Skip to content

Docs: split T25 into T22/T24–T26 after PR #107, renumber the post-v1 tasks T22–T31 [skip ci] - #114

Merged
michel2323 merged 2 commits into
mainfrom
ccr-babfc15f-z6psn2
Oct 9, 2026
Merged

michel2323 merged 2 commits into
mainfrom
ccr-babfc15f-z6psn2

Conversation

@michel2323

@michel2323 michel2323 commented Oct 9, 2026 •

Copy link
Copy Markdown
Member

Refs #96, #108, #109, #110, #112, #53.

Merge after #113 (T23). #113 appended the T23 Report inside the region this PR rewrites. #113 is merged and this branch is rebased onto main (deeab78) with the one TASKS.md resolution: the T23 Report from main kept verbatim inside the renumbered T22–T31 layout (headings T22–T31 and External each appear once; the T23 section is identical to main).

Task: owner restructuring between tasks (no task id); docs only.

What was built: TASKS.md post-v1 section rewritten from T22 on. The former T25 (performance pass) carried six deliverables from the #107 prototypes; it becomes T22 ordering chooser on supernodal schedule depth (#108), T24 partitioned-inverse fused solve solve_alg = "algo1" (#82 solve half), T25 split factorization (tiled SYRK, chunked TRSM, split regime-C fronts), T26 segmented fused factorization with dependency counters (#82, #109, #110, #75 remainder, #96 re-measure). T23 unchanged. Hybrid memory → T27, robustness → T28 (with the FP32 finding), non-uniform batch → T29, ND tree export/ordering cache → T30, new T31 for the oneAPI/Metal remainder of T23 (#112, #53), External unchanged. Each new task names the bench prototype it adopts, the measured numbers, the negative results not to re-run, tests and Report asks. PLAN.md, PERFORMANCE.md, bench/README.md: task-id references follow; stale status line and M1 row fixed. STATE.md §8 records the restructuring and the GitHub steps.

Tests: none (documentation only, [skip ci]).

Deviations from PLAN.md / TASKS.md: PLAN.md edited as the owner's re-evaluation between tasks (AGENTS.md allows the owner, not the implementer). The STATE.md §5 recommendation to run the External task before the performance tasks is withdrawn by owner decision.

Follow-up issues opened: none; #96, #108, #109, #110, #112 are labelled triaged; issues #22–#27 renamed, #115–#118 created for T28–T31.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Wzx8WtMEXnAfyzyujdySFh

@michel2323 michel2323 changed the title Docs: split T25 into T22/T24–T26 after PR #107, renumber the post-v1 tasks T22–T30 [skip ci] Docs: split T25 into T22/T24–T26 after PR #107, renumber the post-v1 tasks T22–T31 [skip ci] Oct 9, 2026
michel2323 and others added 2 commits October 9, 2026 14:39
…tasks T22-T30 [skip ci]

The former T25 (performance pass) carried six deliverables from the #107
prototypes, several sessions of work under the one-task-per-session rule.
TASKS.md now has: T22 ordering chooser scoring the supernodal schedule depth
(#108, host only, precondition for the fused solve); T23 backends (unchanged);
T24 partitioned-inverse fused solve, solve_alg = "algo1" (#82 solve half);
T25 split factorization (tiled SYRK, chunked TRSM, split regime-C fronts,
longest-first regime A); T26 segmented fused factorization with dependency
counters, with #109 and #110 inside it plus the #75 remainder and the #96
re-measurement; T27 hybrid memory (was T26); T28 robustness extras (was T27,
with the FP32 finding of #107); T29 non-uniform batch (was T22); T30 ND
partition-tree export and ordering cache (was T24); External unchanged.
Each new task names the bench prototype it adopts, the measured numbers, the
negative results not to re-run, the tests and the Report asks.

PLAN.md, PERFORMANCE.md, bench/README.md: task-id references follow the
renumbering; the stale "empty repository" status line and the M1 row's
ordering sentence are updated. STATE.md: section 8 records the restructuring
and the GitHub steps (issue renames, triage labels, chain resumes with T22).

Refs #96, #108, #109, #110.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Wzx8WtMEXnAfyzyujdySFh
PR #113 delivers the AMDGPU extension and the local-memory cap alone and
opened #112 for the rest of T23. The remainder becomes T31 at the end of the
chain: oneAPI and Metal extensions, the #53 allocation-free :ka fallbacks and
the select_impl order noted in #112. STATE.md section 8 updated.

Refs #112, #53.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Wzx8WtMEXnAfyzyujdySFh
@michel2323
michel2323 force-pushed the ccr-babfc15f-z6psn2 branch from 883e7d5 to 9eb2145 Compare October 9, 2026 14:42
@michel2323
michel2323 merged commit 584e57c into main Oct 9, 2026
michel2323 pushed a commit that referenced this pull request Oct 9, 2026
… note PR #120 under T30 [skip ci]

After #113 (T23) merged, the pipeline chained T24 with the task text that
main still carried before #114 renumbered the tasks (PR #120, closed). The
"Start the next task" step looked for exactly T(NN+1) and stopped when that
id was closed, so T22 -> T24 would have needed a manual label once T22
merges. It now takes the lowest open task issue above the merged one and
applies the same on-hold/external and triage checks. TASKS.md: note under
T30 that the closed PR's branch holds a first draft. STATE.md section 8:
what happened and how the chain resumes (label #22 claude:implement).

Refs #120.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014NfiFN5Ukdcnbh87uff2Dg
michel2323 added a commit that referenced this pull request Oct 9, 2026
… note PR #120 under T30 [skip ci] (#121)

After #113 (T23) merged, the pipeline chained T24 with the task text that
main still carried before #114 renumbered the tasks (PR #120, closed). The
"Start the next task" step looked for exactly T(NN+1) and stopped when that
id was closed, so T22 -> T24 would have needed a manual label once T22
merges. It now takes the lowest open task issue above the merged one and
applies the same on-hold/external and triage checks. TASKS.md: note under
T30 that the closed PR's branch holds a first draft. STATE.md section 8:
what happened and how the chain resumes (label #22 claude:implement).

Refs #120.


Claude-Session: https://claude.ai/code/session_014NfiFN5Ukdcnbh87uff2Dg

Co-authored-by: Claude <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant