Fail the Docker build when the lint stage's linter is not the pin
All checks were successful
check / check (push) Successful in 1m23s

The reordered COPY --from=lint did not make the two stages provably one
toolchain, as the Dockerfile comment, the previous commit message and
TODO.md all claimed. script/bootstrap compares its pin against whatever
PATH resolves, and $GOPATH/bin sits ahead of /usr/local/bin, so any
drift was absorbed: bootstrap rebuilt the pinned version from source,
verified that, and the build went green with the lint stage having
linted at one version and make check having run at another. Bumping the
lint stage image without touching the pin was enough to produce it.

New script/verify-linter-pin fails, naming both versions, unless a given
golangci-lint binary is exactly the version script/bootstrap pins. The
build stage runs it on the binary copied out of the lint stage,
immediately after the copy and before bootstrap, so no reinstall can
satisfy it. The pin is read out of script/bootstrap, which stays its
single source of truth; a pin that cannot be read is a hard failure
rather than a skip. The check takes no CHECK_EPOCH because its only
inputs are the copied binary and script/, so Docker invalidates the
layer exactly when a cached result would stop being true.

The linter version is pinned independently in the lint stage's image
digest and in GOLANGCI_LINT_VERSION, with nothing keeping them in sync;
a half-applied bump is now a build failure instead of a silent split.

ENV PATH keeps $GOPATH/bin, but its comment no longer claims a reinstall
is the reason: bootstrap must be able to run and verify what it
installs, and nothing in this image is shadowed by the entry.

Verified: with the lint stage's linter faked to 2.11.0 after the gates
had really run, the build fails at verify-linter-pin naming 2.11.0 and
2.12.2, with bootstrap and the check gate never reached; an unmodified
make docker is green with all three gates run on a fresh epoch and real
test results. A planted unused finding still fails at the lint stage
with gate check absent from the log; the image still fails
TestScanHardlinkRunFailsTogether under --user 0:0 and passes as uid
1000, both with the Go test cache disabled; and a second build serves
bootstrap, the verify layer and the dependency layers CACHED while the
gates go cold.
This commit is contained in:
clawbot
2026-08-09 15:29:13 +00:00
parent 3a183aa64b
commit 5ca68804ac
4 changed files with 177 additions and 27 deletions

View File

@@ -32,10 +32,12 @@ FROM golang@sha256:56961d79ea8129efddcc0b8643fd8a5416b4e6228cfd477e3fd61deb2672c
# We never build or run as root. Create an unprivileged user and point
# HOME and the Go caches at its home so go build/test and golangci-lint
# can write their caches when we drop to it below. $GOPATH/bin is on
# PATH because that is where script/bootstrap's `go install` lands: if
# the linter copied in below ever stops matching bootstrap's pin,
# bootstrap reinstalls it and then verifies the pin against what PATH
# resolves, which can only succeed if that directory is searched.
# PATH because that is where script/bootstrap's `go install` lands: a
# tool bootstrap installs must be runnable afterwards, and bootstrap
# verifies its own installs against what PATH resolves, so leaving that
# directory unsearched would make any install it performs both unusable
# and self-reported as shadowed. Nothing in this image is shadowed by
# it: the directory does not exist until bootstrap runs.
RUN adduser -D -u 1000 builder
ENV HOME=/home/builder
ENV GOPATH=/home/builder/go
@@ -52,22 +54,40 @@ WORKDIR /src
# so it is what forces BuildKit to finish fmt-check and lint before
# compilation and tests start. Remove it and the fail-fast design
# dies silently: the build stops gating on lint and still exits 0.
# - It is what keeps the two stages on one toolchain. script/bootstrap
# version-checks whatever PATH resolves against its pin, so copying
# the lint stage's binary in first means every build now compares
# the lint stage's linter to that pin and fails loudly if they ever
# drift apart. Bootstrap installing its own linter here instead
# would restore exactly the two-independent-toolchains problem the
# copy prevents (and cost a from-source build of the linter).
# - Together with the check below it is what keeps the two stages on
# one toolchain: `make check` here runs the very binary the lint
# stage ran, not a second one that happens to agree. Bootstrap
# installing its own linter here instead would restore exactly the
# two-independent-toolchains problem the copy prevents (and cost a
# from-source build of the linter).
COPY --from=lint /usr/bin/golangci-lint /usr/local/bin/golangci-lint
# Fail the build, naming both versions, unless the binary that just
# arrived from the lint stage is the version script/bootstrap pins.
#
# Nothing else enforces that. The linter version is pinned in two
# independent places — the lint stage's image digest above and
# GOLANGCI_LINT_VERSION in script/bootstrap — and bumping one alone is
# an easy mistake. Without this check that mistake is invisible:
# bootstrap below would see a version that is not its pin, quietly
# rebuild the pinned one from source into a directory that is on PATH,
# verify that, and exit 0. The build would go green with the lint stage
# having linted at one version and `make check` at another, which is
# precisely the divergence the copy above exists to prevent.
#
# It runs here, before bootstrap, so that a reinstall cannot satisfy it,
# and it needs no CHECK_EPOCH: its only inputs are the copied binary and
# script/, so Docker invalidates this layer exactly when a cached result
# would stop being true.
COPY script/ script/
RUN script/verify-linter-pin /usr/local/bin/golangci-lint
# Install development prerequisites the same way a developer does,
# rather than duplicating the installs inline. script/ and the
# dependency manifests are copied first, and nothing else is, so this
# layer stays cached until the scripts or the dependencies change —
# rather than duplicating the installs inline. Only script/ (copied
# above) and the dependency manifests are copied first, nothing else, so
# this layer stays cached until the scripts or the dependencies change —
# bootstrap ends in `go mod download`, which is why there is no separate
# invocation of it here.
COPY script/ script/
COPY go.mod go.sum ./
RUN script/bootstrap