core: preserve truncation in signed integer division

[vera-rubin]

Signed integer division by a power-of-two constant rounded negative quotients down. This keeps Div for signed integers and keeps the shift rewrite for unsigned ones. The regression test and the existing core tests pass.

Fixes #3003

— ChipLab

🍍

[github-actions]

✅ Bench vs main — no regressions

Reference: 2026-10-14 morning nightly run (0d old) · full report → run

Speed — evaltime · prefill · decode

no inference-speed regressions

Improvements

Δ	metric	device	main → PR
🟢 -15.6%	mobilenet_v2_1 evaltime · pass_mt	i9-9900k	13.2 ms → 11.1 ms

---

Bench numbers come from shared CI hardware and do produce false positives: a flagged
row is not on its own evidence of a real regression, and a clean report is not proof
there is none. Contributors do not need to chase what is reported here — leave the
reading of it to the maintainers, who will say if something needs action.

[github-actions]

✅ Bench vs main — no regressions

Reference: 2026-10-14 morning nightly run (0d old) · full report → run

Speed — evaltime · prefill · decode

no inference-speed regressions

Improvements

Δ	metric	device	main → PR
🟢 -15.6%	mobilenet_v2_1 evaltime · pass_mt	i9-9900k	13.2 ms → 11.1 ms

---

Bench numbers come from shared CI hardware and do produce false positives: a flagged
row is not on its own evidence of a real regression, and a clean report is not proof
there is none. Contributors do not need to chase what is reported here — leave the
reading of it to the maintainers, who will say if something needs action.

[kali]

reds are not yours. I'm dealing with them on the side.

Thanks for the PR.