Skip to content

ggml-cpu(s390x): guard VXE-only repack helpers - #28775

Merged
taronaeo merged 1 commit into
ggml-org:masterfrom
bernardladenthin:s390x-guard-vxe-helpers
Sep 13, 2026
Merged

ggml-cpu(s390x): guard VXE-only repack helpers#28775
taronaeo merged 1 commit into
ggml-org:masterfrom
bernardladenthin:s390x-guard-vxe-helpers

Conversation

@bernardladenthin

@bernardladenthin bernardladenthin commented Sep 11, 2026

Copy link
Copy Markdown
Contributor

Overview

On s390x without VXE/VXE2, the repack helpers vxe_dot_acc, vxe_splat_granule and vxe_fold are still compiled even though their vector intrinsics are unavailable, breaking the build.
This wraps the three helpers in the same #if defined(__VXE__) || defined(__VXE2__) guard as their call sites.

Additional information

Reported in #28667, ; @taronaeo confirmed the missing guards there.

Requirements

  • I have read and agree with the contributing guidelines
  • AI usage disclosure: YES - AI-assisted, but manually written, manually reviewed and manually committed. I take full responsibility for the changes.

@taronaeo taronaeo added the merge ready A maintainer can use this label to indicate that they consider the changes final and ready to merge. label Sep 11, 2026
@bernardladenthin bernardladenthin mentioned this pull request Sep 11, 2026
Merged
@github-actions github-actions Bot added the ggml changes relating to the ggml tensor library for machine learning label Sep 11, 2026
This was referenced Sep 13, 2026
Merged
Merged
taronaeo added a commit to taronaeo/llama.cpp-s390x that referenced this pull request Sep 13, 2026
@taronaeo
ggml-cpu: temporarily add ggml-org#28775 patch until its merged
d464525 
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
@taronaeo taronaeo mentioned this pull request Sep 13, 2026
Merged
@taronaeo
taronaeo merged commit 6978052 into ggml-org:master Sep 13, 2026
29 of 30 checks passed
taronaeo added a commit to taronaeo/llama.cpp-s390x that referenced this pull request Sep 13, 2026
@taronaeo
Revert "ggml-cpu: temporarily add ggml-org#28775 patch until its merged"
5da27d1 
This reverts commit d464525.

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
taronaeo added a commit to taronaeo/llama.cpp-s390x that referenced this pull request Sep 13, 2026
@taronaeo
tests: add non-vxe build to tests
390879d 
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

ggml-cpu: add unused macro to fix ci

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

Revert "ggml-cpu: temporarily add ggml-org#28775 patch until its merged"

This reverts commit d464525.

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
@bernardladenthin bernardladenthin mentioned this pull request Sep 13, 2026
Merged
5 tasks done
ggerganov pushed a commit that referenced this pull request Sep 14, 2026
@taronaeo
tests(s390x): add non-vxe build to tests (#28776)
be2c6d7 
* tests: add non-vxe build to tests

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

ggml-cpu: add unused macro to fix ci

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

Revert "ggml-cpu: temporarily add #28775 patch until its merged"

This reverts commit d464525.

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

* ggml-cpu: revert back to upstream/master

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

---------

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
@ggerganov ggerganov mentioned this pull request Sep 14, 2026
Merged
@BrewTestBot BrewTestBot mentioned this pull request Sep 14, 2026
Merged
1 task done
pl752 pushed a commit to pl752/llama.cpp that referenced this pull request Sep 15, 2026
@bernardladenthin @pl752
ggml-cpu(s390x): guard VXE-only repack helpers (ggml-org#28775)
f7b502d
pl752 pushed a commit to pl752/llama.cpp that referenced this pull request Sep 15, 2026
@taronaeo @pl752
tests(s390x): add non-vxe build to tests (ggml-org#28776)
3d8ac1b 
* tests: add non-vxe build to tests

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

ggml-cpu: add unused macro to fix ci

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

Revert "ggml-cpu: temporarily add ggml-org#28775 patch until its merged"

This reverts commit d464525.

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

* ggml-cpu: revert back to upstream/master

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

---------

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
@bernardladenthin bernardladenthin mentioned this pull request Sep 15, 2026
Merged
5 of 6 tasks
vaiju1981 pushed a commit to vaiju1981/java-llama.cpp that referenced this pull request Sep 15, 2026
@claude
feat: upgrade llama.cpp from b10908 to b10909
f87c2dc 
Second and final chunk, reaching the latest upstream release b10909.

One commit (#28164, "metal : single-source fusion table + fusion debug
rework"): 26 files, 111 KiB, and zero files on the priority-ordered
API-compatibility list. The change is confined to the Metal backend plus
upstream's own test and CI scaffolding; the `tests/*` files are applied
but never compiled here, since a FetchContent subproject builds with
`LLAMA_BUILD_TESTS=OFF`.

All ten patches apply at pristine b10909 and all four standing
drop-checks report "still required". `0012`'s `tests/CMakeLists.txt` hunk
was checked against this chunk's edit to the same file and does not
collide.

`0013` was filed upstream during this bump as ggml-org/llama.cpp#28775
and is approved but not yet merged, so it is still required here.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AnNYn8W1xuVxVJtyL34GyH
vaiju1981 pushed a commit to vaiju1981/java-llama.cpp that referenced this pull request Sep 15, 2026
@claude
feat: upgrade llama.cpp from b10938 to b10948, drop patch 0013
bceb5a4 
Upstream merged this project's own PR ggml-org/llama.cpp#28775
("ggml-cpu(s390x): guard VXE-only repack helpers", commit 6978052),
first tagged at b10948, so patches/0013 is now redundant: pristine
b10948 already wraps vxe_dot_acc / vxe_splat_granule / vxe_fold in the
#if defined(__VXE__) || defined(__VXE2__) guard the patch added, and
`git apply` of the patch fails "does not apply" exactly as the
fail-loud applier is designed to signal. Dropped, not refreshed, per
the 0009 precedent at b10280.

The rationale that outlives the patch -- why build-linux-s390x is a
scalar (non-VXE) cross build, and why -DGGML_VXE=ON is the wrong
response to a future VXE compile error -- moves into a "0013 was
dropped at the b10948 bump" note in CLAUDE.md; the s390x job's own
comment in publish.yml now points there.

Verified against pristine b10948: patches 0001-0012 apply clean, 0013
fails; the three standing drop-checks still say "still required"
(0001 no common_params_parse_main in common/arg.h; 0010 vocab_type
still uncast at server-context.cpp:4554; 0012 bare
splits[i] /= split_sum at llama-model.cpp:1491). A fresh
`cmake -B build -DBUILD_TESTING=ON` configures clean through the real
FetchContent path, stamping nine patches at head 5f436dddb (= b10948),
with extraction unchanged at 138 CLI / 57 request / 15 trainer names.
With the real s390x-linux-gnu-g++ cross toolchain, upstream's guarded
repack.cpp compiles clean both with the CI job's scalar flags and with
-mvx -mzvector -march=z15.

The rest of the range touches no file on the API-compatibility review
list: no common/, include/, tools/server/ or tools/mtmd/ changes, so
the three mechanical server-contract greps have no input to compare.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SuJySZsHGCaxFq5Jo5cEtc
quimmedes pushed a commit to quimmedes/cafe-llama.cpp that referenced this pull request Sep 16, 2026
@bernardladenthin @quimmedes
ggml-cpu(s390x): guard VXE-only repack helpers (ggml-org#28775)
93f4c82
quimmedes pushed a commit to quimmedes/cafe-llama.cpp that referenced this pull request Sep 16, 2026
@taronaeo @quimmedes
tests(s390x): add non-vxe build to tests (ggml-org#28776)
2e130a9 
* tests: add non-vxe build to tests

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

ggml-cpu: add unused macro to fix ci

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

Revert "ggml-cpu: temporarily add ggml-org#28775 patch until its merged"

This reverts commit d464525.

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

* ggml-cpu: revert back to upstream/master

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

---------

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
zsogitbe pushed a commit to zsogitbe/llama.cpp that referenced this pull request Sep 17, 2026
@bernardladenthin @zsogitbe
ggml-cpu(s390x): guard VXE-only repack helpers (ggml-org#28775)
b8a0eb3
zsogitbe pushed a commit to zsogitbe/llama.cpp that referenced this pull request Sep 17, 2026
@taronaeo @zsogitbe
tests(s390x): add non-vxe build to tests (ggml-org#28776)
15cad31 
* tests: add non-vxe build to tests

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

ggml-cpu: add unused macro to fix ci

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

Revert "ggml-cpu: temporarily add ggml-org#28775 patch until its merged"

This reverts commit d464525.

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

* ggml-cpu: revert back to upstream/master

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>

---------

Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@CISC CISC CISC approved these changes
@taronaeo taronaeo taronaeo approved these changes
@ggerganov ggerganov Awaiting requested review from ggerganov ggerganov is a code owner

Assignees

No one assigned

Labels

ggml changes relating to the ggml tensor library for machine learning merge ready A maintainer can use this label to indicate that they consider the changes final and ready to merge.

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@bernardladenthin @CISC @taronaeo