Skip to content

Bump HRX: llama.cpp 8dafc4a050ba + hrx-system 2ab3114966a4 - #451

Closed
bong-water-water-bong wants to merge 1 commit into
mainfrom
bump-hrx/8dafc4a050ba-2ab3114966a4
Closed

bong-water-water-bong wants to merge 1 commit into
mainfrom
bump-hrx/8dafc4a050ba-2ab3114966a4

Conversation

@bong-water-water-bong

Copy link
Copy Markdown
Collaborator

Moves the HRX submodules along our line. AMD's tested pair (ROCm/ggml-staging-automation) is llama.cpp dbe1d6e1a27e + hrx-system 61c152ac4b40; the workflow merges it into our branches (a conflict fails the run and names the files).

from to
third_party/llama.cpp (1bit-MONSTER/llama.cpp 1bit/hrx-vulkan-patched) 8d9ce2c1aa60 8dafc4a050ba
third_party/hrx-system (1bit-MONSTER/hrx-system 1bit/main) 2ab3114966a4 2ab3114966a4

hrx-system HELD at 2ab3114966a4 by config/hrx-system-hold.config (AMD's tested pin is 61c152ac4b40).

llama.cpp commits new to the engine:

CI here builds without HRX. Before merging, on Strix Halo run the long-prompt correctness
probe and test-backend-ops -b HRX0; record the exact validated hrx-system pin in
config/hrx-long-prompt-validated-pin.config:
python3 tests/hrx_long_prompt.py
cmake -B build -G Ninja -DONEBIT_HRX=ON && cmake --build build --target onebit && for d in hrx cpu; do tests/serve_e2e.sh build/1bit <Qwen3-0.6B Q4_K_M .gguf> $d; done

@github-actions

Copy link
Copy Markdown

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

🎫 Ticket compliance analysis 🔶

422 - Partially compliant

Compliant requirements:

  • The test is now buildable as the include statement has been removed

Non-compliant requirements:

  • The test body still references the deleted symbol ggml::hrx::common_symmetric_i4_lowrow_mul_mat_eligible and the registration suite.host_case("symmetric_i4_consumer_qualification", ...)

Requires further human verification:

  • The test body changes need to be verified to ensure the functionality is properly replaced or removed

104 - Partially compliant

Compliant requirements:

  • The llama.cpp submodule has been updated to 8dafc4a050ba
  • The registry now reflects the new llama.cpp commit

Non-compliant requirements:

  • The PR description mentions that the test needs to be updated to remove references to the deleted symbol, but this is not done in the diff

Requires further human verification:

  • The removal of the 32768 context default for --device hrx needs to be verified in the codebase
  • The documentation updates need to be verified
  • The full context flash attention behavior needs to be confirmed
⏱️ Estimated effort to review: 3 🔵🔵🔵⚪⚪
🧪 No relevant tests
🔒 No security concerns identified
⚡ Recommended focus areas for review

Missing Test Fix
The test file tests/hrx-backend-test.cpp still references the deleted symbol ggml::hrx::common_symmetric_i4_lowrow_mul_mat_eligible and the registration suite.host_case("symmetric_i4_consumer_qualification", ...). These references need to be removed or replaced to ensure the test builds and runs correctly.

@bong-water-water-bong

Copy link
Copy Markdown
Collaborator Author

Superseded by #453, which pins llama.cpp 7227a7711ba0, one commit ahead of 8dafc4a050ba (fork #105). The hw_gate failure here is the missing KLD base logits on strixhalo plus the 27B server killed at 16:17 when the #315 soak started. It gets re-run after the soak.

@bong-water-water-bong
bong-water-water-bong deleted the bump-hrx/8dafc4a050ba-2ab3114966a4 branch October 23, 2026 11:04
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant