1345 Commits
Author SHA1 Message Date
Valiantsin DzerakhandClaude Opus 5.5 3d386fe37e Merge task/inteluhd: dynamic alpha to coverage and vertex input off by default on Intel's Windows driver for integrated GPUs (#970, #992)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-10 09:13:52 +03:00
Valiantsin DzerakhandClaude Opus 5.5 905e2e720d Turn dynamic alpha to coverage and dynamic vertex input off by default on Intel's Windows driver for integrated GPUs (#970, #992)
UHD Graphics (i3-1215U, i3-1305U) and Iris Xe drew coloured noise on foliage with dynamic alpha to coverage and black
patches over the scenery with dynamic vertex input; players confirmed that vk_dyn_a2c = 0 and vk_dyn_vertex = 0 clear
both. vk_caps.intel_igpu_win marks Intel's proprietary Windows driver on an integrated GPU, and both defaults follow it.
An explicit ini value still wins. Arc discrete GPUs and Mesa's ANV keep both on: no report there, and the ANV report
(#989) had both off with its glitch still present.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-10 09:01:54 +03:00
Valiantsin DzerakhandClaude Opus 5.5 729b00a1c0 Merge task/sdl3420: ship SDL 3.4.20 in the release for pads without motion calibration (#912)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-10 08:58:32 +03:00
Valiantsin DzerakhandClaude Opus 5.5 2fe1ea1716 Bump the bundled Windows SDL to 3.4.20 (#912)
SDL 3.4.20 uses default Switch controller stick and IMU calibration when the
pad has none, which fixes the infinite gyro readings on the Beitong Asura 2 Pro+.
The pins in tools/make_release.py (version, zip and devel sha256, SDL3.dll sha256)
and the Windows rows of THIRD_PARTY_NOTICES.md follow. The sniper sysroot row stays.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-10 08:53:49 +03:00
shenjack ec62494bd9 Merge remote-tracking branch 'origin/dev' into pr/v110-patch-parity 2026-11-10 13:39:41 +08:00
Valiantsin DzerakhandClaude Opus 5.5 32f86a3296 Merge task/cheatc0: run cheat opcodes C0-C3 in the data cheat VM (#973)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-10 08:07:24 +03:00
Valiantsin DzerakhandClaude Opus 5.5 14b1b1aece Run cheat opcodes C0-C3 in the data cheat VM (#973)
The VM refused C0 (register conditional), so cheats like C046B400 FFFFFFFF were greyed out as unsupported. C0 now
runs with dmnt's operand types 0-5, widths and conditions; C1 and C2 save, restore and clear registers, C3 moves
registers to and from the static registers. tools/cheats.py gives the same opcode lengths and accepts the same opcodes.

skip() took the end of a nested block at depth 1 for the else of the failed outer block; it now counts depth the way
dmnt does, and 2x000000 is an else only for x = 1.

tests/cheats_vm_check.py runs tools/cheats.py and runtime/cheats.c on cheats of its own, the two from the issue
included.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-10 08:05:01 +03:00
ipavlin98 a4b46fa7a9 Keep the targets a frame renders in the table that tells per-frame passes from one-shot ones
The table of the frames each target was last rendered into was direct-mapped with 1024 entries. In a 1.4.3 field run about 16 entries rendered the frame before were taken by another subresource each frame, the scene's colour target among them, so every draw there whose pipeline was still building waited for it as one-shot, 50 to 400 ms, instead of skipping a frame. Sets of 8 ways that evict the way rendered longest ago lost none in the same run.
2026-11-09 21:22:38 +03:00
Valiantsin Dzerakh 8aba3647e0 Up version 0.3.14 2026-11-09 19:57:07 +03:00
Valiantsin DzerakhandClaude Opus 5.5 0238f3069a Merge task/pgostale: warn when a PGO profile was trained with other runtime headers
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 18:50:36 +03:00
Valiantsin DzerakhandClaude Opus 5.5 2d2c6782af Record the runtime headers a PGO profile was trained with and warn when they change
The floating-point helpers in simd.h and simd_fp.h inline into every lifted function before llvm-profdata matches
the profile, so the fpcheap change left pgo-v143 with a fifth of its weight while recomp/ stayed the same and
make_release saw nothing. pgo.py merge now reads the instrumented build's Ninja deps log and stores the digests of
the runtime/ headers the generated objects include (cpu.h, simd.h, simd_fp.h); make_release warns when one differs
or a record has none. docs/building.md says to retrain after a change to those headers, and build_game.py names the
libiconv-2.dll that MSYS2's SDL3.dll needs in deps-mingw/bin.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 18:50:36 +03:00
Valiantsin DzerakhandClaude Opus 5.5 bdcfb42c07 Add contributor and agent guides for pull requests, running and debugging
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 14:44:59 +03:00
Valiantsin DzerakhandClaude Opus 5.5 f9ed079614 Merge task/handheldscene: render the handheld scene at the asked size on 1.4.3
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 14:38:44 +03:00
Valiantsin DzerakhandClaude Opus 5.5 bed6962e09 Render the handheld scene at the asked size on 1.4.3
1.4.3 scales the scene size it reads from the patch page by canvas width over the console mode's base width, which is
1280 in handheld, not 1920. A handheld 1280x720 output at render_scale 100 put 1920x1080 on the page, so the game
rendered a 1920x1080 scene while the DynamicTextureAllocator heap was sized for 1280x720, and heavy scenes ran it out
(crash at main+0x1c2590). scene_scale now uses the mode's base for res_init and res_dlss_mode. In handheld the game
composes a 1280x720 canvas unless the output patch is on, so the output patch is switched on with the scene patch there
and the canvas is the output the factor and the heap assume. Docked is unchanged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 14:38:03 +03:00
itaskaev2 59eb026bd7 Merge branch 'dev' into carry/dynamic-res-100 2026-11-09 09:55:24 +01:00
itaskaev2 2e5d59df83 Merge remote-tracking branch 'origin/dev' into carry/sync-ahead
# Conflicts:
#	docs/architecture.md
#	runtime/cfg.c
#	runtime/vk_sync.c
2026-11-09 09:54:45 +01:00
Valiantsin DzerakhandClaude Opus 5.5 34b39ceca0 Merge task/memcommit: take 64 MiB blocks from the driver for buffers and other images
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 11:08:31 +03:00
Valiantsin DzerakhandClaude Opus 5.5 29cb79086f Take 64 MiB blocks from the driver for buffers and other images
Since the textures moved to a pool of their own, the default device-local
blocks hold mirrors and other buffers. With VMA's 256 MiB blocks one block
held 53 MiB of them after 10 minutes in the field, and the headroom thread
kept half a block of each device-local type ahead. TOTK_VMA_BLOCK_MIB
(default 64, 0 = VMA's 256) sets the size; the headroom spare follows it.

On the RX 9070, 10 minutes in the field: VMA blocks 3582 -> 3252 MiB,
process commit 8614 -> 8327 MB, slow frames 13 (2.8 s) -> 9 (1.1 s).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 10:51:49 +03:00
Valiantsin DzerakhandClaude Opus 5.5 a78268935a Merge task/memcommit: wait out a short out-of-memory spell instead of ending the game
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 10:15:50 +03:00
Valiantsin DzerakhandClaude Opus 5.5 cf44e9013c Merge task/memoom: wait out a short out-of-memory spell instead of ending the game
Conflicts: the image retry goes through vk_texture_image, sweep_pressure keeps the released-texture cap on the new
drop helpers, and vk_headroom_release also frees the texture pool's anchor.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 10:12:55 +03:00
Valiantsin DzerakhandClaude Opus 5.5 4ea472e864 Merge task/memcommit: textures in a pool of their own with 64 MiB blocks
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 10:11:51 +03:00
Valiantsin DzerakhandClaude Opus 5.5 8e05d2b409 Give textures a pool of their own with 64 MiB blocks
Textures shared the default 256 MiB blocks with the mirrors and other buffers, and a block went back to the driver
only once everything in it was gone. On the RX 9070 after a few minutes of roaming the field, 3 of 9 device-local
blocks were 13-27% full, held by a hundred small textures and buffers each, and blocks held 0.8-1.4 GB more than their
allocations; where device memory takes process commit, that commit backs nothing.

Own-memory textures now go to a VMA pool with 64 MiB blocks (TOTK_TEX_BLOCK_MIB, 0 = the default pools). Images of a
quarter block or more, images the driver wants a dedicated allocation for, and images the pool's memory type does not
take get memory of their own; the requirements are read before the image is made, since VMA checks neither for a
custom pool. The headroom thread keeps the pool's next block ahead like the default pools'.

Same 10-minute route: 807 MiB of blocks beyond their allocations instead of 1037, most pool blocks 75-100% full, slow
frames not up. Probes pass on v143.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 10:11:44 +03:00
Valiantsin DzerakhandClaude Opus 5.5 a79ff17314 Merge task/memcommit: count device memory against system memory where it takes process commit, BC3 for ASTC on integrated GPUs, cap the released-texture cache by bytes
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:58:10 +03:00
Valiantsin DzerakhandClaude Opus 5.5 0005c01cf5 Remove the stray brace in vk_mem_census that stopped vk_memory.c from compiling
79c6951 left the closing brace of a debug block in the per-heap loop, so the function ended early and the build failed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:57:32 +03:00
Valiantsin DzerakhandClaude Opus 5.5 34b340a208 Cap the released-texture cache by bytes
Released textures stayed until 2048 of them were kept or VRAM passed 90% of the budget. On the RX 9070 a 10-minute
field run kept 818 MiB of them, while almost every take-back came within a second. TOTK_TEX_RELEASED_MIB (default
256) runs the same drops as under VRAM pressure while the released entries hold more own memory than that. On the
same route the run ended with 1.1 GB less process commit and 1 GB less in VMA blocks; 15 dropped textures were
created again and the slow frames did not change.

Also documents TOTK_MEM_RESERVE_MIB.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:57:32 +03:00
Valiantsin DzerakhandClaude Opus 5.5 213bc6af6a Wait out a short out-of-memory spell instead of ending the game
Crash logs ended in fatal errors when the system's commit ran out for a moment: a staging buffer (-2), a guest commit
with the system at 24442 of 24447 MB, image creation (-2). Windows grows a system-managed page file and some drivers
charge device memory to the process commit, so the same request often succeeds a moment later.

oom.c holds the wait nvn_api_cmd.c realloc_wait already used: it requests a release, retries every 10 ms for up to
30 s, logs the first 8 with the commit numbers. rt_oom_request sets a flag that vk_texture_new_submit takes (one atomic
load per submit): it drops up to 256 released textures of the two kinds sweep_pressure drops (overwritten, clean
mirrored), never GPU-written content, and frees the headroom blocks. vk_oom_release does at once what needs no submit:
headroom blocks, finished upload batches and deferred objects.

- memory.c commit: waits and retries without frz_lock between tries; every caller waits, boot ones too.
- vk_staging and the draw-data chunks (vk_exec_draw.c scratch): release, then wait; the open upload batch is not
  submitted, its caller still records into it.
- create_tex: release and one retry; no wait under the texture lock, the texture is asked for again at its next use.
- the texture write-back buffer: release, then wait.
- the headroom thread keeps no spare block for 30 s after a failure.

TOTK_TEST_OOM=<site>:<n> (commit, staging, image, chunk) fails the first n attempts; probe-kernel forces 3 refused
commits, probe-vkexec 2 staging buffers, 1 image and 1 chunk, and both pass through them.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:53:30 +03:00
Valiantsin DzerakhandClaude Opus 5.5 5e479f593b Remove the stray brace in vk_mem_census that stopped vk_memory.c from compiling
79c6951 left the closing brace of a debug block in the per-heap loop, so the function ended early and the build failed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:51:04 +03:00
Valiantsin DzerakhandClaude Opus 5.5 79c6951688 Measure whether device-local memory takes process commit instead of assuming it per vendor
c838d96 counted system memory for every discrete GPU but NVIDIA's, an assumption. An RTX 5080 log (#398) showed 10.4 GB
of process commit at frame 1200, about what the RX 9070 holds with its VRAM counted, and the RTX 3060 Ti in #670 ran
out of system commit (17.8 GB) at 4K with its VRAM full. vk_mem_host_probe now allocates 128 MiB of device-local memory
once after device creation and compares the process commit before and after; half the size or more turns the system
memory cap on. The log line carries the measured amount, so user logs show what each driver does.

RX 9070: "a 128 MiB device-local allocation took 128 MiB of process commit".

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:39:54 +03:00
Valiantsin DzerakhandClaude Opus 5.5 a689922180 Decode ASTC to BC3 by default on integrated GPUs
The size rule for TOTK_ASTC_BC looked at the largest device-local heap only. On an integrated GPU that heap is system
RAM reported as VRAM (10.6 GiB on the Ally X, 14.9 GiB on the 880M), so RGBA8 stayed on exactly where memory is
shortest. vk_mem_uma() now turns BC3 on there as well.

7900X iGPU, field save, same exe, frame 1800: RGBA8 commit 7047 MB, GPU 49-53 ms, 16.8-18.4 fps; BC3 commit 6155 MB,
GPU 42.6 ms, 21.6-21.8 fps.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:20:09 +03:00
Valiantsin DzerakhandClaude Opus 5.5 d8090a85fe Keep headroom blocks for device-local memory types only
Headroom woke for every memory type that got new device memory, the staging pool's own 64 MiB blocks included, and
then kept a 256 MiB spare block in that type's default pool. On the RX 9070 the default pools of host memory types 1
and 3 held 4 MiB, beside 512 MiB of spare blocks. Textures and mirrors, whose block allocations headroom moves off the
render thread, live in device-local types.

RX 9070, field save, 60 s: heap 0 VMA blocks 192 MiB instead of 704, driver usage 1384 MiB instead of 1896; 60 fps,
1% low 51.9.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:05:09 +03:00
Valiantsin DzerakhandClaude Opus 5.5 c838d9692f Count system memory in the device-local budget where device memory takes host memory
The paths that give memory back under pressure (headroom's spare block, released textures, pool mirrors) watched the
Vulkan heap budget only. On a UMA GPU that budget is gigabytes of the same RAM the game runs in, and on Windows the
AMD driver charges device-local allocations to the process commit one to one (RX 9070: VRAM use and private bytes
grew by the same 1164, 256 and 256 MiB). Crash logs show the result: the process at 4.4-10.5 GB of commit, 16 GB
machines at 80-94% RAM, std::bad_alloc inside amdvlk64.dll (#285), allocation failures on Intel UHD.

vk_mem_host_room (vk_memory.c) is what the system can still give beyond a reserve (TOTK_MEM_RESERVE_MIB, 1536):
commit, and on a UMA GPU physical memory too; Linux overcommits, so there only a UMA GPU's available memory counts.
NVIDIA is left out, not measured. vk_mem_vram caps the budget at use plus that room, and headroom keeps its spare
block only while four blocks fit in it.

RX 9070, field save, 60 s with the room near zero from the start (TOTK_MEM_RESERVE_MIB=21000): commit 7124 MB instead
of 8965, VRAM 2631 MiB instead of 3781, released textures 153 instead of 1866; mirrors refused for 423 MiB of pools,
GPU 4.45 ms a frame instead of 3.34 ms, 60 fps either way. Without pressure nothing changes. The census logs the
room, the device class line the mode.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 09:02:36 +03:00
Valiantsin DzerakhandClaude Opus 5.5 6e9ca1ebfd Log the process commit in crash reports and in the memory census
A driver that runs out of memory faults or throws std::bad_alloc instead of returning an error: #285's log shows
bad_alloc inside amdvlk64.dll under gpl_link, and nothing in the report told that apart from a driver bug. The crash
report now prints the process and system commit, as rt_fatal does for our own failed allocations; --probe-kernel
checks the line.

The census at frames 600 and 1800 logs the commit as a baseline and, per heap, VMA's blocks and allocations beside
the driver's usage and budget, so a report shows whose memory grew.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 08:53:57 +03:00
Valiantsin DzerakhandClaude Opus 5.5 717e68f601 Merge task/crashinfo: crash reports carry host registers, code bytes and memory at the register values
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 08:20:01 +03:00
Valiantsin DzerakhandClaude Opus 5.5 06d70d0dfa Report host registers, code bytes and memory at a crash
The crash report named the lifted function and printed the guest registers, but the lifted code keeps guest
registers in host locals and writes them back only at calls, so the dump showed stale values and the faulting
instruction stayed unknown. Logs from #1005, #967 and #998 could not be traced further than the function.

The report now adds the x86 registers of the faulting thread (Windows and Linux), 64 bytes of machine code around
rip with the offset into the lifted function, so the instruction can be found in a local build of the same version,
64 bytes of memory around every register value that points at readable memory, and the first 128 bytes of the
guest stack. --probe-kernel checks the new lines with a child that faults with known values in rbx and x19.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 08:19:31 +03:00
Luxury 2907f4efc3 Stop the installer space check at an unavailable drive root 2026-11-08 23:09:58 -05:00
shenjack 68b918933f Check every SIMD register for integer-only patch preservation 2026-11-09 09:13:09 +08:00
shenjack d2b1c084fa Keep the texture offset attribution in the component table 2026-11-09 08:59:35 +08:00
shenjack c8eb315c1e Describe 1.1.0 field of view support and strengthen patch preservation checks 2026-11-09 08:44:44 +08:00
shenjack 89364414ea Check the 1.1.0 patch contracts and document their coverage 2026-11-09 08:39:10 +08:00
shenjack 6966cee0d4 Bring the core frame and camera patches to game 1.1.0 2026-11-09 08:37:24 +08:00
Valiantsin DzerakhandClaude Opus 5.5 7e440be7ca Merge pull request #978: draw amiibo file names in any script in the F1 menu (KyokoMiki)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 21:13:04 +03:00
Valiantsin DzerakhandClaude Opus 5.5 e17a27acba Merge pull request #623: report heap corruption from the vectored handler (ipavlin98)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 21:13:04 +03:00
Valiantsin DzerakhandClaude Opus 5.5 35968c1efb Merge pull request #990: Linux AppImage: ignore the user site-packages so installer.gui imports (jay23606)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 21:13:00 +03:00
Valiantsin DzerakhandClaude Opus 5.5 169d057add Merge pull request #660: keep DLSS off the inventory screen (mictony6)
Ported onto vk_dlaa_plan (#794): the scene end in the inventory also records its submission and op (end_fence,
end_pos) so chunked replays see it at the same point. DLSS-only: everything sits behind dl_active().

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 20:50:18 +03:00
Valiantsin DzerakhandClaude Opus 5.5 0704543249 Merge pull request #947: sync shadow pools from the per-submission write-watch snapshots (corne-ac)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 20:41:23 +03:00
Valiantsin DzerakhandClaude Opus 5.5 bdb41dff7e Merge pull request #979: PC prompts for every game resolution the setup lays the UI out for (Wintego)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 20:36:02 +03:00
Wintego 62b8e9dee5 Draw the PC prompts' Shift as an arrow over its name
The arrow alone read as a direction (review of #585). The key keeps the width of the game's own glyph, so the name goes
under a solid arrow rather than on a wider key.
2026-11-08 20:35:55 +03:00
Valiantsin DzerakhandClaude Opus 5.5 5e56bfc7f3 Merge task/chunked: record large submissions in chunks on several threads (#628, #794), off by default
replay_workers stays 0. On the 7900X 4 threads cut the replay by a quarter (town 6.7 -> 5.0 ms, fight2 5.3 -> 3.9 ms)
with no fps change: the game's own threads bound the frame there. Weak CPUs may gain.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 20:24:08 +03:00
Valiantsin DzerakhandClaude Opus 5.5 35b3d37ea3 Make a chunk's new texture entries follow replay order, and add a knob that records chunks out of order
entry_of decides at record time whether a new texture takes the content of a live GPU-written texture at its storage
(alias_copy) or the storage bytes. In a chunked replay the source's GPU ownership is set when a chunk records its
render, so the choice followed the order the threads got to the chunks: a later chunk could take the old bytes before
an earlier chunk recorded the render, and a mirrored texture keeps them. Now a chunk that makes an entry while a live
colour texture of its size sits at the storage waits until the chunks before it are recorded (vk_exec_chunk_settle,
the texture lock released for the wait), skips sources the scan revives at a later op, and ignores GPU ownership a
later chunk set in this submission. slot_resolve and vk_texture_by_id_locked no longer read tslots through a pointer
taken before the wait.

TOTK_REPLAY_SKEW_US (debug) makes the even chunks wait before they record, so the chunk after each one records
first in every submission. With it at 3000 us and 4 threads the field save runs clean (grass drawn, 0 validation errors); the alias path
did not fire there (0 copied, 0 filled from storage), so this does not reproduce the missing grass reported on
RADV, and the fix there stays unconfirmed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 19:53:12 +03:00
jay23606 9d07157b3e Linux AppImage fails with "No module named installer.gui" when the PyPI 2026-11-08 11:45:21 -05:00