Skip to content

smoke tests: 27 packages (a–l, batch 3) - #1368

Merged
bryan-minimal merged 1 commit into
mainfrom
smoke/bin-a-l-3
Nov 7, 2026
Merged

bryan-minimal merged 1 commit into
mainfrom
smoke/bin-a-l-3

Conversation

@bryan-minimal

Copy link
Copy Markdown
Member

Adds a Standalone smoke test to 27 packages that ship an OutputBin and had no Standalone test (a–l list, batch 3 of 3).

package test
helm /usr/bin/helm version
hex-patch /usr/bin/hex-patch --version
hexhog /usr/bin/hexhog --version
hexyl /usr/bin/hexyl --version
hyperfine /usr/bin/hyperfine --version
icu /usr/bin/uconv --version
imgcatr /usr/bin/imgcatr --version
iproute2 /usr/sbin/ip -V
jdtls /usr/bin/jdtls --help
jqfmt bash -c '/usr/bin/jqfmt -h; rc=$?; [ $rc -lt 126 ]' (test_deps: bash-bootstrap)
json-server /usr/bin/json-server --version
just /usr/bin/just --version
kittyview /usr/bin/kittyview --version
lcms2 bash -c '/usr/bin/transicc -h; rc=$?; [ $rc -lt 126 ]' (test_deps: bash-bootstrap)
lerna /usr/bin/lerna --version
libcap /usr/sbin/capsh --print
libfuse /usr/bin/fusermount3 --version
libsvtav1 /usr/bin/SvtAv1EncApp --version
libtasn1 /usr/bin/asn1Parser --version
libtool /usr/bin/libtool --version
libvips /usr/bin/vips --version
libvmaf /usr/bin/vmaf --version
libvpx bash -c '/usr/bin/vpxdec --help; rc=$?; [ $rc -lt 126 ]' (test_deps: bash-bootstrap)
libwebp /usr/bin/cwebp -version
libx265 /usr/bin/x265 --version
llvm /usr/bin/clang --version
lsd /usr/bin/lsd --version

How these were verified

Locally, in a fresh min session activate . --loadout pkgbuild on this tree (a new session name for every run): min package build <pkgs> (cached artifacts, no rebuild; tests are not in the spec hash), then min check --packages <pkgs> with the names space-separated. All 81 tests across the three a–l batches report standalone tests...Pass, and no other check fails. pkgs CI shows standalone tests...Skip for these because nothing rebuilds, so the local run is the real evidence.

Negative control: in a fresh session with all the tests applied, hyperfine's test was swapped for /usr/bin/hyperfine --version; exit 7. min check reported smoketest: ... had exit code 7 / standalone tests...Fail. So the tests do run and a failure is reported. The real failures below (exit 127, loader errors) also came through as Fail.

Wrapper-style tests: where a command needs a shell (a sh script like freetype-config, or a tool whose help flag exits non-zero by design, wrapped as cmd; rc=$?; [ $rc -lt 126 ] to prove only that exec and the loader work), the test uses test_deps = [bash-bootstrap], not base. bash-bootstrap's only runtime_dep is glibc. base drags in zlib, zstd, libmagic, libstdc++ and others, and that hid real missing runtime_deps: libgd and diffoscope both passed with base and failed without it.

Skipped (no test added)

package reason
bash-bootstrap, gawk-bootstrap, ghc-bootstrap, llvm-bootstrap bootstrap packages
binutils-2.46-glibc, gcc-15.2.0-glibc, glibc-bedrock-2.42, go-1.4, jikes-1.22 bootstrap rungs (} target multi-arch shape)
hex0 stage0
curl, file, htop, ghostscript, libidn2, libgd, diffoscope, dejagnu, llama.cpp the smoke test fails because of a real packaging bug (findings below). The test should land with the fix.

Findings (separate PRs, since runtime_deps are hashed and trigger a rebuild)

The missing libraries were found by an in-test ldd run under test_deps = [bash-bootstrap], with the result read back through exit codes, because min check drops test output.

package symptom missing
curl curl --version exit 127 libzstd.so.1 (zstd), same as batch 1 (#820)
file file --version exit 127 libzstd.so.1, libz.so.1 (zstd, zlib), same as batch 1
libidn2 idn2 --version exit 127 libunistring.so.5 (libunistring), same as batch 1
htop htop --version exit 127 libcap.so.2 (libcap)
ghostscript gs --version exit 127 libbz2.so.1.0 (bzip2)
libgd gdtopng exit ≥126 without base libstdc++.so.6, libgcc_s.so.1 (subsetOf gcc ["libgcc","libstdcpp"])
diffoscope --version exit 1 without base: ImportError: failed to find libmagic file (libmagic)
dejagnu runtest --version exit 1: ERROR: unable to find expect expect
llama.cpp llama-cli --version exit 127 its OWN libs: libmtmd.so.0, libllama.so.0, libllama-common.so.0, libggml*.so.0. build.sh runs install -vm755 build/bin/*.so, which ships only the unversioned names, so the SONAME files are never installed. llama-cli and llama-server cannot start.

Tests ran on the Mac's VM (arm64). amd64 is not exercised locally.

🤖 Generated with Claude Code

Add a Standalone smoke test (--version or equivalent) to 27 packages that
ship an OutputBin and had none. Tests are not in the spec hash, so nothing
rebuilds. Verified locally with min check in a fresh session.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@bryan-minimal
bryan-minimal requested a review from a team as a code owner November 6, 2026 21:09
@coderabbitai

coderabbitai Bot commented Nov 6, 2026

Copy link
Copy Markdown
Contributor

Warning

Review limit reached

  • Run on-demand review

This review includes 27 billable files and costs up to $6.75.

  • Ask an admin to make reviews automatic

Open in CodeRabbit

Reviews can continue after your included limit without a manual trigger. An admin must approve usage-based billing.

Or wait 57 minutes for your next included review.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available. Your 65 included PR review attempts over the past 7 days set your current allowance at 1 review per hour.

Learn how review limits work.

Review configuration:

Run configuration

Commits

Reviewing files that changed from the base of the PR and between 7a198f0 and d8f88ba.

Files selected for processing: 27

  • packages/helm/build.ncl
  • packages/hex-patch/build.ncl
  • packages/hexhog/build.ncl
  • packages/hexyl/build.ncl
  • packages/hyperfine/build.ncl
  • packages/icu/build.ncl
  • packages/imgcatr/build.ncl
  • packages/iproute2/build.ncl
  • packages/jdtls/build.ncl
  • packages/jqfmt/build.ncl
  • packages/json-server/build.ncl
  • packages/just/build.ncl
  • packages/kittyview/build.ncl
  • packages/lcms2/build.ncl
  • packages/lerna/build.ncl
  • packages/libcap/build.ncl
  • packages/libfuse/build.ncl
  • packages/libsvtav1/build.ncl
  • packages/libtasn1/build.ncl
  • packages/libtool/build.ncl
  • packages/libvips/build.ncl
  • packages/libvmaf/build.ncl
  • packages/libvpx/build.ncl
  • packages/libwebp/build.ncl
  • packages/libx265/build.ncl
  • packages/llvm/build.ncl
  • packages/lsd/build.ncl
  • Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

@norrietaylor norrietaylor left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Scope matches the title. The diff adds one smoketest to each of the 27 listed packages. It changes no build_deps, runtime_deps, outputs or build_args. Clobber none. Each hunk only adds a tests block, adds to the minimal.ncl import line, or adds a bash-bootstrap import. The diff removes no identifier from an import line. Correctness I checked each test path against the package outputs globs. All 27 binaries are in a declared output (usr/bin/* or usr/sbin/*). In libcap, the tests block is on the outer spec and not inside replace_on_cycle. The three wrapper tests use /usr/bin/bash, and bash-bootstrap ships that path. They also follow the rc -lt 126 shape in packages/edgedelta/build.ncl:68. Import lines keep the lowercase-first order. One weak test: jdtls --help never starts the JDK. Not checked I did not run the tests. That needs a build and min check. The PR body says the author ran them locally on arm64 only.

Comment thread packages/jdtls/build.ncl

tests = {
smoketest = standaloneTest "/usr/bin/jdtls --help",
},

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

low · jdtls --help does not start the JDK. The wrapper in build.sh:26 execs python3 /usr/share/jdtls/bin/jdtls. The upstream launcher parses arguments with argparse. argparse prints help and exits 0 before it builds the java command line. So the test proves only that sh, sha1sum, cut and python3 work. It passes if jdk leaves runtime_deps or the JDK cannot load. The other tests in this PR prove the shipped binary loads, so this one is weaker than its neighbours. Add a second command to cmds that reaches the JVM, or say in a comment that this test covers only the wrapper. jdtls is the only wrapper-launched JVM tool in this batch.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants