SPIR-V leaves a register read before any write undefined and Mesa folds undefined values freely. The unconditional patch 0014 guarded against that and was taken out in #701 because the whole picture shook with DLSS on NVIDIA. Patch 0016 brings the zero start back behind a zero_locals job flag, which vk_caps_driver turns on for RADV, ANV and NVK. TOTK_ZERO_LOCALS overrides it, and gfx_shader_set_host mixes it into the translator identity so the variant has its own cache keys.
FXAA_FS gains the flag-on hash of the game's FXAA shader, so vk_dlaa_skip_draw knows it on Mesa too. A game run with TOTK_ZERO_LOCALS=1 on an RX 9070 produced that hash; a TOTK_FXAA=0 run logged no skip with either hash, so the switch itself is not confirmed.
All 29736 stored 1.4.3 inputs translate with the flag off and on, 0 spirv-val failures. FXAA hashes to 5c43ae18948d52cb with the flag off and aad42238801d31d8 with it on. Flag-on output matches the old translator byte for byte on a sample of 312 shaders. Not run on a Mesa driver.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Patch 0014 gave every function-local variable of a translated shader an OpConstantNull initializer. On an RTX 4080
SUPER with game 1.4.3 that made the whole picture move by about 0.9 output pixels a frame whenever DLSS or DLAA runs,
with a still camera (#612). The shake follows the DLAA jitter (TOTK_DLAA_JITTER=0 removes it) and goes away with the
v0.3.7 runtime and a translator built without this one patch. Patches 0012, 0013 and 0015 stay.
docs/maxwell.md 3.12 now says that registers read before any write have no fix for now; a zero start for Mesa only, or
for the shaders that need it, is still open.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Merged on top of #566. The camera motion is #594's: its blocks in camb also feed the per-object pass, and
dlaa_mv.comp stays as on main, fed the identity and the reprojection made on the CPU in double. #566's
vk_usnap_host goes, vk_usnap_bytes serves the same. #566's scaling-draw swap, camera rune guard and output-size
scaling pass stay in vk_dlaa.c. Translator patch 0012 becomes 0015 after dev's 0012-0014.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Solid scene draws whose position reads an object's own uniform blocks (bones c5, place c7, c6) are drawn a second time
with a vertex shader variant (translator patch 0012) that computes last frame's position from last frame's blocks; the
game writes those into two copies that take turns, and the uniform snapshot keeps both (vk_mvprev.c). The draws are
listed during the replay and recorded through the recorder thread with repeated state skipped. Blocks whose last-frame
copies measured wrong (c9 and c11 of walls and cliffs) keep this frame's; positions without an object block (terrain,
grass) keep the camera's motion, which measured as good.
The camera motion is reprojected in double precision on the CPU: the float inverse made noise on far surfaces and a
still camera that DLSS followed into a wobble. DLSS gets Ultra Performance (a third of the output, also in the F1 menu)
and NVIDIA's recommended preset per mode (K, M for Performance, L for Ultra Performance; TOTK_DLSS_PRESET overrides).
Test switches: TOTK_MV_CHECK measures the motion vectors and DLSS's output against the picture, TOTK_DLAA_TIMES times
DLSS and the motion passes on the GPU. A reactive mask pass for transparents is included but off (TOTK_DLSS_REACTIVE):
its mask covers sky and fog too.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- BFE / BFI (patch 0012): a field past bit 31 is cut there, one that starts past it extracts 0 (signed: bit 31
repeated) and inserts nothing, as PTX bfe / bfi. Clamped only where the field is not a constant inside the word.
1.4.3's Depths terrain vertex shaders (42 variants, 134 extracts) compute the position as ((2y + x) * 8) & 0xff,
up to 248; AMD's bitfield instructions take it modulo 32. UltraTears saw the Depths walls and ground missing on
RADV (Steam Deck, RX 570).
- SHF by 0 (patch 0013) gives the unshifted word instead of shifting the other one by 32. Every register SHF of
1.4.3 pairs its word with 0, so 1.4.3 results do not change.
- Function-local variables start at 0 (patch 0014): a register read on a path that never wrote it is no longer an
undefined value Mesa may fold.
All 12,454 1.4.3 shaders translate and pass spirv-val. RX 9070, Depths save: the same picture before and after.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Merged with applezra's follow-ups on the same branch: the sun shadow filter narrowed by the shadow map scale
(translator patch 0011), anisotropic filtering forced only on tiling samplers, and TOTK_DUMP_UBO with a vec4 count.
Conflicts with dev: cfg.c (names, display section, defaults next to aspect and device_profile), menu.cpp (the new rows
after the aspect row), vk_texture.c (the pool alias check takes scale == 1 and dev's bind_pool_memory size argument),
build_xlate.py (patches 0005-0011 are ours; the Linux note kept), totk.ini.example (both sections). The contact pass
release went into gfx_present.c, the shared half of the old gfx_window.c.
aniso defaults to game: the picture stays as before until a player raises it in the menu.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The deferred lighting shaders soften the sun shadow with four compares around the receiver, offset by element 56 .xy of
their lighting block. The offset is in UV and the game sized it for its own 960x960 cascades, so with shadow_scale above
1 it spans that many more texels and the near cascade shows four separate copies of each edge. Translator patch 0011
multiplies those reads by the support buffer's new shadow_spread field in fragment shaders that sample a 2D array
shadow texture, leaving fp_c4 (the cascade matrices) alone. support_fill writes 1 / shadow_scale there, so the spacing
stays the game's in texels; at shadow_scale 1 the shaders behave as before. The support buffer grows to 1440 bytes and
build_xlate.py expects eleven patches.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The deferred lighting shaders take the sun shadow as four explicit-LOD compares on the cascade array. Translator patch
0010 turns every explicit-LOD compare on a 2D array texture into a PCSS lookup when the support buffer's new pcss
field has x above 0: eight raw depth fetches within pcss.w find the blockers, (receiver - blocker) * pcss.x gives the
penumbra width clamped to pcss.y..pcss.z, and eight compares on a disc of that radius average to the result. With x at
0 the original single compare runs behind a branch. Radii are in UV, so they hold at any shadow_scale.
The support buffer grows to 1424 bytes; support_fill writes the field from pcss (percent of the default width, 0 to
200, default 0; TOTK_PCSS_K overrides the factor at 100 percent, default 0.5). build_xlate.py expects ten patches. The
setting is in totk.ini.example and on the Screen tab of the menu (English and Russian).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
AMD and Intel round a gather coordinate that sits between two texels the other way from Maxwell, so shadow-map
filtering picked the neighbouring texel and drew dark dashes along wall and floor edges that moved with the camera
(docs/maxwell.md 2.10; BotW had the same on AMD). vk_caps.c passes the device's subTexelPrecisionBits to the translator
for every non-NVIDIA GPU, the job carries it as gather_bias, and totk_xlate answers QueryHostGatherBiasPrecision with
it, which turns on Ryujinx's own coordinate bias (TexturePass.InsertCoordGatherBias). The answer joins the translator
identity, so those shaders get their own cache keys. TOTK_GATHER_BIAS=0 turns it off.
The point form of a sampler slot is built once (vk_sampler_point_by_id_locked). Translator patch 0009 rounds the 2DArrayShadow reference in Ryujinx's SPIR-V backend.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>