Skip to content

Update meson to 1.11.2 - #435

Merged
twitchyliquid64 merged 2 commits into
mainfrom
update-meson-1.11.2
Jul 21, 2026
Merged

Update meson to 1.11.2#435
twitchyliquid64 merged 2 commits into
mainfrom
update-meson-1.11.2

Conversation

@gominimal-pkgmgr-mgr

Copy link
Copy Markdown
Contributor

Update meson 1.10.21.11.2

Source: github:mesonbuild/meson
Release: https://github.com/mesonbuild/meson/releases/tag/1.11.2
Changelog: mesonbuild/meson@1.10.2...1.11.2
Released: 5 days ago (2026-07-11)

Pkgscan: clean — diff against the prior version surfaced no newly-introduced suspicious patterns.

Components changed

CycloneDX component delta (declared materials — the package's own version, not a dependency-tree diff)
Component Old New
~ meson 1.10.2 1.11.2
~ meson-upstream 1.10.2 1.11.2

Changes

Old New
Version 1.10.2 1.11.2
SHA256 4f3c6fe1d163fb6d... 09cc2faedc61262f...
Size 5.0 MB 5.0 MB
Source gs://minimal-staging-archives/mesonbuild/meson/1.10.2.tar.gz gs://minimal-staging-archives/mesonbuild/meson/1.11.2.tar.gz
  • License: Apache-2.0 (source: GitHub + tarball)

Quality suggestions

  • Missing tests block. This package has no standalone tests, so the buildbot will only verify compilation — not functional correctness. Consider adding a minimal smoke test (e.g., a --version or small round-trip invocation) as part of this PR so future bumps catch regressions. See packages/python/build.ncl for a simple example.

Created by pkgmgr

@bryan-minimal

Copy link
Copy Markdown
Member

Moving back to draft — this meson bump is blocked on a FoundationDB / gcc-15 incompatibility that isn't meson's fault.

What happens

meson 1.10.2 → 1.11.2 looks trivial, but 22 packages build-depend on meson (glib among them), so bumping it busts the content-addressed cache of that whole closure and forces rebuilds — including FoundationDB (pulled in transitively via glib). FDB then FTBFS'd after a ~65-minute build.

Why FDB fails (not meson)

  • FoundationDB officially targets GCC 13 (its build docs use gcc-toolset-13 or Clang). Minimal ships gcc 15.2.0 — ~2 major compiler versions ahead of what FDB supports.
  • FDB 7.3.69 was last verified here long ago (foundationdb: fix test hang #131), on an older gcc. Since then gcc moved to 15.2, but FDB's cached artifact masked the incompatibility — nothing rebuilt it. This meson bump busted that cache and re-exposed a latent break. Nothing about meson or FDB changed; the compiler moved out from under a package already on borrowed time.
  • Same gcc-15.2 that broke abseil in the Update stacked: 4 packages #375 stacked-bundle incident. gcc-15's stricter C++20 conformance (CWG DR2237 + tighter header transitivity) FTBFS's a lot of old C++.

Note: the template-id not allowed for constructor lines in the buildbot log are warnings, not the failure (FDB isn't -Werror here) — the actual hard error: is a separate, truncated line (likely a gcc-15 missing-#include).

Path forward (separate from this PR)

FDB-vs-gcc15 is its own task, not a one-liner. Options: patch FDB for gcc-15 (add missing includes / small source fix), build it with its documented clang + libc++ path, bump FDB to a newer version if one supports gcc-15, or drop FDB if it's not worth chasing. Tracking that separately.

This PR stays draft until FDB's gcc-15 story is resolved (or FDB is decoupled), so a routine meson bump doesn't detonate a 65-min FDB rebuild. Meta: meson is effectively base-soup-class (22 reverse-deps) and its bumps should be build-gated (#480) rather than blindly PR'd.

@bryan-minimal
bryan-minimal marked this pull request as draft July 16, 2026 23:28
…aks build)

The meson 1.11.2 bump cascades (meson -> glib -> ...) and forces a
foundationdb rebuild, which exposed a pre-existing breakage that the
build cache had been masking. meson itself is fine.

foundationdb's bindings/CMakeLists.txt guards the binding tester with
only:

  if(NOT WIN32 AND NOT OPEN_FOR_IDE)
    package_bindingtester()
    package_bindingtester2()
  endif()

so on Linux it runs UNCONDITIONALLY -- it honours neither the
WITH_*_BINDING flags nor BUILD_TESTING=OFF, both of which we already set.
Its prepare_binding_test_files step then fails:

  FAILED: [code=1] bindingtester.touch
  Error copying directory from "/build/bindings" to
    "/build/build/bindingtester/tests": No such file or directory

Comment both calls out, mirroring how this build.sh already disables
flowbench. The existing Python stub-file workaround is left in place --
it was fighting this same target, and removing it is unverified.

Diagnosed from the full on-disk build log at
/var/lib/buildbot/logs/<run_id>/ -- the bounded error_message in the
run.failed event is only an ~8KB window and cut off the FAILED: line.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@bryan-minimal
bryan-minimal marked this pull request as ready for review July 20, 2026 21:17
@twitchyliquid64
twitchyliquid64 added this pull request to the merge queue Jul 20, 2026
@github-merge-queue
github-merge-queue Bot removed this pull request from the merge queue due to no response for status checks Jul 20, 2026
@twitchyliquid64
twitchyliquid64 added this pull request to the merge queue Jul 20, 2026
@github-merge-queue
github-merge-queue Bot removed this pull request from the merge queue due to failed status checks Jul 20, 2026
@twitchyliquid64
twitchyliquid64 added this pull request to the merge queue Jul 21, 2026
@github-merge-queue
github-merge-queue Bot removed this pull request from the merge queue due to no response for status checks Jul 21, 2026
@twitchyliquid64
twitchyliquid64 added this pull request to the merge queue Jul 21, 2026
Merged via the queue into main with commit 03b449b Jul 21, 2026
9 checks passed
@twitchyliquid64
twitchyliquid64 deleted the update-meson-1.11.2 branch July 21, 2026 03:46
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