sycl : fix test-backend-ops CI break && restore Kronecker product FWHT support (#28016) by philip-jingxin · Pull Request #28254 · ggml-org/llama.cpp

philip-jingxin · 2026-09-02T16:07:35Z

Overview

Reintroduces the sycl Kronecker product FWHT support originally merged in #28016 (and reverted in #28184 due to CI build failure).

The failure was due to tests/test-backend-ops.cpp containing an unused variable warning (unused variable 'M'), becoming an error with LLAMA_FATAL_WARNINGS=ON

Additional information

Tested with the following commands

cmake -B build_error   -DGGML_SYCL=ON   -DCMAKE_C_COMPILER=icx   -DCMAKE_CXX_COMPILER=icpx   -DCMAKE_BUILD_TYPE=Debug   -DLLAMA_FATAL_WARNINGS=ON   -DLLAMA_ALL_WARNINGS=ON   -DLLAMA_BUILD_TESTS=ON
cmake --build build_error --target test-backend-ops -j

Requirements

I have read and agree with the contributing guidelines
AI usage disclosure: YES - Gemini was used to generate the build options

…, 768, 12…" (ggml-org#28184) This reverts commit c845263.

Titaniumtown · 2026-09-03T07:02:23Z

Running the workflows this time :p xd

Titaniumtown · 2026-09-03T07:03:51Z

@philip-jingxin can you fix the trailing whitespace at the end of tests/test-backend-ops.cpp. Thanks!

arthw

It's good job!
I test it on Arc770 and it's passed.

Thank you!

arthw · 2026-09-04T03:10:43Z

@philip-jingxin
There is code format issue shown in CI task: EditorConfig Checker / editorconfig (pull_request)Failing after 27s.

Could you fix them?

Thank you!

…backend only

philip-jingxin · 2026-09-04T13:53:49Z

@philip-jingxin There is code format issue shown in CI task: EditorConfig Checker / editorconfig (pull_request)Failing after 27s.

Could you fix them?

Thank you!

Hi @arthw I just fixed the bug. Thanks for reviewing!

philip-jingxin · 2026-09-04T13:57:06Z

@philip-jingxin can you fix the trailing whitespace at the end of tests/test-backend-ops.cpp. Thanks!

@Titaniumtown Sorry for the delay, was quite busy during the day. I've cleaned up the trailing space. Thanks for pointing it out!

Titaniumtown · 2026-09-05T02:36:49Z

Thank you for following through and fixing it!

philip-jingxin · 2026-09-05T14:21:05Z

Thank you for following through and fixing it!

@Titaniumtown Thanks! Happy to contribute!

ggerganov

@Titaniumtown For changes in non-SYCL code like here in test-backend-ops.cpp wait for an additional review from another maintainer before merging.

ggerganov · 2026-09-06T07:18:52Z

+#ifdef GGML_USE_SYCL
+                const bool is_kronecker =
+                    ((n_cols % 12 == 0) && is_pow2(n_cols / 12)) || ((n_cols % 20 == 0) && is_pow2(n_cols / 20));
+#else
+                const bool is_kronecker = false;
+#endif


Special-casing the backends like this is not allowed - the user logic should not depend on what ggml backend has been compiled. Please follow-up with a PR that removes all #ifdef GGML_USE_SYCL directives from test-backend-ops.cpp.

Hi @ggerganov, just followed up in #28490. Removed the test special-casing and implemented Kronecker operations for CPU.

Titaniumtown · 2026-09-06T17:56:58Z

@ggerganov Alright. sorry. I will make sure to wait on you next time.

…T support (ggml-org#28016) (ggml-org#28254) * Reapply "sycl : add Kronecker product FWHT support for sizes 384, 640, 768, 12…" (ggml-org#28184) This reverts commit c845263. * tests : fix unused variable M in test-backend-ops * tests: fix trailing space error and isolate kronecker tests for sycl backend only

…T support (ggml-org#28016) (ggml-org#28254) * Reapply "sycl : add Kronecker product FWHT support for sizes 384, 640, 768, 12…" (ggml-org#28184) This reverts commit 79dc68c. * tests : fix unused variable M in test-backend-ops * tests: fix trailing space error and isolate kronecker tests for sycl backend only

…T support (ggml-org#28016) (ggml-org#28254) * Reapply "sycl : add Kronecker product FWHT support for sizes 384, 640, 768, 12…" (ggml-org#28184) This reverts commit c845263. * tests : fix unused variable M in test-backend-ops * tests: fix trailing space error and isolate kronecker tests for sycl backend only

…gml-org#28254, ggml-org#29243) (#302) * sycl : fix test-backend-ops CI break && restore Kronecker product FWHT support (ggml-org#28016) (ggml-org#28254) * Reapply "sycl : add Kronecker product FWHT support for sizes 384, 640, 768, 12…" (ggml-org#28184) This reverts commit c845263. * tests : fix unused variable M in test-backend-ops * tests: fix trailing space error and isolate kronecker tests for sycl backend only (cherry picked from commit 4d91760) * sycl: FWHT kernels for block widths above 512 (ggml-org#29243) The SYCL FWHT covers 64 to 512 via the standard butterfly network, plus 384/640/768/1280 via the Kronecker/Paley construction added separately in Hadamard hint can produce (1024, 2048, 4096, 8192); those still fall through to the default case and run as a dense GEMM against the materialized rotation tensor, correct but O(n^2) instead of O(n log n). fwht_kernel_wide runs one row per work-group instead of per sub-group, so each work-item keeps N/NT values rather than N/WARP_SIZE. Butterflies below the sub-group width still shuffle; those up to the work-group width go through work-group local memory; the rest stay in registers. Same butterfly and sign convention as the existing narrow kernel. ggml's SYCL backend registration (dpct::dev_mgr) unconditionally requires a GPU-labeled platform to exist and throws before any op-level test can run, so test-backend-ops could not be exercised on this box (a GPU-less pod) even via the CPU device. Verified instead with a standalone harness: the same kernel body run through a real SYCL CPU device (Intel oneAPI DPC++ 2026.1, OpenCL CPU backend), checked against an independent recursive-doubling Hadamard reference, cross-validated by first running the existing unmodified narrow kernel through the identical harness and confirming it passes (rules out a reference-convention bug before trusting a pass on the new code). Random-input results for all four widths, single- and multi-row: N=1024 NT=256 rows=1 max_abs_err=1.7e-07 max_rel_err=4.9e-04 PASS N=2048 NT=256 rows=1 max_abs_err=1.9e-07 max_rel_err=2.0e-04 PASS N=4096 NT=256 rows=1 max_abs_err=2.0e-07 max_rel_err=1.4e-04 PASS N=8192 NT=256 rows=1 max_abs_err=2.5e-07 max_rel_err=3.8e-03 PASS N=1024 NT=256 rows=7 max_abs_err=2.4e-07 max_rel_err=1.0e-03 PASS N=2048 NT=256 rows=5 max_abs_err=3.0e-07 max_rel_err=9.4e-04 PASS N=4096 NT=256 rows=3 max_abs_err=2.7e-07 max_rel_err=1.7e-03 PASS N=8192 NT=256 rows=2 max_abs_err=2.5e-07 max_rel_err=1.9e-03 PASS This covers the kernel algorithm itself; it does not exercise the ggml dispatch/supports_op integration end to end, which needs a real GPU (or a SYCL GPU plugin) to get past backend registration. test-backend-ops build is verified: fwht.cpp recompiles with zero warnings as part of ggml-sycl. (cherry picked from commit c829670) --------- Co-authored-by: Jingxin (Philip) Li <philipaslee@gmail.com>

…T support (ggml-org#28016) (ggml-org#28254) * Reapply "sycl : add Kronecker product FWHT support for sizes 384, 640, 768, 12…" (ggml-org#28184) This reverts commit c845263. * tests : fix unused variable M in test-backend-ops * tests: fix trailing space error and isolate kronecker tests for sycl backend only

philip-jingxin added 2 commits September 2, 2026 23:43

Reapply "sycl : add Kronecker product FWHT support for sizes 384, 640…

1bfab34

…, 768, 12…" (ggml-org#28184) This reverts commit c845263.

tests : fix unused variable M in test-backend-ops

dda553d

philip-jingxin requested review from a team and ggerganov as code owners September 2, 2026 16:07

github-actions Bot added testing Everything test related ggml changes relating to the ggml tensor library for machine learning SYCL https://en.wikipedia.org/wiki/SYCL - GPU programming language labels Sep 2, 2026

philip-jingxin mentioned this pull request Sep 2, 2026

sycl : add Kronecker product FWHT support for sizes 384, 640, 768, 1280 #28016

Merged

arthw approved these changes Sep 4, 2026

View reviewed changes

tests: fix trailing space error and isolate kronecker tests for sycl …

0a84ac6

…backend only

Titaniumtown approved these changes Sep 5, 2026

View reviewed changes

Titaniumtown merged commit 4d91760 into ggml-org:master Sep 5, 2026
23 of 29 checks passed

ggerganov reviewed Sep 6, 2026

View reviewed changes

philip-jingxin mentioned this pull request Sep 6, 2026

ggml-cpu : implement Kronecker and clean up SYCL test special-casing #28490

Open

ggerganov mentioned this pull request Sep 10, 2026

tests : drop SYCL special-casing in test-backend-ops.cpp #28688

Merged

ggerganov mentioned this pull request Sep 14, 2026

ggml : bump version to 0.24.0 ggml-org/ggml#1627

Merged

BrewTestBot mentioned this pull request Sep 14, 2026

ggml 0.24.0 Homebrew/homebrew-core#310890

Merged

1 task done

Lonny154 mentioned this pull request Sep 17, 2026

cuda: add Kronecker FWHT support #29051

Open

bri-prism mentioned this pull request Oct 1, 2026

sycl: FWHT for widths above 512 and Kronecker sizes (port of upstream #28254, #29243) PrismML-Eng/llama.cpp#302

Merged

Navigation Menu

sycl : fix test-backend-ops CI break && restore Kronecker product FWHT support (#28016) - #28254

sycl : fix test-backend-ops CI break && restore Kronecker product FWHT support (#28016)#28254
Titaniumtown merged 3 commits into
ggml-org:masterfrom
philip-jingxin:fix/sycl-fwht-kronecker-ci

philip-jingxin commented Sep 2, 2026

Uh oh!

Titaniumtown commented Sep 3, 2026

Uh oh!

Titaniumtown commented Sep 3, 2026

Uh oh!

arthw left a comment

Uh oh!

arthw commented Sep 4, 2026

Uh oh!

philip-jingxin commented Sep 4, 2026 •

edited

Loading

Uh oh!

philip-jingxin commented Sep 4, 2026

Uh oh!

Titaniumtown commented Sep 5, 2026

Uh oh!

Uh oh!

philip-jingxin commented Sep 5, 2026

Uh oh!

ggerganov left a comment

Uh oh!

ggerganov Sep 6, 2026

Uh oh!

philip-jingxin Sep 6, 2026

Uh oh!

Titaniumtown commented Sep 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

philip-jingxin commented Sep 2, 2026

Overview

Additional information

Requirements

Uh oh!

Titaniumtown commented Sep 3, 2026

Uh oh!

Titaniumtown commented Sep 3, 2026

Uh oh!

arthw left a comment

Choose a reason for hiding this comment

Uh oh!

arthw commented Sep 4, 2026

Uh oh!

philip-jingxin commented Sep 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

philip-jingxin commented Sep 4, 2026

Uh oh!

Titaniumtown commented Sep 5, 2026

Uh oh!

Uh oh!

philip-jingxin commented Sep 5, 2026

Uh oh!

ggerganov left a comment

Choose a reason for hiding this comment

Uh oh!

ggerganov Sep 6, 2026

Choose a reason for hiding this comment

Uh oh!

philip-jingxin Sep 6, 2026

Choose a reason for hiding this comment

Uh oh!

Titaniumtown commented Sep 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

philip-jingxin commented Sep 4, 2026 •

edited

Loading