llvm-toolchain-22 (1:22.1.8-2) [PTS] [DDPO]

COMMITS: VCS has seen 23 commits since the debian/1%22.1.8-1 tag

Git: https://salsa.debian.org/pkg-llvm-team/llvm-toolchain.git -b 22
JSON
Enter a full Vcs-* header, supported schemes are: Vcs-Bzr, Vcs-Cvs, Vcs-Darcs, Vcs-Git, Vcs-Hg, Vcs-Mtn, Vcs-Svn.
For Vcs-Git append -b branch to select the branch name; append [subdir] if the debian/ directory is not in the repository root.
The next upload will reset the information from the source package headers.
Branch: 22
Path: debian/changelog
Repo size: 2850816
Browser: https://salsa.debian.org/pkg-llvm-team/llvm-toolchain/tree/22
Last scan: 2026-10-06 03:13:09+00
Next scan: 2026-10-12 14:08:00+00
Merge requests: 2

Debian changelog in Git:

llvm-toolchain-22 (1:22.1.8-2) unstable; urgency=medium

  [ Sylvestre Ledru ]
  * d/rules: retry the documentation build serially when sphinx's parallel
    workers die (bare "EOFError" from multiprocessing after a successful
    compile, trixie amd64/arm64 2026-09-02/03)
  * d/rules: set SCCACHE_IDLE_TIMEOUT=0 so the sccache server stays alive for
    the whole build; otherwise it idles out (default 600s) during the quiet
    final-link phase and the libfuzzer step auto-spawns a default-config server
    writing to $HOME/.cache/sccache, which fails with HOME=/nonexistent
  * d/rules: use the shared GCS sccache cache on arm64 too (add arm64 to the
    GCP arch filter)
  * d/rules: Disable sccache on the releases whose glibc is too old for the
    binary the build host provides. /opt/sccache/sccache comes from
    mozilla/sccache's latest release and, on s390x (the only arch without a
    static musl build), needs glibc >= 2.34 since v0.17.0. On focal
    (glibc 2.31) it cannot even start, and the build died at --start-server
    with "version `GLIBC_2.34' not found" in stamps/configure.
  * d/rules: Drop llvm-mt from the install list when it wasn't built.
    Upstream only builds llvm-mt when it detects libxml2 at configure
    time; on Ubuntu focal/arm64 the libxml2 symbol check fails and the
    tool is silently skipped, which made dh_install --fail-missing abort
    on usr/lib/llvm-22/bin/llvm-mt and usr/bin/llvm-mt-22.
  * Add a test for the wasi path detection. Thanks Max Gilmour for the work
    (Closes: #1138702)
  * d/control.in: libfuzzer-X.Y-dev now depends on libclang-rt-X.Y-dev.
    -fsanitize=fuzzer links libclang_rt.fuzzer*.a (and
    libclang_rt.ubsan_standalone.a) from the compiler-rt resource
    directory, which lives in libclang-rt-X.Y-dev. As it was only a
    Recommends of libclang-common-X.Y-dev, it was not pulled into build
    chroots and the link failed. Closes: #1147573.
  * d/debian-llvm-testsuite.bats: don't skip the libFuzzer compilation test
    when the compiler-rt libraries are missing, fail instead.
  * d/libmlir-X.Y.links.in: new, keep a /usr/lib/llvm-X.Y/lib symlink to the
    libMLIR SONAME now installed in /usr/lib/<triplet>. libmlir-X.Y-dev
    ships /usr/lib/llvm-X.Y/lib/libMLIR.so as a relative symlink to it and
    the mlir tools have that directory in their RUNPATH, so both were left
    dangling.
  * d/debian-llvm-testsuite.bats: link the fuzzer runtime from the
    compiler-rt resource directory instead of the removed
    /usr/lib/llvm-X.Y/lib/libFuzzer.a, and check that it is present.
  * d/control.in: document that libfuzzer-X.Y-dev is now a dependency
    package: the fuzzer runtime (libclang_rt.fuzzer*.a) and
    FuzzedDataProvider.h come from the compiler-rt resource directory in
    libclang-rt-X.Y-dev, which is what clang links for -fsanitize=fuzzer.
    This also avoids the empty-binary-package lintian warning.

  [ Talha Can Havadar ]
  * d/p/amdgpu-fix-simulated-trap-continuation-block.diff: cherry-pick
    llvm/llvm-project@9429a1e809fd (PR #174774,
    "[AMDGPU] Fix insertSimulatedTrap to return correct continuation block")
    Fixes a codegen crash on gfx1100/gfx1101/gfx1102 (RDNA3 desktop:
    RX 7900/7800/7700/7600) where clang aborts with
    "fatal error: error in backend: Unsupported instruction : <MCInst NNNN ...>"
    from the AMDGPU AsmPrinter whenever a kernel contains a block that ends in
    llvm.trap() with no successors — an unlowered pseudo reaches the MC layer.
    Notably affects pytorch-rocm 2.12's Loss.hip (FTBFS). Reported upstream as
    llvm/llvm-project#208912. Fix landed on main branch on 2026-01-21 (LLVM 23);
    release/22.x was cut 8 days before and never received it. This cherry-pick
    lets 22.1.x carry the fix without waiting for a 22.1.9 or 22.2.0 point
    release upstream. (Closes: #1141868)

  [ Matthias Klose ]
  * Fix removing RUNPATH, broken in Nov. 2025. Closes: #1142548.
  * Stop building, shipping and testing the obsolete libFuzzer.a.
    Closes: #1147573.
  * Install libmlir shared library in /usr/lib. Closes: #1142778.
  * libmlir: Remove provides, conflicts and replaces. Not needed.
    Closes: #1137244.

 -- Matthias Klose <doko@debian.org>  Sat, 25 Jul 2026 18:43:43 +0200

This branch is 23 commits ahead of tag debian/1%22.1.8-1

Git log:

commit 11ccc521a6affe0548d36f01f35a9b69f62da393
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sat Sep 26 22:16:04 2026 +0200

    prepare upload

commit 9a4825cf64dec027788e32c43cf147e2e8448f5e
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sat Sep 19 11:29:41 2026 +0200

    Align with the 23 branch: libMLIR compat symlink and the fuzzer bats test
    
    * d/libmlir-X.Y.links.in: since 9e8cfc60 installs libMLIR.so.* in
      /usr/lib/<triplet>, keep a symlink under /usr/lib/llvm-X.Y/lib:
      libmlir-X.Y-dev ships libMLIR.so there as a relative symlink to the
      SONAME, and the mlir tools have that directory in their RUNPATH.
    * d/debian-llvm-testsuite.bats: the explicit-linking part of "Test libFuzzer
      functionality" still passed the removed
      /usr/lib/llvm-X.Y/lib/libFuzzer.a; use
      $(clang-X.Y -print-resource-dir)/lib/linux/libclang_rt.fuzzer-<arch>.a.

commit dcf9028c54cd7a21b196b7e0e04b90fcda5eddc3
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sat Sep 19 11:21:36 2026 +0200

    d/control.in: document libfuzzer-X.Y-dev as a dependency package
    
    Since libFuzzer.a is not built anymore, the package ships no files: the
    fuzzer runtime (libclang_rt.fuzzer*.a) and FuzzedDataProvider.h live in the
    compiler-rt resource directory shipped by libclang-rt-X.Y-dev, which is what
    clang links for -fsanitize=fuzzer. Saying so in the long description also
    keeps lintian's empty-binary-package check quiet.

commit 2e20b2ccb96712100aeccbd9ccfa6284c733e220
Author: Matthias Klose <doko@ubuntu.com>
Date:   Mon Sep 14 09:22:13 2026 +0200

      * Stop building, shipping and testing the obsolete libFuzzer.a.
    
    (cherry picked from commit 2dd1b1137662a606395e17cc6043a8ea2bb58ee5)

commit 3e2219f851d7a9a70e4b2843b0a7f1c0bccfa67b
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sun Sep 13 09:49:51 2026 +0200

    d/control.in: libfuzzer-X.Y-dev now depends on libclang-rt-X.Y-dev. -fsanitize=fuzzer links libclang_rt.fuzzer*.a (and libclang_rt.ubsan_standalone.a) from the compiler-rt resource directory, which lives in libclang-rt-X.Y-dev. As it was only a Recommends of libclang-common-X.Y-dev, it was not pulled into build chroots and the link failed. Closes: #1147573.
    
    (cherry picked from commit e35fb1de551488bd0ca4521000bcc694c4805bd8)

commit aa4b1db5ec18cf26a615a68cfba843b972e06e50
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sat Sep 19 10:32:39 2026 +0200

    fix the changelog

commit 35d4510e85f35c6e25f95a3e17d162c2b42613f4
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sat Sep 19 10:32:03 2026 +0200

    remove patch already applied

commit 74ad34414543b9416ce57a7dbbf380a579aadafd
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sat Sep 19 10:01:40 2026 +0200

    Fix a wasi path detection. Thanks Max Gilmour for the patch (Closes: #1138702)

commit dbbe5520a45e6096482c0c9f5fcbf4f23557cf83
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Mon Sep 7 10:49:53 2026 +0200

    d/rules: retry the documentation build serially when sphinx workers die
    
    Every docs-* target runs sphinx-build -j 16 on the 32-core agents and
    ninja starts all six at once; when a forked worker dies the parent only
    reports a bare EOFError from multiprocessing/connection.py (trixie
    amd64/arm64, 2026-09-02/03, always a *-man target). Drop the doctrees,
    reconfigure with LLVM_SPHINX_THREADS=1 and rebuild the docs with ninja
    -j1 before failing, so a real docs error shows its traceback.

commit 2bda99476e808bf363314c770a5a602d9208391d
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Thu Aug 13 23:08:43 2026 +0200

    d/rules: disable sccache on old suites, drop llvm-mt when not built
    
    Two independent focal build failures:
    
    * On focal s390x the /opt/sccache/sccache binary provided by the build host
      needs glibc >= 2.34 (since sccache 0.17.0; s390x is the only arch without a
      static musl build), while focal ships 2.31. The server could not start and
      stamps/configure died with "version `GLIBC_2.34' not found". Follow the
      house *_DISABLE_DISTRO idiom and build those out-of-support suites without
      sccache.
    
    * On focal arm64 upstream did not build llvm-mt: it is only built when the
      libxml2 detection succeeds at configure time, and the symbol check fails
      there, so dh_install --fail-missing aborted on usr/lib/llvm-22/bin/llvm-mt
      and usr/bin/llvm-mt-22. Drop the entries when the binary is absent, as
      branch 23 already does (80feaa5d8de7).

commit 55c1752d0a3500521ae2d0bc596a30ba50356508
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Mon Aug 3 22:03:22 2026 +0200

    d/rules: only enable sccache if the binary actually runs
    
    The s390x sccache release is a glibc build (upstream ships no working musl
    one for that arch), linked against glibc >= 2.34. On focal (2.31) it fails
    to load, so --start-server returned non-zero and the build FTBFS at
    configure. `test -e` was not enough of a check; probe with --version and
    build without sccache when it does not run.

commit e4a509c8953db99c759b612eb2c93c53c340590e
Merge: 7d1679b 4db1872
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Wed Jul 29 13:40:36 2026 +0000

    Merge branch 'mr/update-miscompilation-fix' into '22'
    
    update patch for miscompilation fix
    
    See merge request pkg-llvm-team/llvm-toolchain!226

commit 7d1679b56a05c29162e47e06652766c922347858
Author: Matthias Klose <doko@ubuntu.com>
Date:   Sat Jul 25 20:02:50 2026 +0200

      * libmlir: Remove provides, conflicts and replaces. Not needed.
        Closes: #1137244.

commit 9e8cfc60183533e74d956a8bb7784f9ef3d1c7e7
Author: Matthias Klose <doko@ubuntu.com>
Date:   Sat Jul 25 19:59:34 2026 +0200

      * Install libmlir shared library in /usr/lib. Closes: #1142778.

commit 08ab5e633c807d035fbcb061b412cacca92f69ba
Author: Matthias Klose <doko@ubuntu.com>
Date:   Sat Jul 25 18:48:01 2026 +0200

    collapse changelog entry

commit 1ac3ab7d9991c3018b3ad6e1e8ccce7af3372ab1
Author: Matthias Klose <doko@ubuntu.com>
Date:   Sat Jul 25 18:46:11 2026 +0200

      * Fix removing RUNPATH, broken in Nov. 2025. Closes: #1142548.

commit 4db187225f195daacb2bca41fb38d221d137eeb8
Author: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Date:   Tue Jul 21 13:32:30 2026 +0200

    update patch for miscompilation fix
    
    removing the big test hunk that makes no difference for the package build.
    
    Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>

commit 69e41c30c8cc8fa3cba0bfc58a9a4af600e33695
Merge: 4b66b8b c9673e8
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Fri Jul 24 20:00:51 2026 +0000

    Merge branch 'mr/rustc-1.97-miscompilation' into '22'
    
    cherry-pick fix for miscompilation
    
    See merge request pkg-llvm-team/llvm-toolchain!224

commit c9673e8f6828adfdf496cbd66b69b7fa748c7245
Author: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Date:   Tue Jul 21 13:32:30 2026 +0200

    cherry-pick fix for miscompilation
    
    this is https://github.com/llvm/llvm-project/issues/208611 and
    https://github.com/rust-lang/rust/issues/159035 upstream
    
    Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>

commit 4b66b8b493796a31e1ce4a5101bdfbf000225f71
Merge: fa997e3 4d8cd1b
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Sat Jul 11 21:41:39 2026 +0000

    Merge branch '22' into '22'
    
    #1141868 d/p/amdgpu-fix-simulated-trap-continuation-block.diff: cherry-pick...
    
    See merge request pkg-llvm-team/llvm-toolchain!223

commit 4d8cd1b8b6389f80a91bd15bbed11c2f311eb546
Author: Talha Can Havadar <havadartalha@gmail.com>
Date:   Sat Jul 11 23:02:11 2026 +0200

    d/changelog: update changelog with AMDGPU RDNA3 fix

commit 1707a05759a754f895b9db89a570b39b892d41a8
Author: Talha Can Havadar <havadartalha@gmail.com>
Date:   Sat Jul 11 18:16:55 2026 +0200

    d/p/amdgpu-fix-simulated-trap-continuation-block.diff: cherry-pick llvm/llvm-project@9429a1e809fd (PR #174774, "[AMDGPU] Fix insertSimulatedTrap to return correct continuation block")
    
    Fixes a codegen crash on gfx1100/gfx1101/gfx1102 (RDNA3 desktop:
    RX 7900/7800/7700/7600) where clang aborts with
    "fatal error: error in backend: Unsupported instruction : <MCInst NNNN ...>"
    from the AMDGPU AsmPrinter whenever a kernel contains a block that ends in
    llvm.trap() with no successors — an unlowered pseudo reaches the MC layer.
    Notably affects pytorch-rocm 2.12's Loss.hip (FTBFS). Reported upstream as
    llvm/llvm-project#208912. Fix landed on main branch on 2026-01-21 (LLVM 23);
    release/22.x was cut 8 days before and never received it. This cherry-pick
    lets 22.1.x carry the fix without waiting for a 22.1.9 or 22.2.0 point
    release upstream.
    
    Closes: #1141868
    Gbp-Dch: full

commit fa997e39aca00815da81e28c31dd231eb3c74f5f
Author: Sylvestre Ledru <sylvestre@debian.org>
Date:   Wed Jun 24 12:34:29 2026 +0200

    d/rules: keep sccache server alive (SCCACHE_IDLE_TIMEOUT=0) + use GCS cache on arm64
    
    The libfuzzer step calls sccache directly with no backend env, long after the
    server was started in configure. With the default 600s idle timeout the server
    dies during the sccache-quiet final-link phase, so libfuzzer auto-spawns a
    default-config server that writes to $HOME/.cache/sccache and fails with
    HOME=/nonexistent. Set SCCACHE_IDLE_TIMEOUT=0 on both --start-server calls so
    the configured server (GCS or local) stays up for the whole build. Also add
    arm64 to the GCP arch filter so it uses the shared GCS cache like amd64.