessary rebuilds during\nno-op syncs.\n\nCapture the directory mtime prior to modifying .jiri_cache_hash and\nrestore it via os.Chtimes in CacheStamp.Write and CacheStamp.Remove.\n\nTest: go test -v -run \"TestCacheStamp_PreservesDirectoryMtime\" ./project\nBug: 557237866\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: I7c22206b0adc70c18041b9c54fb34c7d432fe2ef\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1843808\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "a6c9452273b0e9ce7e7ea19ce2e6aa579b37cc05",
      "tree": "ce8b6f96524605fddcf0705ffee328bd62a8146d",
      "parents": [
        "7b382d7de3a55eec4aba82d2c9caffaeaa40fc9f"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Sep 24 10:56:11 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Sep 24 10:56:11 2026 -0700"
      },
      "message": "[project] Make flag, metadata, and history writes idempotent\n\nDuring jiri update and sync invocations, Jiri unconditionally wrote\npackage flag files, project flag files, prebuilt JSON metadata, and\nthe update_history/latest hardlink even when their contents had not\nchanged.\n\nDownstream build tools (GN and Ninja) frequently declare these files\nas inputs. Overwriting them bumped their modification timestamps,\ntriggering redundant build graph re-evaluations and recompilations\nduring no-op syncs.\n\nCheck whether existing files have byte-identical contents before\nwriting or relinking, preserving timestamps on no-op syncs.\n\nTest: go test -v -run \"TestSafeWriteFile_Idempotent|TestWriteUpdateHistorySnapshot_Idempotent|TestWritePackageFlags|TestWriteProjectFlags\" ./project\nBug: 557237866\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: If87653fce6847ab6e378a5fd1b41ca443acbfdc4\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1843807\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "7b382d7de3a55eec4aba82d2c9caffaeaa40fc9f",
      "tree": "3d9d202fb7600318a05b99c015a04e201de00530",
      "parents": [
        "58fff65a1daa0b094bd5dd9f948c2ac0e5aae136"
      ],
      "author": {
        "name": "Adam Barth",
        "email": "abarth@google.com",
        "time": "Wed Sep 23 11:06:23 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Sep 23 11:06:23 2026 -0700"
      },
      "message": "[cache] Handle raw CIPD instance IDs in manifests\n\nWhen a manifest \u003cpackage\u003e entry specifies a raw CIPD instance ID in its\nversion attribute instead of a tag or ref, it is not recorded in\njiri.lock and therefore has len(pkg.Instances) \u003d\u003d 0 after manifest\nloading.\n\nPreviously, collectWorkspaceDestinations treated any package with\nlen(pkg.Instances) \u003d\u003d 0 as an unpinned package and invoked\ncipd ensure-file-resolve. Because cipd ensure-file-resolve omits\npackages whose version is already a raw instance ID from the generated\n.version file, collectWorkspaceDestinations constructed a\nWorkspaceDestination with 0 instances, wiping the destination directory\nand writing a .jiri_cache_hash stamp with no instances.\n\nAdditionally, CacheStamp.MatchesDestination returned true when comparing\nan empty .jiri_cache_hash (0 instances) against a WorkspaceDestination\nwith 0 instances, causing MigratePackageCacheToStandalone to detach the\nempty directory in-place instead of queuing the package for standalone\nCIPD extraction on the first pass.\n\nFix both issues by:\n- Adding cipd.IsInstanceID to detect raw CIPD instance IDs and populating\n  PackageCacheInstance directly in collectWorkspaceDestinations without\n  spawning cipd ensure-file-resolve.\n- Skipping destinations with 0 instances and rejecting empty instance\n  lists in CacheStamp.MatchesDestination and WorkspaceDestination.\n\nFixed: 565366056\nTest: go test ./cipd/... ./project/...\nChange-Id: I713900378d2d00d1e3ec70a2448d2c14a14c5dcb\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1842245\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "58fff65a1daa0b094bd5dd9f948c2ac0e5aae136",
      "tree": "7351aa335605a5476e1304919abcc59969f05869",
      "parents": [
        "014e6d57cc235ba13d7c529d4e7dce50e04745e1"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Tue Sep 22 12:58:49 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Sep 22 12:58:49 2026 -0700"
      },
      "message": "[cache] Implement instance-based CIPD package cache\n\nReplace the legacy package-group cache architecture with a canonical\ninstance-level CIPD cache and workspace hardlink forest pipeline.\n\nPipeline stages:\n- Stage 1 (Legacy Adoption \u0026 Immediate Retirement): Adopts legacy\n  package-group cache entries without network downloads by hardlinking\n  from .cipd/pkgs/\u003cidx\u003e/\u003cid\u003e into instances/\u003cid\u003e and replicating CIPD\n  metadata. Retires legacy group directories (.jiri_root/packages/\u003chex64\u003e)\n  immediately upon adoption under .cache.lock.\n- Stage 2 (Cache Synchronization): Collects unpinned and pinned\n  instances across the main checkout and all active worktrees into a\n  single master ensure file. Invoking \u0027cipd ensure\u0027 validates,\n  downloads, and natively prunes unreferenced instances under\n  instances/\u003cid\u003e.\n- Stage 3 (Workspace Forest Projection): Constructs clean hardlink\n  forests in workspace directories from target cache instances.\n  Resolves multi-instance collisions in manifest order and writes\n  .jiri_cache_hash as an atomic stamp.\n- Concurrency: The entire pipeline (Stages 1, 2, and 3) is covered by\n  a single fail-fast lock (.jiri_root/packages/.cache.lock), preventing\n  cross-worktree pruning/linking races and workspace collisions.\n- Standalone Compatibility: Workspaces with package-cache\u003dfalse reuse\n  cached instances without downloading, detaching .jiri_cache_hash in\n  O(1) time.\n- Pruning: Repurposes PackageGC to clean obsolete prebuilt directories,\n  removing findUsedCachePaths and relying on native CIPD pruning during sync.\n\nBug: 557237866\nTest: go test -v -count\u003d1 -run \"^TestPackageCache\" ./project\nTest: go test ./...\nTAG\u003dagy\nCONV\u003d33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: I35ad939424f06c70c8abea86892a3eeec44797f7\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1812230\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "014e6d57cc235ba13d7c529d4e7dce50e04745e1",
      "tree": "a0d518c90d775c8d289d9148666aa850b9edc07d",
      "parents": [
        "dd9588a9065a9f089b889739cf111cbf928510c8"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Thu Sep 17 10:14:47 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Sep 17 10:14:47 2026 -0700"
      },
      "message": "[cipd] Make \"Fetching CIPD packages\" non-persistent\n\nNow that CIPD has decent progress output it\u0027s not helpful for Jiri to\noutput a semi-persistent \"Fetching CIPD packages\" log that keeps getting\nre-printed intermittently between the logs from the CIPD client.\n\nChange-Id: I9ded16d8d91036ea9b023c611823d6148948cc1c\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1832369\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "dd9588a9065a9f089b889739cf111cbf928510c8",
      "tree": "66ee17307c77ec2ffb6e0a3c85368ba41a8a1aa9",
      "parents": [
        "cc22b8a6381c4a8d33e70ef5d5cf3bd0f45f1cc4"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Sep 14 15:00:26 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Sep 14 15:00:26 2026 -0700"
      },
      "message": "[cache] Preserve hardlinks on package cache opt-out\n\nPreviously, when opting out of the package cache (package-cache\u003dfalse)\nor restoring a cached package into a cache-disabled workspace,\nrestorePackageFromCache deleted destDir and called osutil.CopyPath.\nFor large checkouts (such as Fuchsia\u0027s ~30 GB of CIPD prebuilts across\n206 packages), this resulted in nearly 5 minutes of disk copying.\n\nBecause CIPD packages are marked read-only and CIPD package rolls\natomically replace files using temporary staging directories,\npreserving hardlinks does not risk cache mutation or corruption.\nFurthermore, in POSIX filesystems, when PackageGC prunes unreferenced\nentries from the cache, unlinking merely decrements the inode link count\nwithout affecting workspace files.\n\nUpdate restorePackageFromCache to:\n- Preserve existing hardlinks in-place and simply remove\n  .jiri_cache_hash if destDir already matches the expected cache\n  version.\n- Use osutil.LinkPath instead of osutil.CopyPath when restoring into\n  a missing destDir, falling back to CopyPath only if hardlinking\n  fails (e.g. cross-device).\n- Update tests S14 and S15 to verify that hardlinks are preserved and\n  restored without full disk copying.\n\nTiming benchmark on Fuchsia (206 packages, ~30 GB):\n- Migrate INTO cache (package-cache\u003dtrue): ~5.0s (Rename + LinkPath)\n- Migrate OUT of cache (package-cache\u003dfalse):\n  - Before (CopyPath): 4m 46.7s\n  - After (Hardlinks): 2.7s (\u003e105x faster)\n- Warm-cache no-op fetch: 2.2s (fast-path hash check)\n\nBug: 557237866\nTest: go test -v -run \"TestPackageCacheState.*\" ./project\nTest: go test ./...\nTAG\u003dagy\nCONV\u003d33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: I4ff19a77d6c41dae062b75813efdd1d6d3fc366b\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1812210\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "cc22b8a6381c4a8d33e70ef5d5cf3bd0f45f1cc4",
      "tree": "a5ef9013d9ae98afec09e466c300fd73a20c3012",
      "parents": [
        "cc3b5a6ac33e0b0fb2439001195474e096baead3"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Sep 14 14:58:00 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Sep 14 14:58:00 2026 -0700"
      },
      "message": "[cache] Fix package rolls, worktree GC, and prebuilt cleanup\n\nPreviously, several issues affected Jiri package caching:\n\n1. Package rolls and cache migration: When rolling a package while\npackage caching was enabled, Jiri renamed the destination directory\n(containing old binaries) into the cache directory for the new\nversion. In addition, turning off package caching failed to detach\nhardlinks and triggered redundant remote CIPD downloads.\n\n2. Worktree GC preservation: findUsedCachePaths only inspected\nsymlinks and ignored hardlinks (directories with .jiri_cache_hash).\nAs a result, PackageGC deleted cache entries in .jiri_root/packages\nthat were still actively required by worktrees.\n\n3. Obsolete prebuilt cleanup: Jiri never removed obsolete hardlinks,\nsymlinks, or old packages from prebuilt/ when packages rolled or\nwere removed from the manifest.\n\nThis change resolves these issues:\n- Refrains from migrating stale hardlinks or dirty directories into\n  the cache, validating standalone destinations before migration.\n- Properly detaches hardlinks and removes .jiri_cache_hash when\n  package caching is disabled.\n- Partitions packages in fetchPackagesWithoutPackageCache to skip CIPD\n  downloads when packages can be restored from cache.\n- Updates findUsedCachePaths to detect .jiri_cache_hash so worktree\n  dependencies in the cache are preserved during PackageGC.\n- Implements garbageCollectObsoletePrebuilt to clean up obsolete\n  hardlinks and symlinks in prebuilt/ during package updates and GC.\n- Adds comprehensive package cache documentation in\n  docs/package_cache.md and a consolidated test suite in\n  project/package_cache_test.go.\n\nTest: go test -v ./project -run \"TestPackageCache.*\"\nTest: go test ./...\nTAG: agy\nCONV: 33255e46-db78-4991-9df6-59598d4eb4e5\nChange-Id: I09bb2e12f6aab057cde20afa7ac82187eb459c68\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1802200\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "cc3b5a6ac33e0b0fb2439001195474e096baead3",
      "tree": "1d24554b35448a6a86dd1b1d85e71cf802a71347",
      "parents": [
        "c68eb341c2fcb2a3d46b61d29c7960ed2736b681"
      ],
      "author": {
        "name": "Oliver Newman",
        "email": "olivernewman@google.com",
        "time": "Thu Sep 10 11:59:42 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Sep 10 11:59:42 2026 -0700"
      },
      "message": "[worktree] Make package GC logs quieter\n\nThis log is noisy and unreadable since the names are all hashes rather\nthan actual package names, so lower it to debug level so it\u0027s hidden by\ndefault.\n\nChange-Id: I1da990bfa71ccd6b43e330ab0432c7e1aebe3e11\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1813296\nFuchsia-Auto-Submit: Oliver Newman \u003colivernewman@google.com\u003e\nReviewed-by: Ina Huh \u003cihuh@google.com\u003e\nCommit-Queue: Oliver Newman \u003colivernewman@google.com\u003e\nPresubmit-Verified: fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com \u003cfuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com\u003e\n"
    },
    {
      "commit": "c68eb341c2fcb2a3d46b61d29c7960ed2736b681",
      "tree": "805061584289d10721ef9660b90dc7d9f4a0f4f9",
      "parents": [
        "873d88fefc271bc223b1beeca4967d002ab23408"
      ],
      "author": {
        "name": "global-integration-roller",
        "email": "global-integration-roller@fuchsia-infra.iam.gserviceaccount.com",
        "time": "Tue Sep 01 10:10:23 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Sep 01 10:10:23 2026 -0700"
      },
      "message": "[cipd] Update Jiri\u0027s version of the cipd CLI\n\nThis is an automated change that updates Jiri\u0027s pinned version of the\n\u0027cipd\u0027 command-line tool. It\u0027s important to use an up-to-date version\nthe cipd CLI in case because the CIPD backend doesn\u0027t guarantee support\nfor arbitrarily old versions of the cipd CLI.\n\nThis change does not require review, just a Code-Review+2 and\nCommit-Queue+2.\n\nHowever, this change should only be submitted by a human during normal\nworking hours. Due to Jiri\u0027s self-updating functionality Jiri changes\nroll out automatically to users the next time they run \u0027jiri update\u0027,\nso it\u0027s important to remain online shortly after submitting this\nchange to respond in case users report issues.\n\nGenerated by \u0027scripts/update_cipd.sh\u0027.\n\nRoller-URL: https://cr-buildbucket.appspot.com/build/8671844116339472481\nCQ-Do-Not-Cancel-Tryjobs: true\nChange-Id: Ia6a245f0ad1b0c9edfd0b86ddfecdc999138ad3f\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1798698\nCommit-Queue: Rob Mohr \u003cmohrr@google.com\u003e\nReviewed-by: Rob Mohr \u003cmohrr@google.com\u003e\n"
    },
    {
      "commit": "873d88fefc271bc223b1beeca4967d002ab23408",
      "tree": "e08ae875c48969e185e78a967eedd3316b6c102a",
      "parents": [
        "65782c4fa7f040e87f9c5fe5106a076a2f9aa232"
      ],
      "author": {
        "name": "Erick Tryzelaar",
        "email": "etryzelaar@google.com",
        "time": "Mon Aug 31 13:00:44 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Aug 31 13:00:44 2026 -0700"
      },
      "message": "[jiri] Add support for other git clone filters\n\nThis allows jiri to configure the git repositories to support all filter\ngit options by deprecating the `-partial` argument, and adding `jiri\ninit -partial-filter $FILTER_SPEC ...` option. It uses `blob:none` by\ndefault if `-partial` is not specified for backwards compatibility.\n\nHere\u0027s the size differences for a `jiri update -fetch-packages\u003dfalse\n-run-hooks\u003dfalse -autoupdate\u003dfalse ...`:\n\n* Normal: 16 GiB\n* `blob:none`: 7.3 GiB\n* `tree:0`: 6.2 GiB\n\nSee this link for other filter specs we could use:\nhttps://git-scm.com/docs/git-clone#Documentation/git-clone.txt---filterfilter-spec\n\nFixed: 391895414\nChange-Id: I501c2207bba12f2501628f09937cd0622b35b63a\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1774735\nCommit-Queue: Erick Tryzelaar \u003cetryzelaar@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\n"
    },
    {
      "commit": "65782c4fa7f040e87f9c5fe5106a076a2f9aa232",
      "tree": "ffeecc6ab8d32729f89b449c53f9625782b31d74",
      "parents": [
        "8b5feec5bb6595d4551be5f4b9c8e85f3f2bd586"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Aug 31 10:15:06 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Aug 31 10:15:06 2026 -0700"
      },
      "message": "[worktree] Allow remove with missing parent project\n\nWhen removing a worktree, Jiri was failing if a project in the worktree\ncould not be found in the parent workspace. This happens if the parent\nworkspace has already been updated to a configuration that does not\ncontain the project.\n\nSince the parent project is missing, we cannot clean up its git\nworktree metadata anyway. Jiri now logs a warning and proceeds with\nthe rest of the worktree cleanup.\n\nTest: integrationtests:TestWorktreeRemoveParentProjectNotFound\n\nTAG: agy\nCONV: 542027ae-c0a0-43c4-94a9-843c71e35d0a\nChange-Id: I8967379f89ead442c1196f684e597271c4febde4\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1740070\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "8b5feec5bb6595d4551be5f4b9c8e85f3f2bd586",
      "tree": "1cd02ca02690dbc9ff54de2e4fd58302d91c0d5a",
      "parents": [
        "f3bb16503afbd04e6bfcfed7b64b8704817ff1c3"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Mon Aug 31 10:14:07 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Mon Aug 31 10:14:07 2026 -0700"
      },
      "message": "[worktree] Fix path sync bug in WorktreeSync\n\nWhen a project path is changed in the parent repo, the worktree sync\nfails because the worktree\u0027s local gitlink becomes broken and it is\nskipped during local projects scanning.\n\nFix this by repairing the broken gitlink and renaming the directory\nin the worktree before scanning local projects, using the parent\u0027s\nlocal projects as the source of truth for the path change.\n\nTest: integrationtests:TestWorktreeSyncPathChange\n\nTAG: agy\nCONV: 542027ae-c0a0-43c4-94a9-843c71e35d0a\nChange-Id: Icb5f8caebddabbe6df1276d6894f7d2b028a6bfc\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1740069\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "f3bb16503afbd04e6bfcfed7b64b8704817ff1c3",
      "tree": "76708e52591a60d0a0597a8c129c8bbcdefe5887",
      "parents": [
        "39631c30056d1828634436d9edb124091e0fc22c"
      ],
      "author": {
        "name": "global-integration-roller",
        "email": "global-integration-roller@fuchsia-infra.iam.gserviceaccount.com",
        "time": "Tue Aug 04 10:59:00 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Aug 04 10:59:00 2026 -0700"
      },
      "message": "[cipd] Update Jiri\u0027s version of the cipd CLI\n\nThis is an automated change that updates Jiri\u0027s pinned version of the\n\u0027cipd\u0027 command-line tool. It\u0027s important to use an up-to-date version\nthe cipd CLI in case because the CIPD backend doesn\u0027t guarantee support\nfor arbitrarily old versions of the cipd CLI.\n\nThis change does not require review, just a Code-Review+2 and\nCommit-Queue+2.\n\nHowever, this change should only be submitted by a human during normal\nworking hours. Due to Jiri\u0027s self-updating functionality Jiri changes\nroll out automatically to users the next time they run \u0027jiri update\u0027,\nso it\u0027s important to remain online shortly after submitting this\nchange to respond in case users report issues.\n\nGenerated by \u0027scripts/update_cipd.sh\u0027.\n\nRoller-URL: https://cr-buildbucket.appspot.com/build/8674380837804721665\nCQ-Do-Not-Cancel-Tryjobs: true\nChange-Id: Iee9d1e669201937ca93f183e7c316596491dc6d3\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1741709\nReviewed-by: Ina Huh \u003cihuh@google.com\u003e\nCommit-Queue: Ina Huh \u003cihuh@google.com\u003e\n"
    },
    {
      "commit": "39631c30056d1828634436d9edb124091e0fc22c",
      "tree": "7c7ce93213e3419b1270519fb5855f94b556e6a0",
      "parents": [
        "7a6b5e63a0ab9df1c213400b89af49cf7a854138"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Jul 30 15:03:11 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Jul 30 15:03:11 2026 -0700"
      },
      "message": "[worktree] Align worktree with parent\u0027s sync point\n\nWhen syncing a worktree, align it with the parent\u0027s last update\nsnapshot (sync point) instead of its current local revisions. This\nprevents uncommitted local commits in the parent tree from being\nincorrectly propagated to the worktree during sync.\n\nIf the parent has never been updated (and thus has no snapshot), it\nfalls back to the previous behavior of using the parent\u0027s current\nlocal revisions.\n\nHow it was tested:\n1. Added integration test TestWorktreeSyncMirrorsSyncPoint which\n   creates a local commit in parent and verifies worktree does not\n   get it during sync. Verified it fails without the fix and passes\n   with the fix.\n2. Manually verified using a script that sets up custom git repos,\n   creates a local commit in parent, and runs jiri worktree add\n   (which triggers sync). Confirmed it fails without the fix and\n   passes with the fix.\n\nTest: integrationtests/worktree_test.go (TestWorktreeSyncMirrorsSyncPoint)\nTest: manual verification using custom git repos\nBug: 540829453\nTAG: agy\nCONV: fa884e20-c4bc-45c5-8982-14de582bc883\nChange-Id: I1570722d1bd37d1f6965dd5efdbdc475299fc598\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1734490\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "7a6b5e63a0ab9df1c213400b89af49cf7a854138",
      "tree": "e28476fbb419ba8c0e2c6f8bb174b2368383dcca",
      "parents": [
        "24dfffcdd6c93c15d8fa86e01f9d259db86fb8d6"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Jul 30 15:02:31 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Jul 30 15:02:31 2026 -0700"
      },
      "message": "[worktree] Use hard links for packages everywhere\n\nUse hard links instead of symlinks for packages in all setups\n(both main tree and worktrees) for consistency. This resolves the\nRBE boundary issue where reclient fails to upload toolchain symlinks\nthat point outside the worktree.\n\nA metadata file `.jiri_cache_hash` is written to the package\ndirectory to track the current cache hash and avoid redundant\nreconstruction.\n\nSince hard-linked packages are real directories (unlike symlinks),\nwe cannot use `os.Readlink` to verify if they match the expected\ncache version. The `.jiri_cache_hash` file stores the package group\nhash, allowing Jiri to quickly verify the state and skip the\nexpensive recursive hard-linking process on subsequent updates/syncs\nif the package is already up-to-date.\n\nBug: 528310441\nTest: unit tests in project/ and osutil/, integration tests updated\nTAG\u003dagy\nCONV\u003d1387eccb-3543-436f-88a0-38702658138a\nChange-Id: Ia646ab8c9a32ac5c5f46adb54acc37cfc136505d\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1731737\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "24dfffcdd6c93c15d8fa86e01f9d259db86fb8d6",
      "tree": "ebbd42bae7dce5888deb8a2778cc5762a081ee3e",
      "parents": [
        "d60e1a9065f0f1cce1ec26578b6999daa902101a"
      ],
      "author": {
        "name": "Aidan Wolter",
        "email": "awolter@google.com",
        "time": "Thu Jul 30 07:01:28 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Jul 30 07:01:28 2026 -0700"
      },
      "message": "[worktree] Fix jiri upload inside worktree\n\nResolve symlinks in IsLocalProject, ProjectAtPath, and upload CWD.\n\nWhen running jiri from a symlinked worktree directory, path\ncomparisons between the symlinked CWD and resolved git paths\nwould mismatch, causing jiri to fail to find the project.\n\nResolving symlinks ensures path comparisons match consistently.\n\nBug: 534850642\nTest: Manual upload test in worktree\nTAG\u003dagy\nCONV\u003d6118192f-33a1-4138-b784-f3912e29a284\nChange-Id: I06c0fd6db1beec3daad074d1b119ac003290af9f\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1731576\nFuchsia-Auto-Submit: Aidan Wolter \u003cawolter@google.com\u003e\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Aidan Wolter \u003cawolter@google.com\u003e\n"
    },
    {
      "commit": "d60e1a9065f0f1cce1ec26578b6999daa902101a",
      "tree": "bf137be02c23acd3d9dca48d3d2c583da46053c3",
      "parents": [
        "00ec690d560a4cae6b16349cb974b26b8972eb75"
      ],
      "author": {
        "name": "Ina Huh",
        "email": "ihuh@google.com",
        "time": "Fri Jul 24 15:57:08 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Fri Jul 24 15:57:08 2026 -0700"
      },
      "message": "[jiri] Rewrite remote to use https:// when changing remote.\n\nChange-Id: I59f59f058c5fbcf3ea0ab158116e8aa16f3f14d2\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1724539\nCommit-Queue: Ina Huh \u003cihuh@google.com\u003e\nReviewed-by: Rahul Bangar \u003crahulbn@google.com\u003e\n"
    },
    {
      "commit": "00ec690d560a4cae6b16349cb974b26b8972eb75",
      "tree": "eca2663796983324e11ac8f7ac0ad1735aa1580a",
      "parents": [
        "eca9dd3ecbe0dabefbacf6e14904a489a55f2cc8"
      ],
      "author": {
        "name": "Ina Huh",
        "email": "ihuh@google.com",
        "time": "Thu Jul 23 06:29:07 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Thu Jul 23 06:29:07 2026 -0700"
      },
      "message": "[project] Set remotebranch to parent import remotebranch by default.\n\nIf the import manifest specifies a remote branch, we should use that for\nany projects that don\u0027t specify a revision or remote branch as opposed\nto always defaulting to main.\n\nBug: 537870330\nChange-Id: Idba650be39dedd3c18cfefd463961f2991573350\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1718793\nReviewed-by: Carver Forbes \u003ccarverforbes@google.com\u003e\nCommit-Queue: Ina Huh \u003cihuh@google.com\u003e\nFuchsia-Auto-Submit: Ina Huh \u003cihuh@google.com\u003e\n"
    },
    {
      "commit": "eca9dd3ecbe0dabefbacf6e14904a489a55f2cc8",
      "tree": "44f4572ab55741030587f2be0f54eb4f45720a4b",
      "parents": [
        "fbe3c83efca80639122f31915f09322c27f2cdc6"
      ],
      "author": {
        "name": "Dustin Green",
        "email": "dustingreen@google.com",
        "time": "Wed Jul 15 08:49:37 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Wed Jul 15 08:49:37 2026 -0700"
      },
      "message": "[project] project root being jiri root beats -gc\n\nAvoid deleting the root of the source tree even if -gc.\n\nAs of when this CL was posted, this could arise when using \u0027fx sync-to\u0027\nto a release branch on a freshly created fuchsia source tree with\nglobal-integration added and the subsequent jiri update done.\n\nThere\u0027s likely at least one more change needed elsewhere to avoid this\nCL\u0027s backstop coming into play in the first place. The intent of this CL\nis to \"safe\" the footgun ASAP. This CL isn\u0027t a complete fix on its own,\nbut we probably want to keep this backstop even after the main fix.\n\nChange-Id: Ie3948db7f828401b6c7cb00a55e3e5203892d379\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1707714\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nFuchsia-Auto-Submit: Dustin Green \u003cdustingreen@google.com\u003e\nCommit-Queue: Dustin Green \u003cdustingreen@google.com\u003e\n"
    },
    {
      "commit": "fbe3c83efca80639122f31915f09322c27f2cdc6",
      "tree": "5185aa3288d9b923a46014a6d31210dc57737c8c",
      "parents": [
        "683519d5c521adb59bbd4648801a642795082ade"
      ],
      "author": {
        "name": "Garratt Gallagher",
        "email": "garratt@google.com",
        "time": "Tue Jul 14 01:05:14 2026 -0700"
      },
      "committer": {
        "name": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "email": "fuchsia-internal-scoped@luci-project-accounts.iam.gserviceaccount.com",
        "time": "Tue Jul 14 01:05:14 2026 -0700"
      },
      "message": "[cache] Fix git directory detection in bare repositories\n\nWhen using a shared cache, jiri creates bare repositories.\nSome git commands (like git config) fail with\n \"fatal: not in a git directory\"\nwhen run inside a bare repository when safe.bareRepository\u003dexplicit\n\nThis fix adds a check to gitutil.New which\nsets GIT_DIR\u003d\".\" in the environment if it\ndetects that the git repo is bare.\nThis environmental variable is then set for\nall commands executed by that instance.\n\nThis prevents jiri from incorrectly assuming the\ncache is corrupt and clearing it when updates fail.\n\nTest: Manually verified that jiri update works with\nshared cache enabled on multiple workspaces.\n\nChange-Id: Ifec37d5adc4dc76448ca530d66098139120a19dc\nReviewed-on: https://fuchsia-review.googlesource.com/c/jiri/+/1573741\nReviewed-by: Oliver Newman \u003colivernewman@google.com\u003e\nCommit-Queue: Garratt Gallagher \u003cgarratt@google.com\u003e\n"
    }
  ],
  "next": "683519d5c521adb59bbd4648801a642795082ade"
}
