Merge pull request #78501 from bhipple/feature/fetchcargo-tar-gz

fetchcargo: use flat tar.gz file for vendored src instead of recursive hash dir

doc/languages-frameworks/rust.section.md

@@ -37,7 +37,7 @@ rustPlatform.buildRustPackage rec {
   };
 
   cargoSha256 = "17ldqr3asrdcsh4l29m3b5r37r5d0b3npq1lrgjmxb6vlx6a36qh";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   meta = with stdenv.lib; {
     description = "A fast line-oriented regex search tool, similar to ag and ack";
@@ -59,12 +59,19 @@ When the `Cargo.lock`, provided by upstream, is not in sync with the
 added in `cargoPatches` will also be prepended to the patches in `patches` at
 build-time.
 
-When `verifyCargoDeps` is set to `true`, the build will also verify that the
+Setting `legacyCargoFetcher` to `false` enables the following behavior:
-`cargoSha256` is not out of date by comparing the `Cargo.lock` file in both the
+
-`cargoDeps` and `src`. Note that this option changes the value of `cargoSha256`
+1. The `Cargo.lock` file is copied into the cargo vendor directory.
-since it also copies the `Cargo.lock` in it. To avoid breaking
+2. At buildtime, `buildRustPackage` will ensure that the `src` and `cargoSha256`
-backward-compatibility this option is not enabled by default but hopefully will
+   are consistent. This avoids errors where one but not the other is updated.
-be in the future.
+3. The builder will compress the vendored cargo src directory into a tar.gz file
+   for storage after vendoring, and decompress it before the build. This saves
+   disk space and enables hashed mirrors for Rust dependencies.
+
+Note that this option changes the value of `cargoSha256`, so it is currently
+defaulted to `false`. When updating a Rust package, please set it to `true`;
+eventually we will default this to true and update the remaining Rust packages,
+then delete the option from all individual Rust package expressions.
 
 ### Building a crate for a different target
 

pkgs/applications/editors/hexdino/default.nix

@@ -11,8 +11,8 @@ rustPlatform.buildRustPackage {
     sha256 = "11mz07735gxqfamjcjjmxya6swlvr1p77sgd377zjcmd6z54gwyf";
   };
 
-  cargoSha256 = "0qa8ypp5a7sf1gic482zh3i6s94w6k6bgmk5ynfvwi7g49ql7c4z";
+  cargoSha256 = "06ghcd4j751mdkzwb88nqwk8la4zdb137y0iqrkpykkfx0as43x3";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   buildInputs = [ ncurses ];
 

pkgs/applications/version-management/git-and-tools/git-workspace/default.nix

@@ -15,9 +15,9 @@ rustPlatform.buildRustPackage rec {
     sha256 = "0pl5z0gx2ypkrgq7vj1cxj5iwj06vcd06x3b3nh0g7w7q7nl8pr4";
   };
 
-  cargoSha256 = "0jbsz7r9n3jcgb9sd6pdjwzjf1b35qpfqw8ba8fjjmzfvs9qn6ld";
+  cargoSha256 = "1z4cb7rcb7ldj16xxynrjh4hg872rj39rbbp0vy15kdp3ifyi466";
 
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   buildInputs = with stdenv; lib.optional isDarwin Security;
 

pkgs/build-support/rust/README.md

@@ -0,0 +1,45 @@
+# Updated fetchCargo behavior
+
+Changes to the `fetchcargo.nix` behavior that cause changes to the `cargoSha256`
+are somewhat disruptive, so historically we've added conditionals to provide
+backwards compatibility. We've now accumulated enough of these that it makes
+sense to do a clean sweep updating hashes, and delete the conditionals in the
+fetcher to simplify maintenance and implementation complexity. These
+conditionals are:
+
+1. When cargo vendors dependencies, it generates a config. Previously, we were
+   hard-coding our own config, but this fails if there are git dependencies. We
+   have conditional logic to sometimes copy the vendored cargo config in, and
+   sometimes not.
+
+2. When a user updates the src package, they may forget to update the
+   `cargoSha256`. We have an opt-in conditional flag to add the `Cargo.lock`
+   into the vendor dir for inspection and compare at build-time, but it defaults
+   to false.
+
+3. We were previously vendoring into a directory with a recursive hash, but
+   would like to vendor into a compressed tar.gz file instead, for the reasons
+   specified in the git commit message adding this feature.
+
+
+## Migration plan
+
+1. (DONE in this PR) Implement `fetchCargoTarball` as a separate, clean fetcher
+   implementation along-side `fetchcargo`. Rename `verifyCargoDeps` (default
+   false) to `legacyCargoFetcher` (default true), which switches the fetcher
+   implementation used. Replace `verifyCargoDeps = true;` with
+   `legacyCargoFetcher = false;` in Rust applications.
+
+2. Send a treewide Rust PR that sets `legacyCargoFetcher = true;` in all Rust
+   applications not using this (which is ~200 of them), with a note to
+   maintainers to delete if updating the package. Change the default in
+   `buildRustPackage` to false.
+
+3. Go through all Rust src packages deleting the `legacyCargoFetcher = false;`
+   line and re-computing the `cargoSha256`, merging as we go.
+
+4. Delete the `fetchcargo.nix` implementation entirely and also remove:
+  - All overrides in application-level packages
+  - The `fetchcargo-default-config.toml` and conditionals around using it when
+    no `$CARGO_CONFIG` exists
+  - This README.md file

pkgs/build-support/rust/default.nix

@@ -1,4 +1,4 @@
-{ stdenv, cacert, git, rust, cargo, rustc, fetchcargo, buildPackages, windows }:
+{ stdenv, cacert, git, rust, cargo, rustc, fetchcargo, fetchCargoTarball, buildPackages, windows }:
 
 { name ? "${args.pname}-${args.version}"
 , cargoSha256 ? "unset"
@@ -14,13 +14,13 @@
 , cargoUpdateHook ? ""
 , cargoDepsHook ? ""
 , cargoBuildFlags ? []
-, # Set to true to verify if the cargo dependencies are up to date.
+  # Please set to true on any Rust package updates. Once all packages set this
-  # This will change the value of cargoSha256.
+  # to true, we will delete and make it the default. For details, see the Rust
-  verifyCargoDeps ? false
+  # section on the manual and ./README.md.
+, legacyCargoFetcher ? true
 , buildType ? "release"
 , meta ? {}
 , target ? null
-
 , cargoVendorDir ? null
 , ... } @ args:
 
@@ -28,20 +28,27 @@ assert cargoVendorDir == null -> cargoSha256 != "unset";
 assert buildType == "release" || buildType == "debug";
 
 let
+
+  cargoFetcher = if legacyCargoFetcher
+                 then fetchcargo
+                 else fetchCargoTarball;
+
   cargoDeps = if cargoVendorDir == null
-    then fetchcargo {
+    then cargoFetcher {
         inherit name src srcs sourceRoot unpackPhase cargoUpdateHook;
-        copyLockfile = verifyCargoDeps;
         patches = cargoPatches;
         sha256 = cargoSha256;
       }
     else null;
 
+  # If we're using the modern fetcher that always preserves the original Cargo.lock
+  # and have vendored deps, check them against the src attr for consistency.
+  validateCargoDeps = cargoSha256 != "unset" && !legacyCargoFetcher;
+
   setupVendorDir = if cargoVendorDir == null
     then ''
       unpackFile "$cargoDeps"
-      cargoDepsCopy=$(stripHash $(basename $cargoDeps))
+      cargoDepsCopy=$(stripHash $cargoDeps)
-      chmod -R +w "$cargoDepsCopy"
     ''
     else ''
       cargoDepsCopy="$sourceRoot/${cargoVendorDir}"
@@ -54,9 +61,14 @@ let
   ccForHost="${stdenv.cc}/bin/${stdenv.cc.targetPrefix}cc";
   cxxForHost="${stdenv.cc}/bin/${stdenv.cc.targetPrefix}c++";
   releaseDir = "target/${rustTarget}/${buildType}";
+
+  # Fetcher implementation choice should not be part of the hash in final
+  # derivation; only the cargoSha256 input matters.
+  filteredArgs = builtins.removeAttrs args [ "legacyCargoFetcher" ];
+
 in
 
-stdenv.mkDerivation (args // {
+stdenv.mkDerivation (filteredArgs // {
   inherit cargoDeps;
 
   patchRegistryDeps = ./patch-registry-deps;
@@ -95,14 +107,13 @@ stdenv.mkDerivation (args // {
     ''}
     EOF
 
-    unset cargoDepsCopy
     export RUST_LOG=${logLevel}
-  '' + stdenv.lib.optionalString verifyCargoDeps ''
+  '' + stdenv.lib.optionalString validateCargoDeps ''
-    if ! diff source/Cargo.lock $cargoDeps/Cargo.lock ; then
+    if ! diff source/Cargo.lock $cargoDepsCopy/Cargo.lock ; then
       echo
       echo "ERROR: cargoSha256 is out of date"
       echo
-      echo "Cargo.lock is not the same in $cargoDeps"
+      echo "Cargo.lock is not the same in $cargoDepsCopy"
       echo
       echo "To fix the issue:"
       echo '1. Use "1111111111111111111111111111111111111111111111111111" as the cargoSha256 value'
@@ -112,6 +123,8 @@ stdenv.mkDerivation (args // {
 
       exit 1
     fi
+  '' + ''
+    unset cargoDepsCopy
   '' + (args.postUnpack or "");
 
   configurePhase = args.configurePhase or ''

pkgs/build-support/rust/fetchCargoTarball.nix

@@ -0,0 +1,81 @@
+{ stdenv, cacert, git, cargo, python3 }:
+let cargo-vendor-normalise = stdenv.mkDerivation {
+  name = "cargo-vendor-normalise";
+  src = ./cargo-vendor-normalise.py;
+  nativeBuildInputs = [ python3.pkgs.wrapPython ];
+  dontUnpack = true;
+  installPhase = "install -D $src $out/bin/cargo-vendor-normalise";
+  pythonPath = [ python3.pkgs.toml ];
+  postFixup = "wrapPythonPrograms";
+  doInstallCheck = true;
+  installCheckPhase = ''
+    # check that ./fetchcargo-default-config.toml is a fix point
+    reference=${./fetchcargo-default-config.toml}
+    < $reference $out/bin/cargo-vendor-normalise > test;
+    cmp test $reference
+  '';
+  preferLocalBuild = true;
+};
+in
+{ name ? "cargo-deps"
+, src ? null
+, srcs ? []
+, patches ? []
+, sourceRoot
+, sha256
+, cargoUpdateHook ? ""
+, ...
+} @ args:
+stdenv.mkDerivation ({
+  name = "${name}-vendor.tar.gz";
+  nativeBuildInputs = [ cacert git cargo-vendor-normalise cargo ];
+
+  phases = "unpackPhase patchPhase buildPhase installPhase";
+
+  buildPhase = ''
+    # Ensure deterministic Cargo vendor builds
+    export SOURCE_DATE_EPOCH=1
+
+    if [[ ! -f Cargo.lock ]]; then
+        echo
+        echo "ERROR: The Cargo.lock file doesn't exist"
+        echo
+        echo "Cargo.lock is needed to make sure that cargoSha256 doesn't change"
+        echo "when the registry is updated."
+        echo
+
+        exit 1
+    fi
+
+    # Keep the original around for copyLockfile
+    cp Cargo.lock Cargo.lock.orig
+
+    export CARGO_HOME=$(mktemp -d cargo-home.XXX)
+    CARGO_CONFIG=$(mktemp cargo-config.XXXX)
+
+    ${cargoUpdateHook}
+
+    cargo vendor $name | cargo-vendor-normalise > $CARGO_CONFIG
+
+    # Add the Cargo.lock to allow hash invalidation
+    cp Cargo.lock.orig $name/Cargo.lock
+
+    # Packages with git dependencies generate non-default cargo configs, so
+    # always install it rather than trying to write a standard default template.
+    install -D $CARGO_CONFIG $name/.cargo/config;
+  '';
+
+  # Build a reproducible tar, per instructions at https://reproducible-builds.org/docs/archives/
+  installPhase = ''
+    tar --owner=0 --group=0 --numeric-owner --format=gnu \
+        --sort=name --mtime="@$SOURCE_DATE_EPOCH" \
+        -czf $out $name
+  '';
+
+  outputHashAlgo = "sha256";
+  outputHash = sha256;
+
+  impureEnvVars = stdenv.lib.fetchers.proxyImpureEnvVars;
+} // (builtins.removeAttrs args [
+  "name" "sha256" "cargoUpdateHook"
+]))

pkgs/development/compilers/rust/default.nix

@@ -25,12 +25,18 @@
       inherit rustc cargo;
     };
 
+    fetchCargoTarball = buildPackages.callPackage ../../../build-support/rust/fetchCargoTarball.nix {
+      inherit cargo;
+    };
+
+    # N.B. This is a legacy fetcher implementation that is being phased out and deleted.
+    # See ../../../build-support/rust/README.md for details.
     fetchcargo = buildPackages.callPackage ../../../build-support/rust/fetchcargo.nix {
       inherit cargo;
     };
 
     buildRustPackage = callPackage ../../../build-support/rust {
-      inherit rustc cargo fetchcargo;
+      inherit rustc cargo fetchcargo fetchCargoTarball;
     };
 
     rustcSrc = callPackage ./rust-src.nix {

pkgs/development/tools/documentation/mdsh/default.nix

@@ -11,8 +11,8 @@ rustPlatform.buildRustPackage rec {
     sha256 = "1a9i6h8fzrrfzjyfxaps73lxgkz92k0bnmwbjbwdmiwci4qgi9ms";
   };
 
-  cargoSha256 = "0rarpzfigyxr6s0ba13z00kvnms29qkjfbfjkay72mb6xn7f1059";
+  cargoSha256 = "1fxajh1n0qvcdas6w7dy3g92wilhfldy90pyk3779mrnh57fa6n5";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   meta = with stdenv.lib; {
     description = "Markdown shell pre-processor";

pkgs/tools/misc/broot/default.nix

@@ -11,8 +11,8 @@ rustPlatform.buildRustPackage rec {
     sha256 = "13b1w9g68aj3r70w9bmrmdc772y959n77ajbdm2cpjs5f4kgfpak";
   };
 
-  cargoSha256 = "0vzpyymylzxjm613lf5xr6hd21ijkl3vwq4y6h1q3as41phw2sqb";
+  cargoSha256 = "0zrwpmsrzwnjml0964zky8w222zmlargha3z0n6hf8cfshx23s4k";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   nativeBuildInputs = [ installShellFiles ];
 

pkgs/tools/misc/wagyu/default.nix

@@ -11,8 +11,8 @@ rustPlatform.buildRustPackage rec {
     sha256 = "1646j0lgg3hhznifvbkvr672p3yqlcavswijawaxq7n33ll8vmcn";
   };
 
-  cargoSha256 = "10b96l0b32zxq0xrnhivv3gihmi5y31rllbizv67hrg1axz095vn";
+  cargoSha256 = "16d1b3pamkg29nq80n6cbzc4zl9z3cgfvdxjkr2z4xrnzmkn1ysi";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   meta = with lib; {
     description = "Rust library for generating cryptocurrency wallets";

pkgs/tools/package-management/nix-du/default.nix

@@ -9,8 +9,8 @@ rustPlatform.buildRustPackage rec {
     rev = "v${version}";
     sha256 = "149d60mid29s5alv5m3d7jrhyzc6cj7b6hpiq399gsdwzgxr00wq";
   };
-  cargoSha256 = "18kb4car5nzch3vpl6z1499silhs3fyn8c6xj3rzk94mm2m9srg4";
+  cargoSha256 = "1a6svl89dcdb5fpvs2i32i6agyhl0sx7kkkw70rqr17fyzl5psai";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   doCheck = true;
   checkInputs = [ graphviz ];

pkgs/tools/security/fido2luks/default.nix

@@ -19,8 +19,8 @@ rustPlatform.buildRustPackage rec {
   buildInputs = [ cryptsetup ];
   nativeBuildInputs = [ pkg-config ];
 
-  cargoSha256 = "1i37k4ih6118z3wip2qh4jqk7ja2z0v1w8dri1lwqwlciqw17zi9";
+  cargoSha256 = "0rp4f6xnwmvf3pv6h0qwsg01jrndf77yn67675ac39kxzmrzfy2f";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   meta = with stdenv.lib; {
     description = "Decrypt your LUKS partition using a FIDO2 compatible authenticator";

pkgs/tools/system/tre-command/default.nix

@@ -11,8 +11,8 @@ rustPlatform.buildRustPackage rec {
     sha256 = "1fazw2wn738iknbv54gv7qll7d4q2gy9bq1s3f3cv21cdv6bqral";
   };
 
-  cargoSha256 = "0m82zbi610zgvcza6n03xl80g31x6bfkjyrfxcxa6fyf2l5cj9pv";
+  cargoSha256 = "1m3ccp5ncafkifg8sxyxczsg3ja1gvq8wmgni68bgzm2lwxh2qgw";
-  verifyCargoDeps = true;
+  legacyCargoFetcher = false;
 
   meta = with stdenv.lib; {
     description = "Tree command, improved";