WIP: fix(linux): RX_VK_NO_DMABUF leaves the guest window anonymous memory, so its pools can be host-pointer imports #2100

Draft
dyst wants to merge 2 commits from dyst/Longshot:fix/no-dmabuf-window into main
Member

Summary

On Linux, RX_VK_NO_DMABUF=1 is meant to hand the GPU guest memory as host pointers instead of dma-bufs. On RADV it
gave no imports at all: every pool was copied, and the game crawled after a save loaded.

The cause is in reserve_guest (host/hos_main.c). It makes the guest window a memfd's shared mapping whenever
/dev/udmabuf opens, whatever RX_VK_NO_DMABUF says. With that switch the renderer makes no dma-bufs
(gpu/vk/vk_memory.cpp import_dma_buf) and asks for host-pointer imports instead. RADV refuses a shared mapping's
pages as host pointers (VK_ERROR_INVALID_EXTERNAL_HANDLE, as reserve_guest's own comment says), so nothing was
imported.

Now, with RX_VK_NO_DMABUF on, the window is plain anonymous memory, and the log says so:
mem: RX_VK_NO_DMABUF: the guest window is anonymous memory, for host-pointer imports.

  • The switch is read with env_on, the rule the renderer reads it by, so RX_VK_NO_DMABUF=0 keeps the memfd window and
    the dma-bufs.
  • HOS_GUEST_MEMFD still decides when it is set.
  • With neither set, nothing changes.
  • longshot_env.txt is applied before the window is made (main calls load_longshot_env before game_main), so the
    switch works from there too.

To see it on main (Linux, RADV, /dev/udmabuf usable): start with RX_VK_NO_DMABUF=1 and load a save. The log shows
pools not imported and no live imports, and the game runs at a few fps for a while after the save loads.

Found by Mooseca in HeyListen, a fork of Longshot's history:
https://github.com/Sacrappy/HeyListen/pull/75. There, RX 9070 XT on RADV 26.2.6: 170 pools refused and 0 imported with
the switch, 2-5 fps for ~15 s after the save loads; with their fix, 0 refused and about 2 GB imported. Their fix tests
getenv() != NULL, so RX_VK_NO_DMABUF=0 would also have turned the memfd off; this one takes the renderer's rule.

Related: #1164 falls back to host pointers when udmabuf rejects
a pool. Over a memfd window that fallback meets the same RADV refusal, so the two touch the same case from two sides.
This PR does not change vk_memory.cpp.

Read first: reserve_guest in host/hos_main.c, then tests/test_no_dmabuf_window.py.

Testing

Tested:

  • The game-free check tests/test_no_dmabuf_window.py, on Windows (it cuts the condition out and runs everywhere).
    • It runs 13 cases: udmabuf opened or not; HOS_GUEST_MEMFD unset, on or off; RX_VK_NO_DMABUF unset, on (1,
      yes, on) or a word that says off (0, off, empty).
    • It fails on main for the bug: with the switch at 1, yes or on the window is still a memfd.
    • It passes here with gcc and clang (llvm-mingw).
    • With the fix's key line undone it fails, and with getenv() != NULL in its place (HeyListen's reading) it fails
      on 0, off and empty.
  • Every check that reads the changed files (tests/run_all.py --changed), on Windows: 42 pass, 8 are skipped on
    Windows. 10 fail, and they fail the same way with main's host/hos_main.c: test_crash_record,
    test_host_guest_writes, test_profiling_mode, test_slow_frames, test_display_startup_failure,
    test_mods_list, test_vibration_menu, test_crash_dialog, test_perf_report, test_lift_cache.

Not run in the game: the change is in the Linux-only branch of reserve_guest, and my PC is Windows with an RTX 3080.
Not compiled on Linux here either; CI's Linux build does that.

Still to test:

  • Linux, RADV (radv), /dev/udmabuf usable: each build with RX_VK_NO_DMABUF=1 and with nothing set. Load a save,
    Hyrule Field then the Market. Expected:

    • main with the switch: pools not imported and a few fps for a while after the load;
    • here with the switch: the new log line, the pools imported as host pointers, play as without the switch;
    • with nothing set: both builds the same, the window a memfd's and guest memory as dma-bufs.

    (@Beanstalk)

  • Linux, NVIDIA (linux+nvidia), the same with RX_VK_NO_DMABUF=1: host-pointer imports on anonymous memory, as
    HOS_GUEST_MEMFD=0 gives today (@Beanstalk, or by hand).

Measurements

A fix to a switch's start-up; the default is unchanged, so there is no speed to measure on main's defaults. The
in-game evidence (the imports and the frame times after a save loads, with the switch) comes from the run above.

Before: main @ 936d4e009 After: this PR @ 2f236c54c
tests/test_no_dmabuf_window.py fails: 3 of 13 cases (RX_VK_NO_DMABUF=1, yes, on: memfd window) passes, 13 of 13
Linux RADV in the game, RX_VK_NO_DMABUF=1 not run yet not run yet

Risks

Only runs with RX_VK_NO_DMABUF on change. Their guest window is anonymous memory, which is what HOS_GUEST_MEMFD=0
already gives. Those runs lose the dma-buf path, which is what the switch asks for, so on AMD their submissions are
slower than with dma-bufs; anyone who wants the memfd anyway sets HOS_GUEST_MEMFD=1. A revert is enough. Nothing
changes after updating for anyone who doesn't set the switch.

Checklist

Ready for review

  • One topic, on current main; the title is a Conventional Commit.
  • Tested on my own PC, in the game for a change the game runs, as Testing says. (Linux-only code; my PC is Windows: the farm run above)
  • mise run test passes (any failure that also fails on main is named). (only the checks that read the changed files were run: 10 failures, all also on main, named above)
  • Measured against main, or Measurements says why there is nothing to measure.
  • No game files, build output or hand-edited generated files; third_party/ untouched; nothing GPL-incompatible.

When it applies

  • A fix: a game-free test that fails on main and passes here.
  • Host code builds on Windows (llvm-mingw) and Linux, with UTF-8 paths. (the changed lines are Linux-only; not compiled on Linux here: CI)
  • AI assistance recorded with Assisted-by: / Co-Authored-By: trailers.

Ready to merge

  • Everything in Still to test has run, its results in Tested.
  • Tier 1 hardware it can affect is tested (docs/dev/standards.md, "Hardware tiers").
  • Approved in review, every comment answered.
  • CI passes on the head; on current main, measured again or a comment says why the numbers still hold.

🤖 Generated with Claude Code

## Summary On Linux, `RX_VK_NO_DMABUF=1` is meant to hand the GPU guest memory as host pointers instead of dma-bufs. On RADV it gave no imports at all: every pool was copied, and the game crawled after a save loaded. The cause is in `reserve_guest` (`host/hos_main.c`). It makes the guest window a memfd's shared mapping whenever `/dev/udmabuf` opens, whatever `RX_VK_NO_DMABUF` says. With that switch the renderer makes no dma-bufs (`gpu/vk/vk_memory.cpp` `import_dma_buf`) and asks for host-pointer imports instead. RADV refuses a shared mapping's pages as host pointers (`VK_ERROR_INVALID_EXTERNAL_HANDLE`, as `reserve_guest`'s own comment says), so nothing was imported. Now, with `RX_VK_NO_DMABUF` on, the window is plain anonymous memory, and the log says so: `mem: RX_VK_NO_DMABUF: the guest window is anonymous memory, for host-pointer imports`. - The switch is read with `env_on`, the rule the renderer reads it by, so `RX_VK_NO_DMABUF=0` keeps the memfd window and the dma-bufs. - `HOS_GUEST_MEMFD` still decides when it is set. - With neither set, nothing changes. - `longshot_env.txt` is applied before the window is made (`main` calls `load_longshot_env` before `game_main`), so the switch works from there too. **To see it on main** (Linux, RADV, `/dev/udmabuf` usable): start with `RX_VK_NO_DMABUF=1` and load a save. The log shows pools `not imported` and no live imports, and the game runs at a few fps for a while after the save loads. Found by Mooseca in HeyListen, a fork of Longshot's history: https://github.com/Sacrappy/HeyListen/pull/75. There, RX 9070 XT on RADV 26.2.6: 170 pools refused and 0 imported with the switch, 2-5 fps for ~15 s after the save loads; with their fix, 0 refused and about 2 GB imported. Their fix tests `getenv() != NULL`, so `RX_VK_NO_DMABUF=0` would also have turned the memfd off; this one takes the renderer's rule. Related: https://git.magicbean.dev/MagicBeanDev/Longshot/pulls/1164 falls back to host pointers when udmabuf rejects a pool. Over a memfd window that fallback meets the same RADV refusal, so the two touch the same case from two sides. This PR does not change `vk_memory.cpp`. Read first: `reserve_guest` in `host/hos_main.c`, then `tests/test_no_dmabuf_window.py`. ## Testing **Tested:** - The game-free check `tests/test_no_dmabuf_window.py`, on Windows (it cuts the condition out and runs everywhere). - It runs 13 cases: udmabuf opened or not; `HOS_GUEST_MEMFD` unset, on or off; `RX_VK_NO_DMABUF` unset, on (`1`, `yes`, `on`) or a word that says off (`0`, `off`, empty). - It fails on `main` for the bug: with the switch at `1`, `yes` or `on` the window is still a memfd. - It passes here with gcc and clang (llvm-mingw). - With the fix's key line undone it fails, and with `getenv() != NULL` in its place (HeyListen's reading) it fails on `0`, `off` and empty. - Every check that reads the changed files (`tests/run_all.py --changed`), on Windows: 42 pass, 8 are skipped on Windows. 10 fail, and they fail the same way with `main`'s `host/hos_main.c`: `test_crash_record`, `test_host_guest_writes`, `test_profiling_mode`, `test_slow_frames`, `test_display_startup_failure`, `test_mods_list`, `test_vibration_menu`, `test_crash_dialog`, `test_perf_report`, `test_lift_cache`. Not run in the game: the change is in the Linux-only branch of `reserve_guest`, and my PC is Windows with an RTX 3080. Not compiled on Linux here either; CI's Linux build does that. **Still to test:** - Linux, RADV (`radv`), `/dev/udmabuf` usable: each build with `RX_VK_NO_DMABUF=1` and with nothing set. Load a save, Hyrule Field then the Market. Expected: - `main` with the switch: pools `not imported` and a few fps for a while after the load; - here with the switch: the new log line, the pools imported as host pointers, play as without the switch; - with nothing set: both builds the same, the window a memfd's and guest memory as dma-bufs. (@Beanstalk) - Linux, NVIDIA (`linux+nvidia`), the same with `RX_VK_NO_DMABUF=1`: host-pointer imports on anonymous memory, as `HOS_GUEST_MEMFD=0` gives today (@Beanstalk, or by hand). ## Measurements A fix to a switch's start-up; the default is unchanged, so there is no speed to measure on `main`'s defaults. The in-game evidence (the imports and the frame times after a save loads, with the switch) comes from the run above. | | Before: main @ 936d4e009 | After: this PR @ 2f236c54c | |---|---|---| | `tests/test_no_dmabuf_window.py` | fails: 3 of 13 cases (`RX_VK_NO_DMABUF=1`, `yes`, `on`: memfd window) | passes, 13 of 13 | | Linux RADV in the game, `RX_VK_NO_DMABUF=1` | not run yet | not run yet | ## Risks Only runs with `RX_VK_NO_DMABUF` on change. Their guest window is anonymous memory, which is what `HOS_GUEST_MEMFD=0` already gives. Those runs lose the dma-buf path, which is what the switch asks for, so on AMD their submissions are slower than with dma-bufs; anyone who wants the memfd anyway sets `HOS_GUEST_MEMFD=1`. A revert is enough. Nothing changes after updating for anyone who doesn't set the switch. ## Checklist ### Ready for review - [x] One topic, on current `main`; the title is a Conventional Commit. - [ ] Tested on my own PC, in the game for a change the game runs, as Testing says. (Linux-only code; my PC is Windows: the farm run above) - [ ] `mise run test` passes (any failure that also fails on `main` is named). (only the checks that read the changed files were run: 10 failures, all also on `main`, named above) - [x] Measured against `main`, or Measurements says why there is nothing to measure. - [x] No game files, build output or hand-edited generated files; `third_party/` untouched; nothing GPL-incompatible. ### When it applies - [x] A fix: a game-free test that fails on `main` and passes here. - [ ] Host code builds on Windows (llvm-mingw) and Linux, with UTF-8 paths. (the changed lines are Linux-only; not compiled on Linux here: CI) - [x] AI assistance recorded with `Assisted-by:` / `Co-Authored-By:` trailers. ### Ready to merge - [ ] Everything in Still to test has run, its results in Tested. - [ ] Tier 1 hardware it can affect is tested (docs/dev/standards.md, "Hardware tiers"). - [ ] Approved in review, every comment answered. - [ ] CI passes on the head; on current `main`, measured again or a comment says why the numbers still hold. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
reserve_guest (host/hos_main.c) makes the guest window a memfd's shared mapping wherever /dev/udmabuf opens, whatever
RX_VK_NO_DMABUF says. With that switch the renderer makes no dma-bufs and hands the GPU host pointers, and RADV refuses
a shared mapping's pages as host-pointer imports: no pool is imported. The check cuts the condition that chooses the
memfd window and runs it with longshot_env.h's rule for each case. It fails here: with RX_VK_NO_DMABUF=1, yes or on
the window is still a memfd.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
fix(linux): RX_VK_NO_DMABUF leaves the guest window anonymous memory, so its pools can be host-pointer imports
Some checks failed
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been skipped
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been skipped
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been skipped
Third-party parts / The same as upstream (pull_request) Has been skipped
Website / Website on Cloudflare (pull_request_target) Has been skipped
CI / Conventional Commit title (pull_request) Successful in 21s
PR description / Description follows the template (pull_request_target) Successful in 27s
Guide screenshots / Guide images and local UI source review (pull_request) Failing after 31s
CI / Workflow lint (actionlint) (pull_request) Successful in 36s
Third-party parts / As their notes say (pull_request) Successful in 38s
beanstalk/approval can't be tested as asked: see the farm's comment
CI / Website build and checks (pull_request) Has been cancelled
2f236c54c0
The switch says host pointers only, and the renderer makes no dma-bufs with it (vk_memory.cpp import_dma_buf), but
reserve_guest still made the window a memfd's shared mapping wherever /dev/udmabuf opened. RADV refuses a shared
mapping's pages as host-pointer imports (VK_ERROR_INVALID_EXTERNAL_HANDLE), so no pool was imported and every one was
copied: 2-5 fps for ~15 s after a save loads on an RX 9070 XT, by HeyListen's measurement of the same code.

Now the window is anonymous memory when RX_VK_NO_DMABUF is on, read by env_on as the renderer reads it, and the log says
so. HOS_GUEST_MEMFD still decides when it is set. The default, with neither set, is unchanged.

Found by Mooseca in HeyListen: https://github.com/Sacrappy/HeyListen/pull/75. Their fix tests getenv() != NULL; this
one takes the renderer's rule, so RX_VK_NO_DMABUF=0 keeps the memfd window and the dma-bufs.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
Author
Member

@Beanstalk only RADV on Linux, with /dev/udmabuf usable: this changes only what RX_VK_NO_DMABUF=1 does at start-up. Run main and this PR each with RX_VK_NO_DMABUF=1 set, and each with nothing set. Load a save in Hyrule Field, stand a minute, then go to the Market. With the switch, main should log pools 'not imported' and run at a few fps for a while after the load; this PR should log 'mem: RX_VK_NO_DMABUF: the guest window is anonymous memory, for host-pointer imports', import its pools and play as it does without the switch. With nothing set the two builds should match (the window a memfd's, guest memory as dma-bufs). Report the import lines ('not imported', 'imports: ... live'), the frame times in the first 30 s after the load, and the 'vk: guest memory as dma-bufs' line of each run.

@Beanstalk only RADV on Linux, with /dev/udmabuf usable: this changes only what RX_VK_NO_DMABUF=1 does at start-up. Run main and this PR each with RX_VK_NO_DMABUF=1 set, and each with nothing set. Load a save in Hyrule Field, stand a minute, then go to the Market. With the switch, main should log pools 'not imported' and run at a few fps for a while after the load; this PR should log 'mem: RX_VK_NO_DMABUF: the guest window is anonymous memory, for host-pointer imports', import its pools and play as it does without the switch. With nothing set the two builds should match (the window a memfd's, guest memory as dma-bufs). Report the import lines ('not imported', 'imports: ... live'), the frame times in the first 30 s after the load, and the 'vk: guest memory as dma-bufs' line of each run.
Member

Possibly related, for context:

  • PR #1164 WIP: fix(vk): fall back to host pointers on udmabuf rejection and retry mirror allocation — Open fallback PR touches same udmabuf/host-pointer import path; same RADV limitation
Possibly related, for context: - PR #1164 WIP: fix(vk): fall back to host pointers on udmabuf rejection and retry mirror allocation — Open fallback PR touches same udmabuf/host-pointer import path; same RADV limitation
dyst force-pushed fix/no-dmabuf-window from 2f236c54c0
Some checks failed
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been skipped
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been skipped
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been skipped
Third-party parts / The same as upstream (pull_request) Has been skipped
Website / Website on Cloudflare (pull_request_target) Has been skipped
CI / Conventional Commit title (pull_request) Successful in 21s
PR description / Description follows the template (pull_request_target) Successful in 27s
Guide screenshots / Guide images and local UI source review (pull_request) Failing after 31s
CI / Workflow lint (actionlint) (pull_request) Successful in 36s
Third-party parts / As their notes say (pull_request) Successful in 38s
beanstalk/approval can't be tested as asked: see the farm's comment
CI / Website build and checks (pull_request) Has been cancelled
to b04ee7b0cb
Some checks failed
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been skipped
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been skipped
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been skipped
Third-party parts / The same as upstream (pull_request) Has been skipped
Website / Website on Cloudflare (pull_request) Has been skipped
CI / Conventional Commit title (pull_request) Successful in 7s
PR description / Description follows the template (pull_request_target) Successful in 9s
Guide screenshots / Guide images and local UI source review (pull_request) Failing after 11s
CI / Workflow lint (actionlint) (pull_request) Successful in 11s
Third-party parts / As their notes say (pull_request) Successful in 13s
CI / Website build and checks (pull_request) Failing after 49s
Website / Website on Cloudflare (pull_request_target) Has been skipped
beanstalk/approval approved by dyst (reviewed)
beanstalk/linux+radv 1/1 fail
2026-11-21 19:01:44 +00:00
Compare
Author
Member

@Beanstalk (again for b04ee7b, rebased past #2098's build fix) only RADV on Linux, with /dev/udmabuf usable: this changes only what RX_VK_NO_DMABUF=1 does at start-up. Run main and this PR each with RX_VK_NO_DMABUF=1 set, and each with nothing set. Load a save in Hyrule Field, stand a minute, then go to the Market. With the switch, main should log pools 'not imported' and run at a few fps for a while after the load; this PR should log 'mem: RX_VK_NO_DMABUF: the guest window is anonymous memory, for host-pointer imports', import its pools and play as it does without the switch. With nothing set the two builds should match (the window a memfd's, guest memory as dma-bufs). Report the import lines ('not imported', 'imports: ... live'), the frame times in the first 30 s after the load, and the 'vk: guest memory as dma-bufs' line of each run.

@Beanstalk (again for b04ee7b, rebased past #2098's build fix) only RADV on Linux, with /dev/udmabuf usable: this changes only what RX_VK_NO_DMABUF=1 does at start-up. Run main and this PR each with RX_VK_NO_DMABUF=1 set, and each with nothing set. Load a save in Hyrule Field, stand a minute, then go to the Market. With the switch, main should log pools 'not imported' and run at a few fps for a while after the load; this PR should log 'mem: RX_VK_NO_DMABUF: the guest window is anonymous memory, for host-pointer imports', import its pools and play as it does without the switch. With nothing set the two builds should match (the window a memfd's, guest memory as dma-bufs). Report the import lines ('not imported', 'imports: ... live'), the frame times in the first 30 s after the load, and the 'vk: guest memory as dma-bufs' line of each run.
Member

Beanstalk test farm · ❌ failed on linux+radv

Commit b04ee7b · approved by dyst (reviewed) · dashboard

⚠️ Note: The farm cannot enforce that /dev/udmabuf is usable; without it, the requested main-side failure and default dma-buf path may not be exercised.
⚠️ Note: The first 30-second measurement starts after teleporting to Hyrule Field, not immediately after the Kokiri save loads. Log matches are reported but cannot be made conditional on the switch in these separate scenarios.

Summary, written by AI (gpt-6.1-terra) from the numbers below:

The change failed on Linux RADV due to a picture difference on the RX 7900 XTX with the switch unset; the RX 890M report could not be judged. The screenshot difference was 0.0564, above the 0.05 limit. On the RX 7900 XTX, switch-unset Market frame time p99 rose from 36.85 to 43.56 ms (+18.2%), but this comparison was inconclusive because there were too few usable starts. Both PCs timed out in both switch scenarios, leaving all required frame-time comparisons inconclusive.

Done (started 2026-11-21 23:49 UTC, last report 01:25 UTC). Run 2: re-planned by dyst. Comment @Beanstalk to run it again.

Target Result Reports Frame time p99 (ms), main → PR VRAM peak (MB) Crashes
linux+radv ❌ fail (an outlier) 2 1 2 36.8 → 43.6 (+18%) — none

Why it failed

  • linux+radv: switch-unset screenshot.max_diff: limit 0.1

Couldn't be judged: the tester's PC could not finish the test. The farm tries another PC, or the same one after a fix; it says nothing about the PR on its own.

  • linux+radv, report 2: Insufficient frame-time evidence for 4 required comparison(s). See rule results. a step ran out of time: timeout: the save did not load into play within 900 s (then the game was in menu on KokiriForest_Child); a step ran out of time at step
What was tested

What will be tested · AI plan (gpt-6.1-terra)

On Linux RADV, compare main and the PR with RX_VK_NO_DMABUF=1 and with the switch unset. Each run loads the Kokiri save, measures the first 30 seconds in Hyrule Field, stands there for another 30 seconds, then visits the Market. The plan reports frame times and import-related log matches, and fails on crashes, substantial frame-time regressions or picture differences. It takes approximately 45 minutes per tester.

  • Where: linux+radv (1-2 reports).
  • switch-on: from the repository's save kokiri-start (switches RX_VK_NO_DMABUF=1), then teleport to Hyrule Field, measure field-first-30 for 30 s (after 0 s warm-up), take the picture field to compare with main, teleport to Market, measure market for 10 s (after 5 s warm-up), take the picture market to compare with main; 2 runs on each side.
  • switch-unset: from the repository's save kokiri-start, then teleport to Hyrule Field, measure field-first-30 for 30 s (after 0 s warm-up), take the picture field to compare with main, teleport to Market, measure market for 10 s (after 5 s warm-up), take the picture market to compare with main; 2 runs on each side.
  • Measured: crashes, frame times, stutters and hitches, the game's log (an expected line that never appears: could not be judged).
  • The PR fails if: it crashes; or a picture differs from main by more than 5%; or frame time (p99) is more than 30% above main.
  • Time: about 50 min of game runs per tester (8 game starts), plus the two builds.
Measurements on `linux+radv`, for the PR description

Verdict: fail on linux+radv (2 reports): switch-unset screenshot.max_diff failed.

Hardware: 1: linux+amd+radv+dgpu AMD Radeon RX 7900 XTX (RADV NAVI31), 26.2.7 AMD Ryzen 7 5800X3D 8-Core Processor, 32 GB RAM; 2: linux+amd+radv+igpu AMD Radeon 890M Graphics (RADV STRIX1), 26.1.99 AMD Ryzen Z2 Extreme, 32 GB RAM
Scene: switch-on, from the repository's save kokiri-start (RX_VK_NO_DMABUF=1); switch-unset, from the repository's save kokiri-start; each cell's settings and switches as the table names them
Method: the Beanstalk test farm: main and this PR built and run on the same PC in one session, game starts alternating (ABBA); switch-on: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; switch-unset: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; each number the median of the runs, then of 2 reports
How to reproduce: the farm's plan above (@Beanstalk on this PR runs it again)

Before: main @ d76d202 After: this PR @ b04ee7b
switch-unset · market · frame time (p99) (ms) 39.9 43.6 (+9%)
switch-unset · field-first-30 · frame time (p99) (ms) 23.2 19.3 (−17%)

Flags and settings

None: the plan varied no setting, switch or build option.

As Markdown for the PR description:

## Measurements

**Verdict:** fail on linux+radv (2 reports): `switch-unset screenshot.max_diff` failed.

Hardware: 1: linux+amd+radv+dgpu `AMD Radeon RX 7900 XTX (RADV NAVI31), 26.2.7` `AMD Ryzen 7 5800X3D 8-Core Processor, 32 GB RAM`; 2: linux+amd+radv+igpu `AMD Radeon 890M Graphics (RADV STRIX1), 26.1.99` `AMD Ryzen Z2 Extreme, 32 GB RAM`  
Scene: switch-on, from the repository's save kokiri-start (RX_VK_NO_DMABUF=1); switch-unset, from the repository's save kokiri-start; each cell's settings and switches as the table names them  
Method: the Beanstalk test farm: main and this PR built and run on the same PC in one session, game starts alternating (ABBA); switch-on: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; switch-unset: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; each number the median of the runs, then of 2 reports  
How to reproduce: the farm's plan above (`@Beanstalk` on this PR runs it again)

| | Before: main @ d76d202 | After: this PR @ b04ee7b |
|---|---|---|
| switch-unset · market · frame time (p99) (ms) | 39.9 | 43.6 (+9%) |
| switch-unset · field-first-30 · frame time (p99) (ms) | 23.2 | 19.3 (−17%) |

**Flags and settings**

None: the plan varied no setting, switch or build option.
The testers' hardware
Target Report Ran on
linux+radv 1 linux+amd+radv+dgpu AMD Radeon RX 7900 XTX (RADV NAVI31), 26.2.7 AMD Ryzen 7 5800X3D 8-Core Processor, 32 GB RAM
linux+radv 2 linux+amd+radv+igpu AMD Radeon 890M Graphics (RADV STRIX1), 26.1.99 AMD Ryzen Z2 Extreme, 32 GB RAM
<!-- beanstalk:sticky --> ### Beanstalk test farm · ❌ failed on linux+radv Commit `b04ee7b` · approved by dyst (reviewed) · [dashboard](https://beanstalk.magicbean.dev/prs/2100) ⚠️ **Note:** The farm cannot enforce that /dev/udmabuf is usable; without it, the requested main-side failure and default dma-buf path may not be exercised. ⚠️ **Note:** The first 30-second measurement starts after teleporting to Hyrule Field, not immediately after the Kokiri save loads. Log matches are reported but cannot be made conditional on the switch in these separate scenarios. > **Summary**, written by AI (`gpt-6.1-terra`) from the numbers below: > > The change failed on Linux RADV due to a picture difference on the RX 7900 XTX with the switch unset; the RX 890M report could not be judged. The screenshot difference was 0.0564, above the 0.05 limit. On the RX 7900 XTX, switch-unset Market frame time p99 rose from 36.85 to 43.56 ms (+18.2%), but this comparison was inconclusive because there were too few usable starts. Both PCs timed out in both switch scenarios, leaving all required frame-time comparisons inconclusive. Done (started 2026-11-21 23:49 UTC, last report 01:25 UTC). Run 2: re-planned by dyst. Comment `@Beanstalk` to run it again. | Target | Result | Reports | Frame time p99 (ms), main → PR | VRAM peak (MB) | Crashes | | --- | --- | --- | --- | --- | --- | | `linux+radv` | ❌ fail (an outlier) | 2 [1](https://beanstalk.magicbean.dev/reports/rp_1a84cc8026b7ddfc) [2](https://beanstalk.magicbean.dev/reports/rp_07cd4bf981e1bff1) | 36.8 → 43.6 (+18%) | — | none | **Why it failed** - `linux+radv`: `switch-unset` `screenshot.max_diff`: `limit 0.1` **Couldn't be judged**: the tester's PC could not finish the test. The farm tries another PC, or the same one after a fix; it says nothing about the PR on its own. - `linux+radv`, [report 2](https://beanstalk.magicbean.dev/reports/rp_07cd4bf981e1bff1): `Insufficient frame-time evidence for 4 required comparison(s). See rule results. a step ran out of time: timeout: the save did not load into play within 900 s (then the game was in menu on KokiriForest_Child); a step ran out of time at step` <details><summary>What was tested</summary> **What will be tested** · AI plan (`gpt-6.1-terra`) On Linux RADV, compare main and the PR with RX_VK_NO_DMABUF=1 and with the switch unset. Each run loads the Kokiri save, measures the first 30 seconds in Hyrule Field, stands there for another 30 seconds, then visits the Market. The plan reports frame times and import-related log matches, and fails on crashes, substantial frame-time regressions or picture differences. It takes approximately 45 minutes per tester. - **Where:** linux+radv (1-2 reports). - **switch-on:** from the repository's save `kokiri-start` (switches `RX_VK_NO_DMABUF=1`), then teleport to Hyrule Field, measure field-first-30 for 30 s (after 0 s warm-up), take the picture field to compare with main, teleport to Market, measure market for 10 s (after 5 s warm-up), take the picture market to compare with main; 2 runs on each side. - **switch-unset:** from the repository's save `kokiri-start`, then teleport to Hyrule Field, measure field-first-30 for 30 s (after 0 s warm-up), take the picture field to compare with main, teleport to Market, measure market for 10 s (after 5 s warm-up), take the picture market to compare with main; 2 runs on each side. - **Measured:** crashes, frame times, stutters and hitches, the game's log (an expected line that never appears: could not be judged). - **The PR fails if:** it crashes; or a picture differs from main by more than 5%; or frame time (p99) is more than 30% above main. - **Time:** about 50 min of game runs per tester (8 game starts), plus the two builds. </details> <details><summary>Measurements on `linux+radv`, for the PR description</summary> **Verdict:** fail on linux+radv (2 reports): `switch-unset screenshot.max_diff` failed. Hardware: 1: linux+amd+radv+dgpu `AMD Radeon RX 7900 XTX (RADV NAVI31), 26.2.7` `AMD Ryzen 7 5800X3D 8-Core Processor, 32 GB RAM`; 2: linux+amd+radv+igpu `AMD Radeon 890M Graphics (RADV STRIX1), 26.1.99` `AMD Ryzen Z2 Extreme, 32 GB RAM` Scene: switch-on, from the repository's save kokiri-start (RX_VK_NO_DMABUF=1); switch-unset, from the repository's save kokiri-start; each cell's settings and switches as the table names them Method: the Beanstalk test farm: main and this PR built and run on the same PC in one session, game starts alternating (ABBA); switch-on: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; switch-unset: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; each number the median of the runs, then of 2 reports How to reproduce: the farm's plan above (`@Beanstalk` on this PR runs it again) | | Before: main @ d76d202 | After: this PR @ b04ee7b | |---|---|---| | switch-unset · market · frame time (p99) (ms) | 39.9 | 43.6 (+9%) | | switch-unset · field-first-30 · frame time (p99) (ms) | 23.2 | 19.3 (−17%) | **Flags and settings** None: the plan varied no setting, switch or build option. As Markdown for the PR description: ```markdown ## Measurements **Verdict:** fail on linux+radv (2 reports): `switch-unset screenshot.max_diff` failed. Hardware: 1: linux+amd+radv+dgpu `AMD Radeon RX 7900 XTX (RADV NAVI31), 26.2.7` `AMD Ryzen 7 5800X3D 8-Core Processor, 32 GB RAM`; 2: linux+amd+radv+igpu `AMD Radeon 890M Graphics (RADV STRIX1), 26.1.99` `AMD Ryzen Z2 Extreme, 32 GB RAM` Scene: switch-on, from the repository's save kokiri-start (RX_VK_NO_DMABUF=1); switch-unset, from the repository's save kokiri-start; each cell's settings and switches as the table names them Method: the Beanstalk test farm: main and this PR built and run on the same PC in one session, game starts alternating (ABBA); switch-on: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; switch-unset: 2 game starts a side, field-first-30 30 s after 0 s warm-up, market 10 s after 5 s warm-up; each number the median of the runs, then of 2 reports How to reproduce: the farm's plan above (`@Beanstalk` on this PR runs it again) | | Before: main @ d76d202 | After: this PR @ b04ee7b | |---|---|---| | switch-unset · market · frame time (p99) (ms) | 39.9 | 43.6 (+9%) | | switch-unset · field-first-30 · frame time (p99) (ms) | 23.2 | 19.3 (−17%) | **Flags and settings** None: the plan varied no setting, switch or build option. ``` </details> <details><summary>The testers' hardware</summary> | Target | Report | Ran on | | --- | --- | --- | | `linux+radv` | [1](https://beanstalk.magicbean.dev/reports/rp_1a84cc8026b7ddfc) | linux+amd+radv+dgpu `AMD Radeon RX 7900 XTX (RADV NAVI31), 26.2.7` `AMD Ryzen 7 5800X3D 8-Core Processor, 32 GB RAM` | | `linux+radv` | [2](https://beanstalk.magicbean.dev/reports/rp_07cd4bf981e1bff1) | linux+amd+radv+igpu `AMD Radeon 890M Graphics (RADV STRIX1), 26.1.99` `AMD Ryzen Z2 Extreme, 32 GB RAM` | </details>
Author
Member

@Beanstalk test it anyway

@Beanstalk test it anyway
Some checks failed
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been skipped
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been skipped
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been skipped
Third-party parts / The same as upstream (pull_request) Has been skipped
Website / Website on Cloudflare (pull_request) Has been skipped
CI / Conventional Commit title (pull_request) Successful in 7s
PR description / Description follows the template (pull_request_target) Successful in 9s
Guide screenshots / Guide images and local UI source review (pull_request) Failing after 11s
CI / Workflow lint (actionlint) (pull_request) Successful in 11s
Third-party parts / As their notes say (pull_request) Successful in 13s
CI / Website build and checks (pull_request) Failing after 49s
Website / Website on Cloudflare (pull_request_target) Has been skipped
beanstalk/approval approved by dyst (reviewed)
beanstalk/linux+radv 1/1 fail
This pull request is marked as a work in progress.
This branch is out-of-date with the base branch
View command line instructions

Checkout

From your project repository, check out a new branch and test the changes.
git fetch -u 'fix/no-dmabuf-window:dyst-fix/no-dmabuf-window'
git switch 'dyst-fix/no-dmabuf-window'
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
3 participants
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!2100
No description provided.