Repository navigation
Ocarina of Time's prewarm list gets 15 more task / mesh pipelines recorded on a MacBook M1 - #2280
Open
JoseHuertasDev wants to merge 1 commit into
Open
JoseHuertasDev wants to merge 1 commit into
JoseHuertasDev wants to merge 1 commit into
Conversation
…m a MacBook M1: model pipelines behind the task shader a5288d0a6c4971fd_b that this Mac's pso_user.tsv holds and pso_manifest.tsv lacks, built at their first draw in the first hours of a new game (Kokiri Forest, the way out of the forest and Hyrule Field among it; the areas were not noted row by row); 179 M rows in all. The rows were recorded on v131072 with host 04811a6 plus PR 1951 (chain 39eb2f781d07); each one's task, mesh and pixel shader key is in shader_map.tsv with that stage, none repeats a row here or in PR 2017, and the key is 456 bytes as the 164 before. Not measured: the stall each caused and the prepare step's added time (the game's files are not on this Mac any more, so no cold first start was run) Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
2 tasks done
Aznracer
pushed a commit
to Aznracer/Mighty-Green-Kokiri-Kid
that referenced
this pull request
Nov 11, 2026
…ecorded in a sweep of 33 areas and in long play on an M3 Max (not measured on main) 166 pipelines behind the task shader a5288d0a6c4971fd_b and 5 behind three others (b43c34b2a11c2142_b, efaa4fa6a80ba719_b, 2d3240242e2ec411_b); 335 M rows in all. A task / mesh pipeline that no list names is compiled at its first draw (Native mesh pipelines compiled at their first draw). These were recorded on v131072 by a fork whose host writes the same rows (the key without the border field, the length of main's 164 M rows) as it built them. One of them, task a5288d0a6c4971fd / mesh 53dc93ba404ae78c / pixel 1ac5c1ed2feed1bc, froze adult Zora's Domain for 824 ms at its first draw on that fork. Checked against main: every row parses with pso_mesh_skey_parse and is byte for byte its normalized form; none is on main already or repeats another (task key and normalized key compared); every task, mesh and pixel key is in shader_map.tsv with that stage; none is among PR seedds#2280's 15 rows. Tests: test_docs, test_game_profile, test_pso_offline, test_nsp_incremental; t_pso_skey and t_shmap_warm (make -C dm/gfx/tests GAME=0400486024DEC000); the Mac host builds. Not run: a game on main, so neither the stalls the rows avoid nor the prepare step's added time is measured (PR seedds#2083's 41 rows added about 2 s on an M4 Pro); iOS. (cherry picked from commit 81b3971)
2 of 3 tasks
UserDerezzed
pushed a commit
to UserDerezzed/nxnative
that referenced
this pull request
Nov 12, 2026
…ecorded in a sweep of 33 areas and in long play on an M3 Max (not measured on main) 166 pipelines behind the task shader a5288d0a6c4971fd_b and 5 behind three others (b43c34b2a11c2142_b, efaa4fa6a80ba719_b, 2d3240242e2ec411_b); 335 M rows in all. A task / mesh pipeline that no list names is compiled at its first draw (Native mesh pipelines compiled at their first draw). These were recorded on v131072 by a fork whose host writes the same rows (the key without the border field, the length of main's 164 M rows) as it built them. One of them, task a5288d0a6c4971fd / mesh 53dc93ba404ae78c / pixel 1ac5c1ed2feed1bc, froze adult Zora's Domain for 824 ms at its first draw on that fork. Checked against main: every row parses with pso_mesh_skey_parse and is byte for byte its normalized form; none is on main already or repeats another (task key and normalized key compared); every task, mesh and pixel key is in shader_map.tsv with that stage; none is among PR seedds#2280's 15 rows. Tests: test_docs, test_game_profile, test_pso_offline, test_nsp_incremental; t_pso_skey and t_shmap_warm (make -C dm/gfx/tests GAME=0400486024DEC000); the Mac host builds. Not run: a game on main, so neither the stalls the rows avoid nor the prepare step's added time is measured (PR seedds#2083's 41 rows added about 2 s on an M4 Pro); iOS.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What and why
On a MacBook M1 (8-core GPU, 16 GB) first looks at new places still stall on pipelines no list names. This Mac's
pso_user.tsvholds 37Mrows (native task / mesh pipelines); 22 are already inpso_manifest.tsvon main, and this appends the other 15 with a header line, as #2083 did. All 15 are model pipelines behind the task shadera5288d0a6c4971fd_b, the same one as 39 of #2083's rows.They were built at their first draw in the first hours of a new game: Kokiri Forest, the way out of the forest and Hyrule Field among it. The areas were not noted row by row.
The same list also has about 3,500
S/R/Crows main lacks. They are left out here: a longer list lengthens a cold first start's prepare step (#2017), and I cannot measure that on this Mac. Happy to send them separately if they are wanted.The game note's count goes from 164 to 179
Mrows.Checked
#line at the end of the file. No row repeats one on main (no duplicate lines in the file), and none is among theMrows of Ocarina of Time prewarm rows recorded on an iPhone 18 Pro Max (from HookLink) #2017 or games/0400486024DEC000: update Ocarina of Time pso_manifest.tsv, shader_seen.tsv and notes #1732.shader_map.tsvwith that stage; each key is 456 bytes, as the 164 rows before.test_docs,test_game_profile,test_pso_offline,test_nsp_incremental,test_nsp_mac_apppass on this branch (the last one needsSDKROOTset to Xcode's SDK on this Mac, on main too).Docs
Mrow count under "Found on this game"); the symptom index does not change🤖 Generated with Claude Code