Reconciled with #470 (PGXP): gpu.c precise vertices take PGXP only under geometry correction and count it, then native-wide edge recovery; the PGXP triangle stat stays PGXP-only. Stubs and tests from both kept; stamps recomputed.
Master moved the textured-dot test stubs into gpu_exec_stubs.inc; the PGXP stubs move there too. pgxp_note_rect_bypass stub takes (int). Stamps recomputed.
A renderer that keeps a screen X as SX << 16 tests the right edge against
a lui-held W << 16 and the left edge with bltz on the shifted value. No
explicit [widescreen.cull] kind fits either: bltz_sites compares the whole
register with -margin, and a lui at a screen_x site was a build error.
screen_x_sites now also takes
lui rt, W -> rt = (W + margin) << 16
bltz x, outside -> taken while SX < -margin (x = SX << 16)
Both are the identity at 4:3. The helpers are defined for native code and
for overlay shards; their math is in ws_cull_edge.h. The two prototypes are
written only for a config that lists screen_x_sites, so the generated code
of every other title is unchanged.
Two cases that could still draw a triangle with precise and native corners:
- The rectangle shortcut bypass now needs all four vertices of an
axis-aligned textured quad to be dataflow-precise. A quad with only some
precise corners, such as a CPU-built sprite with one corner copied from
SXY, keeps the shortcut and counts as rect_partial; two triangles would
have mixed precise and native corners. R4's facade quads are fully
precise (rect_partial 0 in every R4 run).
- Under preserve-projection, RTPS now applies the GPU's truncation window
itself (pgxp_ppp_accept): an exact projection outside it is not
shadowed, the vertex keeps the IR path's shadow, which always agrees, and
it draws precise instead of being rejected to native at the GPU
(ppp_window_fallback counts these). The exact math is skipped while the
hooks record nothing (pgxp_active(): speculative passes).
Also:
- The prepare_precise_triangle comment no longer claims a native neighbour
can only disagree by the sub-pixel fraction. Under preserve-projection it
can disagree by up to the window.
- gpu_textured_dot_nw_shift_exec_test links with -dead_strip on Darwin,
as isc_store_test does. GNU --gc-sections kept it from linking on macOS,
so the rectangle cases never ran there. The test now covers no, one and
all four precise corners.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A textured quad whose four packet vertices form an axis-aligned rectangle in
integer screen space and in UV takes the 2D rectangle shortcut
(gr_draw_textured_rect), which draws the native integer position with no
perspective. That is right for a CPU-built sprite, but a GTE-projected world
quad can land on such a rectangle too: a facade seen straight on. Its
neighbours are drawn precise, so at a high internal resolution the shortcut
left a background-coloured seam along the shared edges. On R4's Helter
Skelter start straight it showed as an L-shaped sky line along a building
panel at 4K, with tolerance off, both with and without preserve projection.
When geometry or texture correction is on and any of the quad's vertices
resolves to a dataflow-precise position (pgxp_probe_precise_vertex: the same
decision as pgxp_get_precise_vertex, not counted), the quad is drawn as its
two triangles like any other world quad. A CPU-built sprite has no dataflow
shadow and keeps the shortcut. With PGXP off nothing changes. geom_correction
counts the bypassed quads (rect_bypass).
Tests: gpu_textured_dot_nw_shift_exec_test gains the quad case (PGXP off,
on without a precise vertex, on with one); pgxp_test checks that the probe
does not count; gpu_gp0_history_guard_test's gpu.c stubs gain the two new
pgxp entry points.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Two opt-in [video] keys for titles whose build carries the PGXP hooks, both
live-tunable through the TCP pgxp verb. Defaults keep today's behaviour.
pgxp_position_fallback (default true). false skips the position-cache tier
in pgxp_get_precise_vertex and stops gte.cpp filling it. At hook-build
coverage the cache only hands unrelated 3D fractions to CPU-built 2D
polygons (a HUD needle, menus). Beetle and DuckStation leave their vertex
cache off by default.
pgxp_preserve_projection (default false). true makes RTPS/RTPT shadow the
exact projection, from the unshifted MACs and a true divide
(pgxp_project_precise), instead of the GTE's integer-IR one. That removes
the residual wobble of near geometry. Guest SXY, MAC0 and FLAG are
untouched. Only sf=1 vertices with no IR, SZ3 or divide saturation qualify,
and the rest keep the IR path. The exact position can sit a little below the
guest integer or more than a pixel above it, so truncation agreement becomes
the window -4 < precise - native < 5 px, a sanity bound over the IR path's
own error (on R4 every on-screen vertex falls within -1.9 .. +2.6 px). A
packet half at or beyond the GTE saturation limit must still agree exactly.
The tolerance clamp now measures the offset both ways. [widescreen]
precise_nclip reads the same shadows, so in a title that sets both, its
branch sign follows the exact projection too.
Observability: geom_correction reports both switches, ppp_produced, and
geometry-corrected triangles split into all-precise, mixed and native. The
mixed share is the mesh-cracking exposure of G1.1.
Tests: pgxp_test (fallback off, window edges, saturation, two-sided
tolerance, the exact projection and its qualifiers, triangle census);
gte_register_access_test (RTPS/RTPT guest state equals the oracle with the
mode on, and the shadow is the exact projection); video settings test
(defaults, parsing, type check). gpu_gp0_history_guard_test's gpu.c stubs
gain pgxp_note_triangle.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Preserve current segmented dispatch, active-plan mod hooks and software scale limits. Extend function filters through the active plan, retain the last-store pointer with overlay ABI 25, and keep default config fingerprints stable. Refresh BIOS emitter stamps after byte-identical regeneration and reconcile regression fixtures.
Only CPU stores are cache-isolated. Mods, the FMV skip, debug pokes, the
bg2d ring fill and the exception sentinel write guest memory from the host;
they now go through psx_host_write_*, which bypasses isc_store the way DMA
does, so a host store inside a FlushCache window reaches memory instead of
rewriting an I-cache tag. The render-pass checkpoint restores the new depth
counter like the DMA depth. isc_store_test covers it (a lane-0 host store in
tag-test mode would otherwise mark a cold line valid).
Docs: the SCPH-1001 residual starts at kernel init 0x598, and without #429
the fix removes 113 of the 126 OpenBIOS flush cycles.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
psx_mod_render_pass_plan() / psx_mod_render_pass() let a trusted plugin
redraw its game between two logic ticks with the game's own draw code. A
pass is a sandboxed nested guest call:
- time is frozen: cycles are counted (GTE and mult/div deadlines keep
working) but no device advances, no VBlank or device event fires and no
interrupt is delivered; GPU linked-list DMA and delayed completions finish
synchronously (psx_cycles freeze, interrupts, dma);
- stores go straight to RAM/scratchpad, bypassing code-page tracking,
overlay watch and write traces; MMIO is an allow-list (GP0, GP1 DMA mode /
info, GPU and OTC DMA, I_STAT/I_MASK) and SPU, CD, timer and other device
stores are dropped and counted (memory.c);
- afterwards CPU (with the GTE), RAM, scratchpad, I-cache tags, interrupt,
timer, DMA and GPU registers are restored exactly (render_pass.c, a GPU
register checkpoint without savestate side effects), with a guest-cycle
watchdog that rolls a runaway pass back;
- on OpenGL the declared display rect (hr colour + mask stencil, raw mirror,
native-wide band, CPU VRAM rows) is backed up, drawn, captured and rolled
back; writes outside it are journaled and restored, or refused.
The presenter keeps two generations of images (the game's own frame plus
one image per pass phase), promotes one when the flip-aware source sees the
display flip to its rect, and shows the newest image at each output
deadline's phase, crossfading across gaps when passes were shed. Presents
that fall due while passes run are made between them. The plan predicts the
output deadlines of the coming frame and sheds passes to fit a host-time
budget learnt from the presenter's idle time (PSX_RENDER_PASS_BUDGET).
PSX_MOD_FRAME_INTERPOLATION_HOLD repeats the newest frame where no pass
image applies, so a pass-driven title never shows anything later than stock.
Default off: nothing runs unless a plugin calls the API, and the plan
refuses in netplay, rollback, rewind, turbo, self-check resimulation, while
suspended (FMV) or without GL flip-aware interpolation. Pass writes are kept
out of frame_fingerprint. Debug: render_pass_stats, render_pass_dump,
PSX_RENDER_PASS_VERIFY=1 (state + VRAM hashes around every pass).
The freeze API lives in the runtime-only psx_cycle_freeze.h. psx_cycles.h is
part of the overlay codegen hash and stays byte-identical, so titles that
never use passes keep their overlay caches and savestates.
Tests: render_pass_plan_test (deadline-exact phases per rate, shedding,
selection), render_pass_freeze_test (10^6 frozen cycles, exact restore,
watchdog), render_pass_guards (sandbox choke points; the codegen-hashed headers stay
free of the pass API).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A native-wide surface is g_wide_w*S wide and the driver limit bounds it
(1024 native columns at 16x on a 16384 limit, about 38:9 at a 320-px
display). An uncapped Fit to Window aspect could ask for more; wide_fbo_for
then refused the surface and the frame fell back to a 1x CPU present with
no native-wide margins.
update_adaptive_widescreen now narrows the Fit aspect to the widest the
surface holds (gl_renderer_fit_wide_aspect: psx_gl_max_wide_width at the
context's scale, psx_gl_fit_wide_aspect at the current display width,
gpu_ws_display_width). It changes nothing unless the surface would be
refused. The refusal stays for a fixed aspect on a wide display mode; its
comment now gives the right limit and what the frame falls back to.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Two generic explicit cull-site kinds for native-wide titles whose screen-X
rejects the existing keys cannot express exactly:
- bgez_sites: `bgez SX, keep` in a signed per-vertex left-edge chain
(`bgez x0,keep; ...; bltz xN,reject`). Emitted through psx_ws_cull_bgez
(keep while SX >= -margin). bltz_sites alone still rejects a primitive
whose first vertices sit in the revealed band; with both kinds the chain
rejects exactly when every vertex is left of -margin.
- clip_edge_x_load_sites (+ clip_edge_width, default screen_w_imms[0]):
lh/lhu/lw of a clip-rectangle X bound. psx_ws_clip_edge_x moves a loaded 0
to -margin and a loaded width to width+margin and leaves interior bounds
(mirrors, split-screen halves) unchanged. xclip_load_sites is lw-only and
upper-bound-only, so it cannot express this.
Both are identity at 4:3, emit nothing when empty, and fail main-EXE
generation on an opcode mismatch (overlay variants stay vanilla). They enter
the overlay-cache hash only when non-empty, so every existing title keeps
its hash (pinned in the test against the previous layout).
Interpreter parity: the dirty-RAM interpreter now also honours explicit
bltz_sites, bgez_sites, branch_keep_sites and clip_edge_x_load_sites. Its
explicit site lists are stored sorted (binary-search lookup), the cap rises
from 64 to 256 per key, and an over-long list is logged instead of being
silently truncated.
Shared math lives in runtime/include/ws_cull_edge.h. New ctest
ws_cull_edge_codegen_test covers parsing, hash identity, emitted code,
the opcode guards and the helper math.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Matched Ape Escape retail SCPH1001 LLE builds identify 0661ce77 as the visible boot regression: parent 155e269b reaches game code, while 0661ce77 stalls near BIOS frame 375. Revert its four dependent follow-ups too; retain PR #326 readback bounds and PR #328 Vulkan aspect changes.
The row-by-row raster dispatcher amplifies an existing malformed DMA list into massive GL work. Keep the broader CPU/DMA arbitration experiment separate for compatibility testing. Document attribution, validation, and the temporarily restored interlace limitation.
Validation: full Ape target reaches title/menu and frame 11669 on retail LLE; OpenBIOS control reaches frame 3701. Eight focused tests pass; retained GL readback fixture passes 71 checks at both 1x and 4x after linking local SDL with iconv. Beads: beads-iisy; DMA follow-up: beads-eio.3.122.
Matthew Stanley