"unchanged" was too strong for the kernel pool walk

[glslang]

Changelog prose only, correcting a claim #183 made about itself once a live kernel could be walked on that pin.

#183 said "The kernel pool walk is not asked at all and is unchanged." The first half is right and the second is not. A kernel session is never asked about commitment, so the walk classifies exactly as before and its coverage is identical — but the walk_vs site #183 made noisy is silent on a kernel too, and now names the chunks it drops.

Measured

ctf-vm, live 26100 over KDNET, 2026-09-24, on 28dd63d:

633,665 chunks walked (444,836 allocated), coverage: partial, 42.3s

Inside its budget, so the conservative path is the one a kernel takes and it costs nothing — committed_run is gated on is_kernel_target before any engine query. The new shape is the largest on that target:

shape	count
VS chunk at # is # bytes and runs # past the committed extent at #; no span is emitted for it	2,768
region # is only committed through # unreadable space extends # bytes	2,617
everything else, 9 shapes	94

Those 2,768 chunks were dropped before #183 as well, and the site already cleared complete; nothing in the answer said which, or how many. So what rises is PoolDiagnostics::emitted, which is the improvement rather than a regression — and anything thresholding on that count will see it.

Not a before/after against windbg-mcp FOLLOWUPS item 100's table. That run was a different build on a differently-loaded guest (783,042 chunks at 88.1% allocated, against 633,665 at 70.2% here), so its 207 and this 2,617 are two readings of two machine states. What says the unchanged shape is unchanged is the third arm of test_a_gap_with_nothing_behind_it_is_not_worth_a_diagnostic — the kernel case exactly, a source that cannot be asked still emitting it — and the conditional #183 added can only make that diagnostic fire less, never more.

The whole live-kernel tier passed on this pin: 10 tests, 0 failed, 1,263s, guest uptime advancing 6.013s across a 4s detach and every patched byte restored.

🤖 Generated with Claude Code

https://claude.ai/code/session_01MUhLUt9rB6zd42Y25btB3h

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Advanced

Run ID: 87ae0179-5049-48ae-8542-dcb49d560fd7

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}