perf(vk): GCN (RX 400/500): f16 lowered where the GPU has none; no VRAM mirrors or batched submits by default (from hookshot-k10) by NotRexed · Pull Request #2343 · NotRexed/Hookshot

NotRexed · 2026-11-11T15:17:12Z

Ported from deadlynch/hookshot-k10 commit e09059c1, credited as co-author. That fork carries the fixes that got an RX 580 under RADV to play.

f16 only where the GPU has f16 shaders. RADV gives Polaris (RX 400/500) shaderFloat16 = false, yet the translator always emitted f16, and Kokiri Forest drew as coloured noise. f16 is now lowered to f32 there, as Pascal (GTX 10xx, also without f16) needs too. The f16 networks fall back to the linear stand-in, and these programs are kept apart in the program cache (disk_stamp), so other GPUs keep theirs.
GCN: VRAM mirrors and batched submits off by default. That applies on a GPU that pins no stage to 32 lanes and runs wider. An RX 580 ran Kokiri Forest at 20 fps with either on and 24.5 with both off; an RX 590 played with them off in [Tracking] GPUs without mesh shaders (GTX 9xx/10xx, RX 400/500/5000, Iris Xe/UHD): v2.0.6 fallback reports #1001. RX_VK_MIRROR / RX_VK_BATCH still choose.

GPUs with f16 shaders and 32-lane stages (RTX, RDNA) take exactly the same path as before.

Checked

Applied cleanly onto main; the three files compile.
Tested on an RX 580 by the fork's author.
Not yet run here (RTX 4060 Ti, where it changes nothing): a run follows before merging.

🤖 Generated with Claude Code

…16 shaders; no VRAM mirrors or batched submits by default Ported from deadlynch's hookshot-k10 (e09059c1), tested there on an RX 580 under RADV: - RADV gives Polaris no f16 shaders (shaderFloat16 false), yet the translator always emitted f16, and Kokiri Forest drew as coloured noise. f16 is now lowered to f32 where the GPU lacks it (as Pascal needs too), the f16 networks fall back to the linear stand-in, and such programs are kept apart in the program cache. - On a GPU that pins no stage to 32 lanes and runs wider (GCN), VRAM mirrors and batched submits are off by default: an RX 580 ran Kokiri Forest at 20 fps with either on and 24.5 with both off (an RX 590 played with them off, #1001). RX_VK_MIRROR and RX_VK_BATCH still choose. GPUs with f16 shaders and 32-lane stages (RTX, RDNA) are unchanged. Co-authored-by: deadlynch <20408322+deadlynch@users.noreply.github.com> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

hckrman101

Thanks for porting this. The gating looks right to me for the cards that matter most: on RTX and RDNA, sg32_stages is non-zero (minSubgroupSize <= 32 with subgroup size control), so gcn is false and mirrors and batching stay as before. Intel and NVIDIA report 32-lane subgroups, so they're excluded too. The f16 change only applies where shaderFloat16 is false, and the nf16 stamp is only added there, so caches on f16 cards don't change. Nothing has been posted on the thread yet, so these are first notes.

The one blocker is CI. tests/test_amd_subgroup_profiles.py fails on Linux and Windows because of this PR. It copies disk_stamp() from const uint32_t lanes[] to // (the records keep Shader::Info, and the new nf16 block falls inside that slice. The fixture's dev stub has no vk12 (no member named 'vk12'), and its add stub expects every call to be 5 words. The simplest fix is to move the block below the abi[] add, outside the slice. If you'd rather keep it where it is, give the stub a vk12 and let check_stamp accept the extra word. The other red jobs (Python checks part 1/2 on Ubuntu, SPIR-V fixtures, appimage) died in apt with Temporary failure resolving 'azure.archive.ubuntu.com', so a re-run should clear them.

Things to be aware of, but not blocking:

The f16 lowering also turns on by default on any other card without shaderFloat16, including GTX 10xx going by the description. The existing comment at host_info() says Eden's lowering on its own drew Kokiri as noise on an RTX card. Pascal now gets that lowering, the linear stand-in for the networks, and a fresh program cache, and it hasn't been tested yet. A run on a 10xx before merging would be worth it, or at least a line in the description saying Pascal changes too.
The gcn test also matches Vega (the comment says so) and Vega/GCN APUs on RADV, where batching now turns off. The numbers come from an RX 580 (and an RX 590 in #1001), so for Vega and the APUs this is an assumption. That's probably fine since RX_VK_MIRROR / RX_VK_BATCH still override.

Good to merge once the subgroup-profile test is fixed and CI is green, plus the RTX 4060 Ti run you mentioned.

hckrman101 · 2026-11-11T16:29:22Z

                              getenv("RX_VK_WARP64") ? 1u : 0u};
    add(lanes, sizeof lanes);
+    // (k10) f16 lowered to f32 on a GPU without f16 shaders (host_info): only then, so other GPUs keep their caches
+    if (!dev.vk12.shaderFloat16) {


This sits inside the slice tests/test_amd_subgroup_profiles.py takes from disk_stamp() (const uint32_t lanes[] up to // (the records keep Shader::Info). The fixture's dev has no vk12, so it fails to compile on both platforms. Even with one, its add stub asserts bytes == 5 * sizeof(uint32_t). Moving this block after the abi[] add avoids both problems.

hckrman101 · 2026-11-11T16:29:22Z

-    h.support_float16 = !getenv("RX_VK_NO_F16");
+    // (k10) and only where the GPU runs f16 shaders: RADV gives Polaris (RX 400/500) none (shaderFloat16 false), and f16 SPIR-V
+    // there drew Kokiri Forest as coloured noise on an RX 580. Eden's lowering to f32 is what Pascal takes as well
+    h.support_float16 = !getenv("RX_VK_NO_F16") && dev.vk12.shaderFloat16;


This also changes Pascal (and any other card without shaderFloat16) by default, not only Polaris. The comment just above says this lowering on its own gave noise on an RTX card. It would be good to have a GTX 10xx run, or at least to say in the description that Pascal changes too.

hckrman101 · 2026-11-11T16:29:22Z

-        const bool v = e ? atoi(e) != 0 : radv;
-        logf("vk: batched submits %s (RX_VK_BATCH%s)", v ? "on" : "off", !e && radv ? ", on with RADV" : "");
+        // (k10) not on GCN (no stage pinned to 32 lanes, 64-lane waves): an RX 580 ran Kokiri Forest at 24 fps without and 20 with
+        const bool gcn = dev.sg32_stages == 0 && dev.subgroup_size > 32;


Nit: sg32_stages is also 0 when RX_VK_NO_SUBGROUP_PIN is set, so on RDNA that debug switch now also turns batching (and VRAM mirrors) off, which could confuse an A/B test. It would also catch an RDNA driver without subgroup size control. Neither is a big deal. Maybe say it in the comment, or work out GCN from minSubgroupSize instead.

# Conflicts: # gpu/vk/vk_shader.cpp

…ich host_info reads now Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Joey-Magnus

The stub now sets vk12.shaderFloat16 = true. test_amd_subgroup_profiles and test_shader_stamp_cache pass locally. The gate sg32_stages == 0 && subgroup_size > 32 is false on RTX, RDNA, Intel and Arc, and the nf16 stamp word is only added without shaderFloat16, so caches on f16 GPUs stay valid.

Side effect: setting RX_VK_NO_SUBGROUP_PIN on RDNA now also matches GCN and loses mirrors and batching (RX_VK_MIRROR/RX_VK_BATCH still override).

This changes vk_shader.cpp, as do #2309, #2306, #2442 and #2457. Shipping them in one release means players re-prepare once.

A dGPU has its own fast memory and a slower bus to system RAM; stock mirrors every guest pool into VRAM and keeps both copies coherent with captures, shadows and write-backs. At 4K that coherence is CPU-bound: the write-back diff compares ~3.1 GB per 600 submits and mirror syncs move ~137k dirty pages, for 35.6 fps where 1080p runs 60 (issue NotRexed#2099). Under RX_VK_DGPU=1 a pool's first binding makes no mirror (MIRROR_REFUSED, the state iGPU runs use today): reads take the lock-free TLS path and GPU reads/writes land in the imported guest memory, coherent by construction - no mirror sync, no write-back. Every 600-submit window the heaviest guest-memory-read pool with >=1 MB of reads and not write-dominated runs a measured trial (120 frames direct -> 12 settle, held until the mirror is really on -> 120 mirrored, on the recordings' own timestamps); the mirror stays only when the win is both >10% (RX_VK_DGPU_GAIN) and >0.25 ms (RX_VK_DGPU_FLOOR) on present-to-present frame time, so the verdict sees the mirror's CPU cost and its PCIe benefit alike. Trials whose windows did not measure the same workload (zero binds, a >2x bind shift, a mirror that never came) abort with no verdict and no skip line; a pool that aborts three times sits out the session, and iGPUs run no trials. A lost verdict persists to hookshot_mirror_skip.txt beside the game's own state files, keyed to the device and driver (a foreign or header-less file is ignored and re-trialled). RX_VK_DGPU_TRIALS=0 is pure direct. Trials need serial replay (documented); under RX_VK_PARALLEL the policy runs trial-less. Unset = stock exactly: the added counters and logging are gated on the switch. Refused pools read uniform buffers as submitted (NotRexed#1795's per-pool form: the snapshot engages for mirror-less pools at the read, trial-won pools keep in-place mirror binds), so the artifact class the issue documents on mirror-less GPUs is covered under the policy too. Measured on RX 9070 XT / RADV, 4K, Uncapped frame cap, warp-normalized adjacent pairs (Hyrule Field): 70.2->78.0, 71.0->77.0, 59.6->76.6 fps (+9.7% median), draw parity 705 vs 709, GPU 9.2 ms of a ~14 ms frame. Mirror-sync dirty pages 136,872 -> 0; write-back diff 3.06-3.13 GB -> 330 MB or less per 600 submits. Review round 2 (external): RX_VK_DGPU_TRIALS=0's predicate was inverted and now matches the docs (=0 is pure direct); the window's read counter moved off a shared map into the pool (dgpu_read, both direct paths); Settle holds until the mirror is actually live; zero-bind baselines and >2x bind drift between the direct and mirrored windows abort the trial un-blacklisted; keep_line refuses signed and trailing-garbage lines. Review round 3 (hckrman101, Joey-Magnus): refused pools read UBs as submitted (NotRexed#1795 per-pool, live-smoked at 1.26M copies fault-free); verdicts decide on present-to-present frame time instead of GPU-only time (the mirror's CPU and upload-buffer cost is what the frame pays); the policy runs trial-less under RX_VK_PARALLEL and its fast-path counters are parallel-safe; no trials on iGPUs and a 3-abort cap; the skip list carries a device+driver header and ignores foreign files; the test driver uses {cxx}. Rebased onto main's env_on switch rule (NotRexed#2042) and the split-submission fence work (NotRexed#2419); the trials' functions sit outside the write-back observation's compile-span check (NotRexed#2131) and the GCN mirror-less default (NotRexed#2343) composes unchanged (on GCN the switch runs trial-less direct plus the UB snapshot). Co-authored-by: GLM-5.3-Flash <glm-5.3-flash@opencode.noreply.github.com>

A dGPU has its own fast memory and a slower bus to system RAM; stock mirrors every guest pool into VRAM and keeps both copies coherent with captures, shadows and write-backs. At 4K that coherence is CPU-bound: the write-back diff compares ~3.1 GB per 600 submits and mirror syncs move ~137k dirty pages, for 35.6 fps where 1080p runs 60 (issue NotRexed#2099). Under RX_VK_DGPU=1 a pool's first binding makes no mirror (MIRROR_REFUSED, the state iGPU runs use today): reads take the lock-free TLS path and GPU reads/writes land in the imported guest memory, coherent by construction - no mirror sync, no write-back. Every 600-submit window the heaviest guest-memory-read pool with >=1 MB of reads and not write-dominated runs a measured trial (120 frames direct -> 12 settle, held until the mirror is really on -> 120 mirrored, on the recordings' own timestamps); the mirror stays only when the win is both >10% (RX_VK_DGPU_GAIN) and >0.25 ms (RX_VK_DGPU_FLOOR) on present-to-present frame time, so the verdict sees the mirror's CPU cost and its PCIe benefit alike. Trials whose windows did not measure the same workload (zero binds, a >2x bind shift, a mirror that never came) abort with no verdict and no skip line; a pool that aborts three times sits out the session, and iGPUs run no trials. A lost verdict persists to hookshot_mirror_skip.txt beside the game's own state files, keyed to the device and driver (a foreign or header-less file is ignored and re-trialled). RX_VK_DGPU_TRIALS=0 is pure direct. Trials need serial replay (documented); under RX_VK_PARALLEL the policy runs trial-less. Unset = stock exactly: the added counters and logging are gated on the switch. Refused pools read uniform buffers as submitted (NotRexed#1795's per-pool form: the snapshot engages for mirror-less pools at the read, trial-won pools keep in-place mirror binds), so the artifact class the issue documents on mirror-less GPUs is covered under the policy too. Measured on RX 9070 XT / RADV, 4K, Uncapped frame cap, warp-normalized adjacent pairs (Hyrule Field): 70.2->78.0, 71.0->77.0, 59.6->76.6 fps (+9.7% median), draw parity 705 vs 709, GPU 9.2 ms of a ~14 ms frame. Mirror-sync dirty pages 136,872 -> 0; write-back diff 3.06-3.13 GB -> 330 MB or less per 600 submits. Review round 2 (external): RX_VK_DGPU_TRIALS=0's predicate was inverted and now matches the docs (=0 is pure direct); the window's read counter moved off a shared map into the pool (dgpu_read, both direct paths); Settle holds until the mirror is actually live; zero-bind baselines and >2x bind drift between the direct and mirrored windows abort the trial un-blacklisted; keep_line refuses signed and trailing-garbage lines. Review round 3 (hckrman101, Joey-Magnus): refused pools read UBs as submitted (NotRexed#1795 per-pool, live-smoked at 1.26M copies fault-free); verdicts decide on present-to-present frame time instead of GPU-only time (the mirror's CPU and upload-buffer cost is what the frame pays); the policy runs trial-less under RX_VK_PARALLEL and its fast-path counters are parallel-safe; no trials on iGPUs and a 3-abort cap; the skip list carries a device+driver header and ignores foreign files; the test driver uses {cxx}. Follow-up fix: the drift check counted served volume, so a working mirror's read collapse voided 40+ trials a run and trials never converged; binds now count once at the serving path and the drift rule compares the scene's pool set, with a three-no-verdict cap. Rebased onto main's env_on switch rule (NotRexed#2042) and the split-submission fence work (NotRexed#2419); the trials' functions sit outside the write-back observation's compile-span check (NotRexed#2131) and the GCN mirror-less default (NotRexed#2343) composes unchanged (on GCN the switch runs trial-less direct plus the UB snapshot). Co-authored-by: GLM-5.3-Flash <glm-5.3-flash@opencode.noreply.github.com>

A dGPU has its own fast memory and a slower bus to system RAM; stock mirrors every guest pool into VRAM and keeps both copies coherent with captures, shadows and write-backs. At 4K that coherence is CPU-bound: the write-back diff compares ~3.1 GB per 600 submits and mirror syncs move ~137k dirty pages, for 35.6 fps where 1080p runs 60 (issue NotRexed#2099). Under RX_VK_DGPU=1 a pool's first binding makes no mirror (MIRROR_REFUSED, the state iGPU runs use today): reads take the lock-free TLS path and GPU reads/writes land in the imported guest memory, coherent by construction - no mirror sync, no write-back. Every 600-submit window the heaviest guest-memory-read pool with >=1 MB of reads and not write-dominated runs a measured trial (120 frames direct -> 12 settle, held until the mirror is really on -> 120 mirrored, on the recordings' own timestamps); the mirror stays only when the win is both >10% (RX_VK_DGPU_GAIN) and >0.25 ms (RX_VK_DGPU_FLOOR) on present-to-present frame time, so the verdict sees the mirror's CPU cost and its PCIe benefit alike. Trials whose windows did not measure the same workload (zero binds, a >2x bind shift, a mirror that never came) abort with no verdict and no skip line; a pool that aborts three times sits out the session, and iGPUs run no trials. A lost verdict persists to hookshot_mirror_skip.txt beside the game's own state files, keyed to the device and driver (a foreign or header-less file is ignored and re-trialled). RX_VK_DGPU_TRIALS=0 is pure direct. Trials need serial replay (documented); under RX_VK_PARALLEL the policy runs trial-less. Unset = stock exactly: the added counters and logging are gated on the switch. Refused pools read uniform buffers as submitted (NotRexed#1795's per-pool form: the snapshot engages for mirror-less pools at the read, trial-won pools keep in-place mirror binds), so the artifact class the issue documents on mirror-less GPUs is covered under the policy too. Measured on RX 9070 XT / RADV, 4K, Uncapped frame cap, warp-normalized adjacent pairs (Hyrule Field): 70.2->78.0, 71.0->77.0, 59.6->76.6 fps (+9.7% median), draw parity 705 vs 709, GPU 9.2 ms of a ~14 ms frame. Mirror-sync dirty pages 136,872 -> 0; write-back diff 3.06-3.13 GB -> 330 MB or less per 600 submits. Review round 2 (external): RX_VK_DGPU_TRIALS=0's predicate was inverted and now matches the docs (=0 is pure direct); the window's read counter moved off a shared map into the pool (dgpu_read, both direct paths); Settle holds until the mirror is actually live; zero-bind baselines and >2x bind drift between the direct and mirrored windows abort the trial un-blacklisted; keep_line refuses signed and trailing-garbage lines. Review round 3 (hckrman101, Joey-Magnus): refused pools read UBs as submitted (NotRexed#1795 per-pool, live-smoked at 1.26M copies fault-free); verdicts decide on present-to-present frame time instead of GPU-only time (the mirror's CPU and upload-buffer cost is what the frame pays); the policy runs trial-less under RX_VK_PARALLEL and its fast-path counters are parallel-safe; no trials on iGPUs and a 3-abort cap; the skip list carries a device+driver header and ignores foreign files; the test driver uses {cxx}. Follow-up fix: the drift check counted served volume, so a working mirror's read collapse voided 40+ trials a run and trials never converged; binds now count once at the serving path and the drift rule compares the scene's pool set, with a three-no-verdict cap. Rebased onto main's env_on switch rule (NotRexed#2042) and the split-submission fence work (NotRexed#2419); the trials' functions sit outside the write-back observation's compile-span check (NotRexed#2131) and the GCN mirror-less default (NotRexed#2343) composes unchanged (on GCN the switch runs trial-less direct plus the UB snapshot). After the presenter-thread refactor (NotRexed#2325) the present-to-present tick lives in present_finish, which every made present passes exactly once in either threading mode, with a headless twin so offscreen runs stay measurable. Co-authored-by: GLM-5.3-Flash <glm-5.3-flash@opencode.noreply.github.com>

… no VRAM mirror With no VRAM mirror (mirror_on() off) a draw's uniform buffers are bound in the guest pools, and the GPU reads them when it runs the frame: by then the game has written the next frame's per-object matrices over them. On Intel Core Ultra iGPUs the Kokiri depth prepass drew the ground off screen and the fog pass painted sky over it (the white veil, NotRexed#999). Still there on main after NotRexed#2419. RX_VK_UB_SNAPSHOT (copy the uniforms into the stream as submitted) is now on by default on Intel when the pools are not mirrored. Other unmirrored GPUs (AMD APUs, GCN since NotRexed#2343, RX_VK_MIRROR=0) keep it off until measured. RX_VK_UB_SNAPSHOT=0/1 overrides. Assisted-by: Claude Code Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

A dGPU has its own fast memory and a slower bus to system RAM; stock mirrors every guest pool into VRAM and keeps both copies coherent with captures, shadows and write-backs. At 4K that coherence is CPU-bound: the write-back diff compares ~3.1 GB per 600 submits and mirror syncs move ~137k dirty pages, for 35.6 fps where 1080p runs 60 (issue NotRexed#2099). Under RX_VK_DGPU=1 a pool's first binding makes no mirror (MIRROR_REFUSED, the state iGPU runs use today): reads take the lock-free TLS path and GPU reads/writes land in the imported guest memory, coherent by construction - no mirror sync, no write-back. Every 600-submit window the heaviest guest-memory-read pool with >=1 MB of reads and not write-dominated runs a measured trial (120 frames direct -> 12 settle, held until the mirror is really on -> 120 mirrored, on the recordings' own timestamps); the mirror stays only when the win is both >10% (RX_VK_DGPU_GAIN) and >0.25 ms (RX_VK_DGPU_FLOOR) on present-to-present frame time, so the verdict sees the mirror's CPU cost and its PCIe benefit alike. Trials whose windows did not measure the same workload (zero binds, a >2x bind shift, a mirror that never came) abort with no verdict and no skip line; a pool that aborts three times sits out the session, and iGPUs run no trials. A lost verdict persists to hookshot_mirror_skip.txt beside the game's own state files, keyed to the device and driver (a foreign or header-less file is ignored and re-trialled). RX_VK_DGPU_TRIALS=0 is pure direct. Trials need serial replay (documented); under RX_VK_PARALLEL the policy runs trial-less. Unset = stock exactly: the added counters and logging are gated on the switch. Refused pools read uniform buffers as submitted (NotRexed#1795's per-pool form: the snapshot engages for mirror-less pools at the read, trial-won pools keep in-place mirror binds), so the artifact class the issue documents on mirror-less GPUs is covered under the policy too. Measured on RX 9070 XT / RADV, 4K, Uncapped frame cap, warp-normalized adjacent pairs (Hyrule Field): 70.2->78.0, 71.0->77.0, 59.6->76.6 fps (+9.7% median), draw parity 705 vs 709, GPU 9.2 ms of a ~14 ms frame. Mirror-sync dirty pages 136,872 -> 0; write-back diff 3.06-3.13 GB -> 330 MB or less per 600 submits. Review round 2 (external): RX_VK_DGPU_TRIALS=0's predicate was inverted and now matches the docs (=0 is pure direct); the window's read counter moved off a shared map into the pool (dgpu_read, both direct paths); Settle holds until the mirror is actually live; zero-bind baselines and >2x bind drift between the direct and mirrored windows abort the trial un-blacklisted; keep_line refuses signed and trailing-garbage lines. Review round 3 (hckrman101, Joey-Magnus): refused pools read UBs as submitted (NotRexed#1795 per-pool, live-smoked at 1.26M copies fault-free); verdicts decide on present-to-present frame time instead of GPU-only time (the mirror's CPU and upload-buffer cost is what the frame pays); the policy runs trial-less under RX_VK_PARALLEL and its fast-path counters are parallel-safe; no trials on iGPUs and a 3-abort cap; the skip list carries a device+driver header and ignores foreign files; the test driver uses {cxx}. Follow-up fix: the drift check counted served volume, so a working mirror's read collapse voided 40+ trials a run and trials never converged; binds now count once at the serving path and the drift rule compares the scene's pool set, with a three-no-verdict cap. The set rule still voided every trial in a streaming scene, so anchors now match by guest range (a recycled pool is the same anchor), anchor at 4x the floor, void only when a substantial fraction of the scene moves, and the candidate floor rose to 256 MB of reads a window (the verdict floor is unreachable below it - a trial there is churn by construction). Rebased onto main's env_on switch rule (NotRexed#2042) and the split-submission fence work (NotRexed#2419); the trials' functions sit outside the write-back observation's compile-span check (NotRexed#2131) and the GCN mirror-less default (NotRexed#2343) composes unchanged (on GCN the switch runs trial-less direct plus the UB snapshot). After the presenter-thread refactor (NotRexed#2325) the present-to-present tick lives in present_finish, which every made present passes exactly once in either threading mode, with a headless twin so offscreen runs stay measurable. Co-authored-by: GLM-5.3-Flash <glm-5.3-flash@opencode.noreply.github.com>

… no VRAM mirror (#2464) This is my first contribution to Hookshot, so I am happy to change anything about the approach, the style or the scope. I did the investigation and the change with the help of AI, as the commit trailers say. I checked the results on my own machine, described below. On Intel Core Ultra iGPUs, Kokiri Forest is covered by a white veil: the ground is missing and the sky/fog is painted over it. The mechanism is not Intel-specific (any GPU whose buffer pools have no VRAM mirror can read uniform buffers stale), but the new default is limited to Intel for now. ## Why With no VRAM mirror (`mirror_on()` off: integrated GPUs, GCN since #2343, `RX_VK_MIRROR=0`), a draw's uniform buffers are bound straight into the imported guest pools. The GPU reads them when it runs the frame. By then the game has already written the next frame's per-object matrices over the ones that were submitted. In Kokiri, the depth prepass (generic program `ffb53ed102dd9908`) drew the terrain with the wrong `c4` matrices, so the terrain landed off screen. The depth there stayed at far (1.0). The fog/atmosphere composite (`1a7482c0ec9991ea`) then treats every pixel above 0.9999 as sky and painted fog over the ground. The forward pass draws the ground colour correctly, which is why only the fog layer looked wrong. On a discrete GPU the mirror is copied at submission time, so it never sees the rewrite. ## What changed `RX_VK_UB_SNAPSHOT` already copies a uniform buffer into the stream as submitted (`memory_read_as_submitted`). It is now on by default on Intel when the pools are not mirrored (`vendorID == 0x8086 && !pools_mirrored()`). Other GPUs without a mirror (AMD APUs, Steam Deck, GCN, `RX_VK_MIRROR=0`) keep it off until someone measures the cost there. `RX_VK_UB_SNAPSHOT=0` turns it off; `=1` forces it on anywhere. ## How it was found - `RX_VK_DUMP_EACH=ffb53ed102dd9908` + `RX_VK_DUMP_ARGS` read the depth back after each prepass draw. 551 of 558 draws added no depth at all, including 12k-24k-index terrain meshes drawn while the depth was still empty. - A CPU projection of the same draws, using the logged vertices and `c3`/`c4`, put 454 of them on screen with depth ~0.99. - Writing depth with compare ALWAYS, no cull and depth clamp changed nothing. Neither did copying the vertex and index buffers into the stream as submitted. Copying the uniform buffers did: the prepass depth went from ~45,000 far samples (of 57,600) to ~1,000, and the veil went away. ## Testing - **Machine:** Lenovo laptop (type 83DJ). - **CPU:** Intel Core Ultra 7 155H (Meteor Lake). - **GPU:** integrated Intel Arc Graphics (8 Xe-cores), no dedicated GPU; video memory is shared with system RAM. - **RAM:** 32 GB LPDDR5x-7467. - **OS:** Windows 11 Pro, build 26300 (Insider). - **Driver:** Intel 32.0.101.9175, installed by hand (the version from #741). - **Game settings:** output resolution setting 3 (`HOS_OUTPUT_RES=3`), `RX_VK_FIFO=1`, `RX_VK_PREPARE_BACKGROUND=1`. - **Build:** v2.1.0 with this change; then current `main` (3e32949, with #2419) built from source with this branch rebased on it. - **Current main without this change:** the veil is still there (#2419 does not cover it). Numbers and screenshots in the comment below. - **Kokiri, from the house to the village:** - before: white veil over the ground in every wide view; - after: ground, paths, deck and ivy correct, light fog as intended; - frame rate unchanged, ~8.5-11 fps in Kokiri at the same settings. - No AMD or discrete GPU tested; the default there is unchanged. Related: #999 (Intel tracker); report and workaround: #999 (comment) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>

The worklist was behind the forum and the merges. Since v2.1.0, many fix PRs have merged, several rows already shipped, and the most-reported problems in Discord #bug-reporting had no row. This PR brings `docs/worklist.md` back in line, following the rules in the worklist and in `docs/workflow.md`. ## Method - **Forum snapshot.** All 409 #bug-reporting posts as of Nov 11, ~11:10 PM ET (Nov 7-11). Each post is dated by its thread ID, so the submit time is in ET. - **How dates were used.** - v2.1.0 was cut from `main` at about 8:32 PM ET on Nov 10 and published at 9:29 PM ET. A PR merged before the cut counts as shipped; I checked this with `git merge-base --is-ancestor <merge commit> v2.1.0`. - A report sent before a fix shipped and not repeated after it counts as fixed. - A report sent after v2.1.0, or before a fix that merged later, counts as still open or waiting for a release. - **Score.** Severity x reach x confidence. - Reach comes from workflow.md's 2026-11-10 reporter shares. Where the affected subset is unknown, it is marked as an upper bound. - Confidence is 0.7 only where the GitHub tracker has logs from several commenters; otherwise it is 0.4. - **Owners.** New rows have no owner. Existing owners are kept. ## Changes - **Added 20 rows (18 Open, 2 Fixed-waiting).** - Open: RAM/VRAM fill-up to OOM (#993, 5.6; 13 threads after v2.1.0), shader preparation stalls (3.2), AMD Windows crashes (#996, 2.9), CPU-bound fps in Castle Town and Hyrule Field (#995), invisible objects and doors (#1002), build.py stopped on v2.1.0, cutscene audio desync, shadow artifacts, coloured lighting, HDR black screen and flicker, frame generation stutter, the block grid around lights and fog, the v2.1.0 update failing, AF outlines, wrong button glyphs, Steam Deck at 6 fps, Linux AMD slower than Windows, and the FSR 4 tick not rebuilding. - Fixed-waiting: the pick-up counter above 60 fps (#2006) and RADV device lost after the intro (#2362). - **Updated 6 rows.** - The Link-jump row is now the "above a 60 fps cap" row (#1742), covering the Scarecrow's Song softlock, Gohma and the item hanging above Link's head. It goes from S2 to S1 and is re-scored from 1.2 to 3.5. - New forum links or scores on the antivirus, Intel black screen (0.11 to 0.2, #999 has logs), fairy grotto, voices (#2506) and RX 580 (#2343) rows. - **Moved 10 rows to Fixed, waiting for a release.** #2545 (ground garbling), #2451 (RTX 30 crash), #2286, #2275, #2449, #2096, #2406, #2515, #1954 and #2367. - **Deleted 8 rows** whose fix shipped in v2.1.0: #1792, #1922, #1819, #1863, #1447, #1451, #1769-#1810 and the HDR10 overlay. ## Not done - PRs listed as "candidate" fixes are matched by title and description only. None is confirmed against the reports. - Long forum threads were judged from their summaries. - `tests/run_all.py` was not run (docs-only change). Co-authored-by: Joey Magnus <joey.magnus@cxalloy.com>

main's NotRexed#2464 copies uniform buffers on Intel without mirrors; this keeps that and extends the default to every integrated GPU without mirrors (the Steam Deck), not to discrete GPUs without them (GCN since NotRexed#2343). The reads use env_on; a copy that could not be read is not kept for reuse. The event-wait default (e5d5054) is taken out here, to go separately with the design note. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>

hckrman101 reviewed Nov 11, 2026

View reviewed changes

hckrman101 mentioned this pull request Nov 11, 2026

perf(vk): a PC-like dGPU policy (RX_VK_DGPU=1) - direct guest-memory binds, mirrors only where a measured trial proves them #2338

Open

NotRexed and others added 2 commits November 11, 2026 15:17

Merge remote-tracking branch 'origin/main' into port/k10-gcn-f16-mirrors

fe05ed1

# Conflicts: # gpu/vk/vk_shader.cpp

test: the subgroup-profile fixture's dev has vk12 (shaderFloat16), wh…

4e97363

…ich host_info reads now Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Joey-Magnus approved these changes Nov 11, 2026

View reviewed changes

hckrman101 merged commit 1f8c02f into main Nov 11, 2026

github-actions Bot mentioned this pull request Nov 11, 2026

chore(main): release 2.2.0-rc.1 #2071

Open

Joey-Magnus mentioned this pull request Nov 12, 2026

docs(worklist): triage #bug-reporting forum (Nov 7–11) #2657

Merged

mooki000 mentioned this pull request Nov 12, 2026

[Tracking] GPUs without mesh shaders (GTX 9xx/10xx, RX 400/500/5000, Iris Xe/UHD): v2.0.6 fallback reports #1001

Open

Navigation Menu

Repository navigation

perf(vk): GCN (RX 400/500): f16 lowered where the GPU has none; no VRAM mirrors or batched submits by default (from hookshot-k10) - #2343

perf(vk): GCN (RX 400/500): f16 lowered where the GPU has none; no VRAM mirrors or batched submits by default (from hookshot-k10)#2343
hckrman101 merged 3 commits into
mainfrom
port/k10-gcn-f16-mirrors

NotRexed commented Nov 11, 2026

Uh oh!

hckrman101 left a comment

Uh oh!

hckrman101 Nov 11, 2026

Uh oh!

hckrman101 Nov 11, 2026

Uh oh!

hckrman101 Nov 11, 2026

Uh oh!

Joey-Magnus left a comment

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

NotRexed commented Nov 11, 2026

Checked

Uh oh!

hckrman101 left a comment

Choose a reason for hiding this comment

Uh oh!

hckrman101 Nov 11, 2026

Choose a reason for hiding this comment

Uh oh!

hckrman101 Nov 11, 2026

Choose a reason for hiding this comment

Uh oh!

hckrman101 Nov 11, 2026

Choose a reason for hiding this comment

Uh oh!

Joey-Magnus left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants