On Windows 10 a file with a mapped view cannot be deleted, and FreeType's
Windows stream keeps the face's file mapped, so removing SST-Bold.otf threw
and ended guest_system_font. Moving it out of the font directory leaves
nothing at the original path and works on every host.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The validator takes the geometryShader and sampleRateShading enables from the
context (set where #295 enables them), so a device without them refuses the
draw instead of creating a module that needs features it never enabled.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Pixel shaders that export a coverage mask write gl_SampleMask, an
int[1] (or uint[1]) output. The validator rejected array interface
types; it now spells them elem[N] and accepts SampleMask as int[1] or
uint[1].
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
The recompiler emits SampleId and Layer fragment inputs (pixel ancillary
data) and a FragDepth output with the DepthReplacing execution mode
(depth export), but the validator refused them, so those draws were
skipped. They are now accepted, as are the Geometry and SampleRateShading
capabilities those inputs need in fragment shaders. The device enables
both features when it has them.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
The draw key mixed all 32 user data words of every stage. It now mixes only the
words a stage's RSRC2 counts, so draws that differ only in words their programs
never read share an entry. The RSRC2 registers stay in the key.
APS5_DRAW_KEY_ALL_USER_WORDS=1 mixes all 32 again.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
A0-A3 (MOV AL/eAX/rAX to and from moffs) have no ModRM byte and carry an
8-byte absolute address in 64-bit mode, or 4 bytes with a 0x67 prefix.
The decoder added nothing for the address, so the linear scanner decoded
the address bytes as code.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The Linux patcher appends the DT_RUNPATH string after the copied
.dynstr data but kept DT_STRSZ and the .dynstr section size at the size
of the copied data, so DT_RUNPATH pointed at the first byte past the
declared string table.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
In 64-bit mode REX.W takes precedence over the 0x66 operand-size prefix,
so the immediate of 81, C7, 69, 68, A9, the 05..3D accumulator forms and
the F7 TEST form stays imm32. The decoder only applied this to B8+r and
returned a length two bytes short for the others.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
PPSA01325 stops on sceFontGlyphDefineAttribute right after generating a character glyph: it sets attribute 0x11 to 0 on that glyph, deletes it without using it, and draws the character with sceFontRenderCharGlyphImageHorizontal. The meaning of the attributes is unknown, so the call checks the glyph and ignores the rest; TechnicalDebt lists it under silent stubs.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01UTjpbMX7BUFMZZaxJE98zV
Astro's Playroom (PPSA01325) fails eager binding on three imports. The two voice chat group requests follow the offline model of the other sceVoiceChatRequest* calls and return SIGNED_OUT; sceRudpActivate, whose signature is unknown, is declared as a throwing stub.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01UTjpbMX7BUFMZZaxJE98zV
Recorder::KeepDrawSnapshot keeps nothing larger than the pool's byte
budget, and the storage-buffer pool's budget was 256 MiB (the draw
input pool, for vertex and index data, has 1 GiB). A read-only storage
buffer a draw reads in place from an import is copied into a snapshot
per draw; one larger than 256 MiB was never kept, so every draw copied
it again even when the guest had not touched it.
PPSA14632 (LEGO Horizon Adventures) binds a 576 MiB direct-memory buffer
to many draws: each one memcpy'd all 576 MiB (about 100 ms), 176 times
in 60 s. With the budget at 1 GiB, as for draw inputs, the snapshot is
reused while the buffer is unchanged (394 reuses and 41 copies in 90 s),
and Graphics::Draw drops from about 7 ms to 1 ms per draw packet.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Add header-only support for end-to-end relinker tests: a minimal PS5
executable builder, a PT_LOAD mapper that applies segment permissions,
and a posix_spawn runner for the relinker binary. Every
core/relinker/tests/e2e/*.cpp becomes its own test, so new tests do not
touch CMake.
The first test relinks an executable with EXTRQ and INSERTQ sites using
--to-intel, maps the output and calls the patched code, which runs both
stubs from their Linux segment and returns through their branches.
Co-Authored-By: Claude <[email protected]>
On Linux ngs2_atrac9 aborted without any output once TestParseFile was part of
its main(). The exception that TestCorruptSuperframeAtPageEnd throws in the
Ngs2 module and catches in main() has to resume through a cleanup landing pad
inside main(). That resume goes through the unwinder statically linked into the
test executable, which then fails with _URC_FATAL_PHASE2_ERROR and aborts
silently. Whether main() gets such a pad depends on what the compiler inlines
into it, so adding any function with non-trivial cleanups to that main() is
enough to trigger it.
Move the PCM and file parse tests into a new ngs2_parse executable and restore
Ngs2Atrac9.cpp to its upstream content, so the test that catches exceptions is
unchanged. The new executable has no catch and never relies on that path.
A depth-only pass can leave a stale pixel program address in
SPI_SHADER_PGM_LO/HI_PS while never setting the pixel registers it does
not need. The draw then failed with "missing register at DWORD 0x1b3"
(SPI_PS_INPUT_ENA), although the hardware never launches that program:
with no color written (CB_TARGET_MASK & CB_SHADER_MASK == 0), no depth,
stencil or mask export (SPI_SHADER_Z_FORMAT 0 and those enables of
DB_SHADER_CONTROL clear), and neither KILL_ENABLE, EXEC_ON_HIER_FAIL,
EXEC_ON_NOOP nor EXEC_IF_OVERLAPPED set, the pixel shader has no effect
and is not executed.
PixelProgramUnset becomes PixelProgramSkipped: true for a zero program
address as before, and for such an inert pixel stage. Both use the
driver's null pixel program; the precheck and the decode agree.
PPSA14632 (LEGO Horizon Adventures) issues an HTILE expand this way
(DB_RENDER_CONTROL DEPTH/STENCIL_COMPRESS_DISABLE, CB disabled, depth
test and writes off) once per run.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Adrià