Skip to content

Repository navigation

ci: unbreak the Rust quality lane on main #1346

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
justinchuby merged 2 commits into from
Aug 19, 2026
Merged

ci: unbreak the Rust quality lane on main #1346

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9e889e4
style(cuda): rustfmt the split-KV workspace match arm
justinchuby Aug 18, 2026
f4c6767
ci: unbreak the Rust quality lane on main
justinchuby Aug 18, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions .github/scripts/verify_documented_env_vars.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -74,6 +74,9 @@
"NXRT_SQNBIT_PREFILL_MIN": "referenced in prose about a proposed tuning knob",
"ONNX_GENAI_BASE_URL": "client-side variable read by external tooling, not by this workspace",
"ONNX_GENAI_SD_PACKAGE": "referenced in prose about a proposed packaging layout",
"ONNX_GENAI_DECODE_GEMV_PROBE_ROWS": "throwaway ceiling-probe flag in "
"2026-08-18-multirow-gemv-ceiling-probe.md, whose method section states the "
"probe was reverted after measuring and is not shipped",
}

# Uppercase only: the lowercase `onnx_genai_*` names in the docs are Prometheus
Expand Down
10 changes: 4 additions & 6 deletions crates/onnx-runtime-ep-cuda/src/kernels/standard_attention.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2125,12 +2125,10 @@ impl StandardAttentionKernel {
.saturating_mul(q_heads as u64)
.saturating_mul(q_seq as u64);
let split_bytes = match split_cfg {
Some((num_splits, _)) => attention_split_scratch_floats(
num_splits,
total_rows_u,
v_head_size as u64,
)
.saturating_mul(std::mem::size_of::<f32>()),
Some((num_splits, _)) => {
attention_split_scratch_floats(num_splits, total_rows_u, v_head_size as u64)
.saturating_mul(std::mem::size_of::<f32>())
}
None => 0,
};
let mut ws = if capture_workspace_eligible {
Expand Down
Loading