PS5 dumps carry no playgo-chunkdefs.xml (only sce_sys/playgo-scenario.json, which
lists no chunks), so the valid set fell back to chunk 0 and every query for another
chunk returned SCE_PLAYGO_ERROR_BAD_CHUNK_ID. A title that asks for all of its chunks
in one scePlayGoGetLocus call then never sees them as installed and waits forever.
Without definitions the chunk table is unknown and a dump is fully installed, so any
chunk id is accepted and reported as local. With playgo-chunkdefs.xml present,
validation is unchanged. scePlayGoGetChunkId still reports the known set.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Coherent (GLC/DLC) global loads were emitted as Volatile loads through
physical pointers. Volatile does not make another workgroup's stores
visible within the same dispatch, so a workgroup spinning on a flag
written by another one (decoupled look-back scans, as in RADV's
acceleration-structure build) never saw it on NVIDIA. Aligned coherent
dword reads now use OpAtomicLoad at device scope; the byte path keeps
the old reads for unaligned addresses.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
VLC (PPSA85300) imports daemon (Wn6I3wVATUE) and fails to load because
no library exports it. Forking into the background is not supported,
so it throws if called. VLC 3.0 calls daemon only when its daemon
option is set, and the title starts libVLC with --ignore-config and
without --daemon.
A job's inputs were joined into one vector before decoding, copying every
input buffer at its stated size. PPSA36082's FMOD submits Opus run jobs
whose one input buffer states 0xFFFFFE74 bytes (a remaining size that
went below zero) ahead of a single packet, so each such job copied about
4 GiB and sceAjmBatchStart took about a second; the mixer missed its
deadline and the title pushed silence. A job with one input buffer now
decodes from that buffer in place, reading only the packets it decodes;
jobs with several input buffers are joined as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Drop the retry on <name>.prx after libs/<name>.prx fails. The test now
runs from its own directory holding libs/libSceFiber.prx, copied from
the libSceFiber output, so the single path resolves in the test too.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
A stack base need not be page aligned (it is not on Linux), so rounding the
range down first could leave the stack and miss the stack path. The range the
guest asked for is now checked against the guest stack, and the rounded
pages are protected, as FreeBSD rounds the address down.
A thread that exits with part of its stack protected left those pages to the
next thread that reused the stack. Protected stack pages are recorded, and a
thread's stack is made readable and writable again when it is unregistered at
exit.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
On Windows a guest thread runs on its host thread's stack, which is not a
registered guest allocation, so mprotect on any of its pages failed with
EINVAL. A guest that guards the bottom of its own stack (Dolphin's JIT
read-protects 64 KiB there to catch stack overflows) got an error where the
PS5 changes the protection.
A range that lies inside a live guest thread stack, with no GPU access
requested, now takes the native protection directly. Ranges that cross the
stack's bounds are still rejected.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Link libc, configure Windows unwinding for the test so its runtime_error
checks are caught, and run it from libSceFiber's output directory so
openHostModule finds the unpatched libSceFiber.prx next to the modules the
test already loads.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01K9MzdVqDf6wHUKzNymQSze
Return a host module handle for a loaded sysmodule id so callers can
sceKernelDlsym after sceSysmoduleLoadModule. Unknown or null inputs
throw like neighbouring internal load paths; unloaded ids return
0x805A1001. Unblocks Prospero Win past NID D8cuU4d72xM at load.
Co-Authored-By: Composer <noreply@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
#3158 imports a run of address-adjacent registered ranges as one span.
On Windows a shared mapping in the span needs a read-write alias, and
MapAlias refuses one unless the range is a single contiguous run of
views of one section. Ranges that are adjacent in guest address can be
separate views, so the refusal escaped from importAllocation and
agc_driver_dcc_keys_first_write failed on main with "read-write alias
of shared guest memory ...: the range is not one contiguous run of
views of a section".
A span whose alias is refused is now remembered, and its ranges are
imported one at a time as before spans; a single range that is refused
still throws.
Claude-Session: https://claude.ai/code/session_012SLQbhAiexxW6TtjYToW8a
A shader store through the BDA table records its 4 KiB page in the fault
buffer, and the driver marks those pages written after the use. The page
number was truncated to 32 bits, so a store above 16 TiB marked a page at
another address. And the driver scanned all 4096 slots after every use,
even when nothing was stored.
Slots now hold a 128 KiB chunk (its number plus one, 32 bits for any
address below 512 TiB) and a mask of its 4 KiB pages. A shader that
claims a slot appends it to a list, with a count. The driver reads the
count, decodes only the claimed slots, marks each run of pages once and
clears only those slots. A writable BDA range past 512 TiB is refused
when the table is built.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Since #1454, a libSceLibcInternal import that no native library exports
on Windows falls back to the title's own libc.prx. The loader shim that
ps5-native-app-boilerplate builds into homebrew titles exports 1874
functions at three addresses whose code is `xor eax, eax; ret` or
`pxor xmm0, xmm0; xor eax, eax; ret`. Such an import then binds to a
function that returns 0, and the title starts instead of stopping with
`unresolved ELF import`.
The Windows guest module writer records the exported functions whose
code is one of these two sequences, and the entry stub leaves the guest
libc.prx out of the search for those names. A guest export with code of
its own still provides a missing libSceLibcInternal function, and native
libraries keep their priority.
A draw or dispatch whose shaders have guarded descriptor slots (or a rect
list) but no BDA table gets a fault buffer, and with it NeedsCompletion(),
which kept its resources out of the resource cache, and NeverReusable
skipped the cache lookup for it. In PPSA36082 nearly every draw in the hub
is such a build, so the cache served none of them.
Such an object has no table to revalidate, no stores through BDA and no
copied writes, so it is now reusable. Each use of a reused object checks
the fault record when its batch completes (CheckFaults) instead of
running the once-only write-back; draw recipes over such a template
register the same check, and dispatch recipes still skip templates with
completion work. NeverReusable now refuses what noteReusable refuses: a
BDA table, or a fault buffer with APS5_LOOP_GUARD set (its fault resets
the buffer).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Extend the independently Windows-tested WAIT_REG_MEM candidate 056534a9 with a focused shutdown regression. Keep labels unsatisfied on graphics and two compute queues beyond the configured diagnostic, then require prompt shutdown without executing the commands or pending submissions behind those waits.
The production cancellation path already calls CheckFailure/checkStopping and catches ProcessShutdown in the worker; no runtime behavior is changed. This test does not explicitly open a Vulkan device.
Validation: Linux/GCC 13.3, C++20, -O2 -Wall -Wextra -Werror object compilation against the production headers; conventions and whitespace pass. CMake configuration, runtime linking and execution, and native Windows execution were not performed in this environment. Runtime verification remains pending. Depends on the persistent-wait candidate for #2832/#2985; maintainer policy approval is pending.
AI-assisted: yes.
With the f32 denormal field of FLOAT_MODE set to keep, v_ldexp_f32
flushed subnormal inputs and results to zero. Keep subnormal inputs and
round a result that lands in the subnormal range in the f32 round mode
of FLOAT_MODE. An overflow follows the round mode as well, with or
without denormals kept: the largest finite value instead of infinity
when the mode rounds that sign toward zero. This matches RDNA2
(gfx1036) on 2163 rows in each round mode. Denormal modes 1 and 2 now
throw.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Unsubscribe an event queue from the output mode events that
sceVideoOutAddOutputModeEvent registers, the same way the flip and
vblank events are removed. Deleting a registration that does not exist
succeeds, and a null queue or an invalid handle throws.
Co-Authored-By: Gemini 3.8 Flash <noreply@google.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
The bundled libc.prx of PPSA21703 calls sceKernelGetModuleInfoFromAddr
with a 0x1A8-byte stack buffer and never sets st_size, so the field holds
whatever was on the stack (983040 in one run) and the call threw. The
console cannot require the field on input. The full size is still
written back. The flags check is unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
vkQuake's NGG culling vertex shaders load only the vertex ID
(SPI_SHADER_PGM_RSRC2_GS ES_VGPR_COMP_CNT 0), and DecodeShaderStages still
required 3 for a program routed to the mesh path. The merged ES inputs keep
their positions whatever the count, and the mesh entry loads all of them,
so the count is now checked only for real geometry programs, as the GS
VGPR count already was.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Apply the three-case test correction offered by Scramblrrrr in 36b26122a5458028f76fe7c7710333fcfca80f6d on issue #2832. Preserve all current-main test additions outside those three functions. Credit the supplied implementation to its contributor.
Independent Windows Vulkan results and the timeout-return negative control are reported in issue #2832. Local validation: exact three-function source match, preservation of unrelated Submit.cpp content, conventions and whitespace. Native Windows/GPU execution was not performed here. Shutdown cancellation and persistent-wait policy remain unresolved.
AI-assisted: yes.
Co-authored-by: Scramblrrrr <70121301+Scramblrrrr@users.noreply.github.com>
Retain the original WAIT_REG_MEM fix from #2985 and resolve its test-registration conflict with main d7cfa728. Preserve both upstream test blocks and the original commit ancestry.
Local conventions and whitespace checks pass. Windows/GPU results cited in issue #2832 belong to their reporting contributors. No local GPU execution.
The original fix predates llroundf, which was added in #2933. Use the same guest range and exception behavior for the fourth wrapper and include it in the edge-case checks.
FreeBSD amd64 converts the rounded value with a plain cast, so NaN, infinities and values outside the 64-bit range give 0x8000000000000000 with FE_INVALID raised. The Windows host llround returns 0 for them.
(cherry picked from commit 0ee2741534bdfcc9d818b1d93f2eb793a68f111e)
FreeBSD's memset_s returns EINVAL for every constraint violation and writes nothing when smax or n exceeds RSIZE_MAX. The guest returned ERANGE when n exceeded smax and wrote up to min(n, smax) bytes even when smax or n exceeded RSIZE_MAX.
(cherry picked from commit bf08837cf929e40f1b646df82e140a33076d5587)
Following the ImageRead case, every image operation was checked against
the two lists the masked-lane walk uses: accesses guarded by a trailing exec
operand, and lane-local values.
- ImageWrite carries its exec last like StoreBuffer, and EmitWriteOp stores
only inside an if on it (bindless tables and FMASK refuse the store). It
joins the exec-guarded list.
- A gather with an LOD of its own (gather4_l, gather4_lz) or outside a pixel
shader reads only its lane's address, like the explicit-LOD samples the
walk already follows. A pixel gather without one still keeps the select,
since on the GPU it takes its LOD from the quad.
- ImageQueryDimensions returns the size of its lane's level of the image and
is lane-local.
Left out on purpose:
- ImageQueryLod and implicit-LOD samples read their quad.
- ImageBvhIntersectRay has a trailing exec operand, but its emitter ignores
it and records faults in every lane.
- Image, buffer, shared and address atomics are guarded by their exec in the
emitter too, but the test from 60c65521 pins them as keeping the select,
and no refused shader has needed them so far.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
An ImageRead carries a trailing exec argument like the buffer loads, but it
was in neither the read-only-where-active list nor the lane-local list, so a
masked select feeding its address was kept.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Kena: Bridge of Spirits (PPSA01802) refuses ten pixel shaders with
packed pixel ancillary input has an unsupported live use (used by SelectU32 ? ?)
at any visit limit. All ten have 30 to 33 phis on loop back edges (reported
on Linux with an RX 9070 XT).
Two rules stopped the proof at a loop, and the first one fires before the
second is ever asked:
- The walk from the select gave up at the first use inside a loop. A value
derived from the select reaches a loop phi as soon as its register is
written inside the loop, because the masked write becomes
Select(loopExec, new, phi). A loop that leaves exec alone needs nothing
more than this.
- An exec that a loop narrows and carries around its back edge is a phi
inside the loop, and such phis counted as not implying the write's exec.
The walk now follows uses into loops. A phi inside a loop implies the
write's exec when all of its incoming values do, where the phi itself counts
as implying it while its back-edge values are checked. This is sound because
each rule of the proof takes a lane bit of one value from a value computed
earlier: a phi's value on each iteration is the incoming value computed
before the edge was taken, a set lane of an And is set in both operands, a
set lane of a ballot is a lane whose predicate held. If any value in the
proof had a lane set where the write's exec is clear, the earliest such
value would take that lane from an earlier value of the proof that is set
there too, which contradicts it being the earliest. Two conditions keep this
true:
- The write's exec must be computed once, outside every loop. Otherwise a
phi can carry a ballot of it from an earlier iteration, which is not the
value the select used. Masks computed inside a loop keep the old rules.
- A result may only be kept if the assumptions it used held. When a value
that was assumed turns out not to be implied, every result recorded since
it was assumed is dropped. Results that say "not implied" are kept; they
only make the proof refuse more.
A lane that is turned back on inside the loop makes the back-edge exec wider
than the write's exec, so the phi is not implied and a read under it keeps
the select. A lane that is turned back on but never reads the derived value
does not.
New tests cover a loop that carries the write with exec saved, narrowed and
restored (read inside and after the loop), a loop that leaves exec alone,
and the helper-only ancillary select feeding such loops. The refusing tests
each fail when the matching check is removed: a revived lane that reads the
write inside the loop (ignoring back-edge values), a result proven under an
assumption that later fails (keeping it), and a write mask computed inside
an earlier loop (allowing such masks). Two existing tests expected a loop to
keep the select although every read is under the write's own exec; they now
expect it to go, next to new counterparts with a wider exec.
On the Taxi Chaos 2 requests, compute shaders with a loop inside a branch now
lose the branch's masked writes: 0x194000b400 removes 207 selects instead of
36, 0x19002bd200 111 instead of 28, and seven more gain 5 to 43. In
0x16400cf700 the 26 new removals are writes under the branch condition
tid < count; the loop's exec starts as a ballot of that condition, only loses
lanes at the latch, and every store and atomic inside the loop runs under it.
No proof on these requests had to drop a result.
Walks now go on into loops, so the pass takes longer on loop shaders, but it
stays below main. Time in MaskedSelectEliminator per prepare (median of 3,
the pass runs twice per prepare), 54 Taxi Chaos 2 requests, one run:
shader previous commit this commit main 5edae353
0x1ac0413000 334 ms 349 ms 424 ms
0x19002bd200 38 ms 103 ms 112 ms
0x19c016aa00 9 ms 74 ms 16 ms
0x194000b400 13 ms 31 ms 32 ms
all 54 547 ms 805 ms 977 ms
0x19c016aa00 removes nothing more; its walks only end later inside its
loops. The 144 requests of the NVIDIA dump go from 1283 to 1589 ms and gain
removals in the copies of the same shaders. 197 of the 198 requests prepare,
as before.
Related: #2548
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
LEGO Horizon Adventures (PPSA14632) still refuses a pixel shader with
packed pixel ancillary input has an unsupported live use (used by SelectU32 ? ?)
although it has the form the previous commit handles: the only whole-word
reader is SelectU32(IEqual32(HelperInvocation, 0), BitFieldUExtract(x, 0, 11),
ancillary), the exec words are Ballot(not helper) & Ballot(true), and there
are no image ops, stores or loop phis. About 10,250 IR values derive from the
word, so the walk from that select stops at its fixed limit of 4096 visited
values and the select is kept. With the limit raised to 8192 the shader
prepares (reported on Linux with an RX 9070 XT).
A walk may now take max(4096, 2 * instructions) steps. A step is a value
visited for the first time or an implication evaluated for the first time.
The walk only visits values derived from the select, at most one per
instruction, so a walk is only cut short when the implications it evaluates
cost as much as the program itself. On the 198 Taxi Chaos 2 requests no walk
needs more than 0.79 * instructions steps. One walk stays linear in the
program, O(instructions + operands): each value is expanded once, and with
the per-walk memo each implication is evaluated once.
The walk is also cheaper, so that the larger limit costs nothing:
- visited values are marked in a table indexed by value id instead of a hash
set, and the table, the work list and the memo are reused across walks;
- blocks outside loops are looked up in a table indexed by block id instead
of a hash set. This lookup ran once per use and was most of the pass's time.
Time in MaskedSelectEliminator per prepare (median of 3, the pass runs twice
per prepare), 54 Taxi Chaos 2 requests:
shader before 40a8a5a7 this commit main 5edae353
0x1ac0413000 689 ms 285 ms 401 ms
0x19002bd200 102 ms 34 ms 100 ms
0x18c07bf000 191 ms 52 ms (refused)
all 54 1294 ms 484 ms 919 ms
0x1ac0413000 and its three copies in the NVIDIA dump now remove 659 selects
instead of 625: their walks no longer stop at 4096. No other request changes
what it removes. 197 of the 198 requests prepare before and after; the other
one fails the same way in both (v_readlane_b32 of lane 63 on a 32-lane host
subgroup).
Related: #2548
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Taxi Chaos 2 (PPSA30279) pixel shader 0x18c07bf000 is refused when it is
registered, and the title stops:
packed pixel ancillary input has an unsupported live use (used by SelectU32 ? ?)
The shader reads the ancillary VGPR (v2) once, with v_bfe_u32 v3, v2, 16, 11,
which #1904 lowers to Layer. The next write to v2, v_rcp_f32 v2, s33, runs
before any kill under the initial exec and becomes
SelectU32(IEqual32(HelperInvocation, 0), rcp, ancillary). The raw word
therefore survives only in helper lanes. The shader has no s_wqm, no DPP,
readlane or swizzle, no stores, and only *_lz samples and gathers. Every
helper-lane value that comes from v2 is dropped by a later exec select, a
compare that is ANDed with exec before its ballot, or an export whose exec
excludes helper lanes.
MaskedSelectEliminator exists to remove selects like this one, but it could
not prove the case. The kill paths save and restore exec, so the later execs
are phis at block merges, and the AND chains the kills build go deeper than
its eight-level limit.
When it checks whether a later exec implies the write's exec, it now:
- accepts a phi when the phi and each incoming value lie outside loops and
every incoming value implies that exec;
- pairs low and high exec words merged by phis in one block, predecessor by
predecessor;
- memoises results for each select and follows up to 64 levels. A cycle
counts as not implied.
Once the select is gone, nothing reads the ancillary word, so no bit of it
is produced. A select whose false lanes can still be observed is kept and
the shader is still refused. That covers a data test that leaves the word in
non-helper lanes that later export, an export under the full wave, and an
unmasked ballot.
Related: #2548
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
When the dynamic symbol count is absent from the SCE or SysV hash metadata, derive it from the GNU hash buckets and chain. Keep all table reads within mapped file-backed ranges.
Carried from issue #3224, commit 1a10126089882a2306318f18d19d05b04a764d6c; original author preserved.
The fp16 overflow shader translates to 64-bit float instructions, so a device without shaderFloat64 cannot run this test. Skip before recompiling on unsupported devices.
Carried from issue #3230, commit 922d4015e05c34f616912e1913854007a5d689bf; original author preserved.
SharedCodeHeaders draws on the initial context, whose PA_CL_CLIP_CNTL
selects the [-1, 1] clip space with viewport Z scale 1 and offset 0, so
the draw asks for a viewport depth range of [-1, 1]. Devices without
VK_EXT_depth_range_unrestricted, such as Intel GPUs on Windows, reject
that and the test fails before it reaches registration. Set
DX_CLIP_SPACE_DEF as the other draw tests do.
(cherry picked from commit 3c3bcf8f80cceaba4aa1f8742c32251fcc143e6b)
A 2D surface and a one-slice 2D array over the same memory are separate
storage images, since the storage key holds the dimension. pendingAlias
only took a pending image of the same dimension, so a refresh of one
view wrote the other view's results back and uploaded them again.
Sonic Origins Plus (PPSA05324) switches between the two views of the
same surfaces about 50 times a frame in its opening animation.
pendingAlias now counts a 2D surface and a one-slice 2D array as the
same dimension. The rest of the shape test (size, array range, mips,
tiling, element size, image layers and depth, tracked units) and the
conditions of the device copy (uncompressed keys, a host import, the
surface's write stamps) are unchanged.
agc_driver_storage_one_slice_alias stores through the 2D view twice and
loads through the array view after each store. It checks the texels
read through the array, that the 2D image is still pending after the
array image's refresh (where the imported block is write-watched), and
the guest memory after the final flush. Without the change the pending
check fails on lavapipe.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
DT_SCE_STRTAB, DT_SCE_SYMTAB, DT_SCE_RELA and DT_SCE_JMPREL are offsets
into the PT_SCE_DYNLIBDATA segment, as the code analysis and the guest
image reader already read them, but the pipeline read them as file
offsets. That only works when the segment starts at file offset 0, as in
the test fixtures. A PS4 executable (CUSA00900) keeps the segment at
0x53e9b40, so the relinker read its relocations from the code and failed
with "Dynamic string offset is outside DT_STRSZ". Without the segment the
values are still read as file offsets.
(cherry picked from commit 5dbf89a57e096f2891ae786f44c3f57935af2285)
Shaders with f64 instructions now declare DenormPreserve for 64-bit floats
when the device reports shaderDenormPreserveFloat64. Without it Mesa ANV
flushed the subnormal result of v_fma_f64 to zero.
(cherry picked from commit 7df6c709de994cea0019707d45b6e7b96698fad1)
The buffer fill and copy fast paths returned before the partial last group
was applied, so a PARTIAL_TG_EN dispatch of the fill kernel filled whole
groups, and a zero last group did not throw. The partial sizes are now
decoded before the fast paths, which are skipped when the last group is
partial; the dispatch then runs the shader with the partial thread counts.
PartialGroupThreads validates all three axes before an empty dispatch
returns, and throws when (groups - 1) * full + last does not fit in 32 bits
instead of wrapping. Validate rejects PARTIAL_TG_EN combined with
USE_THREAD_DIMENSIONS (initiator 0x63).
agc_driver_pattern_buffer_fill now runs the fill kernel with a 16-of-64 last
group through the driver and checks that 80 dwords are written.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RADV sets the draw and dispatch indirect base with SET_BASE to the
argument buffer address plus the command's offset, which Vulkan requires
only to be a multiple of 4; vkQuake's base ends in 0x14. The validator
required 8-byte alignment. The records are read as 32-bit words, so a
4-byte-aligned base is now accepted.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RADV emits 0 as the start-instance location of DRAW_INDIRECT and
DRAW_INDEX_INDIRECT when the vertex shader does not read the base instance,
and as the draw-index location of the multi forms when draw IDs are off.
The validator accepted only 0x280 as "no register". Location 0 is now
accepted and decoded as 0x280; vkQuake's draws use it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RADV clears ORDER_MODE in the dispatch initiator for a dispatch that
needs its waves launched in order, and the validator required it set.
In-order launch is the stricter guarantee and the driver does not reorder
waves either way, so a direct or indirect dispatch now needs only
COMPUTE_SHADER_EN; vkQuake's map load issues one.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RADV writes every COPY_DATA memory destination with DST_SEL 5
(COPY_DATA_DST_MEM in Mesa's sid.h), the current form of the deprecated
GRBM-synchronized DST_SEL 1 that was already accepted. vkQuake's timestamp
writes use it. It is now validated, resolved and executed as memory like 1
and 2; GDS and performance-counter destinations are still rejected.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Where the GPU has no LOAD_CONTEXT_REG_INDEX, RADV loads an image's clear
values into DB_STENCIL_CLEAR/DB_DEPTH_CLEAR and CB_COLORn_CLEAR_WORD0/1
with COPY_DATA from memory to a register (DST_SEL 0), one or two dwords.
The validator rejected every register destination.
A COPY_DATA from memory to one or two context registers is now accepted.
It is not a memory store, so the packet drains like other COPY_DATA the
driver cannot record on the GPU, and executing it reads the source then and
writes the registers. Register, GDS, immediate and clock sources, and
destinations outside the context registers, are still rejected.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RADV sets PARTIAL_TG_EN in the dispatch initiator for an unaligned direct
dispatch, with each axis's last-group size in the upper half of
COMPUTE_NUM_THREAD_X/Y/Z, and TUNNEL_ENABLE on every GFX10 dispatch. The
validator rejected PARTIAL_TG_EN, and DecodeComputeStageInfo read the
whole register, so a partial size would have widened the workgroup.
A direct dispatch with PARTIAL_TG_EN now runs (groups - 1) * full + last
threads on each axis through the dispatch-size limit USE_THREAD_DIMENSIONS
already uses, or plain groups when every last group is complete; a zero
or oversized last group throws. The workgroup size is NUM_THREAD_FULL.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The builder threw on a null indirect base, while the PM4 executor accepts
SET_BASE with address 0 (it only requires the alignment and address bits) and
an indirect draw that later reads the base is checked there. Titles clear the
base with 0, as with sceAgcDcbSetIndexBuffer, which already accepts 0 to
unbind. A non-null base is still checked for 8-byte alignment.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A recorded draw counted as reading and writing any guest memory when its
resources needed completion work, which a fault buffer alone (guarded
descriptor slots, rect lists) also needs. Such a draw reaches no guest
memory through it: the buffer is the draw's own, written only on a fault.
Its pass hazards now come from its known ranges and images like any other
draw; address-based draws, draws through the BDA table and draws with
copied writes still count as touching any guest memory.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Port FreeBSD 11 lib/libc/stdlib/random.c (freebsd-src 4ab2e064): the
additive feedback generator with the default 128-byte TYPE_3 table,
good_rand seeding, state sizes rounded down to 8, 32, 64, 128 or 256
bytes, setstate returning the previous state, and initstate returning
null below 8 bytes without setting errno. Calls are serialised by one
lock; a null state throws.
The import audit of the homebrew catalog on main 262c0316 lists them as
absent libSceLibcInternal imports:
- random: PPSA74126 Stremio Plus, PPSA85300 VLC, PPSA99169 RetroArch,
PPSA99177 PS5VR, PPSA99420 PS5MC, PPSA99505 Jelly5, PPSA99515 Emby5,
PPSA99810 PSBox Cloud Gaming
- srandom: PPSA85300, PPSA99169, PPSA99420, PPSA99810
- initstate, setstate: PPSA99177, PPSA99420, PPSA99505, PPSA99515
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
An 8_8 target with SWAP_ALT keeps the x export in its first channel and
the alpha export in its second, and the color block blends that second
channel with the alpha equation. Blending into such a target threw.
With one equation for color and alpha the draw blends as before. With a
separate alpha equation the pipeline draws twice: the first pass writes
the first channel with the color equation, the second writes the second
channel with the alpha factors and operation moved into its color
equation. The split is refused, by throwing, for draws that write depth
or stencil, use per-face depth bias, or whose pixel shader writes memory,
since those would repeat; factors of the destination alpha throw since
the target keeps it in its second channel. Pipeline libraries are not
used for split pipelines.
A plain image_sample that says DIM 2D on a 1D T# takes its texel and its LOD
from the first coordinate alone, measured on RDNA2 (gfx1036) with the
hardware oracle. Astro's Playroom draws with such a pair.
The expected rows were measured with Vulkan alpha-to-coverage on a Radeon 660M (gfx1035) under RADV_DEBUG=noatocdithering, which programs DB_ALPHA_TO_MASK 0xaa01: an alpha of at least one half, 2 and +inf keep the pixel; below one half, -0, negatives, -inf and NaN drop it.
Covers code, stack and an anonymous page, and an unmapped address. On Linux the test runs again with APS5_NO_PROCMAP_QUERY so the line scan stays covered.
DB_ALPHA_TO_MASK 0xaa01, the enable bit with the four offsets at 2 and no OFFSET_ROUND, reaches the pixel stage, which keeps a pixel when its MRT0 alpha is at least one half. Dithered or rounded settings, and an MRT0 that is missing, swizzled, packed or exported without a floating-point alpha, still throw. God of War Sons of Sparta (PPSA28997) stopped on it a few minutes into play.
It carried its own copy of the /proc/self/maps line scan and is called on the same Unity stack walks as sceKernelGetModuleInfoForUnwind. It now uses the shared HostMapping lookup.
A pixel shader whose stage info asks for alpha-to-coverage declares SampleMask, starts with every sample covered and, at its MRT0 export, clears the mask when the alpha is below one half or NaN, running with late fragment tests. The shader keeps running, as a zero coverage mask does on the GPU. The flag is part of the static shader identity and of the request serialization, now version 16.
On Linux every call opened /proc/self/maps and scanned it line by line with sscanf. Unity titles call it on each stack walk, and with about 900 mappings God of War Sons of Sparta (PPSA28997) spent most of its main thread there. The covering mapping is now asked of the kernel with PROCMAP_QUERY (Linux 6.11), on one descriptor opened once, as GuestMemory does; older kernels and APS5_NO_PROCMAP_QUERY keep the line scan.
AnyPS5's libc heap registered every block with GPU access, so every
address-based build imported or mirrored each malloc block the title had
made. The console's libc.prx maps its heap with protection 3 (CPU read
and write, no GPU bits), so the GPU never sees it there either.
Heap blocks are now registered without GPU access; reallocating keeps the
block's access. The AvPlayer fallback for a title that passes no texture
allocator stands in for that allocator, which hands out memory the GPU
reads, so it allocates through a new GuestHeapAlignForGpu.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A title calls our exports on stacks it sized for the console. With lazy
binding the first call through a PLT slot runs the dynamic linker's symbol
lookup on that stack: on the 2 KiB fiber context Astro's Playroom uses it
wrote below the context into the neighbouring heap chunks.
The import probe only checked whether a GPU read of an imported range
reports its pages as written. lavapipe runs command buffers on CPU
threads, so its writes into an import go through the page tables and
are collected as CPU stores. In agc_unit_shadow the whole publish's
copy into the import is then collected as a CPU store over unit 3, and
the next partial publish drops the slab's newer bytes as stale: "the
partial publish did not copy exactly the partly covered unit".
The probe now also fills the scratch range on the GPU and collects
again. When the read or the fill reports pages as written, host and
dma-buf imports are compared. lavapipe reports 0 of 4 pages after the
read and 4 of 4 after the fill; NVIDIA reports 0 after the fill.
The probe change is the first commit of #1046.
The title id now comes from GetAppTitleId, so param.json has one parser and
its validation (a missing file or titleId throws). Another pid returns EPERM
and a null info returns 0 without writing, as shadPS4 does. app_id stays 1,
the running title's id in libSceSystemService, which is inferred, not
measured.
guest_app_info runs the call without param.json in a child process and
checks that the field between app_id and the title id is left alone.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
sceKernelGetAppInfo fills the app id and the title id of the calling
process: app id 1, the running title's id in libSceSystemService, and the
titleId of /app0/sce_sys/param.json at offset 0x10, up to 9 characters and a
terminator. Only those fields are written, since callers declare the
structure with different sizes. Another pid returns ESRCH and a null info
EFAULT.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The AMM ranges were reserved address-less on first use, so they took the
lowest free address of the arena, directly after the title's last mapping,
where the title's next hinted maps go. They are now reserved at a fixed
place, the last 64 GiB of the guest arena, which maps without MAP_FIXED and
the hints derived from them do not reach; a reservation that cannot be made
there throws.
Fixes#3146
Split at the first delimiter like FreeBSD's strsep: empty fields are
returned, and the cursor becomes null after the last token.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Share rand's generator step with rand_r, which advances the caller's
state instead of the global one, as FreeBSD's rand and rand_r share
do_rand.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Measured on a PS5 (firmware 13.60):
- sceRudpInit rejects a null pool or a size of 0 or less with
INVALID_ARGUMENT, and a pool under 0xC58 bytes or not 8-byte aligned
with MEMORY (0x80770007).
- sceRudpGetStatus rejects a null pointer or a byte count of 0 or above
0xF8 with INVALID_ARGUMENT and reports the live context count at 0x64.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Export sinh, cosh, coshf, asinh, asinhf, acosh, acoshf and atanh,
forwarding to the C++ standard library like tanh and tanhf.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
InitialContextRegisters built the clear-state context registers on every
call: about 200 emplace calls from its list and loops, with the register
vectors growing several times on the way to offset 0x3b8. Each
QueueState starts from it, so it ran for every CLEAR_STATE, and
ResolveShaderAbi and ResolveGraphicsStagesAbi construct a QueueState on
each call, which sceAgcLinkShaders, sceAgcCreatePrimState and
sceAgcCreateInterpolantMapping make while a title draws.
The registers are now built once into a table that is never destroyed,
so a queue cleared while the process exits still finds it, and each call
returns a copy of the table.
A call takes 71 ns instead of 746 ns, and constructing a QueueState 885
ns instead of 1378 ns (200000 calls each, GCC 13 -O2). On the title
screen of PPSA21402 (Balatro) the two functions took 44 to 47 of about
2300 CPU samples per 10 s and no longer appear in the profile.
sceNgs2ParseWaveformFile read everything from the offset to the end of
the file into a zero-filled buffer and parsed that, although parsing
uses only the RIFF chunk headers and the fmt, fact and smpl payloads.
A waveform at an offset in a sound bank cost a read of the rest of the
bank on each call.
ReadChunks now takes the bytes through an accessor. Waveform data passes
its buffer; a waveform file seeks to and reads the headers and those
payloads alone, with the same checks against the length from the offset
to the end of the file. The format dispatch both use moves to
ParseWaveform.
ngs2_parse adds a bank of two waveforms in one file, the second with a
chunk after its samples, and parses each at its offset.
PPSA21402 (Balatro) parses its five 33 MB music tracks at start: one
call took 8.3 to 26.7 ms and allocated 33 MB, and now takes 0.01 to
0.04 ms with the same result.
An unregistered graphics program is read raw only while the shader registry
is empty; with shaders registered it throws, as before. A draw that uses a
raw program is noted, and RegisterShader throws once that has happened,
since a title that registers shaders after drawing unregistered programs is
not understood. The draw cache revalidates a raw program by reading it raw.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The POSIX alias takes (attr, int), returns EINVAL (22) for a null or
uninitialized attribute like the other pthread_attr setters, and
stores the value through scePthreadAttrSetsolosched, returning its
result in errno form. The signature and the stored int follow the SCE
twin in KytyPS5 and sharpemu; PS5PCEM implements the alias the same
way. The value is not applied to scheduling, recorded in TechnicalDebt.
Returnal (PPSA01285 v03.000.005) stops on it after eager binding
(#2872, #1155):
what(): pthread_attr_setsolosched_np_nid_postfix not implemented
LEGO Horizon Adventures (PPSA14632, #734) and Palworld (PPSA23126,
#2598) import it.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The register draw key identifies an unregistered program only by its
address, and a key hit reuses the entry's decode without reading the
program again. A different program uploaded to the same address with
the same registers would run the old compiled shader. A hit now checks
the decode's unregistered programs against guest memory and drops the
entry when one changed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Use the upstream CMASK implementation and remove the import requirement from image clears. Propagate storage-image creation failures. Test repeated clears, copied write-back, and failure without changing guest data.
AI-assisted: yes.
guest_filesystem creates symbolic links for the Windows lstat checks.
Without administrator rights or Developer Mode, CreateSymbolicLinkW
fails with ERROR_PRIVILEGE_NOT_HELD even with
SYMBOLIC_LINK_FLAG_ALLOW_UNPRIVILEGED_CREATE, and the test aborted there
(reported as a stale guest errno). Skip those checks with a note in that
case; any other failure to create the first link still fails the test.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#2839 decides f32 product tininess after rounding, in all four f32
round modes, from the exponents and a UMulHi and IMul of the
significands, and builds the flushed value from them. To nearest even,
the mode of shaders without a known float mode and of the 0xc0 most
titles use, one multiply decides it: one factor times 2^64, which is
exact below 2^64, times the other factor rounds the exact product once
to 24 bits in the normal range, and 0 < |that| < 2^-62 is tininess
after rounding. A product that rounds up to 2^-126 is then not tiny and
keeps the emitted product, which is 2^-126 there, so nothing has to be
built from the significands.
The flush of v_mul_f32 v5, v6, v7 takes 12 IR instructions instead of
34, counted as the instructions the flush select needs beyond the
product. vector_constant_factor_tininess pins 12 to nearest and 36 for
the directed path of the same instruction, and checks that the
directed modes and v_mad_legacy_f32 (before rounding) keep #2839's
check. Its round-to-nearest branches are removed as unreachable.
The f32 denormals entry says which rounding the check inherits: the
multiply is a plain OpFMul like the product, with no f32 rounding mode
requested, so it relies on the driver rounding f32 to nearest even, as
the products do. F32ConstantFactorTininess names the row a check before
rounding flushes, 0x3f7ffffe * 0x00800001, which keeps the emitted
0x00800000.
Against #2839 to nearest even, a model of both checks agrees on
1,074,977,546 operand pairs (every 24-bit significand at both rounding
thresholds, random pairs, specials) with each other and with an exact
reference. On an RTX 5070 the outputs of #2839's own tininess and
mixed-source shaders are bit-identical with either lowering over
165,298 and 164,008 generated rows in all six float modes tested.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A wait on an equeue that holds socket events now blocks in poll/WSAPoll on
those sockets plus a per-thread loopback wake socket, instead of rechecking
every millisecond. Adding, triggering or deleting an event and closing the
equeue wake the blocked waiters. Closing a socket removes its read and write
events, as on FreeBSD. A read event also reports EV_EOF once the peer has
shut down its sending side.
sceKernelAddReadEvent/DeleteReadEvent and sceKernelAddWriteEvent/
DeleteWriteEvent register EVFILT_READ and EVFILT_WRITE events on guest
sockets. The equeue polls the host socket for readiness on each wait
and reports the readable byte count or the send buffer size in data,
with EV_EOF on hang-up or error.
pthread_cond_wait and pthread_cond_timedwait, and their sce counterparts, threw a
host exception when the mutex was unlocked, locked by another thread, or not yet
initialized, and dereferenced the adaptive static initializer. FreeBSD returns
EPERM for these (EINVAL for a destroyed mutex) before touching the condition.
(cherry picked from commit a78bb8a855ca3b277f128c6a9e52ff3949b9cb80)
The elected lane's inner spin loop never ended on Mesa ANV with the LDS in
a storage buffer. It now tries once and the outer loop elects again; the
update is unchanged.
(cherry picked from commit e894f9092326a849ebe1ce46ff0ed3b25fbb7e9b)
sceVideoOutVrrPegToFixedRate and sceVideoOutVrrUnpegFromFixedRate were
NotImplemented without arguments. Both take the output handle; the peg
also takes two further arguments whose meaning is unknown. VRR is not
modelled and the output always runs at a fixed rate, so both validate
the handle and return 0; non-zero peg arguments throw. Recorded in
TechnicalDebt.
Uncharted: Legacy of Thieves Collection (PPSA05684) stops on main with
"sceVideoOutVrrUnpegFromFixedRate not implemented"; it passes its
output handle to both (and zeros to the peg) and ignores the results.
Since #2133, a 3D image_sample_lz or image_load bound to a 2D-array T#
decodes as 2D and matches the new 2D mode, so it reads layer 0 for any
r and returns zero for z != 0. adriadam10 measured on RDNA2 (gfx1035)
that dim 3D on a 2D-array T# behaves exactly like dim 2D-array, with r
or z as the layer. Decode a 2D-array T# bound to a 3D access as 2D-array
and add a 2D-array mode next to the 2D one.
The same measurement covers image_load_mip on both T# types and
image_get_resinfo, which returns (w, h, 1, levels) on a 2D T# and
(w, h, layers, levels) on a 2D-array T#. Allow those accesses and return
1 as the third resinfo component through the 2D mode. image_get_lod is
not measured and keeps throwing.
Pipeline destroyed its shader modules right after LinkPipelineFromLibraries returned, while the pre-rasterization and fragment shader library parts built from them stay cached and are relinked later with VK_PIPELINE_CREATE_LINK_TIME_OPTIMIZATION_BIT_EXT. On AMD's Windows driver (RX 6700 XT, Adrenalin 25.10.45.05) that link faults inside amdvlk64.dll at varying offsets, reading freed heap memory, in 22 agc tests on a cold driver cache; all of them pass with APS5_NO_PIPELINE_LTO=1. Running the optimized link synchronously or without a pipeline cache did not help, keeping the modules alive did.
When the optimized link is enabled, the library store takes over the modules of each newly built shader part and destroys them in ClearPipelineLibraries after the pending links have finished. Modules of a pipeline whose parts were all cached are destroyed as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Key 0x10 (GFX9_DCC_CLEAR_SINGLE in Mesa) marks a 256-byte block whose texels all equal its first texel. A CB DCC decompress pass over uniform comp-to-single keys now writes that texel over its block and stores uncompressed keys; a fast-clear eliminate over them throws.
When another readable allocation starts right after the buffer's allocation,
the bound range still ends at the first allocation: the bytes past it are
outside the buffer the descriptor's records describe, and binding into the
neighbour would send the range back to a copy on every draw.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RADV rounds a strided buffer descriptor's record count up, so stride * records
can run up to one stride past the end of the buffer's allocation. The guest
buffer region then no longer fits inside the allocation's host import and is
copied on the CPU, whole, for every draw that binds it.
When a strided descriptor ends less than one stride past the end of the
readable registered allocation that contains its base, the range is bound up
to the allocation's end. RetroArch's Dolphin core binds a 48 MiB stream buffer
with stride 20 (records 0x266667, 12 bytes past the allocation) on every draw.
With this change the buffer is bound in place from its import.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The window offset register is already required to be zero, so the enable bit has no effect on a draw that gets this far.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Background registration preparations (#2583) copy the device's
SpirvTarget, whose capability and extension spans point into the
device's shader profile. Driver::stop reset the device without waiting
for the detached pool, so a preparation still queued or running read
freed extension names and agc_driver crashed at exit on RADV.
Driver::stop now closes the pool after joining the queue workers: it
stops queueing and waits until no task is queued or running. A
preparation submitted after that runs on the caller's thread, which
holds its own reference to the device, like the existing fallback for
a pool without threads.
Fixes#2891
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The driver's Log2 has an absolute error that does not shrink with the
result, so near 1.0 the relative error explodes: on an RTX 3080, v_log_f32
of 1 - 2^-24 is 23,947,836 ULP off. Measured on RDNA2 (gfx1036), the
hardware's log2 is within 1 ULP of the correctly rounded value, and an
eight-term series in x - 1 is within 2 ULP of both for |x - 1| <= 1/8.
The lowering keeps the driver's Log2 and takes the series only inside that
band when the two differ by more than 4 ULP, so RADV's native result stays
bit-exact and NVIDIA's is within 6 ULP inside the band.
agc_driver_log_near_one pins 64 inputs per type from the oracle with
per-lane tolerances.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012GLfKH7Sqe51wX254k1XsV
Enable the existing input-only f32 flush policy for division scaling.
RDNA2 measurements pin all three sources and signed zeros; keep the
scaling rules unchanged and retain their separate technical debt.
Co-Authored-By: Codex gpt-6.1-sol <noreply@openai.com>
Resolves a guest path the way FreeBSD 11's realpath does: a relative
path starts from the guest working directory, "." and empty components
are skipped, ".." drops the last component and never climbs above "/",
every other component must exist, and one that is not a directory with
anything after it is ENOTDIR. Backslashes separate components as in the
rest of the path layer. The result is the guest path; the host is
consulted only for existence and type, so host symlinks are followed but
not expanded, and ".." is lexical, as in the rest of the path layer. A
null buffer is PATH_MAX bytes from the application heap, a null path is
EINVAL, an empty one ENOENT, and 1024 bytes or more in the input or the
result is ENAMETOOLONG. On failure the caller's buffer holds the path
that failed.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Fnrk5Dazjb2tLooVSpdRQq
Keep each registered header and its prepared shader state instead of
replacing every registration for the same code address. Resolve explicit
ABIs by header and programs by type, preserving re-registration order.
Co-Authored-By: GPT-6.1-Sol <noreply@openai.com>
The detached alarm thread waits on a condition variable that was a
namespace-scope static. At exit its destructor ran with the thread still
waiting, and glibc's pthread_cond_destroy blocks then, so a process that
called alarm never exited on Linux. The mutex, condition variable, deadline
and started flag now live in one structure allocated once and never
destroyed.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A host import covered exactly one registered range, so a buffer crossing
two ranges had no import and was copied on the CPU at every use, under
the GPU mutex. Titles that map a GPU heap piece by piece (Uncharted maps
1 MiB direct-memory pieces) hit this for every buffer larger than a
piece.
An import now covers the run of address-adjacent registered ranges that
are readable with the same write and GPU access. Runs never overlap, so
imports do not either; an import is retired when any range of its run
changes, and a refused run is remembered by its base. Callers already
address imports relative to their base.
sceFontSetScriptLanguage, sceFontGetScriptLanguage,
sceFontSetTypographicDesign and sceFontGetTypographicDesign threw. With
the signatures SharpProspero declares, a font handle now keeps the
language set for each script and the feature set for each typographic
design, under the handle's lock, and returns them, with 0 for one never
set. They do not change shaping or rendering yet; TechnicalDebt records
the assumptions.
Claude-Session: https://claude.ai/code/session_012SLQbhAiexxW6TtjYToW8a
(cherry picked from commit c4dcbe4403dc635fde2a823dec311a939816cefc)
RetroArch (PPSA99169) stops at alarm on main: `FAIL: unresolved ELF import DQXJraCc1rA from libSceLibcInternal.prx`.
alarm(seconds) arms one process-wide real-time alarm and returns the seconds left on the previous one, rounded up like FreeBSD's alarm.c (any partial second counts); alarm(0) cancels it. When it expires, SIGALRM's disposition as recorded by sigaction or signal applies: ignored, nothing happens; default, the process terminates as on FreeBSD (exit status 128 + 14); a guest handler throws, because asynchronous delivery to a guest handler is not implemented.
The timer lives in libkernel next to the signal dispositions (GuestAlarm_nid_no_patch); libSceLibcInternal, which the titles import alarm from, forwards to it.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The converted thread pointer was set to the TLS block plus the size rounded to max(p_align, 16), while the guest reads its first variable at tp minus the size rounded to p_align. For a PT_TLS with memsz 0x21 and p_align 8 (Resident Evil 4) that is 0x30 against 0x28, so the first variable was read from the second slot. The thread pointer now uses the size rounded to p_align; the block layout is unchanged. WIP: the test for this is still missing.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
A PS4 executable (CUSA00900) keeps 37 MiB of read-only data in its
executable segment after the code. The syscall scan and the call-site
resolution decoded the whole segment linearly, reached a float table and
failed with "Unsupported 0F 79 opcode without SSE4A prefix". In
executables with a PT_SCE_DYNLIBDATA segment they now stop at the end of
the PLT that follows the last function of the unwind table. The PLT is
found by searching forward from that function for its first entry, which
pushes the GOT entry at PLTGOT + 8 and jumps through PLTGOT + 16, since
code without unwind entries can sit in between (680 bytes of start-up
code in an OpenOrbis-built sample). When the executable has PLT
relocations and no such entry follows, the relinker fails instead of
guessing. Executables without the segment are scanned as before.
(cherry picked from commit a21e8fe4c0e6c03189d559e992c0f037c076700d)
libSceCdlgG2P and libSceVrSetupDialog threw from every export. They now
run the common-dialog state machine libSceLoginDialog uses: Initialize,
Open, the status calls, GetResult, Close and Terminate follow it, the
dialog finishes at Open without showing anything, and GetResult reports
that the user cancelled, writing only the leading int32 result. The
signatures are assumed from the other common dialogs, as TechnicalDebt
records. 13 exports move from stubs to implemented.
Claude-Session: https://claude.ai/code/session_012SLQbhAiexxW6TtjYToW8a
(cherry picked from commit 07992adf80bae9b2d7da4b751cd21c040fa1c5b8)
A flat_atomic_* whose address lands in the LDS aperture updated the dword without the lock, so it could lose updates against a ds_add_u64 or a 64-bit flat atomic on the same dwords. It takes the lock like the 32-bit ds atomics when the program has a 64-bit LDS atomic; programs without one emit the same SPIR-V as before.
On RADV the race between ds_add_u64 under the workgroup lock and a
concurrent ds_add_u32 only shows when the lanes get their operands from
per-lane buffer loads: with v_mov_b32 constants or loads from one address
no increment is lost there. The new case has the split-lane kernel's
parity split and loads each lane's (0xffffffff, 0, 1, 0) row from the input
buffer at tid * 16 with buffer_load_dwordx4, so the input buffer now holds
256 rows. With SpirvMemoryInstructions.cpp reverted it fails on RADV and
on the RTX 3080; with the change it passes on both.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Fnrk5Dazjb2tLooVSpdRQq
Even lanes run ds_add_u64 and odd lanes run ds_add_u32 on the slot's high
dword, one pass per dispatch, as in the reported kernel, and the dispatch
is repeated 32 times from one compiled shader.
With ANYPS5_ENABLE_SPIRV_TOOLS=ON the scalar loop around the ds_add_u64 lock loop made
SPIRV-Tools' block merge pass merge the loop header with its multi-block continue
target, and validation after optimization rejected the module. The 16 iterations are
now emitted inline; the expected sums and the divergent exec masks are unchanged.
The hardware serialises 32-bit and 64-bit LDS atomics on one address; the lock that makes the 64-bit forms atomic did not order 32-bit atomics, which lost increments between the two halves of a 64-bit update (#2297).
GuestMemory::WriteChanged stores only the bytes the GPU changed, so that CPU writes to the bytes
around them survive. It found them a byte at a time inside every changed 256-byte block and
stored each run with memcpy. Where changed and unchanged bytes interleave (the low bytes of
floats), every run is a few bytes long, and that loop took 43% of the GPU submission thread in
Disney Epic Mickey: Rebrushed's first level (0.9-1.0 ms per MiB).
Each 16-byte chunk is now compared with SSE2. A chunk with no change is skipped, a chunk that
changed entirely is stored whole, and a mixed one is stored with MASKMOVDQU, which writes only the
bytes selected by the mask. An SFENCE after the loop orders those non-temporal stores before the
stores that follow. Bytes the GPU did not change are still never written.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0134RWCyiyydJBbfKz8g9Bwe
An import whose module matches no needed library has no library name,
and the Windows startup stub counted it against every library, so a
missing library with nothing imported from it was still never skipped.
Every such import in the executables of Record of Lodoss War, Resident
Evil 4, PRAGMATA and Returnal declares a module that exists under
another file name (libSceJson for libSceJson2.prx, for example), and
none declares libSceXml or Party.
The relinker now keeps the declared module names of those imports, for
the executable and each guest module, and the stub counts a host
library as imported from when its name up to the first dot matches one
of them, ignoring case. A module declared as libGuest still keeps a
missing libGuest.xyz required, as guest_module_identity checks.
Imports without a module ID declare nothing and count against no
library. The rule and the title data are from xXPBSXx.
The Windows startup stub stopped at the first DT_NEEDED library it could
not load, even when no import is searched in that library. Titles list
libraries they import nothing from (Road of Death needs libSceXml.prx,
Resident Evil 4 needs Party.prx), so they stopped with GetLastError 126
before running. A host library that no import's search list contains is
now skipped with a message when it is not found; its handle stays null
and is never read. Missing libraries that something imports from, other
load errors and guest modules still stop the start.
AddWaveformBlocks rejects a null data pointer before reaching the silence
path, so TestSilenceBlock aborted and its dataless rejection passed for the
wrong reason.
PPSA11386 registers 3840x2160 tiled display buffers in pixel format
0xC001000622000000, R16G16B16A16 float in KytyPS5 and SharpProspero
(0xC001000600000000 is the blue-first variant), and stops on "VideoOut:
unsupported display pixel format". The buffer is eight bytes per texel; its
tiled layout is assumed to be the 8-byte SW_64KB_R_X equation applied to
full coordinates, as addrlib applies it. The values are taken as linear
light and encoded with the sRGB curve for the 8-bit output. The encoding
compares the raw half-float bits against two tables of 255 thresholds
(color and alpha) that the CPU decode and the color transfer shader share,
so both give the same bytes on any driver; negative and NaN values give 0.
Guest-memory detile, the resident image path and the CPU decode all handle
both formats; DCC metadata on a float buffer throws. Neither the layout nor
the transfer is verified against console frames.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Fnrk5Dazjb2tLooVSpdRQq
A waveform block that declares samples with no data threw as not
fitting them. PPSA04867 adds such a block with flags 0x11 (data_size 0,
num_samples 16385) and terminates the process on the throw, followed by
a fully specified block under the same flags, which takes the ordinary
path. Accept flag 0x10, named SILENCE here: a dataless block under it
renders num_samples frames of silence and finishes like a complete
block, reporting its bytes through decoded_data_size. A dataless block
without the flag keeps throwing. No public reference names the flag;
the assumption is recorded in TechnicalDebt.
Co-Authored-By: Claude Code <noreply@anthropic.com>
The CFG used by unused-filter=1 followed only direct branches, so code reached
through lea, relative-relocated pointers or jump tables was unreachable and its
imports were dropped. It now roots RIP-relative lea targets, relative relocation
targets and relative jump tables, and counts any RIP-relative reference to a GOT
slot as an access. Address-taken targets are walked after the direct control
flow, and a target that does not decode as an instruction is not code.
Comments are reserved for technical debt (docs/dev/TechnicalDebt.md), so
the ones the previous commit added to the recorder, the snapshot keep and
the tests go.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HZHkb8kUwnHAuBBbHywvJf
A ShaderResources from the resource caches is the same object for every
draw or dispatch that uses it, and Recorder::ReusableDrawSnapshot hands
the next draw the same snapshot buffer, so passing their bytes with each
Keep counted one allocation once per use and submitted the batch before
it held the budget. Recorder::KeepBytes counts an owner's bytes once per
batch, and the dispatch, draw and snapshot keeps use it.
The recorded kept bytes test runs a dispatch and a draw that copy a
guest buffer and two draws that snapshot one on the device, and fails
when any of the three sites stops counting or a reused snapshot counts
twice. The BdaResources check pins the copied region it sets up.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HZHkb8kUwnHAuBBbHywvJf
The recorder bounds the bytes a batch keeps alive (KeptBytesBudget) and
waits for older batches past twice that. A recorded dispatch or draw kept
its ShaderResources, and a draw its snapshot buffers, with a size of 0, so
the host-visible copies of guest buffers they own never counted. A title
that binds a 124 MiB writable guest buffer in every draw then held about
ninety copies, 11 GB, until the GPU caught up, and aborted with
VK_ERROR_OUT_OF_HOST_MEMORY from vkAllocateMemory.
GuestBufferMemory reports the bytes of its copied regions, and the
dispatch and draw keeps and the draw snapshots pass their bytes to
Recorder::Keep, so the existing bound applies to them.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
The lookback classifies preceding instructions at their decoded opcode
offset, the same place the forbidden-instruction check looks since #1935,
so a legacy-prefixed mov rax, imm or mov r10, rcx is recognized. The REX
byte stays part of the patterns, so mov r8 and mov r8d cannot masquerade
as rax loads. Anything unrecognized still stops the inference.
Adds prefixed fixtures to both test suites: a prefixed constant stub that
must report syscall number 9, and a prefixed clobbered stub that must
report no number.
AI-assisted: yes. The submitting author is responsible for every changed line.
The syscall check names only the code offset, so a title that links a FreeBSD
libc (the ps5-payload-dev SDK statically links one into every homebrew binary)
stops with no indication of which call it reached. Those stubs load the number
with a constant before the instruction (mov rax, 9 for link), so decode the
nearest constant load into rax and add it to the message:
Forbidden syscall instruction at code offset 0xa: syscall number 9
The number is reported only when it can be established. The constant load must
be the instruction directly before the syscall, or one mov r10, rcx (the SDK's
first-argument move, which does not write rax) in front of it; any other
instruction stops the inference instead of returning a stale value, so
mov rax, 9; add rax, 1; syscall reports no number rather than 9. Only the three
invocation opcodes report a number: sysret is a return and keeps the current
message, as does a stub that loads rax from a register, like the SDK's generic
syscall() wrapper. The check still throws; only the text changes.
Tested: relinker-only build, full ctest on Linux x86-64 (45/45). The C++
syscall_scanner test covers the opcode matrix and the review regression; the
Python test_syscall_number covers the same contract through the harness the
relinker lane already runs on Ubuntu, Windows and macOS, where the C++ fixture
helpers (sys/mman.h, posix_spawn) do not build.
AI-assisted: yes
The registry calls an observer on the mapping thread, after the registry
lock is released, with the ranges a kernel mapping gave GPU access:
direct and flexible maps with it, and fixed remaps with it over a
registered range (a reservation). The mutation looks them up while it
still holds the lock and hands them over weakly, with the registry
generation it ended at, so the observer does not keep a range alive.
The driver imports those ranges on its current device, so an
address-based build finds the import instead of pinning the memory on
the queue that first uses it. A range that gains GPU access through
sceKernelMprotect, and AnyPS5's own libc heap, are still imported on
first use.
The import at map does not reconcile the other imports with the
registry, so its cost does not grow with the number of imports. It skips
a range that is gone, and any range when the imports were reconciled
with a newer registry since; the next lookup imports those as before.
It installs the pin waiter as the other import paths do, so unmapping a
range imported at map retires its import. The live import bytes are kept as a
running total instead of being summed over every import. Once the import
table belongs to the current device the import takes only the table's
lock; the first import on a device (which runs the write-watch probe)
and a device that was just replaced go through the GPU lock. Clearing
the observer waits for calls already running.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
On Linux every executable or guest-module import whose NID a guest
module exports was renamed to the guest-suffixed name, whatever module
the import declares. A title that bundles a module exporting system
NIDs then had those imports bound to it: the libc.prx loader shim of
ps5-native-app-boilerplate exports placeholder functions under the
libSceLibcInternal NIDs, so DOOM PPSA99666's snprintf, vsnprintf,
memcpy and malloc (declared from libSceLibcInternal) returned 0 and the
title never found its WAD (#1497).
An import now binds to a guest export only when the module it declares
is a guest module, as on Windows. Imports without import-module
metadata keep the previous behaviour, and a title's own libc.prx
imported from libc still binds to it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ProsperoRadio (PPSA99001) imports six sceOpusCeltDec functions from
libSceOpusCeltDec and fails to load because no such module exists. It
opens this decoder only as a fallback when libSceOpusDec reports a
retryable error for a CELT packet, which libSceOpusDec here never
returns, so the module is exported without an implementation and each
function throws if it is called, as for sceSaveDataCancel.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Measured on RDNA2 (gfx1035) in review:
- the shared and private apertures are 4 GB windows (the gfx9+ layout of kfd_flat_memory.c), so a flat
address routes by its whole high dword, 0x80000000 or 0x70000000, not by its top 16 bits;
- a dword or short access whose last byte is past the end of LDS reads 0 or writes nothing as a whole;
multi-dword accesses are checked per dword.
The aperture test gets rows that straddle and pass the end of LDS; the TechnicalDebt line names the
public source of the layout and keeps only the PS5's base values as inferred.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Main's StoreAddressU32x2..x4 (one BDA lookup per wide store, b95842597) bypassed the aperture routing:
a flat_store_dwordx2 into the shared aperture went to BDA at 0x8000000000000000 and LDS kept its old
value. The wide store now stores each component through the aperture like the 32-bit store, and keeps
the single BDA lookup outside the apertures.
The aperture test's V# word 3 is 0x31016fac (OOB_SELECT 3, raw), as in main's other execution tests:
with OOB_SELECT 0 its output buffer read as out of range after #639, which was the failure reported on
four drivers.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Sub-dword stores and unaligned dword accesses go byte-wise; 32- and 64-bit FLAT atomics
route to LDS (64-bit under the workgroup lock), scratch (read-modify-write) or BDA.
agc_driver_flat_apertures covers them in 121 rows, with mixed-address waves and wave64.
The aperture bases and the unmodelled operands 235-238 are listed as inferred.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
A compute FLAT access whose address has the console's shared aperture in its high dword (0x8000xxxx)
reads or writes the workgroup's LDS at its low dword, and one in the private aperture (0x7000xxxx)
the lane's scratch; other addresses stay guest memory through BDA. GTA V's ray traversal keeps its
stack in LDS and spills it to scratch through such addresses; treated as guest memory the pushes were
lost and the traversal never ended (device lost). An execution test round-trips both apertures.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
v_cvt_f32_i32 and v_cvt_f32_u32 used the host Convert which always
rounds to nearest even. When FLOAT_MODE bits 1:0 are directed, build
the f32 bits from the integer magnitude so discarded low bits follow
round toward +inf, -inf or zero, as measured on RDNA2 (gfx1036).
Co-Authored-By: Composer <noreply@cursor.com>
Twenty-five of the ninety-seven errno messages differed from FreeBSD releng/11.4 lib/libc/gen/errlst.c (for example errno 0, 30, 47, 62, 70 and 84).
(cherry picked from commit ce14c9ad7dc639335ff72b60122bae2bd0fee302)
Directed FLOAT_MODE modes for v_cvt_f16_f32, v_cvt_f16_i16,
v_cvt_f16_u16 and v_ldexp_f16 were ignored, so results stayed
round-to-nearest-even. Hardware uses FLOAT_MODE bits 3:2 (round16);
the directed path runs only when f16 denormals are preserved.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
v_interp_p1_f32_e64, v_interp_p2_f32_e64 and v_interp_mov_f32_e64
(VOP3Interp opcodes 0x200, 0x201 and 0x202) reach the decoder as VOP3
and were refused as unimplemented VOP3 opcodes. Decode them into the
existing VInterpP1F32, VInterpP2F32 and VInterpMovF32 opcodes with the
VOP3Interp operand layout, so the interpolation translation already
used by the VINTRP encodings handles them.
Encodings and operand layout follow LLVM's gfx10 MC tests for
VOP3Interp_Real_gfx10 (D69348). The interpolation parameter is the low
six bits of src0, the channel bits 6:7, the coordinate (or the
interpolation parameter of the mov form) is src1, and clamp and omod
reach the destination. op_sel, the packed-half selector, a third source
and abs/neg on operands that do not exist throw.
The mesh stage was compiled at the driver's default subgroup width, while the translated mesh program lays wave64 lanes out on, and ballots over, a 32-lane host subgroup. Intel's Windows driver (Arrow Lake, 32.0.101.8991) reports subgroupSize 32 but prefers 16 for mesh shaders, and every agc_driver_mesh draw drew no triangle there. The mesh stage now gets VkPipelineShaderStageRequiredSubgroupSizeCreateInfo{32} when the program's hostSubgroupSize is 32 and the device lists the mesh stage in requiredSubgroupSizeStages, as compute pipelines already do.
Rosetta runs AVX2, FMA, BMI1/2, F16C and MOVBE, but, unlike an Intel
host, it has neither RDSEED, CLWB nor RDPID. --to-rosetta converts
everything --to-intel does and lowers these three as well:
- RDSEED becomes RDRAND (the /6 form of the same opcode), which also
returns a random value and sets CF.
- CLWB only writes a cache line back, so it becomes a NOP of the same
length.
- RDPID becomes a load of 0 through an optional stub: Rosetta's TSC_AUX
is a constant, not a processor index, so every thread reports
processor 0. A site with no movable instruction after it for the jump
is kept.
The flag implies --to-intel and conflicts with --windows. It is
independent of --macos: it only changes what the conversion produces, so
--macos alone leaves these instructions as they are, as on an Intel Mac,
and a title for Rosetta is relinked with --macos --to-rosetta. Intel
output is unchanged: the target is a parameter of the converter, Intel by
default.
Scalar f16 add/mul/sub ignore FLOAT_MODE denorm16 flush bits on main.
Flush inputs when bit 0 is clear and results when bit 1 is clear at
nearest-even. Multiply checks tininess after 11-bit rounding
(0x387ff000); add/sub use 0x38800000.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
The relinker reads its arguments as narrow strings, so on a Windows
system whose ANSI code page cannot represent the directory it cannot
open the input, and the windows_working_directory case failed before
reaching the executable. Convert the case under an ASCII name and rename
the directory before starting the executable.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The --windows-diagnostics walker looked up modules and directories with
GetModuleHandleA, LoadLibraryExA, GetFileAttributesA,
GetModuleFileNameA and GetSystemDirectoryA, so under a directory outside
the ANSI code page it failed with "dependency diagnostics Windows API
failed" instead of naming the missing dependency.
Keep the walker's paths in UTF-8 and convert them at the Windows API
calls, which now use the wide functions. The entry stub passes it, and
prints, its paths as UTF-8 instead of the ANSI code page. The dependency
diagnostics test now builds its fixtures under a non-ASCII directory.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Windows entry stub read its path with GetModuleFileNameA, which
turns every character the ANSI code page cannot represent into '?', and
then used it with SetCurrentDirectoryA and LoadLibraryExA. A converted
executable in such a directory (Japanese on a Western-locale system, an
emoji, a script without an ANSI code page) stopped at start-up with
ERROR_INVALID_NAME.
Read the path with GetModuleFileNameW and enter the directory and load
the libraries with SetCurrentDirectoryW and LoadLibraryExW, with library
paths stored as UTF-16. Messages and the --windows-diagnostics walker
still get the paths in the ANSI code page, as recorded in TechnicalDebt.
Fixes#2640
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Keep WAIT_REG_MEM diagnostic timeouts from completing unsatisfied waits.
Register the existing gated-dispatch capture test with a 1ms diagnostic
timeout and document the missing-producer limitation.
sceSystemServiceParamGetInt always reported English (United States) as
the console language, so a title that follows the console's language,
such as Jelly5 (PPSA99505) with its Arabic text, can only show English.
ANYPS5_LANGUAGE now selects a console language number from 0 to 30
(21 is Arabic). Unset or empty keeps English (United States); any other
value throws at the language query instead of being ignored.
(cherry picked from commit 9bcf196daa01b07a22be876a7f1143149503e054)
Submit the same vertex and pixel draw twice through the driver and
require the second one to hit the draw cache, with graphics pipeline
libraries and with APS5_NO_GPL=1. With pipeline libraries the pixel
stage is cached at its fixed push slot, so a lookup at the stacked
offset misses every draw on its layout. The driver exposes a copy of
its draw-cache counters for the test.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
AI-assisted: yes. Return AnyPS5 without reading the native machine name, hostname environment or network configuration. Preserve FreeBSD bounded-copy behavior, guest ENAMETOOLONG, null queries and success errno. The final tree passes a complete Windows Release build and guest_hostname/guest_errors repeated five times each on main 5edae3536. No new Linux or title test is claimed.
AI-assisted: yes. Verify AnyPS5 rather than the native PC hostname. Cover buffer capacities, guard bytes, maximum size_t capacity, null queries, success errno and thread-local isolation. The old implementation fails this regression; the final fixed tree passes the complete Windows build and five repeated runs each of guest_hostname and guest_errors on main 5edae3536. Linux and title tests of this new revision are pending.
AI-assisted: yes. Moves hostname test registration to avoid #2170 overlap. The full current-main build and focused Windows tests pass; implementation and test source are unchanged.
AI-assisted: yes. Add a deterministic backend that overwrites errno on success; verify restoration on success and retained failure errors. Extend native tests for current-time errno preservation. The new regression fails on original ca87abb and passes with this fix.
AI-assisted: yes. Restore guest errno only on successful utimes forwarding and retain delegated failure errors. Register the controlled-backend regression test. Windows library build and focused tests pass; Linux rerun pending.
PPSA11386 calls sceSharePlayInitialize(NULL, 0x1800) at start-up and stops
on any non-zero result, so the console accepts a null heap. The library
keeps no state and does not use the heap; a zero size still throws, as its
error is unknown. libSceSharePlay gets its own CMakeLists for the test.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
An NGG passthrough program on the vertex path whose s_barrier only orders wave 0's GS_ALLOC_REQ produced a workgroup OpControlBarrier, which the Vertex execution model does not allow; SPIR-V validation rejected the module at the program's first draw. In a Vertex-stage program the barrier now emits nothing when the program has no LDS or GDS access and no memory write or atomic, so no wave can observe another. With any of those it throws.
A comparison sample with a 2D-array address on a plain 2D color T# takes the emulated comparison, whose runtime modes were only the 2D-array and cube ones. No mode matched, and materialization threw that the image descriptor was incompatible with the static runtime image interface. The emulated comparison now also gets a 2D mode, as the native modes of a 2D-array access already have; the slice is ignored.
Resize drained the device before reading the surface extent, so a drawable
size that never matches the surface, as with fractional scaling, drained it
on every present.
(cherry picked from commit b4aae246766a108f78cd9ad29d7c874ea036d481)
Export it next to remainderf, stepping to the adjacent float in the
direction of the second argument.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Export it next to lroundf and llround, rounding halfway cases away
from zero into the 64-bit guest long long.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Both parse like strtol and strtoll in base 10. The guest long is 64
bits wide on every host, so atol uses strtoll rather than the host
atol, whose long is 32 bits on Windows.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Export the double version next to __isnan and __isfinite, returning
1 for either infinity like __isinff.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
PokeMMO Prospero (PPSA98001) stops at siglongjmp (pebqbE5ws8s) from
libkernel. The title links its own FreeBSD sigsetjmp, so siglongjmp
follows FreeBSD's amd64 __siglongjmp and its sigjmp_buf layout: when the
save-mask flag at offset 88 is set it restores the mask saved at 72
through _sigprocmask, then reloads rbx, rsp, rbp, r12-r15 and the x87
control word and returns to the saved address with val, or 1 for 0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The ternary forms share minMaxF16 with v_min_f16 and v_max_f16 and follow
the same rule on gfx1036: a subnormal half input becomes a signed zero when
bit 0 of the f16 denormal field is clear, at IEEE=0 and IEEE=1.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
With graphics pipeline libraries a draw places the pixel shader's push
constants at the fixed fragment slot (Graphics::StagePushOffset) and the
draw cache records that offset, but lookupDraw still expected the stacked
offset after the vertex shader's push bytes. Every pixel stage then missed
as a layout mismatch, so no cached draw or draw recipe was reused and each
draw was captured and materialized again.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
FreeBSD's unlink ignores the file's own permission bits, so it succeeds
on a file the guest made read-only with chmod(0400). On Windows the guest
chmod sets the read-only attribute and _wunlink then failed with EACCES,
which sceKernelUnlink and unlink reported. When the path is a regular file
without write permission, make it writable before the single _wunlink and
restore the read-only attribute if the unlink fails, as remove() does
since #2342.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The mutex attribute now keeps its protocol and priority ceiling the way
FreeBSD's libthr does: the ceiling can be set and read only under
PTHREAD_PRIO_PROTECT (EINVAL otherwise), and setting a protocol resets it
to the highest SCHED_RR priority. Mutexes still do not apply priority
protection; that is listed as a silent stub.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit d10c024f8fdbcdfced0187550f272333264f07df)
A map without MAP_FIXED whose hint lies wholly inside a range still
reserved by sceKernelReserveVirtualRange is placed at the hint, as a fixed
map into the reservation would be. Hints outside reservations and ranges
that are partly mapped keep their hint behaviour. The reservation is
checked under the registry mutation, so it cannot be released and its
address reused between the check and the map (MarcZX100, #2687).
Fixes#2687
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A2cY3SsTYnh4u4ZtEr2Adx
Unity reuses the memory of transient render targets: in FPV Simulator
(PPSA10698) an RGBA8 1920x1080 color texture is sampled at the address
that a D32S8 surface's stencil plane occupied earlier. The lookup found
the stencil plane and threw "sampling the stencil plane ... is not
implemented", which ended the presentation.
9d65003c made depth-plane views in a format that cannot hold the plane
fall through to the ordinary texture path. A stencil-plane view whose
texels are not one byte wide cannot hold the 8-bit stencil plane
either, so it now falls through the same way.
AI-assisted: yes
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Registration prepares the variants a shader's registered state
predicts; a draw or dispatch whose state needs another one threw
"prepared shader artifact is missing for the requested static ABI".
Titles that link their own libSceAgc build that state themselves and
never call the driver's static ABI resolution, so their draws differ
from the registered prediction. The missing artifact is now prepared at use, logged, and kept
with the snapshot like the deferred and null pixel cases. SourceHandleFor
now checks that the request refers to the registered code, as
InvocationFor already did.
Uncharted: Legacy of Thieves Collection (PPSA05684) still misses on
main 197ffac9 (after the fixed push slots of #2661) for its own pixel
shaders: "prepared shader artifact is missing for the requested static
ABI: address=85899411712 stage=5 wave=64 pushOffset=128
pushCapacity=128".
sceVoiceSetMuteFlag was NotImplemented with an unknown signature and
sceVoiceGetMuteFlag was not exported. Signatures follow shadPS4's
voice.h: a port id and a bool, and for sceVoiceGetMuteFlag a port id
and a bool pointer. The flag is stored per port like the volume; an
unknown port throws and a null output pointer is an invalid argument.
Uncharted: Legacy of Thieves Collection (PPSA05684) imports
sceVoiceGetMuteFlag (Pc4z1QjForU), so on main it stops at startup with
"undefined symbol: Pc4z1QjForU".
CB_COLOR_INFO COMP_SWAP 1 (SWAP_ALT) on a two-channel format stores the
x and w exports in its two channels (Mesa si_translate_colorswap,
"X__Y"). COLOR_8_8 UNORM and SNORM targets with it bind as R8G8 with
export mapping 0xec; blending into them still throws, since Vulkan
would apply alpha factors to the wrong channel.
Uncharted: Legacy of Thieves Collection (PPSA05684) stopped in its
first level with "unsupported color format 3 number type 0 component
swap 1".
The specialization pass turned OpCompositeConstruct of extracts from one
composite into OpVectorShuffle without checking the types. When the
source is a struct (the {uint, uint} pairs of extended multiplies and
phis of them), the result is invalid SPIR-V ("The type of Vector 1 must
be a vector type"), which the NVIDIA driver (615.71.09) crashes on in
vkCreateGraphicsPipelines instead of rejecting. The fold now requires
both the source and the result to be vector types; a test builds the
struct and vector cases by hand.
Uncharted: Legacy of Thieves Collection (PPSA05684) crashed in the
driver's SPIR-V compiler right after loading its first level.
sceAvPlayerAddSource queues an auto start for a player created with auto
start or without an event callback. When the title enabled its streams
and called sceAvPlayerStart before the controller thread ran it, the auto
start called Start again on the playing player, which stopped it, released
every video buffer, including the frame the title held, and started the
film over. guest_avplayer occasionally caught this as "video frame is not
a texture allocation"; right after the test's own start, 8 texture
buffers were released in 15 of 15 runs.
The auto start now starts the player only while it is still ready, and
checks that under the player lock together with the start.
TestAutoStartKeepsStartedPlayback starts a callback-less player itself and
checks that no texture buffer is released afterwards and that the frame
it holds stays allocated. It fails on every run without the fix.
The Linux patcher writes its own PT_PHDR and PT_INTERP but also copied
the input's. glibc takes the load bias from the last PT_PHDR, which was
the input's entry with the pre-relink address, so such an executable
crashed at start. Both copies also took program header slots, so an
input with PT_PHDR and PT_INTERP and no spare slots failed to relink.
The MinGW CRT delivers argv to main as ANSI code page bytes, and every narrow string became a filesystem path through the same code page, so a game in a directory with characters outside the code page failed to open (#2692). Read the command line as Unicode (wmain via -municode), keep paths as UTF-8 strings, convert at the filesystem boundaries through Io::NativePath and Io::Utf8Path, and run the --autorun output through CreateProcessW.
(cherry picked from commit 30a5181f8a71b177cf2dc03334e39998938b3472)
v_div_fmas_f32 always quieted source NaNs and let a NaN addend
hide the default NaN from an invalid product. Reuse the existing
bit selection helpers, measured on gfx1036 in four float modes.
Co-Authored-By: GPT-6.1 Sol <noreply@openai.com>
RDNA2 min/max flush half inputs to signed zero when input
preservation is disabled, while output flushing alone keeps them.
Apply this to the measured binary forms and pin the hardware rows.
Co-Authored-By: GPT-6.1 Sol <noreply@openai.com>
The background link-time-optimized link used the Pipeline's own layout, which ~Pipeline destroys right after marking the slot released, and ClearCachedPipelines destroys the cached pipelines before ClearPipelineLibraries waits for pending links. A link already past the released check then created its pipeline with a destroyed VkPipelineLayout. On Intel's Windows driver (Arrow Lake iGPU, 32.0.101.8991) agc_driver_vertex_fetch, agc_driver_pixel_interlock and agc_driver_mesh faulted inside igc-default64.dll every run, and agc_driver_prepared_exports and agc_driver_pixel_barycentric_emulation intermittently; all passed with APS5_NO_PIPELINE_LTO=1. The library store's layout, made from the same create info and used for the libraries, lives until ClearPipelineLibraries has waited for the pending links.
(cherry picked from commit 758eeb9fb255d88c41945b299dc3ec5ac1578331)
Validate timeout arguments and track live requests so unknown or deleted
ids return REQUEST_NOT_FOUND. This supplies the remaining absent load
import alongside the dialog Close export; timeouts follow the existing
offline network model and are not stored.
Co-Authored-By: GPT-6.1-sol <noreply@openai.com>
Eager binding fails when a caller imports sceG2PDialogClose.
Add the missing throwing stub alongside the existing G2P exports
and record the assumed Close signature in TechnicalDebt.
Co-Authored-By: GPT-6.1 Sol <noreply@openai.com>
sceAjmBatchJobDecodeSplit queues a split run job with the stream
sideband, as sceAjmBatchJobDecode does for a single buffer pair.
sceAjmBatchJobSetResampleParametersEx with a zero ratio change per
sample sets the ratio as sceAjmBatchJobSetResampleParameters does; a
non-zero change throws, since its ramp is unknown.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Measure the mixed f16/f32 boundary on gfx1036 and pin both factor
slots, half selection, signs and zero f16 addends in the existing test.
Record this measured scope while keeping legacy mad/mac unchanged.
Co-Authored-By: Codex gpt-6.1-sol <noreply@openai.com>
With f32 denormals flushed, RDNA2 flushes a product only when its value
rounded to 24 bits in the current round mode is below 2^-126. The
tiny-product check tested the exact product, so products that round up
to 2^-126 became zero in v_mul_f32, v_mul_legacy_f32, v_mullit_f32 and
zero-addend f32 FMAs. The separately rounded v_mad_legacy_f32 and
v_mac_legacy_f32 keep the exact check, as they are not measured there.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
v_cvt_f16_u16 SDWA only accepted word and dword source selects. Silent
Hill Townfall (PPSA33286) converts the top byte into the low half of a register
and keeps the high half (src0_sel:BYTE_3 dst_sel:WORD_0
dst_unused:UNUSED_PRESERVE), so its shaders failed to recompile. The
generic SDWA source path already zero-extends byte selects, which is the
u16 value the conversion expects, so the rule now accepts every source
select, with whole and partial destinations.
agc_driver_sdwa_cvt_f16_u16 runs four byte selects with dword, WORD_0
and WORD_1 destinations on the GPU and checks the converted halves and
the preserved bits.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit d55dfe8b33c6ee418f40a92e49236f8679a6470b)
A guest path whose bytes the host cannot convert to a native path made
ResolvePath throw std::filesystem::filesystem_error, which aborted the
process. No file can have such a name on the host, so stat reports
ENOENT, as for any other missing file.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
VLC (PPSA85300) imports execvp (0NJDcQ2B8vE) from libkernel and fails
to load because no library exports it. Replacing the process image is
not supported, so it throws if called, like sceSystemServiceLoadExec
does for executable replacement. VLC itself cannot reach the call: its
fork and vfork fail with ENOSYS, so the child that would call execvp is
never created.
sceKernelSignalSema threw invalid_argument for a count of 0. Marvel's
Wolverine (PPSA03671) signals with a count computed as min(a, b) - c,
which can be 0, and ignores the result, so the throw ended the title a
few minutes in. A zero count now adds no tokens and returns SCE_OK, as
in shadPS4 and Kyty. A null semaphore and a negative count still throw.
guest_kernel_errors signals 0 on a full and on an empty semaphore and
checks that the count is unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 7191d34c4d79622ba71087f0da2b75eb81ad2c23)
The Windows TLS rewriter only patches FS-relative instructions that
CodeInstructionCollector decodes. Code that no symbol, unwind range or
root reaches is covered by a linear sweep that starts after int3 padding
and stops at the first instruction beginning with two zero bytes, so
zero-filled bytes are not read as code.
The .NET NativeAOT homebrew titles pad between functions with zero bytes
instead of int3 and have unwind entries for only part of their code. In
Prospero Vibrate (PPSA99920) the sweep ended at the padding after the
first function. Functions reached only through switch jump tables, which
the control-flow graph does not follow, or through pointers it cannot
resolve kept 105 of 670 FS loads unpatched: 84 `mov reg, fs:[0x28]`
stack-guard reads (65 into rax, 19 into rcx) and 21 `mov rax, fs:[0]`
thread-pointer reads. The title faulted reading address 0x28 within a
second of start.
The sweep now treats a zero run that follows a jump, call, return or
trap and ends at the next 16-byte boundary as alignment padding, and
continues at that boundary. A call counts because a function can end in
a call to a noreturn function. Zero fill after straight-line code, or a
whole zero 16-byte block, still ends the sweep as before.
All 670 FS loads in PPSA99920 are now patched, as are the 102 of 685 in
Prospero Explorer (PPSA99109) and the 103 of 718 in Prospero Multi Tools
(PPSA99110). New tls_function_coverage padding cases: loads behind zero
padding after a return, a call and a jump, with 13, 2, 1 and 5 zero
bytes, are patched; a load behind zero fill after a nop is left
unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 0306c677784b6c4272990ce65f6daf9fb2dbc3ca)
sceKernelMkdir checked that the parent of the resolved path exists, but
for "dir/" std::filesystem::path reports "dir" itself as the parent, so
the call failed with ENOENT and created nothing, while FreeBSD's mkdir
(and the POSIX mkdir export) accept it. Strip trailing separators first,
and report ENOTDIR when the parent exists but is not a directory.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 5f74ddba8b346e6c552cf02554d47382708206f3)
The Linux loader takes the unnamed main image for guest code, as the
relinker's ET_DYN output is. Fedora's GCC does not default to PIE, so the
test executables were ET_EXEC and three loader tests failed.
PPSA11386 imports it; on main it is one of the title's three missing
imports, with sceSslGetCaList and sceSslFreeCaList, and the title does not
call it during its start-up. The signature is unknown, so it is declared
without parameters and throws.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
PPSA11386 imports sceSslGetCaList and sceSslFreeCaList and PPSA15508
imports sceSslUnloadCert; on main all three are missing and the loaders of
both titles stop on them. No root certificate bundle is shipped and no
loaded certificate is kept, so an SSL context holds no certificates:
sceSslGetCaList clears the list and returns NOT_FOUND, as sceSslGetCaCerts
does; sceSslFreeCaList accepts only that cleared list; sceSslUnloadCert has
nothing to drop and is assumed to return 0. A null list returns the
invalid-argument code that sceSslGetCaCerts and KytyPS5 return. An unknown
context and a list sceSslGetCaList did not return throw, as their codes are
not confirmed. Signatures and the SceSslCaList layout follow SharpProspero
and shadPS4.
guest_ssl_tests now links libc, as guest_http2_tests does, so an exception
thrown from libSceSsl unwinds with the same unwinder that resumes it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
MemoryIndexBelongsTo walked every instruction of the program and asked
BufferAccessOf, AddressOpcodeInfoOf and ImageOpcodeInfoOf about each one.
TryMakeTableImage calls it for each of a table handle's eight descriptor
dwords and once more for a material key, and PlanIndirectImages calls
TryMakeTableImage for every image instruction whose handle is not planned
yet, so each such instruction could cost up to nine walks of the whole
program. PlanIndirectImages now indexes the instructions that access
memory by their memory index once, before it plans, and
MemoryIndexBelongsTo looks the index up while that pass runs. The
planning pass adds, removes and changes no instruction, so the index
stays valid until the pass drops it at its end. Outside the pass, where
the function has no caller, it walks the program as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Fnrk5Dazjb2tLooVSpdRQq
Add the guest socket descriptor guard to sceKernelFdatasync before any
native syscall, matching the existing sceKernelFchmod pattern. This
prevents high guest socket descriptors (e.g. 0x10000000) from reaching
the Windows CRT _commit, which would trigger the invalid-parameter
handler, and ensures the correct SCE error is returned on all platforms.
Refactor fdatasync_nid_postfix to delegate to sceKernelFdatasync via
PosixResult, removing the duplicated socket guard and native syscall.
Extend GuestFilesystem tests to cover:
- sceKernelFdatasync on a valid guest socket returns SCE EINVAL (0x80020016)
- sceKernelFdatasync on a closed guest socket returns SCE EBADF (0x80020009)
- sceKernelFdatasync on a high invalid descriptor (0x7fffffff) returns SCE EBADF
- closed-ordinary-descriptor tests now run on Windows (verified _commit on
a closed descriptor returns EBADF without provoking the CRT handler)
A recorded dispatch or draw whose use needs completion work goes into the
copied-writer lists (VulkanDevice's copiedWriters, DrawCopiedWriters).
Fills, buffer copies and indirect argument reads treat every write of a
listed object as a store the CPU makes when its batch is reaped, and wait
for that batch. Completion work is also needed for the BDA fault check
alone, and the written sub-ranges of GPU copies are copied back by the
GPU (RecordCopyBacks) before the listing. Such an object has no CPU
write-back (HasCopiedWrites is false), and its stores are ordered by the
barriers like direct writes, yet it was listed and waited for.
Both lists now take an object only when HasCopiedWrites holds after
MarkGpuWrites. The completion action still runs for the fault check.
agc_driver_copied_writer_copy records a dispatch that reads through
global (BDA) loads and stores through a bound buffer into registered
memory, then copies that output on the GPU (VulkanDevice::CopyBuffer)
before the batch ran. It checks that the copy is a transfer between the
host imports without a sync, and the copied bytes. Without the change the
copy waits for the dispatch's batch.
PPSA14632 (LEGO Horizon Adventures): every one of the DMA_DATA copy
waits counted over 75 s came from such writers. Queue 0's GPU-mutex hold
for copies drops from 3.5 s to 30 ms per 10 s, and presents rise from
about 102 to 109 per 10 s.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
An upload of a storage image up to 1 MiB reads the surface's guest bytes
into a baseline (#1599), which a later write-back compares against for
blocks the CPU stamped. When a unit shadow, another image's pending
results or a recorded GPU store covers the surface, that read goes
through the flush hook: the shadow is published and the worker waits for
the GPU, although the upload itself runs on the GPU from the import.
The baseline is now taken only when nothing of that kind overlaps the
surface. Otherwise the image has none, and its write-back keeps the CPU's
bytes for stamped blocks, the rule before #1599. TechnicalDebt's entry
names the condition.
agc_storage_baseline_settled creates a storage image over a watched host
import with a recorded store pending over part of it and checks that the
upload did not submit the open batch; without the change the baseline
read syncs on that store. agc_unchanged_cpu_stamps (#1599) still passes.
PPSA14632 (LEGO Horizon Adventures) refreshes such images about 870
times per 10 s from queue 0 dispatches: the hook's recorded-store syncs
go from about 3400 waiting 4.0 s to about 630 waiting 0.3 s per 10 s,
and a direct dispatch's device call from 356 to 107 us.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A build's sampled image elements often repeat the element just before them
in plan order (the same eight descriptor words). Each still took its own
stage-A record (decode, collect, DCC key read, cache entry) and its own
stage-B lookup.
The lookups run in plan order and nothing runs between two consecutive ones,
so a repeated element's lookup would find the state the previous lookup just
left: the entry it validated or made, current against the same collect. Such
an element now binds the previous element's texture (when the depth-compare
flag matches too, which is part of the cache key), and stage A gives it a
record of the decoded descriptor only. An element repeating an earlier but
not the immediately preceding one still takes its own lookup, as does any
element after a storage image lookup.
A resource build decodes every image element's eight descriptor words,
describes the surface (a mip layout with a vector of levels) and finds the
sampled texture's cache entry by hashing its key, once in stage A and again
in stage B. Most elements repeat a small set of descriptors from build to
build.
Each worker now keeps a small direct-mapped memo by descriptor words of the
decoded resource, its guest size and view component mapping (pure functions
of the words) and of the cache entry the words last found. The entry is
reused only while the sampled texture cache has erased no entry since it
was found (an erase counter bumped by eraseTexture), so the list iterator is
still valid and still holds the same key; anything else takes the hash
lookup.
Buffer and DeviceBuffer created their VkBuffer at the pool's size class, so
a robust read past the requested size stayed in bounds and returned the
bytes an earlier user of the pooled memory left there instead of an
out-of-range result. Each VkBuffer now has the requested size, created with
the pool's usage, over memory of its class (or its slab slot); one taken
from the pool for a different size is recreated over the same memory at
the same offset. agc_buffer_pool follows a pooled buffer by its memory, so
its class and reuse cases hold with the exact-size buffers.
Intersect copied fixed-point sets without allocating each intermediate.
Preserve sorted results and the existing analysis iteration order.
Compact oversized storage only when publishing a changed set.
Keep an allocating control switch outside the cache key.
Compare complete fixed points over empty intersections, loops and DAGs.
Register the same equivalence fixtures for the allocating control.
A DMA_DATA copy larger than the store limit runs as a transfer between
host imports. When the destination is not registered guest memory (the
GPU can neither read nor write it, so no import exists), the transfer
was refused and the packet drained the device: every batch recorded so
far was submitted and waited for, then the CPU copied.
Such a copy only has to see the source's results. When CopyBuffer
refused for want of an import (path 2) and nothing can still store over
the destination (VulkanDevice::StoresPendingOver: a recorded write, a
label, a copied write-back, a storage image's pending write-back or a
unit shadow), the packet now flushes storage results pending over the
source and waits for the batches that write it (Recorder::SyncThrough),
then copies on the CPU without a drain. Anything else still drains as
before. The [sync] line counts these copies.
agc_driver_tests copies a buffer the GPU just wrote into host memory,
once directly and once behind a label into the destination, and checks
the bytes and that the copy did not drain.
PPSA14632 (LEGO Horizon Adventures) copies 256 KiB of GPU results into
host memory this way, about 50 times per 10 s: device drains drop from
about 50 to none per 10 s.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
TestUnsyncedVideoKeepsUpWithAudio built its player without an event
callback, which enables auto-start (autoStartEnabled = init.auto_start ||
callback.event_callback == nullptr). The controller thread then enables the
streams and starts playback shortly after the source is added, racing the
test's own sceAvPlayerEnableStream. When the controller won,
Source::EnableStream saw started == true and the test failed with "enable
video failed".
The test now passes an Events object, as the other tests that drive the
player by hand do.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Jelly5 (PPSA99505) imports sceAudioOutExGetMonitorInfo (PhnNN0UEV6Y),
sceAudioOutExConfigureOutput (VcE+gXSwFXI), sceAudioOutExOpen
(6X6dp+07h4U) and sceAudioOutExClose (0TfjSulCV2A) from libSceAudioOut
and fails to load because no library exports them. It calls them only
for its optional HDMI bitstream output, and their signatures are
unconfirmed, so they are exported without an implementation and throw
if they are called, like sceSaveDataCancel.
Without MAP_FIXED the address is a hint, as for FreeBSD's mmap: the reservation goes at the hint when that range is free, else at the first free range above it. The hint was dropped, so every hinted reservation landed at the lowest free address.
RetroArch (PPSA99169) reserves its cores' direct-memory heap near 0x2000000000 outside the GPU window (0x200000000-0x300000000) and refuses a range inside it; with every reservation at 0x200400000 the heap was never made, and Dolphin's first aligned allocation returned null.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Every draw into a resident target refreshes the target's storage image
(StorageTexture::Refresh): a pending-registry scan for an alias,
FlushPending over the surface, the collect, the key proof, a
whole-surface UnchangedSince, four vectors and a generation update over
every 64 KiB unit, nearly always finding the image current.
A refresh that finds the image current now keeps a proof: the image's
generation and uploaded keys, its key proof's generation and the pending
registry serial. The next refresh is answered from it when the image's
own state is still that, the surface and its key range are unstamped
since those generations (CollectWrites over both, then one
UnchangedSinceAll), and no other image has results pending over the
surface: the registry serial is unchanged, or the registry is scanned and
holds none (a newly pending image, a flush in progress). The proof is
taken only under GuestMemory::GpuMutex, without a pending alias, with
stable keys and when no other pending image overlaps the surface; any
full refresh drops it. A CPU write, a driver or GPU store, a key store,
another image's pending results or a different image therefore always
refresh in full, and the image's own pending results (a draw into it)
keep the proof.
agc_driver_recorder_tests refreshProofTests: an unchanged image and one
holding its own pending results are proved; a CPU write, a driver store
into the surface, a key store, another image's pending results, a new
image of the surface and a fast clear of the keys each refresh in full.
A geometry (NGG) strip draw on the mesh path assigned every primitive the next index positions without looking for the restart index: a triangle spanning a restart fetched vertex 0xFFFF and was drawn to wherever that read landed, and the strip parity never reset, so after an odd-length strip every triangle's winding flipped and face culling dropped it.
For an indexed strip draw with restart enabled and the restart index present, the draw now records each position's strip start (one past the last restart at or before it) and passes the table after the mesh arguments, flagged in draw word 3. The mesh prologue reads the start of its primitive's last vertex: a primitive whose start lies past its first vertex spans a restart and is made degenerate, and the others take their parity from their strip's start (or the draw's first index). The restart vertex itself fetches vertex 0. The table is built from the index bytes the draw already reads (its copy, or the guest memory an in-place import reads), next to #1968's restart-aware highest-index scan.
The new Mesh.cpp test draws quads as four-index strips each followed by a restart (16- and 32-bit indices, and an indirect draw starting mid-buffer) with front faces culled: every quad must render whole and the gaps between them stay clear. Without the table the first check fails ("a triangle crossed the restart after quad 0").
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#2096 keeps a copy of a storage image's guest bytes so that Refresh can
compare memory the write watch does not cover. The host-import upload
path takes that copy only when the range is unwatched, but the DCC
fast-clear path took it on every upload. Reading the whole surface goes
through the flush hook, so each 64 KiB block of a surface that a unit
shadow covers recorded a publish and waited for the GPU.
The clear path now resolves the surface's host import first (the import
can unwatch the range, as a compared host-pointer import does) and takes
the copy only when the range is then unwatched, like the import path.
Write-watched memory, such as direct memory behind a dma-buf import,
skips it; a range that is unwatched later re-uploads instead of
comparing, as before #2096.
agc_driver_dcc_keys_first_write checks that a fast-cleared upload keeps
the copy exactly when its range is unwatched at that point.
PPSA14632 (LEGO Horizon Adventures) on Linux, RX 9070 XT: the flush hook
waited ~60,000 times per 10 s for these publishes (6.4 s), now ~3,500,
and presents rise from ~24 to ~100 per 10 s.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A stale storage snapshot is revalidated by comparing every tracker block
stamped since it was taken. A title that binds buffers a dispatch writes
in place every frame (the whole declared range stamped, few pages really
changed) makes the graphics worker compare gigabytes per second.
Compares of 1 MiB or more are now split by block across a small pool of
helper threads (hardware threads / 8, from 1 to 4). The compare is memory
bound, so the worker's share roughly halves rather than dividing by the
thread count. A caller that finds the pool busy compares on its own
thread. Results and the patch-in-place rules are unchanged.
agc_driver_graphics_tests parallelCompareTests: per-span results on the
pool and serial paths, and two threads comparing at once.
A stale storage snapshot was compared whole with the guest bytes and, on
any difference, copied whole into a new buffer, although few of the
tracker blocks stamped since the copy usually hold different bytes.
Only the 64 KiB tracker blocks stamped since the snapshot's generation
are compared now. When some differ and the cache holds the only
reference (no recorded draw binds the buffer), those blocks are copied
into it in place; otherwise the range is copied again as before.
agc_driver_recorder_tests drawSnapshotPatchTests: changed bytes in
several blocks and a whole changed block patch the snapshot nothing else
holds, an unchanged one is reused, and one a recorded draw still holds is
copied again with its bytes untouched.
A written V# element bound in place in a host import makes every load a
PCIe round trip. A particle update that loads and stores a 10 MiB pool
(2048x64 threads over 131072 records of 80 bytes), above the 2 MiB written
window, took about 4.8 ms of GPU per frame.
An element the shader also reads (DescriptorBinding::bufferRead), written by
a direct dispatch whose threads times the V# stride cover at least half of
it, now takes device-local staging up to 16 MiB: that dispatch drops to
about 0.06 ms plus 0.9 ms of bulk copies. The coverage test keeps a kernel
that updates about 12k records of a 16 MiB pool in place, where staging all
of it would cost 1.3 ms of copies for a 0.08 ms kernel. Write-only
elements, indirect dispatches and raw (stride 0) buffers keep the 2 MiB
window.
A storage write-back copied the image into a linear buffer and retiled
that buffer; an upload detiled into a linear buffer and copied it into
the image. With storage image access without a format, the retile now
reads each texel from the image through an unsigned integer view of the
element size, and the detile writes it there, so the linear buffer, its
copy and one barrier per move go away. The texels move bit for bit as
before, through the same tiling equations and windows.
This covers the block-unit windows of write-backs and uploads and the
whole-surface direct upload, for 2D, array and cube images.
agc_driver_recorder_tests singlePassTests: a first upload, a write-back
of new texels in every mip and a second upload over the stored memory
leave the images and guest memory byte for byte equal to the copies
through a linear buffer, in plain and write-watched memory.
DescriptorBinding gains bufferRead next to bufferWritten and bufferAtomic,
filled from the resource tracker's per-buffer read flag and kept in the
shader disk cache (format 24).
A recorded draw continued the open render pass only after a draw that
wrote no memory, so every draw with a storage write ended the pass and
paid two barriers. A draw now continues the open pass when its accesses
have no hazard (PassHazards) against those of the draws since the last
barrier. Address-based draws, draws through buffer device addresses and
draws with copied writes count as reading and writing any guest memory.
A draw of other attachments with no hazard ends the pass without a
barrier and begins its own without one, and a hazard break records the
begin barrier alone, towards every graphics stage, instead of the pass
end barrier plus the begin barrier. A pass with a depth attachment or
attachment proxies, or one followed by work after the pass, keeps both
barriers. APS5_FULL_BARRIERS keeps every barrier, as for the other merges.
PassHazards keeps the guest ranges, images and attachments the draws
since the last barrier read and wrote, and answers whether a new draw has
a RAW, WAW or WAR hazard against them. Address-based reads and unknown
writes count as touching any guest memory; attachments of the current
pass are ordered by rasterization and only conflict with other passes or
with sampled and storage uses. Unit tests in agc_pass_hazards.
The Windows sceKernelPread stopped going through pread when it learned to
return errors, and with it lost the random device read, so a pread on
/dev/urandom reached ReadFile on a handle that is not a disk and failed
with ESPIPE. Read random devices first, as the Linux path does.
A pread into a read-only guest mapping now returns EFAULT instead of
throwing, matching Linux and the AIO read beside it, so the Windows
guest memory test checks for that return.
Read rtkit's MaxRealtimePriority and RTTimeUSecMax before touching the
limit, keep a hard limit that already satisfies rtkit, lower the soft
limit to at most half of the hard one, and log when the hard limit is
lowered. rtkit checks the hard limit, so it cannot be left alone when it
is above RTTimeUSecMax; the side effect is listed in TechnicalDebt.md.
Log the policy rtkit applied, SCHED_RR by default.
A draw's storage snapshot was copied again whenever the write tracker
reported a store over its range, although most such stores leave the
bytes as they were: in one title screen about 10k of 14k stale lookups per
10 s held the same bytes (3.5 GiB of copies).
A stale storage snapshot is now compared with the guest bytes, after the
caller's collect of the range, before it is dropped. When they are equal
it takes that collect's generation and is reused; any store after the
collect is stamped newer and fails the next check, as for a fresh copy. A
registry change still drops it without reading, and vertex and index
snapshots are dropped as before.
materializeRegisterClear read the target's DCC keys with CurrentDccKeys on
every draw into a resident target with metadata: a full AllKeysEqual scan
of the key range (52 KiB for a 2432x1368 RGBA8 target), about 440k scans
per 10 s and 7.2% of the queue 0 worker's CPU time in one level.
ProvedCurrentDccKeys answers CurrentDccKeys from a proof kept with the
resident image (DccRangeProof: the key range, the keys and the write
generation collected before the scan), the rule ProvedClearKeys already
applies to a storage image's own keys: the proof holds while the key range
is unstamped since that generation (CollectWrites first, then
UnchangedSince), and a scan becomes the proof only when it read the bytes
(not the pending store memo) and no recorded work still writes them. Every
key writer stamps the range, so a store, a pending result or another key
range always scans again. The proof is used under GuestMemory::GpuMutex
only, and not with APS5_NO_KEY_FAST_PATH.
agc_driver_recorder_tests: an unchanged range is proved, and a key store
stamp, a CPU write, another range or address, and a pending key fill each
force a scan; the fill's result is proved once its batch finished.
Every small host buffer (size classes below 1 MiB: draw and dispatch data,
copied regions, snapshots) had a VkDeviceMemory of its own. When the pool
missed, the worker paid a vkAllocateMemory and vkMapMemory, and the release
thread a vkUnmapMemory and vkFreeMemory, for each one. The frees also stall
the worker's vkQueueSubmit in the driver.
A host-visible buffer of a size class now gets its own VkBuffer bound into a
slot of a shared block (2 MiB, or 8 slots for the large classes) per memory
type, size class and device-address use, mapped once. Releasing it destroys
only the VkBuffer and returns the slot; a block is freed when it empties and
its slab already keeps an empty one. A slot is used only when the buffer's
alignment and the non-coherent atom divide its size class, so the slot
offset is valid for the bind and Invalidate covers exactly the slot. The
pool still retains released buffers by class as before; only what backs
them changed.
A sample whose T# is chosen in one if/else and whose S# is chosen in a later one has descriptor Phis in two blocks. The web walk required every Phi of the tuple to sit in one block, so it gave up and the plan threw "GetImageResource dword 0 is not a valid runtime value". The walk now splits at the Phis of the latest block in program order, and Phis of earlier blocks pass through each edge unchanged until the walk reaches their block.
The hold no longer polls every millisecond and has no switch to turn it
off. A held submission waits on the driver's condition variable until no
flip is ahead of it, or until queue 0 waits for memory that only the
title can store. The queue 0 worker reports the second case from its
memory and rewind waits, and the worker that reaches a flip wakes the
holders.
agc_driver_flip_tests: a flip held behind a rendering wait is released
when the worker goes on to wait for a label only the title stores, with
no further submission.
A draw whose stages bind a BDA table or a fault buffer builds resources
that need completion work, so noteReusable never keeps them and the shared
resource cache never holds an entry such a draw could hit. Building the
content key and probing the cache for each of them was pure cost. Those
draws now skip both (ShaderResources::NeverReusable) and count as
uncacheable.
A title that paces itself on pending flips assumes what hardware gives it:
the command processor consumes a submitted buffer at once, so a pending flip
is a frame whose commands have been read and that only waits for the display.
The queue worker is slower than that. With three flips pending it can still be
recording the oldest frame while the title rewrites that frame's slot of a
three-deep ring of per-frame memory for the frame after next, and the draws
are then recorded with the later frame's (or half-written) constants.
A submission on queue 0 that carries a flip now waits until the worker has
reached the flip of every earlier one. The wait ends early when queue 0 is
blocked on memory that no running or pending work writes, since only the
title can end that wait. APS5_NO_FLIP_HOLD=1 submits as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017MPYE7P9ob12aAE1H4JtHF
Throw when /proc/self/maps cannot be read or parsed, and when the kernel
places a reservation chunk at another address (MAP_FIXED_NOREPLACE
treated as a hint). Remove the unused PageSize and keep the POSIX-only
headers in the non-Windows block.
An unmapped range inside the guest arena stays a PROT_NONE host
reservation, so host munlock succeeds on it. Check the guest mapping
state first, as msync does, and return ENOMEM for ranges the guest
has not mapped or reserved.
CheckUnmappedArenaRangeStaysReserved uses mmap, PROT_READ, PROT_WRITE,
MAP_FIXED_NOREPLACE and EEXIST directly, so the Windows build fails to compile it.
The probe exists to show that an unmapped arena range still rejects a hostile
MAP_FIXED_NOREPLACE mapping, which is Linux-specific behaviour.
Guard the function and its call site with __linux__. The other tests in the file
compile on both platforms.
The Linux implementation of GuestArenaReset released the range in addition to
resetting the memory, while the Windows implementation only resets the memory and
leaves accounting to the caller. Two call sites already release explicitly after a
reset, so every one of them released the same range twice: sceKernelMunmap through
ArenaForget, and GuestHeap::release through GuestArenaRelease.
The two releases acquire the arena lock separately. Contains is span-based, so the
second release still runs, and Release erases every overlapping entry. A thread that
allocates the range between the two calls therefore has its entry erased, and the arena
hands the same guest address range out twice.
Delete the release from the Linux reset so both platforms restore the documented
contract. Every existing caller already pairs the reset with an explicit release.
The arena constructor was Windows-only, so on Linux Available() was false: the guest heap fell back to the host allocator and direct memory landed wherever the kernel pleased. Titles that require a low virtual address rejected the mapping and re-issued the identical request, which surfaced as a start-up assert inside their own allocator.
Seed the window from the mappings that already exist instead of reserving it up front, because a PROT_NONE reservation of the whole range starves the kernel's own top-down mmap, which picks addresses just below the stack inside that range. Commit, reset and set-protection become portable so the arena owns its used ranges on both platforms, and fixed mappings are checked against the arena's used ranges rather than against EEXIST from a reservation.
The window-placement and no-overwrite checks in guest_memory were compiled only for Windows; they now run on Linux too, and the window check fails without this change.
Fixes#692
Since guest threads get exactly the stack size they request, host code
that a guest thread calls runs on that small stack too. sceAgcCreateShader
reaches Driver::RegisterShader, whose frame alone is about 50 KiB, so a
title whose job threads ask for 48 KiB overflows the stack (#2603).
Each guest thread now gets a fixed 1 MiB of host stack below the range it
asked for. The guest still sees the requested size and address range
through scePthreadAttrGet and sceKernelIsStack; only host calls made from
the thread use the extra room. Running the prepared-shader driver tests on
a limited main-thread stack needs between 128 and 192 KiB for shader
registration, Vulkan setup included, so 1 MiB leaves several times that.
Switching to a separate host stack at each HLE entry point would need a
trampoline on every export that may go deep, handling of stack-passed
arguments and unwinding across the switch, and a list of which exports
qualify. A margin below the guest range covers every host call with no
per-export list.
On Windows the committed size was the requested size, and the margin
below it depended on the executable's default stack reserve; requests at
or above that reserve had none. The commit now includes the same margin.
macOS does not run guest threads.
guest_pthread_stack now also calls host code that uses 256 KiB of stack
below each thread's guest range; without the margin it crashes on the
16 KiB thread.
Fixes#2603
scePthreadCancel and pthread_cancel threw, scePthreadTestcancel did
nothing and the cleanup handler exports did not exist.
A cancel request is now recorded on the thread and acted on at
scePthreadTestcancel, the condition variable waits, scePthreadSem and
sem waits, scePthreadJoin and the sleeps, and under their POSIX names.
A thread blocked in one of the waits is woken; a condition wait takes
its mutex back before the thread exits. The thread exits with
PTHREAD_CANCELED through scePthreadExit, which now runs the handlers
of __pthread_cleanup_push_imp and __pthread_cleanup_pop_imp.
scePthreadSetcancelstate and scePthreadSetcanceltype reject unknown
values, and pthread_setcanceltype is added.
Constants, the handler node layout and the order of events follow
FreeBSD libthr. A sleep is not interrupted, file and socket calls are
not cancellation points and the asynchronous type acts like the
deferred one except at the two set calls (TechnicalDebt). New test:
guest_pthread_cancel.
Astro's Playroom (PPSA01325) calls sceRudpActivate during start-up, where the throwing stub stopped it. The host has no reliable-UDP stack to activate, so the export returns success and does nothing, matching the OK-returning stubs shadPS4 keeps for the whole library. The guest_offline_imports test now checks the success path with and without sceRudpInit, and still checks that the sceRudpGetStatus sibling throws and writes nothing.
Leave the Windows identity fields at their stat values when
GetFileInformationByHandle fails on a handle whose basic information
could be read, instead of failing the whole call.
st_ino was always 0 and st_dev differed between stat (drive number) and
fstat (0), so the usual a.st_dev == b.st_dev && a.st_ino == b.st_ino
identity check gave wrong answers. Read the volume serial, the file index
and the link count from the file handle, which works the same for path and
descriptor lookups.
Fixes#2296
ProsperoStore (PPSA99000) imports kill from libkernel and stops at it on
current main. Its crash reporter resets a crash signal to SIG_DFL and
sends it again with kill(getpid(), signal).
kill follows FreeBSD's kern_kill and killpg1 for the only guest process:
EINVAL for signals outside 0 to 128, signal 0 only checks the target,
and the caller's own pid, 0 and minus its pid deliver through raise.
Any other pid, and -1, which skips the caller, report ESRCH.
The registry records whether a range was mapped or protected with GPU
access (0x10/0x20), and an address-based build leaves out ranges without
it, as the console's GPU page tables do. The main image is registered
without GPU access; the title's own mprotect with GPU bits adds it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The hardware turns anisotropy off under ANISO_OVERRIDE when the T#'s BASE_LEVEL equals its LAST_LEVEL. The view's level count is taken after LAST_LEVEL is clamped to MAX_MIP, so a T# with BASE 0, LAST 3, MAX_MIP 0 counted as single-level here while the hardware keeps anisotropic filtering for it (measured by @adriadam10 on the hardware oracle). The pairing now reads BASE_LEVEL and LAST_LEVEL from word 3 (DescriptorSingleLevel).
The per-draw path gets a test, agc_driver_image_sample_aniso_override: one ANISO_OVERRIDE sampler shared by two images in a dispatch, with single-level, mipmapped, mixed, and LAST_LEVEL > MAX_MIP descriptors, checking which pairings run (with their samples) and which are refused as mixed. With the old view-based check the LAST_LEVEL > MAX_MIP pairing is refused.
applyAnisoOverride reads the APS5_NO_SAMPLER_CACHE switch through the same cached flag as addImageBinding instead of calling getenv per sampler per draw.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
With ANISO_OVERRIDE the texture unit disables anisotropic filtering when the
sampled view has a single level. Each draw now records, per sampler, whether
its paired images are single-level or mipmapped. A sampler with the bit and an
anisotropic filter whose images are all single-level is bound as the same S#
with point or linear filtering in place of the anisotropic filters, taken from
the sampler cache. One sampler descriptor serves every image paired with it,
so a sampler paired with both kinds in one draw still throws. Samplers of
mipmapped images keep their anisotropy, as on the hardware.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Mesa sets ANISO_OVERRIDE (S# word 2, bit 29) on every GFX10+ sampler
unless aniso_single_level is on; it turns anisotropic filtering off for
single-level views. Samplers carrying it were rejected. Accept and
ignore it: host samplers are built from the S# alone and shared between
images, so single-level views keep the sampler's anisotropy. The gap is
recorded in TechnicalDebt.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Route oversized compute LDS through a per-workgroup storage buffer.
Retain native shared memory for programs that fit the device limit.
Carry the spill stride through prepared shader results and dispatches.
Preserve the stride in the full-result codec and its round-trip fixture.
Keep the Linux RecompileResult size guard aligned with the added field.
Exercise interleaved workgroups and the device-limit boundary.
This topic contains only the LDS change and its reviewer-requested repair.
The base includes upstream's separately merged NaN result correction.
VOP2 0x06 measured the same as VOP3 0x140 on gfx1010 (RX 5700 XT) in
wave32 and wave64, IEEE 0 and 1 and f32 denormal modes 0 to 3: the
product is rounded before the add, f32 denormal sources, product and
result are flushed in every mode, and a zero factor gives a +0 product.
It now shares the v_mad_legacy_f32 translation with the accumulator as
the addend, and f32DenormalFlushFor returns 3 for it as well. A new
execution test pins the measured results of its VOP2 and VOP3 forms.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The guest_interfaces test now follows guest_thread_callback_abi instead of guest_errors, and the TechnicalDebt entry follows the libc SSE4a trap emulation entry instead of ending the Functional list, where other open pull requests add theirs.
A formatted buffer load or store through a GPU-selected descriptor still
emitted the switch over every buffer format inline, at each access. A
Ghost of Yotei compute pass with ~300 such accesses compiled to 37M SPIR-V
words, and its pipeline took 40 s or more to create.
The switch is now emitted once per component count, direction, d16,
coherence, kind and stop mode, as a SPIR-V function that takes the
descriptor fields, the offsets and the access's PC. When faults stop the
invocation, the helper returns whether one did and the caller stops there.
BDA faults record the PC from BdaInstructionPc, the access's own PC or the
one a shared helper takes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The --deferred-undecodable mode registers shaders through a Vulkan
device but, unlike --deferred, did not check for one first. On the
Windows CI runner, which has no Vulkan driver, vkCreateInstance returns
VK_ERROR_INCOMPATIBLE_DRIVER and the test fails instead of skipping, so
the windows job on main has failed since e5b32bb4. It now returns
VulkanTestSkipped like the other modes.
CB_BLENDn_CONTROL factors 15-18 (BLEND_SRC1_COLOR, BLEND_INV_SRC1_COLOR,
BLEND_SRC1_ALPHA, BLEND_INV_SRC1_ALPHA) threw "unsupported blend
factor". They blend MRT0 with the pixel shader's MRT1 export as the
second source, as RADV and radeonsi program it.
DecodeState maps them to the VK_BLEND_FACTOR_SRC1_* factors and accepts
them only with pixel exports to slots 0 and 1, color writes to slot 0
alone, an MRT1 export with MRT0's export format and shader mask, an
ADD, SUBTRACT or REVERSE_SUBTRACT operation and an identity component
order. The blend is draw state, so the specialized module of the
registered pixel shader moves its MRT1 output to location 0 index 1,
which ValidateShaders requires; pipelines need dualSrcBlend, now enabled
when present. The request field dualSourceBlend is in the module key,
the result memo, the draw key and the serialized request (version 15).
Drop IFF_POINTOPOINT from an address without a peer and compute an IPv4 broadcast address the host omits from the netmask, since FreeBSD always provides the address when the flag is set. Check zero port, flowinfo and padding in the test.
Return -1 with the FreeBSD errno for a null list and for expected host and allocation failures, as getcwd does. On Windows, give an unspecified prefix length a host mask, leave PPP and tunnel adapters without broadcast or point-to-point flags since their peer address is not reported, fall back to if<index> for long NDIS names, and bound the adapter query retries. Compare every entry with the host's own list in the test.
List the host's IPv4 and IPv6 addresses in FreeBSD's struct ifaddrs layout, with FreeBSD sockaddrs and interface flags, from getifaddrs on Linux and GetAdaptersAddresses on Windows. The list is one allocation, so freeifaddrs is a single free.
Move the image unregistration out of the Module destructor into
dlclose_nid_postfix. It runs while the handle is still registered, so a
failure (the image stays leased by GPU work) propagates to the caller and
leaves the module loaded and usable. Images shared by another open
handle stay registered until the last handle closes.
guest_memory now checks that dlclose under a held lease throws and keeps
the module, its symbols and its protection.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A .guest.prx image registered by sceKernelMprotect stayed in the guest
allocation registry after dlclose freed it, so its dead ranges blocked
later mappings at that address and could reach GPU builds.
The Windows module destructor now unregisters the image before
FreeLibrary. Unregistering waits for outstanding leases on the image like
unmap and protect do; if the wait fails the module stays loaded.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
On Windows, sceKernelMprotect accepted only the executable's image and threw
"Memory protection of a foreign image is not supported" for every other
image. Bundled title modules are relinked into DLLs named <module>.guest.prx,
so a title that reprotects its own module pages failed. Dying Light 2's
bundled libNRD does this at start-up (#1794).
DoMprotect now also accepts images whose file name ends in .guest.prx, and
registers their committed pages like the executable's. Other images are
still refused.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
On Windows, sceKernelPread and sceKernelPwrite called the POSIX pread and
pwrite wrappers, which throw on every failure, so a bad descriptor, a
negative offset or a write to a read-only descriptor aborted the process
instead of returning an error as the Linux build does. ftruncate and
sceKernelFtruncate threw on any host failure on both platforms.
Give the Windows sceKernelPread and sceKernelPwrite the same checks and
SceErrorFromErrno returns as the Linux ones, and accept a zero-length
transfer with a null buffer. Report a positioned transfer on a handle
opened without the needed access as EBADF instead of EIO. Return EINVAL
from ftruncate for a negative length and for a descriptor opened without
write access, as FreeBSD does, and pass other host errors to the guest.
Fixes#2423
Generate guest PRX dependencies with adjacent and separated repeats, equal names at distinct string offsets, and invalid-name controls. Check acceptance for Linux and Windows conversion and first-seen dependency order in ELF outputs.
Depends on the production fix by Hannah Lyons (babyhuey), c1a6696872153b8e41608f4cc8ab9a0d36f1d206. Written with AI assistance.
sceFontGetCharGlyphMetrics and sceFontGetRenderCharGlyphMetrics set the
vertical bearing and advance to zero although the glyph is loaded with
FT_LOAD_VERTICAL_LAYOUT. The library's own text writing steps the pen by
that vertical advance, so every character of vertical text was placed on
the previous one.
The vertical metrics now follow the FreeType driver's metrics of a
rendered glyph: FreeType's vertical advance, and for a face without
vertical metrics a bearing of half the horizontal advance left of the
horizontal bearing and of half the advance left over by the glyph height.
guest_system_font checks every metric of a square glyph at 100 pixels.
(cherry picked from commit 04dd69f806379df8b52cfdeb2979f2595973a9b3)
(cherry picked from commit 11f6f9e954756ee8d4a3fc2be574ab3b57ecd7c5)
WindowsPeWriter stored every PE section with SizeOfRawData equal to its
virtual size, so the zero-filled tail of a PT_LOAD segment (p_memsz above
p_filesz, the .bss) was written to the executable byte for byte. A segment
with 256 MiB of .bss produced a 269 MB .exe, while the Linux output of the
same image is under 5 KB.
Each section is now stored up to its last non-zero byte, aligned to the file
alignment, and VirtualSize keeps the full size. The PE loader zero-fills the
part of a section beyond SizeOfRawData (Microsoft PE format, section table:
SizeOfRawData). A section with no non-zero byte gets no raw data and a zero
PointerToRawData.
windows_bss relinks an image with 64 MiB of .bss, checks the section table
and the mapped image, and on Windows runs the executable, which reads the
last word of the .bss.
A bundled module whose file starts with a SELF container magic
(4f 15 3d 1d or 54 14 f5 ee) was skipped like any other non-ELF file,
so the relink succeeded without it and the module was missing at run
time. The guest module scan now throws, naming the module, as
ElfReader already does for a SELF executable.
Fixes#2599
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Palworld (PPSA23126) imports sceSaveDataCancel (e6y3WMBYbiM) from
libSceSaveData and fails to load because no library exports it. Its
signature is unknown, so it is exported without an implementation and
throws if it is called, like sceSaveDataConvert.
A registered compute shader takes its wave size from its header's
specials (dispatch_modifier), and PrepareRegistered read them without a
null check. ps5-opengl registers compute headers without specials, since
it writes the specials pointer only for linked graphics stages, and
passes the wave size in the dispatch initiator instead. Registration
then threw "shader metadata is outside the registered header".
Such a shader is now deferred like the other shaders that can't be
prepared at registration, so it is prepared at its first dispatch with
the wave size from the initiator's CS_W32_EN.
On Windows, getargc and getargv reported only the executable path, read
with GetModuleFileNameA in the ANSI code page, and dropped every other
argument. Linux returns the whole command line from /proc/self/cmdline,
and the relinker's Windows entry stub already passes the guest entry all
arguments converted to UTF-8.
Parse GetCommandLineW with CommandLineToArgvW and convert each argument
to UTF-8 with WC_ERR_INVALID_CHARS, as the entry stub does, so getargv
matches the entry's argv and the Linux behaviour. The Linux path is
unchanged.
DecodeColorTileMode rejected 2D color targets with a nonzero MIP0_DEPTH, so
an array target could only be rendered into through a view of a surface that
declares one slice. radeonsi sets CB_COLOR_ATTRIB3.MIP0_DEPTH to the last layer
of an array target (si_state.c, ac_descriptors.c), and Mesa-built homebrew
emits such targets. The view's slice already moves the target by one slice
size; it is now also checked against MIP0_DEPTH.
On Windows libc exports __cxa_init_primary_exception, std::uncaught_exception
and std::uncaught_exceptions only as _nid_postfix symbols (GCC_GLOBAL_ALIAS
is empty there), so host modules bind libstdc++-6.dll's versions:
- std::make_exception_ptr, which std::promise's destructor uses for
broken_promise, allocated the exception with libc and initialized it with
libstdc++'s __cxa_init_primary_exception, whose header layout differs.
A test that makes one and rethrows it dies with an access violation.
- std::uncaught_exceptions read libstdc++'s thread state, which libc's
throws never touch, so it returned 0 while an exception unwound. The
AGC driver's deferred-label guard relies on it to tell a failed record
from a finished one.
The three are now exported natively, like __cxa_throw and the rest.
__cxa_init_primary_exception also sets the reference count to 0, as the
Itanium ABI does: the exception_ptr created next takes the only
reference, so a made exception is destroyed once instead of leaking.
On Windows std::_Exit ends in ExitProcess, which still runs every module's
DLL_PROCESS_DETACH and TLS callbacks after the other threads are gone. A
static destructor that joins a worker then waits forever: a title that calls
_Exit while libSceVideoOut's flip queue runs (ps5-opengl's retirement
fail-stop) stays alive in FlipQueue's destructor. _Exit and libkernel's _exit
now terminate the process with TerminateProcess, as _exit ends a POSIX
process: no atexit handlers, stream flushes or module destructors.
Guest samplers with DISABLE_CUBE_WRAP (S# word 0 bit 28) are created with
VK_SAMPLER_CREATE_NON_SEAMLESS_CUBE_MAP_BIT_EXT when the device enables
VK_EXT_non_seamless_cube_map; RADV maps the flag to the same bit. The bit
only changes cube filtering, so without the extension the sampler is still
created and only a cube texture sampled through it throws.
The buffer pool's device tier (detiler scratch buffers and the staging
shadows of written guest buffers) retained at most 512 MiB by default.
A title whose working set of these buffers is larger allocates and
frees device memory continually: PPSA14632 (LEGO Horizon Adventures)
had 70761 misses and 69828 evictions per 10 s with 485 of 512 MiB
retained, each a vkAllocateMemory or vkFreeMemory.
The default budget is now an eighth of the largest device-local heap,
and at least the old 512 MiB, so a 16 GiB card retains up to 2 GiB and
an 8 GiB card 1 GiB. APS5_STAGING_POOL_MIB still sets it.
Only the constructor knows the device, so only the budget comes from the
heap. Whether the tier is in use at all (APS5_STAGING_POOL_MIB unset or
not 0) stays a static check, DeviceTierEnabled, which Capacity, Usage and
tierFor share, so the size classes and the tier stay consistent.
On PPSA14632 with 2 GiB the device tier has 1312 misses and 99
evictions per 10 s (1628 MiB retained), and presents rise from about 34
to about 39 per 10 s.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A select of exec and zero gives a mask that is the same for every lane. Reading
it back went through a lane number, which Metal does not have in a vertex
shader, so MoltenVK rejected the shader. The select now stores the lane bit and
marks the pair as a mask, a constant mask word reads as a constant, and the
constant folder removes the bit test on an all-zero or all-one word.
agc_scalar_select_mask runs the translated IR over a wave and pins the lane
values for scc set and clear, a source that differs per lane, the two wave64
halves differing, a destination that is also a source, and constant and
literal sources.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Unknown options are now checked for every pid, with 0x80 and 0x80000040 added, and so are valid combinations such as WNOHANG | WUNTRACED. Each call is repeated with a null status. The TechnicalDebt entry cites kern_wait, which adds WEXITED and WTRAPPED, and lists pdfork.
waitpid returns -1 with ECHILD, or EINVAL for an option FreeBSD does not know, and leaves the status unwritten: AnyPS5 runs the title as one process and exports nothing that creates a child. VLC PPSA85300 stops on main at this import.
FillClearColor refused half and B10G11R11 patterns with a denormal code,
since an implementation may flush them in a clear, so such fills over a
resident image took the CPU path: PPSA01325 fast-clears its 3840x2160
R16G16_SFLOAT target 0x515390000 with a denormal half every frame, a
33 MiB guest write and re-upload (about 85 ms).
ClearKeepsDenormals probes the device once per format: a 1x1 image
cleared with denormal codes (half 0x0001, 0x83ff, 0x0200, 0x8001;
11/10-bit 0x001) and read back. When the bits survive, FillClear takes
the refused pattern again with SmallFloat decoding denormals exactly.
NaNs, float32 denormals and the SNORM minimum stay refused.
agc_driver_recorder_tests clears an R16G16_SFLOAT resident image with a
normal, a denormal and a NaN pattern: the denormal is cleared exactly
when the probe says the device keeps it, and the NaN is refused.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UTjpbMX7BUFMZZaxJE98zV
A shader reported in review binds 40 sampled images of one class, more
than 32. The cap only bounds the compact heap arrays, so no layout or
ABI version changes. The test covers 40 images and a full class through
the constant.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A compute program of PPSA26344 binds 29 sampled images of one class, more
than the 16 a typed sampled heap held, and failed with "shader image heap
capacity exceeded". Heap arrays are declared with their compact length,
so shaders that fitted before keep their layout and SPIR-V.
Bindless image tables keep 16 slots: they were sized by the same constant,
and each slot costs an image entry.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A strided vertex fetch whose index is at or past the descriptor's num_records threw and aborted the title. PPSA01325 draws its fullscreen composite as a 4-vertex strip over a 3-record buffer, so the fourth vertex past the records stopped presentation.
Hardware returns (0, 0, 0, 0) for such a fetch with OOB_SELECT 0 and 1, measured on RDNA2 gfx1036 with tools/hw-oracle in #2241. VertexBufferReadSize now spans the whole fetch, and the staging copy reads the record bytes then zero-fills from num_records * stride onward. Each such fetch gets its own copy, outside the shared copies of its sibling attributes, the host-import in-place path and the draw-snapshot cache. OOB_SELECT 2 and 3 keep throwing past num_records, since each needs its own rule. Recorded in docs/dev/TechnicalDebt.md.
eb06284c7 made a read-only 2D image_load through a 1D T# valid, so the
test no longer pins a refusal. The TechnicalDebt line now says storage
loads still throw.
#1629 makes sceAgcDcbResetQueue throw 'unsupported op bit 0x40' (and for
0x200/0x400). Dreaming Sarah and Astro Bot call it with op bit 0x40 at
start-up, so both abort before the first frame. Main with only #1629
reverted runs Dreaming Sarah normally again (59.9/59.9 fps, no skipped
draws, no flashes).
Measured on RDNA2 (gfx1035) in review: image_store with DIM 2D through a
1D T# writes texel x for every y, including y = 0xffffffff. The T# type
decides the addressing and DIM only how many address VGPRs are read. The
store no longer drops y != 0.
The test now pins the second store at y = 1 with dmask 0x5: texel x holds
its components 0 and 2, components 1 and 3 are 0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A compute program of PPSA26344 writes a 1D storage image with image_store
DIM 2D, and no runtime image mode matched its T#. A write-only 2D storage
access now also gets a 1D mode. A store through it is dropped when y is not
0, as a one-texel-high image bounds it; this mirrors the measured 3D loads
through 2D T#s of #2023 and is listed in TechnicalDebt as unmeasured.
Loads, atomics and image_store_mip through such a pair still throw.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A flip packet names a buffer index; the buffer was looked up when the
command buffer was submitted, so a title that registers its buffers
after submitting the first frame failed with "flip buffer is not
registered". The buffer is now resolved when the GPU reaches the flip,
as the hardware does. A flip whose buffer is still unregistered then
presents black and completes (see TechnicalDebt).
Uncharted: Legacy of Thieves Collection (PPSA05684) submits its first flip before registering its buffers and stopped with "flip buffer is not registered".
AI-assisted: yes. Match the firmware 13.60 probe in PR #1798: names of 32 bytes or more return guest ENAMETOOLONG=63 and preserve the prior name. The complete final change passed a full Windows build and eight focused runtime tests.
AI-assisted: yes. Validated with the final library change on Windows; full build and eight focused runtime tests pass. Expected rename behavior comes from the public firmware 13.60 probe in PR #1798.
In SCE_AVPLAYER_AV_SYNC_MODE_NONE every decoded frame was handed out in
order, so a title that requests video slower than the film's frame rate
fell behind its audio (about 52 s by the end of the PPSA12544 intro).
While the title pulls audio, frames the audio clock has already passed are
now skipped; frames that are not late are handed out as before.
Fixes#1635
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
On RDNA2 (gfx1036) ds_min/max/cmpst(_rtn)_f32 and _f64 flush both operands
when bit 0 of the type's denormal field is clear: field 2 flushes like field
0, field 1 keeps like field 3, in IEEE 0 and 1. The flush decision now tests
that bit. agc_driver_lds_float_atomic_modes adds FLOAT_MODE 0x50 (both fields
1) and 0xa0 (both fields 2), and the TechnicalDebt entry records the measured
fields.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012GLfKH7Sqe51wX254k1XsV
ds_min/max/cmpst_f32 and ds_min/max/cmpst_f64 with their _rtn forms follow the
rule measured on RDNA2 (gfx1036): both operands are flushed to a signed zero
when the type's denormal field flushes, a signalling-NaN data operand is stored
quieted, else a signalling-NaN old value, else a quiet NaN loses, else the
ordered compare decides, and the flushed value is written. The shared f32
path used the host's float compares, so its NaN and denormal results were the
driver's; the shared f64 path lacked the signalling-NaN rules the buffer path
has. Both now use AtomicFloatBits. The translation marks the shared float
atomics with the mode's flush in MemoryInfo.
The new agc_driver_lds_float_atomic_modes pins 32 operand pairs per type in
the titles' mode, with both fields kept and with both flushed. The 64-bit
test's model gets the signalling-NaN rule and one lane kind now feeds the
signalling NaN of its table into min/max.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012GLfKH7Sqe51wX254k1XsV
The f64 unary instructions, v_ldexp_f64, v_div_fixup_f64 and
v_cvt_f64_f32 always quieted a signaling NaN. The hardware does that
only with IEEE_MODE=1, as #1867 measured for the f32 and f16 forms;
with IEEE=0 the NaN comes out unchanged, measured on gfx1036.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The ±6.1e-5 lanes take the driver's cosine of a tiny angle, where RADV
returns the hardware's 0x3f7ffffe, NVIDIA 0x3f7ffffd and Intel ANV
0x3f800000 (reported by Kikoarl on #2380), so the lanes tolerate 2 ULP
instead of 1.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012GLfKH7Sqe51wX254k1XsV
After #1979 the cycle is exact, but the driver's sine of the radians loses
tiny results to its absolute error, and near the zeros of sine at k + 0.5
and of cosine at k ± 0.25 the f32 radians drop the low bits of the cycle.
Measured on RDNA2 (gfx1036), the hardware returns the input times 6.2831826
for every input below 2^-12 of a turn and stays within 7 ULP of the f32 2π
times the distance at the half and quarter turns.
Within 2^-12 of a turn from a zero the lowering now returns the linear term
of the exact distance to that zero, with the hardware's constant at the
integers and the f32 2π elsewhere; everywhere else it keeps the driver's
sine or cosine of the radians. agc_driver_sin_cos_near_zero pins 32 inputs
per type from the oracle with per-lane tolerances.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012GLfKH7Sqe51wX254k1XsV
A context without a device, such as the graphics test's mock, leaves state.device equal to context.device, so refreshImports never loaded destroyBuffer and freeMemory and retiring an import called a null pointer.
The pin waiter now receives the mutation range, drains GPU work and retires every overlapping host import before the unmap callback releases the pages. Deferred import retirement keeps only the device handle and the Vulkan destroy functions, so the static import registry no longer owns a Context or its BufferPool past the device.
A compute shader in Astro Bot samples an image as 2D while the bound
T# is 1D, and the recompiler refused it. Offer a 1D runtime mode for
sampled images used only by image_sample_lz and mip-less image_load,
which reads the texel the first coordinate selects, as measured on
RDNA2 in #2237.
With FP16_OVFL=1 the hardware writes an f16 result that would round to
infinity as the largest finite value of that sign, unless a source
operand is infinite. v_div_fixup_f16 keeps the infinity it writes for a
zero denominator or an infinite numerator. Measured on RDNA2 (gfx1036)
at IEEE=0 and IEEE=1 with f16 denormals kept and flushed.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A 64-thread compute kernel (69 dwords) stores pattern[min(i % s9, 3)],
the pattern in s4..s7, to dword i of a 4-byte-stride buffer for every
i below min(launched threads, s8, num_records), with the group id in s10.
PPSA01325 runs it every frame to clear DCC keys, HTILE and other
metadata. It ran as a shader; a direct dispatch with a period of 1, 2
or 4 dwords, whole 16-byte units and a 16-byte aligned destination now
goes through fillBuffer like the existing fill kernel (clear path,
FillBuffer, key memo). APS5_NO_FILL_HLE turns both off.
tests/execution/DwordPatternFill.cpp checks the matcher and runs the
kernel through Submit for matched and unmatched cases (periods 3 and 5,
partial counts, record clamps, a partial dispatch); with
APS5_NO_FILL_HLE=1 the same expectations hold for the translated kernel.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UTjpbMX7BUFMZZaxJE98zV
Both dominance analyses start from sorted unique sets and intersect them.
Insert the current block ID with lower_bound instead of sorting again.
Preserve every intermediate set and the existing ID-order iteration.
Keep the generic unsorted helper and all graph rewrite logic unchanged.
Compare both analyses against the original algorithms on six raw CFGs.
Cover finite loops, closed SCCs, unreachable islands and permuted IDs.
With SDL's disk driver no pad audio device is open, so AudioOut2MixPorts
runs with no pad output. A vibration port and a mono pad speaker port
are pushed together: only the speaker's data reaches the main mix, and
a vibration port alone leaves it silent.
Without a DualSense audio device every AudioOut2 port was mixed into the
main output, including PAD_VIBRATION ports. Their waveforms drive the
controller's actuators on the console and are never heard; mixed into
the speakers they made PPSA28997's combat about 20 dB louder than its
dialogue. Vibration ports are now left out in that case; pad speaker
ports still go to the main output.
The f16 min/max/min3/max3/med3 now pick NaN results on the bits and
quiet a signaling NaN only with IEEE=1. With IEEE=0, v_min_f64 and
v_max_f64 return the other operand for a NaN, not a quieted NaN. This
completes #1826 for these forms, as measured on RDNA2 (gfx1036).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
With IEEE_MODE=1 the hardware quiets a signaling NaN that v_cubesc,
v_cubetc or v_cubema_f32 returns, keeping the sign the instruction gives
it; with IEEE=0 the bits pass through. Same rule as #1867, measured on
gfx1036 in both modes.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Neither existed under the guest root, so opening them failed and titles
that then read from the descriptor crashed. Opening either for reading
now returns a descriptor whose reads, positioned reads and vector reads
fill the buffer from the host's random device, as
sceRandomGetRandomNumber does. Opening them for writing throws.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The executable is linked at a fixed base with a __PAGEZERO that fills the space below it, and
the guest assumes 16 KB pages. With MH_PIE set dyld slides the image by a multiple of 4 KB,
which breaks that assumption and ends in a SIGBUS (KERN_PROTECTION_FAILURE) in about half of
the launches under Rosetta 2. Clearing the flag keeps the image at its linked address.
test_macos_output.py now reads the header flags and checks that MH_PIE is clear.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Since #1178 the SPI_PS_INPUT_ENA/ADDR emplace calls did nothing, so the registered variant had no pixel inputs, and null pixel draws took the last pixel shader's user SGPR count and the bound targets' export mapping. Registration now overwrites those registers, the draw gives the null program no user SGPRs and an empty export mapping, and the at-draw preparation is left for wave32 only. Root cause found by @Raghav314.
A draw that rasterizes both faces with different front and back depth
bias was rejected, since Vulkan has one depth bias per draw and the
cull mode is pipeline state. Such a draw is now recorded twice in the
same pass: front faces with the back faces culled and the front bias,
then back faces with the front faces culled and the back bias. A face
whose offset is disabled gets no bias.
#2279 made agc_driver_f32_arithmetic_nan fail on NVIDIA (Windows):
v_fmamk_f32 returns a quiet NaN where the signalling one is expected
(lane 1 no float mode v_fmamk_f32 is 0x7fc12345, expected 0x7f812345),
5/5 runs on main and 5/5 passing with #2279 reverted. CI's Windows job
skips the AGC execution tests, so it did not show there.
v_fma_legacy was rewritten as a split multiply and add after #2279; it
keeps that and gets back the invalid-product argument it passed before.
Fixes#2581
scePthreadCreate on Linux started guest threads with std::thread, so every
thread got the host default stack (RLIMIT_STACK, usually 8 MiB) whatever the
title asked for. A thread that asks for more than that and uses it crashes
with SIGSEGV. Once the thread ran, scePthreadAttrGet also reported the host
size instead of the requested one.
Create the thread with pthread_create and a stack of the requested size plus
what glibc keeps at its top (static TLS, the thread control block and the
start_thread frame). That reserve is measured once with a probe thread.
The thread reports [top - reserve - size, top - reserve) as its stack, as
Windows reports [top - size, top), and the creator waits until the range is
set, as on Windows.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
With f32 denormals flushed, as in the 0xc0 float mode titles use, the
hardware flushes all three v_mullit_f32 sources and its product like
v_mul_f32, so a positive denormal limit selects -FLT_MAX and a tiny
product gives zero. The recompiler kept denormals in every mode.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
With IEEE_MODE=1 the hardware quiets each signaling NaN half that
v_pack_b32_f16 packs, after neg and abs; with IEEE=0 it copies the bits.
The recompiler always copied them. Same rule as #1867, measured on
gfx1036 in both modes.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The registered-shader decode ran before the deferral decision, so a single
instruction the decoder does not implement ended the title at registration
instead of deferring the shader to its first use.
Give the decoder's unsupported-instruction failures their own type
(UnsupportedInstructionError): the five family-scoped SOP opcode defaults and
the MIMG opcode lookup (which now names the opcode) raise it, and the decode in
PrepareRegistered defers the shader with that reason when it hits one. Any
other exception still throws, so the corrupt-code contract is unchanged.
PPSA26344 hits this: unsupported MIMG opcode 229 (0xE5), issue #1941. The
failing word is reproducible offline from the title's own shader store
(cache_ps5/all_shaders.xpps, code word 260), which is how the fixture was
checked before it was written.
Scope: this defers the scalar-family and MIMG opcode gaps. The vector (VOP*)
and memory (SMEM/MUBUF/MTBUF/FLAT/DS) opcode lookups, the modifier-level
"not supported" throws and the malformed/truncated throws still abort.
Widening the deferral to those is a policy call for the maintainers.
Stacked on #1836 (the deferral machinery); re-stacked on its current head so
this diff no longer carries #1836's earlier broad catch. Once #1836 lands only
the decode change remains.
Written with AI assistance.
A title that opens a main and an overlay port flips both, and every flip
was presented as a whole frame, so the window alternated between the two
buses: PPSA01736 shows its intro on the main bus and an empty overlay,
and the intro blinked black on every other frame. Overlay flips now
complete and signal their events without being shown. Compositing the
overlay over the main output is not implemented.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The guest filesystem root and the save data directory are taken from the
current directory, so an executable started from another directory does
not find its title files and keeps saves elsewhere. Make the executable
directory the current directory in the Windows entry stub before the
system libraries are loaded.
A pixel shader whose early discards branch to one small ending block
gave that block several conditional predecessors outside any loop, and
structurizing failed. When the regular pass throws, the structurizer
restarts from the original graph and first gives each conditional
predecessor its own copy of such a return block (within the clone word
floor and a budget of four clones per block).
Uncharted: Legacy of Thieves Collection (PPSA05684) stopped at such a pixel shader with "conditional block 15 has no structured merge".
A typed storage-image heap held four elements, and a dynamic-mip storage
image takes four of them for its mip slots. A shader with a dynamic-mip
storage image and any other storage image of the same class needed five
and failed with "shader image heap capacity exceeded". The mip slots
per dynamic image stay four (StorageMipSlots); a storage heap now holds
sixteen elements (StorageHeapCapacity). Heap arrays are declared with
their compact length, so shaders that fitted before are unchanged.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Vulkan counts the framebuffer colour attachments against maxPerStageResources
for the fragment stage. The draw's attachment count now seeds the fragment
stage's resource total in ShaderResources, so a pixel shader whose buffers and
images fit the limit but whose pipeline exceeds it throws instead of creating
an invalid pipeline. The draw resource cache key carries the attachment count,
so a cached build is never reused by a draw with more attachments.
RADV clears depth by rewriting HTILE with its clear_htile_mask kernel
(ZMASK to 0) instead of writing depth memory. Recognize that kernel and
clear the matching host depth surface to DB_DEPTH_CLEAR when it is next
bound; the depth target now records its HTILE address.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The f16 forms of v_fma_mix computed an f32 FMA and converted it to f16,
so a result that lands on an f16 tie after the f32 rounding was rounded
twice. RDNA2 rounds the exact product-sum once. They now use the exact
f64 path of v_interp_p2_f16 and flush f32 sources like v_fma_mix_f32.
They follow the f16 round and denormal fields on hardware, so another
f16 mode throws, as packed f16 does.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
DEPTH_CLEAR_ENABLE (DB_RENDER_CONTROL bit 0, Mesa sid.h) makes the DB
write DB_DEPTH_CLEAR wherever the draw covers. Such a draw is now
rasterized as submitted with an always-passing depth write, no depth
bounds, no depth bias and the viewport depth range pinned to the clear
value, so it touches exactly the pixels the title's geometry covers.
Clearing a read-only or absent depth plane throws, at decode and in the
draw precheck. Depth and stencil clears combine; copy, resummarize and
decompress draws still throw.
Uncharted: Legacy of Thieves Collection (PPSA05684) clears with draws
that set DB_RENDER_CONTROL to 0x3 and stopped with "depth/stencil
clear, copy, resummarize or decompress draws (DB_RENDER_CONTROL) is
unsupported (register 0x0 = 0x00000003)".
libc implements the C++ exception runtime and already defines
exception_ptr's out-of-line members, so host modules that link libc
reference-count libc's exceptions. Code inside libstdc++-6.dll does not
see those definitions on Windows: std::__future_base::_Result_base's
destructor is one of them, and it releases the stored exception with
libstdc++'s own exception_ptr::_M_release. That decrements a count 0xa0
bytes before the object, while libc keeps it 0x80 bytes before, so the
decrement lands in the heap memory in front of libc's allocation and
the exception is never freed.
VideoOutDriver reports the presentation thread's start-up failure
through a std::promise, so video_out_flip_lifetime sometimes ended with
0xc0000374 when SDL's offscreen driver refused the window.
libc now defines the base destructor, with a _nid_postfix twin so that
the NID-patched libc keeps its name, and modules bind it there.
AMD's Windows driver advertises shaderSignedZeroInfNanPreserveFloat16,
yet OpFConvert from f32 to f16 returns +0 for -0 inputs and for negative
values that round to zero, so six f16 execution tests fail there since
the native conversion was added. Conversion never changes the sign, so
the native path now takes the sign bit from the f32 source, as the
emulated path already does.
Fixes#2483
The x86 decoder had no XOP (AMD 8F escape) support and read 8F E8.. as
POP r/m, losing instruction boundaries. Since #2238 roots code addresses
taken by RIP-relative LEA, code analysis reaches CPU-dispatched XOP
paths, such as the vpperm routines in Stremio Plus (PPSA74126), OpenNOW
(PPSA99082) and PPSA99420, and their Windows conversion failed with
"Code analysis: overlapping instruction boundaries". Decode 8F followed
by a map of 8 or more as a three-byte XOP prefix (8F can only be POP r/m
with ModRM.reg 0) with a ModRM operand, an 8-bit immediate in map 8 and
a 32-bit immediate in map 0Ah, in both the length and operand decoders.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A directory opens and tellg reports an enormous size, so the relinker
tried to allocate it and failed with std::bad_alloc. Require a regular
file and a non-negative size before allocating the input buffer.
The unused-filter=1 entry point collector translated DT_INIT_ARRAY and
DT_FINI_ARRAY through PT_DYNAMIC and skipped arrays placed before it, which is
the usual linker layout, so constructors and destructors were not roots and
imports used only by them were dropped. Arrays are now translated through the
PT_LOAD that holds them, slot targets come from RELATIVE relocations when
present, and DT_PREINIT_ARRAY is collected as in the strict filter.
A zero, unaligned, overflowing or out-of-range length reached the physical pool, where a sub-page length silently freed a whole page and fragmented the block map, and an out-of-range length walked the page table past the backing store. Validate alignment and bounds up front and return EINVAL instead.
- DISPATCH_DIRECT and DISPATCH_INDIRECT with the SHADER_TYPE header bit,
which RADV sets on every dispatch, as other packets already may.
- FORCE_START_AT_000 in the dispatch initiator. The driver never offsets
dispatches by COMPUTE_START_X/Y/Z, so it already starts at zero.
- LOAD_CONTEXT_REG_INDEX, LOAD_SH_REG_INDEX and LOAD_UCONFIG_REG_INDEX in
their consecutive-register form (a register offset and a count, DATA_FORMAT
0), which RADV uses to load DB_STENCIL_CLEAR and DB_DEPTH_CLEAR from an
image's clear values. The offset-and-data form keeps its old checks.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RADV builds its own command streams. Its preamble and draws use packet
forms the validator refused:
- CONTEXT_CONTROL that only updates the load and shadow enables to zero,
which RADV emits when it does not shadow registers. It is a no-op.
- SET_UCONFIG_REG_INDEX of VGT_PRIMITIVE_TYPE with index 1, next to the
index-2 VGT_INDEX_TYPE write already accepted.
- WRITE_DATA to memory (DST_SEL 5) from the graphics queue.
- DMA_DATA with DST_SEL NOWHERE, RADV's L2 prefetch of shader code. It
writes nothing and resolves to an empty store.
- EVENT_WRITE PIPELINESTAT_START, PIPELINESTAT_STOP, SQ_NON_EVENT and
VGT_FLUSH. The statistics counters are not modelled, and the events that
would read them stay rejected.
- ACQUIRE_MEM over the full 48-bit range (COHER_SIZE_HI and COHER_BASE_HI).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
VOP3 0x140 is v_mad_legacy_f32 on gfx10.1. Measured on gfx1010 (RX 5700
XT) in wave32 and wave64, IEEE 0 and 1 and f32 denormal modes 0 to 3, it
rounds the product before the add, flushes f32 denormal sources, product
and result in every mode, and gives a +0 product for a zero factor.
The translation multiplies, applies the zero rule and flushTinyProduct,
adds, and keeps the nanResultF32 wrapper. f32DenormalFlushFor returns 3
for the instruction before reading FLOAT_MODE. A sum of two zeros takes
its sign on the bits, as lavapipe returned -0 for +0 + -0. The execution
test pins the measured results and compares NaN bits exactly. RDNA2
(gfx1031), where the encoding is v_fma_legacy_f32 and fused, is noted in
TechnicalDebt.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Importing guest memory collected the write watch over the whole import,
which write-protected every page of it, so pages nothing had collected yet
took a write fault on their first CPU write. The import now collects only
the parts a collect already protects (GuestWriteWatchCollectArmed). The
rest stays unprotected until something collects it, which reports it as
written; a GPU write recorded into such memory (MarkWritten) first collects
the never-collected parts of its 64 KiB blocks
(GuestWriteWatchCollectFresh), so that report is older than the GPU's
results and does not read as a CPU write over them.
Those pages are stamped as written and their CPU pages are noted, but the
block's CPU stamp is left alone, so a GPU label record or a pending storage
image elsewhere in the block does not see a CPU write. StoredOver and
UnchangedSinceCollected read the noted pages whenever a block has them, so
a store to such a page is still seen by a check over that page. The extra
generation step happens only when something was collected; a block whose
write protection fails is unwatched and the write returns no generation.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The padded, exact-fit and invalid-conversion swprintf checks come from #1512.
Co-Authored-By: Noface07 <yuvrajsoni7@gmail.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01ULF8vCbJFyzxVsW3SfiWC5
flock_nid_postfix threw on every failure, so a LOCK_NB request on a lock
held through another descriptor ended the process. FreeBSD flock(2)
returns -1 with EWOULDBLOCK there. Return guest errno 35 when LOCK_NB is
set and the host reports the lock as held: ERROR_LOCK_VIOLATION on
Windows, EWOULDBLOCK on Linux. Other failures still throw.
A formatted buffer load or store whose format comes from a prepared
descriptor emitted the switch over every buffer format inline, at each
access. The switch is now emitted once per resource, component count and
direction as a SPIR-V function taking the index, offsets, record count and
storage buffer offset, and each access calls it. Specialization still folds
the switch to the descriptor's format. Accesses through GPU-selected
descriptors keep the inline switch.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
FreeBSD's __sflags reads the access character, then b, +, x, e and v, and
stops at the first other character without failing, so "rt" opens for
reading. On Windows fopen inserted b after the first character and passed
the rest on, so "rt" became "rbt", the CRT rejected it with EINVAL, and
fopen threw. vkQuake opens history.txt with "rt" at startup. freopen
accepted only a fixed list of modes.
Both now parse the mode like __sflags. An unknown access character, or x
on a read-only stream, returns NULL with EINVAL; x on an existing file
returns NULL with EEXIST; e and v are accepted and ignored.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
An anonymous private mapping with MAP_ALIGNED_SUPER (FreeBSD MAP_ALIGNED(1))
is placed on a 2 MiB superpage boundary. Yamagi Quake II requests it for its
hunk allocations of 2 MiB and more.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
SDL queries the SCHED_OTHER range before it sets a thread priority. It gets
the same 256 to 767 range as SCHED_FIFO and SCHED_RR, which
pthread_setschedparam applies regardless of policy. The console's answer is
unverified and recorded in TechnicalDebt.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A global_store_dwordx2/x3/x4 was split into one StoreAddressU32 per
dword, each with its own write-pointer lookup and its own 4-byte store.
Across a wave those are strided partial writes, which are very slow to
host memory: RADV's clear_htile_mask kernel (one dwordx4 load and store
per thread over 640 KiB) took ~9 ms. Multi-dword stores are now one
StoreAddressU32xN; the emitter probes the whole span once and, when it
is mapped writable and 4-byte aligned, stores it as one vector (16- or
8-byte aligned) or as consecutive dwords, noting each written page.
Anything else takes the per-dword path as before, so faults and
range-spanning stores behave as they did. APS5_BDA_DWORD_STORES=1
restores per-dword stores.
AI-assisted (Claude Code).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
0xffff1000 is a type-3 NOP whose count field is 0x3fff, which the command processor reads as a single dword. The driver read it as a 16385-dword packet and rejected every sceAgcDcbResetQueue envelope. Pm4::FillerPacket now also matches exactly this header, so every packet walker steps over it as one dword.
Refs #1581
The case reached the compared path once host imports kept a guest-byte
snapshot, so it passed without the fix. It now runs over watched memory
without an import on every host, and over a watched host import on Linux
(the font atlas's path, with no valid snapshot), checks the image is on
that path, and runs as its own ctest test.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A write-back keeps the CPU's bytes for a block stamped since the layer's
generation. Stamps can fall on bytes the CPU did not change (PPSA12544's
font atlas: the stamped pages still hold the allocator fill the image was
uploaded from), and the GPU results there were lost. Images up to 1 MiB
keep the guest bytes of their generation point; a stamped range that still
equals them stores the GPU results.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The CFG/GOT filter's entry point collector checked header and symbol
positions as `offset + size > file size`, which wraps for offsets
near 2^64. A section header or symbol table offset there passed the
check, read bytes before the input buffer, and could reach an
in-bounds table through the wrapped address, adding entry points that
keep relocations alive. Check every position by subtraction instead.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A CMASK that was neither all fast-cleared nor all expanded threw, so a
title that fast clears a target and then draws over part of it before the
eliminate could not run past it.
The CMASK is now decoded tile by tile: one nibble per 8x8 tile in
1024x512 metablocks of 4 KiB, addressed with the pipe-aligned
SW_64KB_Z_X CMASK equation for 16 pipes (addrlib
GFX10_CMASK_SW_PATTERN, Gfx10Lib::HwlComputeCmaskAddrFromCoord). Code 0
stores CB_COLOR_CLEAR_WORD into the tile's texels, 0xF keeps them; any
other code, a cleared tile of a DCC target and any other CMASK swizzle
mode throw. Edge tiles are included and the CMASK is then stored as
expanded.
For a resident target whose CMASK is host-imported the clear is a
compute pass recorded in stream order: it reads the nibbles from the
import, stores the clear into the cleared tiles through a UINT view of
the image and fills the CMASK expanded, so nothing waits for pending GPU
writes of the CMASK. Invalid codes it reads set a bit that the next
materialization throws on. Otherwise the CPU reads the CMASK and stores
the cleared tiles into guest memory, or fills the resident image when
every tile is cleared. A CMASK nothing has stored over since it was
expanded is skipped.
The decode requires CB_COLOR_ATTRIB3 FMASK_SW_MODE SW_64KB_Z_X with
CMASK_PIPE_ALIGNED, a 4 KiB aligned writable CMASK and texels of at most
64 bits, and CMASK_BASE_EXT is part of the draw key.
Tests: CMASK sizes and nibble addresses against addrlib; decode
rejections and draw key coverage; CPU eliminate of full, partial,
out-of-range-nibble and DCC decompress cases and invalid codes. The new
agc_driver_cmask_fast_clear GPU test draws over fast-cleared and partly
fast-cleared linear and SW_64KB_R_X targets, eliminates partial, edge
and checkerboard CMASKs through the compute pass and checks that an
invalid code it reads throws.
A wave64 vertex program reduces a float key across the wave with the same shape as the integer half-wave reductions: keys selected under the all-lanes EXEC with -inf (v_max_f32) or +inf (v_min_f32) outside the entry EXEC, four DPP row_shr steps, v_permlanex16_b32, then v_readlane_b32 of lanes 31 and 63, combined by one more v_max_f32. On a 32-lane host subgroup the read of lane 63 threw that it is outside the subgroup.
The matcher now also accepts FPMax32 and FPMin32 steps, with their BitCasts and the f32 denormal flush the float mode may add around operands and results; the flush leaves the -inf and +inf keys unchanged. When the entry EXEC proves that lanes 32-63 hold the constant key, every step there combines that constant with itself. The existing v_max_f32/v_min_f32 lowering returns the operand unchanged for equal operands (NaN included), so lane 63 is exactly that constant, and the read returns it. No subgroup float reduction is emitted. Lane 31, the lanes on a 64-lane subgroup, and the combining v_max_f32 keep their existing lowering. So NaN and signed-zero results are the ones the shader computes today.
On Linux, sceKernelSync called sync(2), which blocks until all
dirty pages are written back. Unrelated host I/O can stall guest's
sceKernelSync, causing the guest_kernel_sync test to stall for minutes.
Share the Windows implementation instead.
Fsync each recorded path and its changed parent directories.
Use st_mtim and coarse realtime mark on Linux in place of FILETIME.
Every draw copied its index range and its vertex ranges out of guest
memory into new host buffers (CopyDrawInput) before the GPU could read
them. A range that lies in a host import is now bound from the import
itself, at its offset in it, the way the console's input assembler reads
guest memory directly.
A range is bound in place only when the batch has no write over it that
could still land (a recorded or in-flight GPU write after the pending
results are flushed, a completion label, a queued label store, a copied
buffer a dispatch or draw writes back) and its offset in the import is a
multiple of what the reader needs: the index size, or the strictest
format alignment of the vertex attributes that share the range
(VertexCopyPlan::alignments; a copy starts its range at offset 0, so
only an in-place range can be misaligned). Otherwise it is copied as
before. The read is noted on the batch (Recorder::ReadKind::DrawInput),
so a CPU store into the range waits for the draw like the other in-place
readers, and the barrier validation sees it. Imports gain vertex and
index buffer usage.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
FreeBSD sys_flock takes LOCK_UN over LOCK_EX over LOCK_SH, ignores other
bits and fails with EBADF when none of the three is set. The operation
reached the host unchanged, so malformed operations differed between
hosts: Linux rejects them with EINVAL and the Windows lock tracking threw.
Decode the operation before either host sees it.
flock mapped each call straight onto LockFileEx/UnlockFileEx, whose
byte-range locks stack per handle. LOCK_UN without a lock failed with
ERROR_NOT_LOCKED and aborted the process, a repeated LOCK_EX blocked
forever, converting between shared and exclusive failed, and a repeated
LOCK_SH needed as many unlocks.
Track the one lock each descriptor holds: a repeated mode and LOCK_UN
without a lock succeed, a conversion releases the old lock before taking
the new one, and close forgets the descriptor's lock.
On Windows the existence check accepts "file/", then _wstat64 rejects
it and FillFileStat throws, aborting the title. On Linux the host
reports ENOTDIR, which became ENOENT. A trailing slash on something
that is not a directory now returns ENOTDIR on both hosts, as on
FreeBSD.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Format the address bytes in network order as dotted decimal into a
per-thread buffer, like FreeBSD's inet_ntoa returns its static one.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
FreeBSD's gettimeofday stores the time only when tp is not null, and
clock_getres the resolution only when res is not null; both succeed
otherwise. Both threw, which aborted a title probing a clock or only
asking for the time zone. Write into a local result instead.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A 3D image access bound to a 2D T# matched no runtime image mode and
threw "image descriptor is incompatible with the static runtime image
interface" (Sonic Origins, #2023). CouchRiv measured RDNA2 (gfx1036)
with the hardware oracle: image_sample_lz with dim 3D ignores r and
reads the (u,v) texel, and image_load reads the (x,y) texel when z is
0 and returns zero otherwise. Add a 2D mode for 3D accesses whose only
uses are those two forms, and make the load return zero for z != 0.
Every other access keeps throwing.
Fixes#2023
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Windows runner has no Vulkan driver, so agc_pipeline_cache (#1331)
fails at device creation: 'AGC graphics: vkCreateInstance: Vulkan
result -9'. Skip with 77 like the other agc execution tests, unless
ANYPS5_REQUIRE_VULKAN is set, as on the Linux job.
video_out_flip_tests writes app0/sce_sys/param.json in the shared test
directory on every run, and every VideoOut window reads it for its
title. Under ctest --parallel one test read the file while another
rewrote it: 'unexpected end of param.json' in video_out_flip_controls.
Give all of them one RESOURCE_LOCK.
The unopened-handle checks call sceVideoOutIsOutputSupported before
sceVideoOutOpen, and that first call builds the VideoOut driver, whose
window creation throws without a Vulkan-capable SDL driver. The throw
left main before the skip, so the test aborted on the Linux runner:
'DisplayWindow: Vulkan support is either not configured in SDL or not
available in current SDL video driver (offscreen) or platform'.
Run those checks inside the same skip.
A jump slot whose r_offset is 0xFFFFFFFFFFFFFFF8 or higher makes
targetGotOrPltAddress + 8 wrap, so the upper lower_bound lands before
the lower one and the call-site loop walks past the end of the map
(SIGSEGV; ASan heap-buffer-overflow in ResolveCallSites).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
guest_video_out_output_mode (#1496) opens VideoOut directly, which needs
a Vulkan-capable SDL window, and aborts on the CI runners that have none.
Skip with 77 like the other VideoOut window tests, unless
ANYPS5_REQUIRE_DISPLAY is set.
AJM's instance and batch tables and APR's file, directory and AMM
state were namespace-scope statics, destroyed by std::exit while the
title's threads keep calling the libraries. A decoder thread then read
a freed instance (an unknown codec), sceAjmBatchWait no longer found
its batch, and APR resolved file ids through a freed vector. The state
is now allocated once and never destroyed.
A port kept only the last PCM buffer set before a push and played it again
at every later push without a new one. PPSA12544 sets one grain of a ring
buffer per call on its object port, often two or three between pushes and
then none for a few, so all but the last of each burst were dropped and the
stale grain repeated: a buzz on loud sounds and after a sound stopped.
Buffers set between pushes are now queued and played in order, at most 16
queued. A push without a new buffer repeats the last one only for a port
that has always been given the same buffer (rewritten in place, as the
timing tests do); a port whose buffers move on plays silence.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Without VK_KHR_fragment_shader_barycentric the recompiler dropped
v_interp_p1_f32 and made v_interp_p2_f32 read the attribute Vulkan
interpolated, but never wrote the I/J VGPRs: a pixel shader that read
them for anything else got 0, and v_interp_mov_f32 of P10/P20 (or P0 of
an interpolated input) needed PerVertexKHR and failed pipeline
validation.
The fallback translation now loads the I/J VGPRs and marks v_interp_p1
and v_interp_p2 with witnesses. After the front-end simplification
(SSA, folding, DCE, read-lane and masked-select elimination),
HostInterpolationChecker requires every p2 to read its J, and its p1's
I, from a barycentric of the parameter's kind, every p1 result to reach
only the matching p2 (through a write masked by the p2's EXEC, whose
inactive lanes keep the old register), no parameter read as raw
vertices and no slot shared by different interpolations. It then
lowers the witnesses to the host-interpolated attribute as before,
reruns DCE and masked-select elimination, and accepts the program only
when no barycentric read remains, so reusing the I/J VGPRs after
interpolating keeps host interpolation and the previous SPIR-V. Any other
program is translated with the barycentric path and marked as emulated:
its barycentrics become vec3 inputs (NoPerspective for the linear
weights) and its per-vertex parameters flat vec4[3] inputs, placed in
the free locations by LayoutBarycentricEmulation. The driver then
inserts a generated geometry stage between the vertex and fragment
shaders that copies the position and parameters, writes the one-hot
weights of each vertex and the three vertices' parameter values, so the
rasterizer interpolates the same weights the extension provides.
The stage needs geometryShader and a vertex shader before it;
tessellation, mesh and rect-list pipelines, other primitive types and
interfaces beyond 32 locations or the device's geometry limits throw.
CompiledShaderArtifact carries the emulation (shader disk cache format 21).
Hardware snaps window corners to 1/256 pixel before interpolating
(measured by yahya308 on an RTX 4070 Laptop and a Radeon 780M in #1547);
lavapipe snaps only for coverage (lp_rast.h FIXED_ORDER 8), so the test
accepts either reference per pixel at the original 1e-4 margin.
Sources:
- v_interp_p1_f32/v_interp_p2_f32/v_interp_mov_f32 semantics (P0,
P10, P20, I/J): AMD RDNA2 ISA, "Parameter Interpolation".
- BaryCoordKHR/BaryCoordNoPerspKHR weights and PerVertexKHR vertex order:
SPV_KHR_fragment_shader_barycentric, VK_KHR_fragment_shader_barycentric.
- Geometry shader triangle input order and provoking vertex: Vulkan
specification, "Primitive Topologies" and "Geometry Shading".
- Interpolation decorations decided by the fragment input: Vulkan
specification, "Interpolation Decorations".
- lavapipe interpolation: Mesa src/gallium/drivers/llvmpipe/lp_rast.h
(FIXED_ORDER 8), lp_setup_tri.c (subpixel_snap for coverage only),
lp_state_setup.c (init_args, plane setup from float positions).
A PIXEL_PIPE_STAT_DUMP event (EVENT_WRITE 0x39, Recorder::DumpSamples)
ended the render pass the previous draw left open: the dump ended the
batch-long occlusion query, copied its result with
vkCmdCopyQueryPoolResults and ran the fold kernel, none of which Vulkan
allows inside a render pass. VulkanDevice::DumpSamplesOnGpu also flushed
the queued store run over the dumped bytes first, which ended the pass
and recorded the run's barriers as well. A title that clears each query
slot with WRITE_DATA and brackets every bounding draw with a begin and an
end dump gets a render pass, two copy barriers and a label run per draw.
Samples are now counted in segments. A segment is one slot of a
per-batch occlusion query pool (128 slots, reset outside any pass when the
batch's first pass needs one). It begins at the first draw recorded after
a pass begins or after a dump, inside that pass's subpass, and ends at the
next dump or right before vkCmdEndRenderPass, so each query begins and
ends in one subpass. A pass whose pool has no free slot left is not
continued, so a new pool can be reset. The draw path that records a whole
pass of its own prepares a slot before its pass and ends its segment
before ending it.
A dump ends the open segment and is queued in the batch's store run with
the number of segments before it; it no longer ends the pass and no
longer flushes the queued stores. The run is recorded where queued stores
already are: at Submit, or earlier when a later command reads or writes
the queued bytes. Its recording copies the results of the segments the
dumps need (one vkCmdCopyQueryPoolResults per contiguous slot range) and
the dump list, then records the queue in groups: a group's stores, then
one dispatch of the SampleDumps kernel, which adds each dump's segments to
the running total and stores the 16 counters in queue order. A store over
bytes a dump of the current group writes starts a new group, so every
byte ends with the value of its last write in program order
(Recorder::QueuedWriteGroups). At Submit, segments no dump consumed are
folded the same way once 64 are pending.
No reader sees a dump later than before: a CPU reader of the dumped bytes
syncs the batch (the dump still notes them as a pending write), and Submit
records the run before the batch ends; a later label or store is queued
after the dump in the same run; GPU work that reads or writes the queued
bytes flushes the run first, as for queued label stores. No reader sees it
earlier either: the dump is recorded after every draw whose samples it
counts, and its segments end at the dump. Without the GPU sample counter
(no buffer device addresses) the batch-long query is kept as before.
Tests (agc_driver_recorder_tests, GPU, run by hand): QueuedWriteGroups
keeps the program order's bytes for 20,000 random store and dump
sequences; a queued dump does not land before its run is recorded, a
reader of other bytes leaves it queued and a reader of its bytes records
it; clear, dump and store orders over shared bytes; two dumps inside one
render pass keep it open (the next draw continues it) and split its three
draws' samples exactly with precise queries; a reader flushing a dump
inside a pass ends the pass; 160 dumps in one pass past a pool's 128
slots; 70 passes without a dump folded at Submit.
Main renders depth and stencil tests without a depth target when DB_Z_INFO and DB_STENCIL_INFO both have FORMAT INVALID (48ebf3195e): with no plane, the DB has nothing to test, and every test passes. The depth bounds test (DB_DEPTH_CONTROL bit 3) reads the same absent depth plane, but such a draw still threw "depth bounds without a depth surface". A title does this when a level loads.
Depth bounds now follow the same rule when both planes are absent. With an absent depth plane beside a present stencil plane, the bounds test still throws ("depth bounds without a depth plane"), as before.
Every direct memory allocation had a memfd of its own, so each one the
GPU used became its own udmabuf and its own imported buffer object. An
imported buffer object cannot be always valid in the GPU VM: amdgpu
validates every one of them at every command submission. PPSA14632
(LEGO Horizon Adventures) imports about 7000 allocations, 5800 of them
64 KiB, and each vkQueueSubmit took about 2.3 ms (5 s per 10 s, held
under the GPU mutex). A standalone test shows the cost follows the
count of imported dma-buf buffer objects, not their size or the count
of ordinary allocations.
- libkernel: on Linux all direct memory lives in one sparse memfd of
DIRECT_MEMORY_SIZE, at file offset = physical address
(PhysicalBacking keeps the file and its offset; SharedBacking returns
that offset). A range allocated again is zeroed through a temporary
mapping, as the old per-allocation memfd was.
- Driver: an import of shared direct memory that lies inside one
64 MiB chunk of the file binds its own VkBuffer into that chunk's
memory, one udmabuf import per chunk (HostImport::chunk holds it, so
it lives until the last buffer bound to it retires). Ranges that
cross a chunk or do not meet the buffer alignment are imported alone
as before.
On PPSA14632 vkQueueSubmit drops from about 2400 us to 46 us on
average, the GPU idles 36 ms instead of 172 ms per present, and
presents go from about 40 to 80-97 per 10 s.
Released direct memory stays committed in the file (see
TechnicalDebt.md): a live chunk import pins its pages.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
macOS has no counterpart to userfaultfd write-protect or
MEM_WRITE_WATCH, so GuestWriteWatch had no backend there and the driver
re-read every guest buffer on each use.
The macOS backend write-protects watched pages. The first write to a
page faults, the SIGBUS/SIGSEGV handler records the page and restores
its protection, and Collect reports the recorded pages and protects
them again. A page is protected before its record is cleared, under the
lock the handler takes, so a write that lands in between is recorded.
The lock hands out turns in order: an unfair one starved writers under
Rosetta.
The protection is real, which adds three entry points:
- GuestWriteWatchBeginHostWrite/EndHostWrite lift it around a system
call that writes guest memory, which would otherwise fail with
EFAULT. GuestArena::HostWrite calls them, as it calls the Windows
equivalent.
- GuestWriteWatchProtectionChanged tells the watch the protection given
to watched pages by mprotect or a new mapping.
- GuestWriteWatchProtection reports that protection, since a watched
page is read-only between a collection and the next write.
Register keeps the recorded protection of a page it already watches:
read from the kernel it would be the watch's own read-only protection,
and the page would never be watched again.
The tests build without the prx libraries and run natively and through
Rosetta.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Retain an immutable decoded program and structured control flow while
preparing the registration variants for one shader.
Match owned code, stage and the complete SWAPPC inputs before reuse.
Translate and plan each variant separately using its current inputs.
Keep cache keys and persisted artifact layouts unchanged.
Provide a cache-neutral off switch for controlled comparisons.
Cover identity, lifetime, failed replacements and variant equivalence.
Main's Linux job fails since #1028 and #1714: guest_video_out_resolution_status aborts when SDL's offscreen
driver has no Vulkan window, and check_skips.py rejects the skips of guest_video_out_wqhd_detection and the
flip tests because "no display" is not among the device limitations it accepts.
The resolution status test now skips with code 77 like its wqhd sibling, the flip tests print the same skip
reason, and check_skips.py accepts "no display" as a device limitation. ANYPS5_REQUIRE_DISPLAY still turns
the skip into a failure.
AI-assisted: yes. Full current-main project build passed. Seven focused Windows tests pass; four shader tests also pass in full and none optimization modes with actual Vulkan execution required. Updated Linux and title runs remain pending.
gmtime, localtime, mktime, strftime and asctime took and returned the
host std::tm. The guest's FreeBSD tm has tm_gmtoff and tm_zone after the
nine calendar fields, so on Windows, where std::tm is 36 bytes, a title
reading those fields read past the thread-local result, and gmtime_s left
them unwritten. Windows gmtime_s also accepts only 1969-12-31 12:00 to
3001-01-19 21:59 UTC, where the guest libc converts any time.
Convert through a GuestTm with the guest layout on both hosts. Compute
the UTC fields from the day count, so every time whose year fits an int
converts, and on Windows take the local offset and zone name from the
CRT. gmtime reports the zone as UTC like the guest libc. localtime and
mktime keep the CRT's range on Windows; recorded in TechnicalDebt.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
As requested in review: a failed sceKernelGetModuleInfoFromAddr no longer
reports an empty answer. Adds a failure test (symbol interposition) and
extends the metadata test (EH record walk, terminator, no-header DSO).
Co-authored-by: jweidenh <38249793+jweidenh@users.noreply.github.com>
Co-Authored-By: Claude <noreply@anthropic.com>
A title that ships its own libc.prx unwinds C++ exceptions with that
libc's libunwind, which asks sceKernelGetModuleInfoForUnwind where each
frame's eh_frame is. On Linux the call always reported no tables, so
no C++ exception thrown in guest code could be caught. For the main
executable and .guest.prx images, report the PT_GNU_EH_FRAME tables
that sceKernelGetModuleInfoFromAddr finds, as Windows does from
.ehmeta. Other host images still report none.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HeC9bA3n74TzJRYv9AgBE8
CB_COLOR_INFO format 6 (COLOR_10_11_11) with NUMBER_UNORM threw in
DecodeColorFormat: Vulkan has only the B10G11R11 float format. Such a
target now renders into an R32_UINT attachment in the layout of
format-30 images (unorm11 R and G in bits 0-21, unorm10 B in bits
22-31), so later image loads of the surface read the same words.
The target format is draw state, so the packing is a specialization of
the registered pixel shader: every float MRT export also computes the
packed word (EmitFormatStoreComponent) behind specialization constant
ExportPackingBase + target, and the specialized module turns a packed
target's output into a uint4 that ValidateShaders expects. The new
request field targetExportPacking is in the module key, the result memo,
the draw key and the serialized request (version 14). Blending, partial
RGB writes, UNORM16/SNORM16/UINT16 exports, DCC and ROUND_MODE throw.
agc_driver_pending_exact_surface writes mip 0 of a 2-mip 256x256 surface and then a 1-mip surface at the same address. The chain's mip 0 lies past the 1-mip surface's bytes, so both images stay pending, the chain first. A lookup of the 1-mip size must return the 1-mip image, and with only the chain pending it must still return the chain. On main the first lookup returns the chain.
Winsock fails send, sendto and sendmsg on a socket shut down for writing
with WSAESHUTDOWN, which neither Windows error map handled, so libkernel
and libSceNet reported EIO. FreeBSD reports EPIPE there, and the Linux
maps already translate the host's EPIPE to 32. Map WSAESHUTDOWN to 32
in both.
agc_driver_graphics registers a D32 and a D16 surface on a mock device and looks up views at their addresses. Views of the surface's own extent in R32F or R16, and T#s that read the raw depth bits, sample the depth plane; a view of another extent, a 2D array view of another extent, an R11G11B10 view, an R32 uint view without a depth layout and an R32F view of a D16 surface return null and take the ordinary texture path. On main the first of these throws in DepthSurface::Sampled.
Measured on a PS5 (firmware 13.60) through dlopen of
libSceVideoOut.sprx on a handle the payload opened.
sceVideoOutInitializeOutputOptions writes 0xFF to byte 2 of the 64-byte
options and returns INVALID_ADDRESS for a null pointer.
sceVideoOutIsOutputSupported checks the handle, then the reserved
arguments, then the mode, then the options. It never answers 0: a mode
the display lacks returns UNSUPPORTED_OUTPUT_MODE, and a value that is
not a mode returns 0x8029001E. Options must keep the initialized words,
except word 3.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Fill takes either info struct, so sceKernelGetModuleInfo reports the
name and segments sceKernelGetModuleInfoFromAddr reports for the same
image, on Linux and Windows.
A sampler with FORCE_DEGAMMA (S# word 0 bit 20) threw at decode, so
every dispatch or draw sampling through one was skipped.
DecodeSamplerResource now records the bit, but accepts it only for a
caller that pairs the sampler with its images. The direct sampler
binding does, and RequireDegammaFormat checks each sampled image
against its paired samplers: only already sRGB formats (8, 8_8 and
8_8_8_8 SRGB, BC1, BC2, BC3 and BC7 SRGB) pass, decoded once, as
shadPS4 (PS4) models it. Other textures and unpaired decodes still
throw; gfx10 behaviour is not measured.
On Windows, nanosleep and this_thread::sleep_for/sleep_until go
through winpthreads, which ends a wait on a 15.6 ms GetTickCount64
step. Blocking sceAudioOutOutput and sceAudioOut2ContextPush calls
returned in bursts about every 16 ms, a blocking sceAudio3dPortPush
ran at about 68 % of real time, and sceAudioInInput without a device
woke anywhere from 1 ms to 31 ms apart. They now sleep with
PreciseSleepUs or the new PreciseSleepUntil.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
sigprocmask keeps the first word of the title's set as it is, where
FreeBSD keeps signal n at bit n - 1 (_SIG_BIT), but Dispatch tested bit
n. A title that blocked SIGTERM with a FreeBSD-built set (bit 14) was
not protected, and a set with bit 15 blocked SIGTERM instead of SIGURG.
Dispatch now tests bit n - 1; the stored mask and the set returned by
sigprocmask are unchanged.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The AMD Windows driver returns 1 for a NaN operand when any float form
of ordered not-equal feeds a ballot on a 32-wide subgroup, which is how
e32 compares are read in wave32. FPOrdNotEqual32 is now emitted as an
unordered not-equal with both operands checked for NaN on their bits,
which gives the same result for every input.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Scott Schuster
Adrià Franch
Cesar Rodriguez
MarcZX100
pushkarreddyy
fma-rc
Achraf Reyani
Badr
Get-fix
Adrià
Hannah Lyons
raghav314
Alex
Claude
Abdulrahman Elkordy
xXPBSXx
DariusHQ
MrMha
Dean
SynnZer0
Leonardo
tototomate123
krishnaraya
Hussein Kadhim
MarcZX100
Fishybeing
AlxcBTW
ArjenS-LA
Nahuel Genchi
ThatsLeo
Chad
Joey Fisher
David
XairRrR
Gustavo Ariel Gamarra Rojas
Scramblrrrr
Dean Galvin
hugodearaujo
gookie
TibixDev
kjhq
overlordxrz-source
poonasor
Zayd Zaari
Daan Dehing
Jan Šlégl
john.maia
PhialsBasement
simonbbby
MathiasDevelopes
StarMrGamer
Renan Pires
Josip Parlov
Michon
David
goncalompontes
mugurc
Doston Abdullayev
Sanjaysrinivas
JAB-dev
QRMOE
ps54all
emirplus
Yassine
LeandroLP23
ctj01
aplatogg
arcwiser
Lexroid
Karan Vora
Tharun Balaji
alfalfa-star
TideOPL
Celegans12
Yuvaraj
boykopovar
Rohan K
virp-mta
Dextroid17
AymanFE
jeffrey12cali
Nicolas Abosch
danzaio
Marcus
Mini
LuisBiceps
Alex Collini
nyavana
Patrick Elsasser
Samuel Jean
Morten Holm
Raihan Abdul Majid
authurtabo-byte
gunerky
EdouardSence
Gjon Gojcevic
Hamidreza
Faffu
eliasbruno124-dev
ale
Sasha Papo
Aryan Sharma
ah
Daniele Giovane
NevesHenry
erick-dsnk
seven
Julio Cacko
Cursor Agent
jweidenh
Pouare514
fisherjoey
turkishcuber
Zurisen
aashishaxo
andre
JohnM
krishalivaibhav
22alfbassett
Oguzhan Ozdemir
validation agent
hazzardouslysolid
russhil
Elijah De Guzman
Mohd Faour
clem-gb
Zheorgyan
hamb3r
Roojool
popcoder565