lix/tests/fixed.nix
Maximilian Bosch bb1a851bcf
build: also throw hash-mismatch errors if buildMode == bmCheck
This actually bit me quite recently in `nixpkgs` because I assumed that
`nix-build --check` would also error out if hashes don't match anymore[1]
and so I wrongly assumed that I couldn't reproduce the mismatch error.

The fix is rather simple, during the output registration a so-called
`delayedException` is instantiated e.g. if a FOD hash-mismatch occurs.
However, in case of `nix-build --check` (or `--rebuild` in case of `nix
build`), the code-path where this exception is thrown will never be
reached.

By adding that check to the if-clause that causes an early exit in case
of `bmCheck`, the issue is gone. Also added a (previously failing)
test-case to demonstrate the problem.

[1] https://github.com/NixOS/nixpkgs/pull/139238, the underlying issue
    was that `nix-prefetch-git` returns different hashes than `fetchgit`
    because the latter one fetches submodules by default.
2021-09-27 15:44:39 +02:00

59 lines
2 KiB
Nix

with import ./config.nix;
rec {
f2 = dummy: builder: mode: algo: hash: mkDerivation {
name = "fixed";
inherit builder;
outputHashMode = mode;
outputHashAlgo = algo;
outputHash = hash;
inherit dummy;
impureEnvVars = ["IMPURE_VAR1" "IMPURE_VAR2"];
};
f = f2 "";
good = [
(f ./fixed.builder1.sh "flat" "md5" "8ddd8be4b179a529afa5f2ffae4b9858")
(f ./fixed.builder1.sh "flat" "sha1" "a0b65939670bc2c010f4d5d6a0b3e4e4590fb92b")
(f ./fixed.builder2.sh "recursive" "md5" "3670af73070fa14077ad74e0f5ea4e42")
(f ./fixed.builder2.sh "recursive" "sha1" "vw46m23bizj4n8afrc0fj19wrp7mj3c0")
];
# Expression to test that `nix-build --check` also throws an error if the hash of
# fixed-output derivation has changed even if the hash exists in the store (in this
# case the hash exists because of `fixed.builder2.sh`, but building a derivation
# with the same hash and a different result must throw an error).
check = [
(f ./fixed.builder1.sh "recursive" "md5" "3670af73070fa14077ad74e0f5ea4e42")
];
good2 = [
# Yes, this looks fscked up: builder2 doesn't have that result.
# But Nix sees that an output with the desired hash already
# exists, and will refrain from building it.
(f ./fixed.builder2.sh "flat" "md5" "8ddd8be4b179a529afa5f2ffae4b9858")
];
sameAsAdd =
f ./fixed.builder2.sh "recursive" "sha256" "1ixr6yd3297ciyp9im522dfxpqbkhcw0pylkb2aab915278fqaik";
bad = [
(f ./fixed.builder1.sh "flat" "md5" "0ddd8be4b179a529afa5f2ffae4b9858")
];
reallyBad = [
# Hash too short, and not base-32 either.
(f ./fixed.builder1.sh "flat" "md5" "ddd8be4b179a529afa5f2ffae4b9858")
];
# Test for building two derivations in parallel that produce the
# same output path because they're fixed-output derivations.
parallelSame = [
(f2 "foo" ./fixed.builder2.sh "recursive" "md5" "3670af73070fa14077ad74e0f5ea4e42")
(f2 "bar" ./fixed.builder2.sh "recursive" "md5" "3670af73070fa14077ad74e0f5ea4e42")
];
}