LookBackScan runs a 43-instruction decoupled look-back written for the
test: one aggregate per block, the block index from an atomic counter,
{flag, value} published with buffer_atomic_swap_x2 and read back with
glc buffer_load_dwordx2. Seven single look-backs and 50 scans of 4096
blocks are checked against a CPU prefix sum; without the recompiler
change the first whole scan fails (3 of 3 runs).
TechnicalDebt records that only glc, aligned, in-bounds dwordx2 loads
in programs with the 64-bit storage view are one atomic load: loads
without glc, and loads in programs without 64-bit buffer atomics, can
still tear against a concurrent swap_x2.
A glc buffer_load_dwordx2 is one 64-bit access on the GPU, so it never
sees half of a buffer_atomic_swap_x2. It was emitted as two dword loads,
which can: the first can read the flag from before the swap and the
second the value from after it.
GTA V gives its BVH nodes their record indices with a single-pass
decoupled look-back prefix sum. Each workgroup publishes a {flag, value}
pair for its block with buffer_atomic_swap_x2 and reads the previous 64
blocks' pairs with glc dwordx2 loads. A torn read counts earlier blocks
twice, two nodes then share records, the BVH is corrupt and the ray
tracing refit hangs. It only happens in scans of more than 64 blocks (the
largest BVHs) and depends on timing.
In a program that has the u64 storage-buffer view (it uses 64-bit buffer
atomics), a coherent raw two-dword buffer load that is 8-byte aligned and
in bounds is now read with one device-scope OpAtomicLoad on that view and
split with OpUConvert and a shift (not OpBitcast to a uvec2, which
crashes NVIDIA's shader compiler). All other loads keep the per-dword
path.
tests/execution/LookBackScan runs the game's scan kernel: the look-back
from given blocks, and 50 whole 125-block scans checked against a CPU
prefix sum. Without this change the first whole scan fails.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
ClearImageMirrors drops the mirror map before the device is destroyed,
but the current address space in Spaces() still held its mirrors, so
they were freed by the cache's static destructor from the driver's DLL
detach, after the Vulkan driver was unloaded, and every exit crashed on
Windows. The teardown now releases the current address space too.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A pixel or vertex shader with a 64-bit image atomic declares Int64Atomics,
Int64ImageEXT and SPV_EXT_shader_image_int64, which graphics shader
validation rejected ("SPIR-V requires unsupported device capability"),
so the draw was dropped. Compute dispatches are not validated this way.
The device records whether it enabled shaderImageInt64Atomics
(imageInt64Atomics in the device state and the graphics Context), and
ValidateShaders accepts the two capabilities and the extension when it
did.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
image_atomic_* with a two-dword data mask (four for cmpswap) is the
64-bit form of an integer image atomic. The decoder rejected it with
"invalid MIMG data mask". PPSA14632 (Unreal Engine) uses
image_atomic_umax this way in a compute shader during its first frames.
Swap, add, sub, signed and unsigned min/max, and, or, xor and cmpswap
now decode with 64-bit data and translate to new ImageAtomic*64 IR
opcodes. An image updated by them gets its own descriptor binding
range, declared as an R64ui image of 64-bit integers, and the module
uses Int64ImageEXT from SPV_EXT_shader_image_int64. The driver enables
shaderImageInt64Atomics when the device offers it and binds an
R64_UINT view of the 32_32 UINT, SINT or FLOAT surface. Without the
feature, recompiling such a shader throws.
64-bit inc, dec and the float atomics still throw, as does any other
access in the same shader to an image it updates with 64-bit atomics.
The new agc_driver_image_atomics64 test runs every operation on 32_32
UINT, SINT and FLOAT surfaces and checks the returned and stored values
against edge cases across the dword boundary.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
DrawRejection read 0x1c4 with read(), which throws when the register is
absent, against the function's rule that a missing register gives no
verdict there; it now reuses the value read with value() for the
DB_SHADER_CONTROL check. Tests pin the Z-export rules: formats 1, 2, 3
and 9 with Z_EXPORT_ENABLE accepted, format 0 rejected, the sample-mask
export accepted with 32_ABGR and rejected with 32_R, formats 4 to 8
rejected, and a missing 0x1c4 accepted.
StageOf, VertexInfo, PixelInfo and ShaderWorkgroupInput were defined
in the anonymous namespaces of up to three SPIR-V backend sources with
identical bodies; they are now declared next to FailEmit.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The resource tracker refused a program's 33rd distinct V# with "buffer
resource limit exceeded". The 32 was a recompiler constant, not a
Vulkan limit: every buffer is one constant-indexed element of the guest
buffers binding, which ShaderResources checks against the device's
storage-buffer limits.
Drop ShaderInfo::MaxBuffers and the tracker's refusal, and size the
SPIR-V emitter's per-buffer offset list from the binding layout. When
the user data and offsets do not fit in the push constants, they all go
to the shader data buffer as before. ShaderResources counted only
storage buffers against maxPerStageResources (200 on Intel UHD) and now
adds the stage's sampled and storage images.
agc_driver_many_buffers reads and writes 52 and 124 buffers, and
agc_driver_graphics rejects buffers plus an image above the limit.
An S# with FILTER_MODE (word0 bits 29-30) set threw in
DecodeSamplerResource. Modes 1 and 2 now map to Vulkan's MIN and MAX
sampler reduction through VK_EXT_sampler_filter_minmax, which the device
enables when it reports both filterMinmax properties; mode 3 throws.
Min/max still throws with anisotropy, a linear mip filter (lavapipe and
NVIDIA disagree there), depth comparison, color-texture compare
emulation, and, in PointFilteredSamplerWord (S# words 0 and 2), when it
filters between texels or mips of an image the recompiler point-samples.
A texture sampled through a linear min/max sampler needs
SAMPLED_IMAGE_FILTER_MINMAX on its view format; the image-sampler pairs
travel in DescriptorBinding::imageSamplers (shader disk cache format 5).
Adds the agc_driver_image_sample_reduction GPU test and extends the
sampler, emulated color compare and disk cache tests.
GFX10 format 6 (8_SINT) T#s threw "sampled image descriptor uses an
unsupported format": neither the recompiler's format table nor the
driver's texture format table had a row for it. Format 6 now reads as
a SINT image through an R8_SINT view, so image_load sign-extends each
texel (0x80 reads -128) and image-table entries of format 6 decode.
8_SINT storage still throws.
RADV (Mesa 26.0 and later) initialises a FREE_DESCRIPTOR_SET pool's
allocator with bo_size + RADV_POOL_HEAP_OFFSET as its size, so a set can
be placed up to 32 bytes past the pool's buffer. The chain pools are
0x26000 bytes, a page multiple, so a set at the end of a full pool has
its last descriptors on an unmapped page: the CPU descriptor write
faults, or the shader's scalar load raises a GPU page fault and the ring
is reset.
Two more storage buffer descriptors make the pool 0x26020 bytes; the
buffer's last page then holds those 32 bytes.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
libSceAvPlayer.native/Export.cpp was a byte-identical copy of
libSceAvPlayer/Export.cpp. Point add_sce_avplayer_library at the guest
directory's Export.cpp, as the Audiodec and Font library functions do, and
remove the copy.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_017pTZqAyVSZy2FZmaqakPHZ
Nine SPIR-V backend sources each defined an identical FailEmit in an
anonymous namespace. Declare it once in SpirvEmitterHelpers.hpp and
define it in SpirvEmitterHelpers.cpp; the message is unchanged.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01MBXJqzgohLeVAeHsn4Jpm7
CB_COLOR_INFO format 1 with number type 6 (8_SRGB) threw at decode.
It now decodes as R8_SRGB with component swap 0; the other swaps and
8_8_SRGB still throw.
A device that cannot attach and blend R8_SRGB (NVIDIA, lavapipe)
renders a resident 8_SRGB target through an RGBA8_SRGB view of an
RGBA8 copy of its R8 image: the image is blitted in before the pass
and back when the pass ends, and the write mask keeps R only, so the
encode, decode and blend are the device's sRGB attachment rules.
APS5_SRGB_ATTACHMENT_PROXY=1 forces the copy. Linear, mipmapped and
3D 8_SRGB targets throw on such a device.
PPSA26344 renders its G-buffer slot 5 and a multiply-blend pass into
8_SRGB targets.
Mount, unmount, param, icon and mount-info calls read and wrote the
global mount slot table without a lock, so two threads mounting at
once could take the same slot and one would later fail to unmount its
own mount point. Guard the table with a mutex and make the initialization
count atomic.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01MBXJqzgohLeVAeHsn4Jpm7
LEGO Horizon Adventures imports 19 functions that no library exports
and no open pull request adds, so it fails eager binding. They are
declared as throwing stubs. The two libSceAjmWVorbis imports have no
known name and are exported by NID from libSceAjm.native.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A title maps direct memory with protection 0x3f2, which adds bits
0x100 and 0x200 to the CPU, GPU and AMPR bits, and the mapping threw.
They are the audio coprocessor (ACP) read and write bits, as named in
the KytyPS5 and etaHEN headers. The coprocessor work runs on the host
CPU, so they map to host read and read/write access like the GPU bits.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
