A render thread a frame behind lowers the plan's ceiling by one (recovering
one per 2 s) instead of tripping the breaker; the guest waiting on the queue
trips it only when the wait was a stall (a quarter of the game frame). The
real-frame cost plans with CPU time; the GPU span is reported only. The
generated-frame blit copies colour only and marks the stencil stale.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Armed at init (session start, render-thread start, a new resolution) and on
a savestate load, game entry or window resize. While armed, an overrun jumps
straight to the highest level whose pure-area prediction is at or below
0.78, then is judged like any step; a fit after a step, an up step, a failed
judgement or 6 s disarm it and single steps fine-tune. A fixed cost only
makes the target dearer than predicted, so a jump never lands below the
level that fits.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Opt-in under the render thread. Composed frames go to a ring of offscreen
slots at the window's size (binds of framebuffer 0 on the composing context
land on the current slot); a present thread with a second, shared context on
the same window waits on the slot's fence, copies it 1:1 and swaps. The
compositor's wait moves off rendering; the composing side waits only while
every slot is queued or on screen. Falls back to direct swaps when the shared
context or sync objects are unavailable.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
[video] frame_generation (opt-in, with render_thread): the render thread keeps
each game frame's replayed draw list, matches the triangles of the last two
(frame_gen.c), and presents in-between frames at the display refresh by
redrawing the newer list into separate surfaces with matched vertices
interpolated. It never runs guest code. Real frames are composed at their
flip and presented after the generated ones, with the same pixels and
order. Generation uses only measured surplus: a plan from render-thread CPU,
generated-frame cost and swap time, a breaker after backpressure or late
guest frames, and a pause while dynamic resolution is over budget. The
render-thread core gains a timed tick hook. Frame-cost queries pause around
generation, so dynamic resolution measures real frames only.
Tests: frame_gen_test, render_thread_test (tick), gl_frame_gen_test (real
frames identical off/on; phase 0/1 equal the endpoints). Docs:
docs/FRAME_GENERATION.md.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The hr (VRAM) surface stayed allocated at the ceiling. On GL-on-Metal every
render pass loads and stores the whole attachment, so its cost followed the
ceiling rather than the level: R4 2P VS at 5x ran 40-45 Hz for its first
seconds with a 10x ceiling and 60 Hz with surfaces allocated at 5x. Allocate
a new hr surface at each step and reseed it from the 1x image and the saved
rects; a failed allocation falls back to the in-place step down, or refuses
a step up.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A native-wide surface was allocated at the dynamic-resolution ceiling and
rendered into its lower-left corner. On GL-on-Metal every render pass loads
and stores the whole attachment, so the wide surfaces cost the ceiling's
bandwidth at every level. Allocate them at the level they render at and
reallocate at each step, rescaling the presented rows and margins from the
old surface. The limit check stays at the ceiling.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
--headless-opengl captured the presented image into the ring every frame
through gl_renderer_capture_display_hires, a sync point, so a headless
run (fp_identity) held the context for most of every frame. The capture
is now recorded at the same point of the frame (gl_renderer_ring_capture)
and done by the render thread; ring readers already sync first. A rect the
GL surface cannot serve is skipped there rather than resolved through a
CPU readback, and ensure_cpu refuses to run on the render thread at all:
its readback would land in the private upload source and mark the CPU
side current without gpu.c's array seeing the pixels.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
One frame behind is the normal pipelined state when the render thread is
the slower side; skipping on it dropped nearly every present at 4K
widescreen. rt_frames_ahead() counts the complete frames queued after
the one being replayed; the backend skips only at two, where the
emulation thread is at the in-flight bound waiting on the render thread.
The unit test checks the count never exceeds the bound and that both
cases occur under a slow consumer.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
With the render thread running, the backend's facade table records every
state change, primitive, fill, copy, CPU->VRAM upload (payload copied)
and native-wide call, and the present, into the command ring; the render
thread replays them through the unchanged glb_* functions. With it off
gl_backend_get returns the existing table and nothing new runs.
Sync points: every other GL entry point reached on the emulation thread
(readbacks, savestate restage, render passes, OpenXR, settings,
diagnostics) drains the ring and takes the context (GL_RT_SYNC); the
emulation thread then runs the rest of that frame directly, exactly as
without the thread, and hands the context back at the frame boundary when
the frame may run asynchronously again (no depth24 display, dual-raster,
frame interpolation, render pass, netplay or OpenXR).
Guest-visible VRAM stays gpu.c's array: while the render thread owns the
context, uploads stage from a private copy that receives each payload in
command order. Guest state the backend reads while drawing (per-primitive
widescreen tags, the native-wide fast-path latch) is captured at record
time; host overlays travel with the present; a present that a newer
recorded frame already supersedes is skipped. The emulation thread answers
projective/wide-present/draw-area queries from mirrors of the recorded
calls. The per-frame display-ring readback is queued (fbo_peek_deferred)
instead of being a sync point.
The GL fixtures link render_thread.c and stub the two facade hooks; the
scale fixture also gains the gpu_ws_background_requires_full_composite
stub it was already missing (it failed to link on master).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Backend-neutral core for the opt-in [video] render_thread: the emulation
thread appends records to a single-producer/single-consumer ring and the
render thread replays them in order with the graphics context current.
rt_frame_end bounds closed frames in flight (backpressure), the render
thread can tell a frame is stale, and rt_acquire is the sync point:
drain, then move the context to the emulation thread until rt_release.
No lock is held while a record executes.
render_thread_test drives it with a fake executor and context token:
ordering and payload integrity across ring wraps, oversize refusal, the
in-flight bound under a slow consumer, stale detection, randomized
acquire/release interleaving and stop while held (TSan clean).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A step controller (dynamic_resolution.c) lowers or raises the GL internal scale from frame-time telemetry between a floor and the configured ceiling; off unless a title or the player enables it. Tested by dynamic_resolution_test (synthetic load) and the real-GL gl_scale_invariance_test step modes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
gl_renderer_set_dynamic_resolution(1) keeps the hr surface and the
native-wide surfaces allocated at the scale context init gave them (the
ceiling) and lets the internal scale step to any integer level from 1 to
it while the game runs. A lower level renders into the surfaces'
lower-left VRAM*S corner; every viewport, scissor, pack, copy and capture
already takes the scale at use, so only allocation and the present's
normalised UVs read the allocated size. Off (the default) nothing is
allocated differently and no step can run.
A step runs after a present, never inside a render pass:
- it takes the top-left sample of every native block of the old surface
(the sample the pack reads) into a 1x image and reseeds the new corner
from it, so the guest-visible VRAM is unchanged byte for byte;
- it rescales the displayed rect and the draw area from their old pixels,
so neither the frame on screen nor the one being drawn shows 1x blocks;
- it rescales the native-wide margins in place, in row bands ordered so
no source row is overwritten before it is read;
- it marks the mask stencil stale (rebuilt at once while mask checking is
on) and drops the render-pass backup and images.
Reallocating per step instead measured 30-106 ms at 7x-10x on an Apple
M4, mostly the first touch of the new surface. Steps are not offered in
the windowed high-resolution mode or with dual raster.
gl_scale_invariance_test gains "steps" runs: rendering after a step into
a ceiling-sized surface equals the fixed-scale run at that level; a chain
of steps (9x down to 1x and back, 5x) leaves the native VRAM unchanged at
every step, the two rects and the wide margins exact nearest rescales of
their old pixels, everything else the old top-left samples, and mask
checking right after a step correct; the window mode refuses steps.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Preserve negotiated content and dual-raster pass gates. Retain both fixture assertion sets and HD-off seams. Regenerate both BIOS with actual native emitter. Fix type-only rollback corrections and exclude extra-pad changes from digital-only promote gates.
The episode kept one peer-baseline / ready / POST latch, so with three or
four seats the first reply answered for everyone, and the one early
BASELINE stash was overwritten by the next seat's. In R4's 4-peer run with
the host loaded every frame, two guests replayed a light-tip episode and
resumed while the host had not opened it; host and guests then waited on
each other (pcap_freeze / wire_hole) until the stall timeout.
- BASELINE, ready-ACK and POST go through rnet_rb_quorum: the episode moves
on only when every other occupied seat has answered, and a digest from
any seat that differs fails the comparison. Early BASELINEs stash per
seat. Two seats behave as before.
- Light tip skips the ready RTT only with one peer; with more, every seat's
ready is how the others know it rolled back to the same load.
- A follower that has not simulated the load tick yet defers and joins when
it gets there (no NACK, no zombie-load refusal for that join).
- Seal, ready and verify timeouts restart while a seat still owing a reply
advances its input tip: they fire only when nothing moves.
- psx_netplay_peer_seats(); pins recomp-net rb_quorum.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Four-peer rollback desynced under load (R4, host behind): the guests
predicted the slow host's pad, hash-confirmed the tick among themselves
through the one shared peer ring, then promoted the host's real pad into
history only (hc-silent promote) while the host simulated it. The host
stayed forked; its recovery episode was then refused as a "zombie load"
behind the guests' (wrongly) agreed tip.
- Hash confirm is a quorum over every occupied peer seat
(rnet_hc_set_peer_mask / note_peer_from, recomp-net) for three or more
seats; two seats keep the one-peer chain.
- A peer more than 4 ticks behind the others' inputs sheds its own view
(psx_mod_render_local_view status FAST_FORWARD, held 60 ticks), so the
presentation-only pass goes before the simulation slips.
PSX_NET_LOCAL_VIEW_SHED=0 keeps it.
- The admit-stall watchdog counts time without progress: a remote input tip
that advances restarts it, so a slow peer is waited for.
PSX_NET_ADMIT_STALL_MS (20000) and PSX_NET_LIVENESS_MS (1500) configure it.
- Pins recomp-net fix/rollback-span-stall.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Load modern texpage/texupload PNG replacements independently of DuckStation source. Integrate pack folders and options with mods, preserve native VRAM, and validate GL rendering, format matching, dump safety, and upload invalidation.
A [[plugin]] may declare netplay = "local_view": it only changes what one
player sees. A netplay session now keeps the player's own features whose every
contribution is such a plugin (no writes, overlays or derived discs) instead of
clearing them, without saving state.toml. They activate as usual, but their
hooks run only inside psx_mod_render_local_view (psx_mod_local_view_scope) and
their vblank/savestate callbacks not at all. With such a plan the widescreen
cull margin is the stock 0 in the shared game and this peer's own margin only
inside the sandboxed own-view draw; squash is not used. The canonical frame,
snapshots and digests stay identical on every peer whatever each player chose,
so nothing is negotiated, and each keeps their own aspect.
Tests: netplay_view_mods_test (flag, plan filter, parked hooks, no vblank,
state.toml untouched), ws_local_view_margin_test (margin scope), render-pass
sandbox asserts the scope around a local view.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Fixes Ape Escape terrain seams: CPU-built subdivision midpoints and lwl/lwr-copied
grid corners drew native next to precise neighbours. Adds pgxp_tri_ring/pgxp_store_ring.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
psx_mod_render_local_view() runs a title's draw code in the render-pass
sandbox on a forward netplay frame and keeps the drawn display rect in the
OpenGL presenter's surface only. The authoritative CPU VRAM, guest RAM,
rollback snapshots and digests keep the canonical frame; the presenter shows
this peer's own image. A committed image cancels any local-view crop.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A title whose netplay mode draws every seat's view into one frame (so all
peers keep identical guest state) can now ask the present path to show only
this peer's rectangle of the display: psx_netplay_present_local_view(x, y,
w, h), usually picked from psx_netplay_local_slot(). The GL dual-raster
present shows it from the internal-resolution surface, the software present
crops its staged frame, both at 4:3, and the headless present-image ring
records the same rectangle. The request is host-only (never serialized or
guest-visible), lapses PSX_NETPLAY_LOCAL_VIEW_HOLD ticks after its last
renewal so menus fall back to the full frame, survives rollback replays
without moving its stamp backwards, and is ignored offline. It wins over
[netplay] local_viewport.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Reports whether local Rewind is enabled and open, the title block
(psx_mod_set_rewind_blocked) and the snapshots held, so scripted headless runs
can check that a blocked title captures no history and refuses to open.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A title calls psx_mod_set_rewind_blocked(1) while it is in a mode that must
not be rewound, such as local split-screen multiplayer, and 0 when it ends.
While blocked, opening Rewind is refused with an OSD note, no history is
captured (the ring keeps what it has; a capture already due is dropped and an
open panel closes unloaded), and a title-allowed direct Rewind shortcut is not
in force: its button is not claimed and reaches the guest pad / title
transform as if Rewind were off. Netplay already refuses Rewind on its own.
The block clears at every mod/session start, so a title that never calls it
is unchanged.
The direct-allowance decision (allowed, and for Rewind: enabled and not
blocked) is now one pure helper, psx_hotkey_direct_active, used for both the
claim and the shortcut press, so an unclaimed Rewind button is never a
one-button shortcut (with Rewind off it falls back to Select + button rather
than popping the "Rewind is off" note on every press).
Tests: psx_hotkey_pad_test covers the allowance and a block taking effect
mid-press; rewind_toggle_combo_test guards the capture/open gates and reset.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Shokunin