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]>
A traced run that ends without a crash or an abort never reports which
imports it called. With APS5_TRACE_CALLS_FILE set, the first traced import
starts a thread that rewrites that file every 10 seconds with every import
called so far, most called first.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The graphics pipeline store kept a fixed 1024 entries and the pipeline journal replayed at most
1024 records, and a replayed pipeline that found the store full was dropped. A title whose
working set is larger rebuilt the rest at the draw on every run, and evicted pipelines it kept
using were rebuilt again and again.
The bound now starts at 1024 and doubles up to 4096: a journal replay that does not fit grows it
instead of dropping the pipeline, and so does eviction churn (a ghost list of evicted keys, checked
every bound/32 evictions: more than half of them came back). It never shrinks during a run;
APS5_PIPELINE_STORE_ENTRIES still fixes it. The journal loads its last 4096 records by default
(APS5_PIPELINE_JOURNAL_ENTRIES). Resident pipelines are bounded by count, not bytes.
agc_driver_recorder_tests --benchmark-pipeline-store-memory [count] (opt-in, not registered)
builds count graphics pipelines with one fragment module each through CachedPipeline and prints
process private bytes and VK_EXT_memory_budget heap usage before and after: with the triangle
shaders about 59 KiB host and no device memory per pipeline on an RTX 4090, 100 KiB host and
8 KiB device-local on a UHD 770.
Tests: agc_pipeline_store_growth (fails on main), agc_pipeline_store_fixed_bound,
agc_pipeline_store_replay_growth (replays on one thread: the mock Vulkan is not thread-safe).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_017SNUYRd7sRKZrMYpquAsnU
strtoull took the valid hex prefix of a token and ignored the rest, so a
malformed token such as "123garbage" selected program 0x123, and an
overflowing token was not detected. A token now has to be a whole 64-bit
hex number (with an optional 0x prefix); any other token is reported and
ignored. The test adds a list with a malformed token whose prefix is the
first program's address and an overflowing token.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
APS5_SKIP_SHADER_FILE names a file of hex program addresses (whitespace
separated, # starts a comment). Draws and dispatches that use a listed
program are dropped as with APS5_SKIP_SHADER. The file is re-read within
500 ms of a change, so passes can be switched off and on while the title
runs.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A depth plane copy (a depth surface sampled as several slices, an integer view or a cube) was
kept until the device went away, with a staging buffer of its own size, and a staging buffer it
outgrew was never freed. Every depth surface a scene ever sampled this way held its copy, so
device memory grew with each new scene (0.4-1.2 GB in RE2 after a level load).
The copies now share their image's ownership with the views they hand out, the recorder keeps
the image and the staging buffer until the recorded copy has run, and past
APS5_DEPTH_PLANE_COPY_MIB (default 256; 0 keeps every copy) the least recently used copies
leave the cache. A copy made again refills from the surface and guest memory as a new one does.
agc_driver_depth_plane_sampling_evicting runs the plane sampling cases with a 1 MiB budget.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Queue-0 DMA_DATA packets drained the draw pipeline unless APS5_PIPELINE_DMA was set, so the packet worker waited on the committer at every small immediate store. Kena PPSA01746 issues 17-23k DMA_DATA packets per 10 s and 11329 of its 20134 pipeline drains per 10 s are opcode 0x50. enqueueDmaPacket now queues DMA_DATA stores and fills unconditionally, and keepsDrawPipeline keeps the pipeline for a copy or fill whose source and destination miss every queued write. The APS5_PIPELINE_DMA switch and its documentation are removed and the tests run the default path.
An EOP flip reservation also ended a frame, and so did the submission
whose EOP release carries the reserved token, so a title flipping with
EOP releases counted each displayed frame twice: a capture of N frames
stopped after about N/2 and could start between the reservation and
the submission that flips. The reservation is now only recorded; a
frame ends at a CPU flip or at a submission that flips.
agc_driver_frame_capture_eop reserves and submits six EOP flips with
APS5_FRAME_CAPTURE_FRAMES=3 and checks that the capture starts after
the first flipping submission and holds three frames of one reservation
and one submission each.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
APS5_FRAME_CAPTURE=<dir> makes the AGC driver record a few frames of a
title: the shader registrations and ABI resolutions, video output and
EOP flip reservations it made since it started, a snapshot of every
readable guest allocation (the title image included) taken at a frame
boundary once the driver is idle, and for the next frames each
submission, flip and suspend point with the thread that made it, its
serial progress and the guest pages the write tracker saw the CPU write
before it. At the end it stores the presented display buffers as
reference images. APS5_FRAME_CAPTURE_AT_FLIP, _TRIGGER and _FRAMES pick
the frames.
agc_frame_replay maps the captured memory at its addresses and replays
the calls of each captured thread on a thread of its own, in captured
order, through the current build of the driver with no title running.
It paces GPU work to the in-flight depth the title had, releases GPU
waits that the title's CPU threads released without calling the driver
(AgcDriverAwaitedWaits), and reports frame times, released waits and
how closely the replayed display buffers match the references.
Tests: agc_frame_capture (file format) and agc_driver_frame_replay
(two captured frames of CPU writes and DMA copies replayed and
compared). Docs: BUILD.md, ENVIRONMENT.md and the libSceAgcDriver debt
file.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Dean