docs(worklist): score the RX 580, Windows 860M and Flatpak Steam gaps #1229

Open
TrashPanda wants to merge 5 commits from TrashPanda/Longshot:docs/worklist-device-gaps into main
Contributor

Type: docs · Risk: low · Tested on: none (documentation) · Needs: none

Why

Four device gaps were missing from the work list, or were scored as if nobody had a log. People picking from the top of Open would not see them.

  • An RX 580 (no mesh shaders) freezes to a black screen after a save load, shows artifacts and driver timeouts, speckled graphics in Kokiri Forest (#1204), and crashes when Link moves. The old row was 0.29, confidence 0.4, "no log yet". #973, #717 and #904 are logs from more than one reporter, so confidence is 0.7 and the score is 0.51 (S1 x 0.145 x 0.7). See #973, #717, #904, #1204.
  • A Radeon 860M on Windows loses the device in Kokiri Forest while the GPU decodes two large ASTC arrays. RX_VK_ASTC_CPU=1 avoids it, then the 4,096 MB iGPU budget thrashes until RX_VK_VRAM_MB=6144. One reporter with a log: score 0.38 (S0 x 0.12 x 0.4). See #1040. The open fix #1296 has not run on an 860M yet.
  • Under Flatpak Steam the guest window stays on the slow path because udmabuf's size limit cannot be read. A ROG Ally X ran at 6-10 fps, and about 30 with the memfd path forced on. Score 0.20 (S2 x 0.17 x 0.4). The open fix is #1155. See #1135.
  • The existing Intel row already covers a black screen after the intro. rc.5 adds a B580 and an A770 that close at the start, so those issues are linked on that row rather than as a second one. See #1154, #1129.

What changed

The Open table in docs/worklist.md only: three device-gap rows scored, and the above-60 row brought up to date. No game code.

  • The RX 580 row moves from 0.29 to 0.51 and names the black-screen, timeout and Kokiri Forest artifact reports.
  • A new 0.38 row for the Windows 860M device loss, pointing at #1296.
  • A new 0.20 row for Flatpak Steam on AMD, pointing at #1155.
  • The Intel 0.2 row gains the rc.5 start-up crashes. Its score is unchanged.
  • The above-60 fps row (3.5) names the fixes that already have authors (#78, #1098, #1224) and leaves the jump loop in dialogue after Gohma unclaimed. Folded in from #1254.

The table stays sorted by score.

Measurements

Verdict: nothing to measure. This is a documentation table. No switch, no build, no frame time.

Flags and settings

None.

Testing

Tested: the Open table was checked by score, highest first, after the edit. No test in tests/ reads docs/worklist.md.

python -I -B tests/run_all.py on Linux (g++ 15.2, clang++ 21.1), this branch @ a03e6f51 and main @ 544229a2: 603 passed, 37 failed, 81 skipped on both. The failures are the same 37 checks on both, for example test_linux_preflight.py, test_target_image_pool.py, test_walk.py. Most skips are Vulkan headers missing on this PC.

Not tested: the 37 checks that also fail on main on this PC. No check reads docs/worklist.md.

Must be tested:

Risks

What could break: a person could pick a row whose score uses the reporter shares from 2026-11-10 (RX 580 reach 0.145, APUs 0.12, Linux 0.17, Intel 0.035). Those are the shares already on this page.

Turning it off: revert the commit.

Compatibility: nothing changes. No saves, options, caches or builder tools.

Issues

See #973
See #717
See #904
See #1204
See #1040
See #1296
See #1135
See #1155
See #1154
See #1129

Checklist

Everyone

  • One topic.
  • The title is a Conventional Commit (type(scope): description) of the status line's type.
  • Why gives the reason plainly and links the issues that fit (the tracker searched).
  • Measurements has a verdict, or one line saying why there is nothing to measure.
  • Testing names what was tested, what wasn't, and what must still be tested, where and how.
  • Rebased on current main (current main @ 544229a2 merged in).
  • python -I -B tests/run_all.py passes (any failure that also fails on main is named). The same 37 checks fail on main @ 544229a2 on this PC (Testing).
  • No game files, dumps, build output or generated files.
  • third_party/ untouched (a new folder with a PROVENANCE.md instead).
  • The license is OK: GPL-3.0-or-later, nothing incompatible and nothing decompiled.

When it applies

  • Docs updated.
**Type:** docs · **Risk:** low · **Tested on:** none (documentation) · **Needs:** none ## Why Four device gaps were missing from the work list, or were scored as if nobody had a log. People picking from the top of Open would not see them. - An RX 580 (no mesh shaders) freezes to a black screen after a save load, shows artifacts and driver timeouts, speckled graphics in Kokiri Forest (#1204), and crashes when Link moves. The old row was 0.29, confidence 0.4, "no log yet". #973, #717 and #904 are logs from more than one reporter, so confidence is 0.7 and the score is 0.51 (S1 x 0.145 x 0.7). See #973, #717, #904, #1204. - A Radeon 860M on Windows loses the device in Kokiri Forest while the GPU decodes two large ASTC arrays. `RX_VK_ASTC_CPU=1` avoids it, then the 4,096 MB iGPU budget thrashes until `RX_VK_VRAM_MB=6144`. One reporter with a log: score 0.38 (S0 x 0.12 x 0.4). See #1040. The open fix #1296 has not run on an 860M yet. - Under Flatpak Steam the guest window stays on the slow path because udmabuf's size limit cannot be read. A ROG Ally X ran at 6-10 fps, and about 30 with the memfd path forced on. Score 0.20 (S2 x 0.17 x 0.4). The open fix is #1155. See #1135. - The existing Intel row already covers a black screen after the intro. rc.5 adds a B580 and an A770 that close at the start, so those issues are linked on that row rather than as a second one. See #1154, #1129. ## What changed The Open table in `docs/worklist.md` only: three device-gap rows scored, and the above-60 row brought up to date. No game code. - The RX 580 row moves from 0.29 to 0.51 and names the black-screen, timeout and Kokiri Forest artifact reports. - A new 0.38 row for the Windows 860M device loss, pointing at #1296. - A new 0.20 row for Flatpak Steam on AMD, pointing at #1155. - The Intel 0.2 row gains the rc.5 start-up crashes. Its score is unchanged. - The above-60 fps row (3.5) names the fixes that already have authors (#78, #1098, #1224) and leaves the jump loop in dialogue after Gohma unclaimed. Folded in from #1254. The table stays sorted by score. ## Measurements **Verdict:** nothing to measure. This is a documentation table. No switch, no build, no frame time. ### Flags and settings None. ## Testing **Tested:** the Open table was checked by score, highest first, after the edit. No test in `tests/` reads `docs/worklist.md`. `python -I -B tests/run_all.py` on Linux (g++ 15.2, clang++ 21.1), this branch @ a03e6f51 and main @ 544229a2: 603 passed, 37 failed, 81 skipped on both. The failures are the same 37 checks on both, for example `test_linux_preflight.py`, `test_target_image_pool.py`, `test_walk.py`. Most skips are Vulkan headers missing on this PC. **Not tested:** the 37 checks that also fail on main on this PC. No check reads `docs/worklist.md`. **Must be tested:** - None. A reviewer can read the four rows against #973, #1040, #1135 and #1154. ## Risks **What could break:** a person could pick a row whose score uses the reporter shares from 2026-11-10 (RX 580 reach 0.145, APUs 0.12, Linux 0.17, Intel 0.035). Those are the shares already on this page. **Turning it off:** revert the commit. **Compatibility:** nothing changes. No saves, options, caches or builder tools. ## Issues See #973 See #717 See #904 See #1204 See #1040 See #1296 See #1135 See #1155 See #1154 See #1129 ## Checklist ### Everyone - [x] One topic. - [x] The title is a Conventional Commit (`type(scope): description`) of the status line's type. - [x] Why gives the reason plainly and links the issues that fit (the tracker searched). - [x] Measurements has a verdict, or one line saying why there is nothing to measure. - [x] Testing names what was tested, what wasn't, and what must still be tested, where and how. - [x] Rebased on current `main` (current main @ 544229a2 merged in). - [x] `python -I -B tests/run_all.py` passes (any failure that also fails on `main` is named). The same 37 checks fail on main @ 544229a2 on this PC (Testing). - [x] No game files, dumps, build output or generated files. - [x] `third_party/` untouched (a new folder with a `PROVENANCE.md` instead). - [x] The license is OK: GPL-3.0-or-later, nothing incompatible and nothing decompiled. ### When it applies - [x] Docs updated.
docs(worklist): score the RX 580, Windows 860M and Flatpak Steam gaps
Some checks failed
PR description / Description follows the template (pull_request_target) Successful in 3s
Website / Website on Cloudflare (pull_request_target) Has been skipped
CI / Conventional Commit title (pull_request) Has been cancelled
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been cancelled
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been cancelled
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been cancelled
CI / Workflow lint (actionlint) (pull_request) Has been cancelled
Guide screenshots / Guide images and local UI source review (pull_request) Has been cancelled
Third-party parts / As their notes say (pull_request) Has been cancelled
Third-party parts / The same as upstream (pull_request) Has been cancelled
Website / Website on Cloudflare (pull_request) Has been cancelled
5195e02d00
The RX 580 row was still at confidence 0.4 with no log. More than one reporter now has a log, and the 860M device loss and the Flatpak Steam slow path had no row.
Author
Contributor

Added #1204 (speckled graphics on an RX 580 in Kokiri Forest) to the 0.51 row. No other open pull request was adding these rows. #1114 is a separate 860M change, a short sleep so the APU is not starved for power, and is not the device-loss row.

Added #1204 (speckled graphics on an RX 580 in Kokiri Forest) to the 0.51 row. No other open pull request was adding these rows. #1114 is a separate 860M change, a short sleep so the APU is not starved for power, and is not the device-loss row.
docs(worklist): link the RX 580 Kokiri Forest artifact report
Some checks failed
PR description / Description follows the template (pull_request_target) Successful in 3s
CI / Conventional Commit title (pull_request) Has been cancelled
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been cancelled
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been cancelled
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been cancelled
CI / Workflow lint (actionlint) (pull_request) Has been cancelled
Guide screenshots / Guide images and local UI source review (pull_request) Has been cancelled
Third-party parts / As their notes say (pull_request) Has been cancelled
Third-party parts / The same as upstream (pull_request) Has been cancelled
Website / Website on Cloudflare (pull_request) Has been cancelled
82146a2081
#1204 is the same card class as the 0.51 row and was not in the first pass.
docs(worklist): link the open 860M fix #1296
Some checks failed
PR description / Description follows the template (pull_request_target) Successful in 4s
CI / Conventional Commit title (pull_request) Has been cancelled
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been cancelled
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been cancelled
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been cancelled
CI / Workflow lint (actionlint) (pull_request) Has been cancelled
Guide screenshots / Guide images and local UI source review (pull_request) Has been cancelled
Third-party parts / As their notes say (pull_request) Has been cancelled
Third-party parts / The same as upstream (pull_request) Has been cancelled
Website / Website on Cloudflare (pull_request) Has been cancelled
a03e6f518d
TrashPanda changed title from WIP: docs(worklist): score the RX 580, Windows 860M and Flatpak Steam gaps to docs(worklist): score the RX 580, Windows 860M and Flatpak Steam gaps 2026-11-17 13:58:51 +00:00
docs(worklist): the above-60 row names the fixes that already have authors
Some checks are pending
CI / Conventional Commit title (pull_request) Blocked by required conditions
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Blocked by required conditions
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Blocked by required conditions
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Blocked by required conditions
CI / Workflow lint (actionlint) (pull_request) Blocked by required conditions
Guide screenshots / Guide images and local UI source review (pull_request) Blocked by required conditions
Third-party parts / As their notes say (pull_request) Blocked by required conditions
Third-party parts / The same as upstream (pull_request) Blocked by required conditions
Website / Website on Cloudflare (pull_request) Blocked by required conditions
PR description / Description follows the template (pull_request_target) Successful in 5s
ead1f4d09d
The 3.5 row said nothing covered the Scarecrow's Song, Gohma or the jump
loop. Scarecrow timing landed in #78, Gohma's ceiling circuit in #1098 and
the battle music after a fight in #1224; the row names them and leaves the
jump loop in dialogue after Gohma unclaimed. Folded in from #1254.

Assisted-by: Claude
Author
Contributor

Folded in #1254's one-row change (ead1f4d0): the above-60 row now names #78, #1098 and #1224 as the fixes that already have authors, leaving the jump loop after Gohma unclaimed.

Folded in #1254's one-row change (ead1f4d0): the above-60 row now names #78, #1098 and #1224 as the fixes that already have authors, leaving the jump loop after Gohma unclaimed.
Some checks are pending
CI / Conventional Commit title (pull_request) Blocked by required conditions
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Blocked by required conditions
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Blocked by required conditions
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Blocked by required conditions
CI / Workflow lint (actionlint) (pull_request) Blocked by required conditions
Guide screenshots / Guide images and local UI source review (pull_request) Blocked by required conditions
Third-party parts / As their notes say (pull_request) Blocked by required conditions
Third-party parts / The same as upstream (pull_request) Blocked by required conditions
Website / Website on Cloudflare (pull_request) Blocked by required conditions
PR description / Description follows the template (pull_request_target) Successful in 5s
This pull request doesn't have enough approvals yet. 0 of 1 approvals granted.
Some workflows are waiting to be reviewed.
This branch is out-of-date with the base branch
You are not authorized to merge this pull request.
View command line instructions

Checkout

From your project repository, check out a new branch and test the changes.
git fetch -u 'docs/worklist-device-gaps:TrashPanda-docs/worklist-device-gaps'
git switch 'TrashPanda-docs/worklist-device-gaps'
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
MagicBeanDev/Longshot!1229
No description provided.