Bump vendored XNNPACK to include qd8_bf16_qb4w subgraph FC - #21400
Open
JakeStevens wants to merge 1 commit into
Open
Bump vendored XNNPACK to include qd8_bf16_qb4w subgraph FC#21400JakeStevens wants to merge 1 commit into
JakeStevens wants to merge 1 commit into
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21400
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New FailureAs of commit 65f9e4a with merge base 1a770b8 ( NEW FAILURE - The following job has failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
Bot
added
the
CLA Signed
JakeStevens
added
module: xnnpack
JakeStevens
force-pushed
the
xnnpack-vendor-bump
branch
from
July 27, 2026 17:20
9eddde7 to
094ceb2
Compare
Bumps backends/xnnpack/third-party/XNNPACK from 1adaa7c to 92a7ad5 (google/XNNPACK master, PR pytorch#10818), adding the qd8_bf16_qb4w fully-connected path to the subgraph layer. Required for bf16 dynamic quant (8da4w) to lower and run via XNNPACK. This XNNPACK revision splits two new source files out of existing translation units, so the ExecuTorch Buck build (which enumerates sources explicitly rather than via CMake) must list them or linking fails with undefined symbols: - src/subgraph/rewrites/cvt_to_fp32.cc (fp16/bf16 -> fp32 fallback rewrite; defines xnn_subgraph_fallback_from_{fp16,bf16}_to_fp32 and xnn_subgraph_alias_fp32_fallback_data) -> added to the 'subgraph' target. - src/xnnpack/init-once.c (defines xnn_init_once_impl / xnn_reset_all_init_guards; in CMake it is part of the xnnpack-hardware-config object lib) -> added to the 'XNNPACK' target next to configs/hardware-config.c. It also adds new AVX512-FP16 micro-kernels (e.g. f16-vapproxgelu) that crash clang-cl 18's codegen on Windows. clang-cl is reported to CMake as Clang (not MSVC), so XNNPACK's own MSVC gate for AVX512FP16 does not apply; disable XNNPACK_ENABLE_AVX512FP16 on Windows in the existing WIN32 block alongside the other options that don't build there.
JakeStevens
force-pushed
the
xnnpack-vendor-bump
branch
from
July 27, 2026 18:00
094ceb2 to
65f9e4a
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Bumps the vendored XNNPACK submodule (
backends/xnnpack/third-party/XNNPACK) from1adaa7cto92a7ad5(currentgoogle/XNNPACKmaster).The key change picked up in this range is google/XNNPACK#10818, which adds the
qd8_bf16_qb4wfully-connected path to the subgraph layer. Before this, XNNPACK's subgraph layer could buildqd8_f32_qb4wandqd8_f16_qb4wdynamic-quant FC subgraphs but had no bf16-scale variant, so a bf16 activation feeding an8da4w(int8-dynamic-activation / int4-weight, blockwise) linear could not be lowered/run through XNNPACK.This bump is a prerequisite for bf16 dynamic-quant (
8da4w) delegation to XNNPACK; the ExecuTorch-side backend changes that consume it are in a follow-up PR.Test plan
google/gemma-3-1b-itwith--dtype bfloat16 --qlinear 8da4w --qembedding 8wviaoptimum-executorch(XNNPACK recipe, custom SDPA + KV cache) — lowering succeeds and the FC nodes delegate asqd8_bf16_qb4w..pteproduces finitebfloat16logits with argmax matching the fp32 and bf16-no-quant baselines.cc @GregoryComer @digantdesai @cbilgin