Repository navigation
Conversation
…try is no longer stepped over An NDSR pack is a list of 8-byte entries, and ndsr_shaders walked two at a time. In the variable-ratio packs of agl (Ocarina of Time, Switch 2) a 4-byte data entry sits between two programs, so the next program fell on an odd entry and was stepped over. The log had `[RED] shader d17eefaff4f0bc23 is not in the shader map` with the dispatch skipped every frame, and foliage and prop edges were hard and stair-stepped: d17eefaff4f0bc23 is the upscaler's SMAA blending weights (in a pass dump it reads the edges and the edge list that f8965bffbdf51d78 writes and a 192x128 lookup texture, and sets the 1920x1080 image a335c1db6270c355 reads with the scene). The walk now takes single entries and takes only control entries (types 0, 3, 6) as programs. The split of Ocarina of Time finds 3 more shaders: d17eefaff4f0bc23, 1fdc85178b60f6ce and 92918e9b7e2f89a4. With d17eefaff4f0bc23, edges were anti-aliased in the game on the stack this was found on. Donkey Kong Bananza's agl_ndsr.pck gave the same 30 keys before and after. Checked: test_split_ndsr_entries.py (fails 4 subtests on main's reader), test_docs.py, test_nsp_chain.py, the split's other tests. shader_split.py is not part of the chain. Star Fox and Mario Kart World were not re-split.
github-actions
Bot
added
packing
3 tasks done
stan69b
pushed a commit
to stan69b/nxnative
that referenced
this pull request
Nov 5, 2026
lunks' seedds#1353 (5ec8a62): ndsr_shaders walked an NDSR pack two entries at a time, so a program after a 4-byte data entry was stepped over. In Ocarina of Time's variable-ratio packs that left d17eefaff4f0bc23 (the shader the scene asked for that no archive held, "is not in the shader map"), 1fdc85178b60f6ce and one more out of the split; seedds#1353 ties them to the SMAA weights. shader_split.py only, not a chain tool. The reference conflict was two new entries side by side (seedds#1299's NDVE pack entry and seedds#1353's); both kept, symptom index regenerated. test_docs and test_split_ndsr_entries pass. Whether d17eefaff4f0bc23 is now in the map is checked after the next split.
1 task done
gnosticgeek
pushed a commit
to gnosticgeek/nxforge
that referenced
this pull request
Nov 5, 2026
…a of Time's SMAA shader d17eefaff4f0bc23)
2 tasks done
1 task done
gnosticgeek
pushed a commit
to gnosticgeek/nxforge
that referenced
this pull request
Nov 6, 2026
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016qteK29T18U8p1c9LsVY8g
3 tasks done
AnkMister
added a commit
to AnkMister/nxnative
that referenced
this pull request
Nov 6, 2026
…ee NDSR keys and facts from the corrected split. The splitter algorithm remains behind DM
Owner
|
Thank you for tracking this down. The missing upscaler shader is handled on main now, so I'm closing this. Final checks on the rebased tree (full shader build, walks, regression) are still running, so please report anything you see. |
gnosticgeek
pushed a commit
to gnosticgeek/nxforge
that referenced
this pull request
Nov 6, 2026
) replaces the pull request heads Upstream landed Ocarina of Time in main (chain adec7d81d970: "holds 60 fps in Kokiri Forest, 4 GB less") and closed the pull requests it folded in, with reworded and partly different code. Mixing a pull request head with its landed version hunk by hunk silently broke things (two GlobalToStorage tables, a duplicate AF2ip, two mtk_pso definitions), so the rule here is one version per feature: - Pull requests upstream closed (seedds#1299, seedds#1300, seedds#1312, seedds#1343, seedds#1345, seedds#1353, seedds#1354, seedds#1359, seedds#1360, seedds#1373, seedds#1428, seedds#1430, seedds#1434): their hunks are taken out of our side before the three-way merge, so upstream's landed version stands alone. Gone with them: bindless pool table patching (reg_bump), the ghost table's per-submission hold, the growing raw pool table and its write-log sizing, seedds#1428's pool_buf spares (upstream's pool_table_take is the same idea), the ixc-based indirect mesh draw (upstream uses mdx), and their tests (t_pool_table, t_mesh_indirect, t_gmem_raw, finalized_bindless_pool, test_bnsh_nvn2, test_ndve). - Open pull requests that upstream's landing supersedes: seedds#1394 (upstream's mtl_pool_addressed) and seedds#1393 pool_skip_retired (upstream's budgeted pool walk; the profile does not use it) are dropped with their tests and docs. seedds#1368's mesh pipeline prewarm is dropped too: upstream kept the background build in nvn_geometry.inc, and the prewarm's mtk_build / mtk_claim / g_warm_mps went with it. - Kept, as upstream does not have them: NXForge's own features (GPU watchdog, present filters and anti-aliasing, aniso floor, quick settings, pipeline archive compaction, upload breakdown, memory report lines) and the still open pull requests seedds#1304, seedds#1305, seedds#1308, seedds#1320, seedds#1331, seedds#1332, seedds#1355, seedds#1356, seedds#1369, seedds#1376, seedds#1377, seedds#1379, seedds#1382, seedds#1390, seedds#1391, seedds#1392, seedds#1414, seedds#1420, seedds#1436, seedds#1439, seedds#1446, seedds#1453. - Combined where both sides changed the same lines: GW_N 1 << 16 with seedds#1320's wide-write intervals (GW_PAGES = SBW_WPAGES, 128 slots), the CPU upload log with upstream's linear upload route, arch_open_end before the compaction limit, cbn_prog in geo_gs_one (seedds#1382). - Translator: upstream's sources and binaries, plus seedds#1414's InstEmitAmpereMem diff and the LDSM SubgroupBarrier in InstEmitAmpereTensor.cs. The binaries and dmsh_chain.lock are upstream's until the Mac rebuilds them (dotnet publish, tool_lock --update-lock, dmsh_chain --header); dmsh_chain.h is regenerated for now. - RED_ONCE keys: upstream took 153 and 154, so the present anti-aliasing lines move to 156 and 157 and seedds#1420's packaged-entry line to 158 (array 159). - Host: hle_net.c and t_net_eventfd.c are upstream's (its eventfd poll waits out the whole timeout, which our 5d7920c fixed another way). - Kept ours: dm/tools/romfs_image.py and its test, so the 26 GB RomFS image is not packed again. dm/tools tests on Linux: same failures as on 875449e (Linux-only quirks), plus test_nlg4 and test_dmsh_pool_arrays, which fail on upstream main too. Not yet built on the Mac. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016qteK29T18U8p1c9LsVY8g
gnosticgeek
pushed a commit
to gnosticgeek/nxforge
that referenced
this pull request
Nov 6, 2026
seedds#1299's and seedds#1353's heads Both pull requests are closed upstream and landed in main in their own form (nvn2_stage, the NDSR entry walk); the merge had kept the heads' versions of the same code beside them, as those hunks did not conflict. NXForge's own readers (HAL .cmp, YCD) stay. dm/tools tests: no change. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016qteK29T18U8p1c9LsVY8g
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
Ocarina of Time (Switch 2,
0400486024DEC000, update v131072; profile in #1299) logs[RED] shader d17eefaff4f0bc23 is not in the shader mapand skips that dispatch every frame. It is the SMAA blending-weights program of agl's upscaler, so foliage and prop edges are hard and stair-stepped. #1299's Open list has it: "one shader the scene asks for (d17eefaff4f0bc23) is in none of the shader archives".It is in the archive. An NDSR pack is a list of 8-byte entries and
ndsr_shaderswalked two at a time. In the variable-ratio packs (agl_ndsr_vr.pck,agl_ndsr_vr_light.pck) a 4-byte data entry sits between two programs, so the next program falls on an odd entry and was stepped over. The walk now takes single entries and takes only control entries (types 0, 3, 6) as programs. The split of Ocarina's files finds 3 more shaders:d17eefaff4f0bc23,1fdc85178b60f6ceand92918e9b7e2f89a4.Reference entry: Programs on odd entries of an NDSR pack.
Relation to open PRs: #1298 changes
records_fromin the same function (short packs) and will conflict on that line; #1255 hands another game's NDSR entry tondsr_shadersand does not touch the walk. Neither changes what this does.Checked
dm/tools/tests/test_split_ndsr_entries.py(synthetic bytes; 4 subtests fail on main's reader, pass here),test_docs.py,test_nsp_chain.py,test_split_nesting.py,test_split_unreadable.py,test_split_gzip.py,test_split_viper.py,test_split_narc.py.lint_python.py(ruff 0.17.2, pyright 1.1.414): no finding in the changed files; it reports 3 new ones intest_nso_match.py, which this change does not touch.agl_ndsr.pckgives the same 30 keys as before. Star Fox and Mario Kart World were not re-split.shader_split.pyis not a chain tool (DMSH_CHAINunchanged);regress.pywas not run.Docs
docs/reference/shaders.mdandpython3 dm/tools/docs_symptoms.pywas run (the symptom index)