NVIDIA's proprietary driver flushes f32 add, multiply and fma denormals by default but not the inputs of floor and ceil, so its startup probe failed only on those two. Drop their software flush only where the device declares DenormFlushToZero 32.
The driver declares no shaderDenormFlushToZeroFloat32, so every f32 add, mul and fma kept the software flush. Run the same probe as on MoltenVK and use the driver's default handling when it matches the emulated flush.
Pixel-stage ballots and EXEC reads count only lanes that are not helper
invocations, so a waterfall loop over a saved whole-quad EXEC picks and
clears live lanes only. In programs that execute s_wqm on EXEC, writing
EXEC from a value keeps a helper invocation active while a live lane of
its quad is, so implicit-LOD sampling after a save/restore still has its
helpers.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Opt-in with APS5_PROFILE_GPU_DRAWS=1 together with APS5_PROFILE_GPU=1: a [gputime] range around each recorded draw command keyed by the fragment program's variant id, and a [drawvariant] line naming the code address behind each variant id once.
An import was evictable 256 lookups after its last use. A frame makes thousands of lookups, so when the imports a frame uses do not all fit the budget, each new one evicted another of the same frame and the whole set was imported again every frame. A flip packet now ends a frame, and only imports not used in the current or the previous frame are evicted; what does not fit is refused and copied, as below the budget. A frame boundary also expires the remembered evicting refusals.
sceNpAppLauncherLaunchApp returned the libSceNpGameIntent codes
0x80553802/0x80553804, but the firmware tail remaps the internal
launch library errors into the launcher family, so the export reports
0x80553A02 before initialize and 0x80553A04 on invalid arguments;
the tests now assert those values and the unrelated GameIntent
constants are untouched.
The two assertions that a valid launch call must throw until
implemented are removed; the production throw and its debt entry
stay, and the validation and lifecycle coverage is unchanged. The
debt file keeps the silent-stub, InitParamInit, InfoInit and
remaining unknown-export entries, updating only the launcher parts.
Recovers the 4.03 firmware signature (int32, const char*, const char*), the uninitialized and invalid-argument errors, the strnlen(0x2001) bound on the optional argument and the initialize/terminate flag. The launch itself still throws.
Generated with [Devin](https://devin.ai)
Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Firmware 13.42's module looks the name up with sceNetGetIfName, which
has none past 9, then issues SIOCGIFFLAGS on the empty name. That ioctl
fails, and the call returns -1 with the flags untouched. Return the
same instead of throwing. Interfaces 1 to 9 still throw, and the test
no longer asserts that for 1 and 4, which the console answers from the
kernel.
Asked for in oneandonlydean's review, from BunnyFl's read of the
module.
Generate version 1 UUIDs as FreeBSD 11's kern_uuidgen does: consecutive
timestamps from the guest CLOCK_REALTIME, a 14-bit clock sequence that is
randomised on first use and incremented when the clock has not passed the
last batch, and a random multicast node. A count outside 1 to 2048 fails
with EINVAL and a null store with EFAULT.
Every shader source keeps up to 256 materialized results of
Recompile(request, capture) in its result memo, and as many patch
recipes in its masked memo, with no limit on their bytes. In Ghost of
Tsushima's opening the memos held about 500 MiB after 5 minutes and
kept growing, at a 13 % hit rate.
Each memo entry now counts the bytes its result's vectors keep (the
bindings' vectors included; the module is shared with the prepared
module cache), plus its patch recipe in the masked memo. A result held
by both tables counts once in each. Past a budget, 256 MiB unless
APS5_RESULT_MEMO_MB sets another (0 sets none), the least recently used
entries of both tables of all sources are dropped until three quarters
of the budget is left. The recency stamps come from one clock for all
sources and both tables so they compare across them. The profile line
also reports the trimmed and held entries and their MiB.
The trim takes its lock blocking and repeats, re-reading the bytes held,
until they are within the budget, so the insertion that pushed them
over has its trim run after any other; it stops when nothing is left to
drop.
SetResultMemoBudget and ResultMemoBytes let agc_shader_result_memo_budget
check that a memo over its budget keeps the most recently used results
and drops the oldest one, with the masked memo off. A one-shot hook
inside the trim (SetResultMemoTrimHook) lets
agc_shader_result_memo_budget_threads insert two results from other
threads after the trim has read the bytes held: the memo must still end
within its budget. agc_shader_result_memo_budget_masked, with the masked
memo on, checks that a first result counts its masked entry too and that
a trim empties the masked memo.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Dean