10 Commits

Include renames

Author	SHA1	Message	Date
Dean	d123adecf0	Merge pull request #11840 from rooklz/fix/ab-split-draw-program-key ... fix(tools): leave draw variants out of ab_split program keys	2026-11-03 01:27:01 +13:00
Adrià	3e29acc380	Merge pull request #13891 from deepak-r-135/fix/ab-split-phase-chronology ... fix(tools): reject invalid A/B phase chronology	2026-11-01 08:14:09 +01:00
deepak-r-135	515b5f7107	fix: reject invalid A/B phase chronology	2026-10-31 16:51:31 +05:30
dr-hoseyn	9cbfb37a25	fix(tools): exclude phase gaps from A/B noise estimates	2026-10-31 14:44:37 +03:30
rooklz	f2304865a0	fix(tools): leave draw variants out of ab_split program keys	2026-10-30 11:26:31 +00:00
AdriàandClaude Sonnet 5.5	eb45af7f40	fix(tools): reject phase index restarts in A/B logs ... Co-Authored-By: Claude Sonnet 5.5 <[email protected]>	2026-10-30 05:21:55 +01:00
Queuze	dd1c41bbfe	fix(tools): reject mixed A/B capture logs	2026-10-30 05:21:46 +01:00
StoneDC	d2f2e9c38d	fix(tools): reject non-finite in-run A/B statistics	2026-10-29 01:39:34 +08:00
Cankiee	22e89180f6	fix(tools): match phase kind when settling	2026-10-27 18:39:03 +01:00
Celegans12	2801497e0c	feat(tools): split the periodic reports of an APS5_AB run by phase ... An APS5_AB run interleaves phases A and B, and every periodic report covers the time since the previous report of its kind, so its numbers need sorting before A and B can be compared. tools/ab_split.py reads the [ab] lines and counts a [present], [gputime], [draw] pipelined or [syncwait] report for a phase only when its whole span lies inside the phase after its settled line. It prints presents per 10 s and frame time per phase, GPU busy and idle time per frame, GPU time per present (also per --program key and --class), committer, drain and hook-wait times, the A and B means, and each B phase against the mean of its neighbouring A phases, which cancels a linear drift. AI-assisted: yes	2026-10-26 12:26:29 +01:00