Commit Graph

1004 Commits

Author SHA1 Message Date
chrg6
9065d1485f Let FSR 3.1 and FSR 4 destroy a context they let go of once the GPU is past its last upscale, instead of waiting for the device.
vk_ffx.cpp context_for (two places) and vk_ffx4.cpp context_for (three) waited for the whole device to be idle, with
the queue lock held, on the thread that replays the game's submissions, before they destroyed an upscale context of
another render size, the least recently used one for a new output size, or (FSR 4) every idle one after a refused
creation: a stall of the present's flip, the completion thread and upload submits at every pause-screen size, reload,
window size, dynamic resolution step or mode change with FSR on.

A context that goes is now let go of (retire_up): it goes into a list (vk_up_defer.h) with the timeline value of the
submission of its last upscale (Up.fence, set after each dispatch) and is destroyed on the same thread once the GPU is
past it, at the points where vk_dlaa.c takes its own list (every evaluation, every pass while none runs, and after
something is let go of). The 90-second trims, vk_ffx_trim_idle and vk_ffx_release_idle (the upscaler off, which no
longer needs retire_idle's wait) let go of theirs the same way. FSR 4's create descriptors, which the API asks to stay
live until the context is destroyed, move from the slot into a calloc'd Made that goes with the context, so the slot
can take a new context while the old one waits. A refused FSR 4 creation now waits on the timeline from the host for
the contexts let go of, no queue touched (vk_upscaler_gpu_wait), and tries again when any went; with
NX2NAVI_UPSCALER_IDLE_WAIT=1 that wait and every let go wait for the device as before. docs/architecture.md says
where the upscalers' releases go.

C-5 is narrowed: on the default path no upscaler waits for the device any more, so these waits no longer touch the
frame generation swapchain's own queues; vk_fg.cpp, vk_fg4.cpp, gfx_present.c, vk_hdr.c, vk_fsr.c and vk_color.c keep
theirs.

Not reproduced before the change (no FSR session log here). Every runtime file compiles, vk_dlaa.c, vk_ffx.cpp and
vk_ffx4.cpp with -Wall -Wextra and no warning; vk_ffx4.cpp's non-Windows stubs compile against vk_ffx.h;
up_defer_check passes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:13:01 -05:00
chrg6
22c86f08a0 Let DLSS free what it lets go of once the GPU is past it, instead of waiting for the device on the replaying thread.
vk_dlaa.c waited for the whole device to be idle, with the queue lock held, before it freed a size's images and
feature (free_size: a third scene size, a DLSS model change; stale_sizes), the features release_kept releases before a
create, a reactive-mask snapshot (free_snap), the HUD-less picture of another size (hudless_from), prebloom's images
(pb_images) and what retire_idle gives back. Those run on the thread that replays the game's submissions, and with the
queue lock held the present's flip, the completion thread's hold release and every upload submit waited as well, at
each pause-screen size, reload, window size, dynamic resolution step or mode change; AUDIT F-4-02 measured 16.4 to
44.9 ms for release_kept's wait together with the create that followed.

Now nothing there waits:
- images only the replay uses go to vk_memory.c's deferred destruction, stamped with the submission being recorded;
- a DLSS feature goes into a list (vk_up_defer.h) with the timeline value of the last submission that created or
  evaluated it (Size.feat_fence) and is released on the same thread once the GPU is past that value, which is the
  rule NGX gives for releasing a feature and keeps every NGX call on one thread;
- a size's motion vectors and the HUD-less picture, which frame generation reads at a present, go into that list
  behind a present gate of 8 presents and only then to the deferred destruction; the device's idle never covered a
  present still to come, so this covers more than the wait did;
- the list is taken at the top of every evaluation, at every pass while no evaluation runs (retire_idle), and where
  something is let go of (vk_upscaler_let_go).
A feature creation refused while released features are still in flight waits on the timeline from the host for those
features (vk_upscaler_gpu_wait, no queue touched), takes them and tries once more, with a log line; a full list waits
for the device (vk_upscaler_settle). release_kept's features whose last frame the GPU still runs are now released after
the create that follows, not before it: none of them is evaluated again, and its log line says how many.

NX2NAVI_UPSCALER_IDLE_WAIT (upscaler_idle_wait in [display], default 0, read at start) set to 1 waits for the device
at each of these places as up to 0.10.1, for comparison: rows in cfg.c's known[], metas[] and sections[], and an entry
in nx2navi.ini.example. retire_idle still waits for the device before the FSR contexts it gives back while there are
any; vk_ffx.cpp and vk_ffx4.cpp follow in the next commit.

Not reproduced before the change: no DLSS session log here (C-6 rests on the code and on F-4-02's mark). Every runtime
file compiles, vk_dlaa.c and cfg.c with -Wall -Wextra and no warning; cfg_rows_check reports 0 problems.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:13:01 -05:00
chrg6
86e6d62d35 Add a list of what an upscaler lets go of, released once the GPU is past it, and its unit check.
runtime/vk_up_defer.h holds, per owner, a DLSS feature, an FSR context or an image frame generation can still read,
each with the timeline value of the last submission that named it and an optional present count. upd_take releases,
in the order they were put, the entries whose value the GPU has reached and whose present count has passed; nothing
in it waits or locks, so the owner checks at its own calls on the thread that makes its other upscaler calls. It is
the piece the next commits use to take the upscalers' device-idle waits off the replaying thread: those waits stall
the present's flip, the completion thread and upload submits with the queue lock held, at every pause-screen size,
reload, window size or mode change. Nothing uses it yet.

tests/up_defer_check.py builds tests/up_defer_check.c with the header as C and as C++ (vk_ffx.cpp and vk_ffx4.cpp
include it) with -Wall -Wextra -Werror and checks both gates, the order, a full list, a null handle and a release that
puts. It passed, and failed with each of three faults put into a copy of the header.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:13:01 -05:00
chrg6
c14988150c Keep the prebloom images' last use when they are made again for another picture size.
vk_dlaa.c pb_images noted the present count in PB.used and then, for a new picture size, pb_destroy cleared the whole
struct with the old images, PB.used with it. The new images were made with PB.used 0, so at the next present
idle_release took them for 600 presents unused (once the run was 600 presents old) and destroyed them without a wait
for the GPU, while the frame that had just drawn into them could still be in flight; the frame after made them again.
With dlss_prebloom = 1 (off by default) that happened at every change of the picture's size, with the log line
"prebloom: the picture-size images went unused for 600 presents: freed" right after it. PB.used is now set again once
the new images are made.

Not reproduced before the change (found by reading the code). Every runtime file compiles, vk_dlaa.c with -Wall
-Wextra and no warning.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:51 -05:00
chrg6
c6b29834eb Give the upscalers' device-idle waits and their releases perf marks of their own, so a slow frame's line names them.
The twelve places where DLSS (vk_dlaa.c) and FSR 3.1 and FSR 4 (vk_ffx.cpp, vk_ffx4.cpp context_for) wait for the
whole device to be idle before they free what the GPU may still use go through one function, vk_upscaler_device_idle,
which NX2NAVI_TRACE=perf times as "upscaler device waits". The releases of DLSS features and of FSR contexts are timed
as "upscaler releases". These waits run on the replaying thread with the queue lock held, so the present's flip, the
completion thread's hold release and upload submits wait with them at each pause-screen size, reload, window size or
mode change; until now only the "DLSS feature creates" mark saw one of them, the wait in release_kept, which it summed
with the create. That mark now times the create alone.

Diagnostics only: nothing waits or frees differently. stutter_map_check, stutter_hitch_check, frame_step_check and
fg_reset_check passed; every runtime file compiles, the changed ones with -Wall -Wextra and no warning.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
61b8a6d62c Wait for room in the flight table in mem_lazy_skip only when there is another buffer's range to give back.
The review found that memory.c mem_lazy_skip's loop over the ranges armed for another buffer waited for a free entry
in the flight table before it looked for one. On its last pass, which finds nothing left to take, the replay could
still wait for another thread's fetch to end when the table was full. Now it looks first, and waits for room only when
it has a range to take and every entry is in use, then looks again. The search is lazy_foreign_locked, shared with the
check before the loop. Not reproduced before the change; lazy_fetch_check passed in all its scenarios (foreign and
flood among them).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
47078fa0cd Have an eviction take the next candidate when a victim was finalized while pool_lock was down, as it did before.
The review found that vk_memory.c mirror_evict picks its victims before it fetches their armed ranges without
pool_lock, and skips a victim vk_pool_fini took off the mirrored list meanwhile. The bytes freed then fell short of
what the caller needed, and the candidates after the victims were never tried, where the loop under one lock had moved
on to the next. Now every candidate's range and mirror is kept, and the loop under the lock goes on through the
candidates until the bytes are freed; a candidate past the first victims has its armed ranges fetched under the lock
there, as every victim's were in 0.10.1. hot_check demotes every pool that qualifies, up to eight, so it has nothing
to make up: its comment says that a candidate finalized meanwhile is not replaced, since its mirror goes with the pool.
Rare, needing a finalize between the two locks; compile check only, as an eviction needs a device. Not reproduced
before the change.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
b5e4243cc4 Name every thaw that fetches under the freeze lock in the lazy write-back's comments, and say what LZ_SLOTS does when it runs short.
The review found that the comments on frz_lock (memory.c, above FRZ_SHIFT), lazy_fly_run and vk_memory.c lazy_fetch
said a fetch runs under frz_lock only in a mapping change's thaw or with NX2NAVI_WB_LAZY_UNLOCKED=0. In fact
thaw_begin_locked's lazy_drop_region_locked runs in every thaw under the lock: a mapping change, the lazy heap's commit,
and thaw_fault on a write lazy_fault did not take. lazy_drop_region_locked's own comment said "a write fault in a
frozen region", but a frozen region holds no armed ranges. And the LZ_SLOTS comment called four slots more than the
threads that touch armed pages together, where the reference session had six such threads and the flight table holds
64. The comments now give that list of thaws, and LZ_SLOTS says that fetches past it wait 20 us at a time for a slot.
Comments only.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
a8d18ef474 Ask an armed region's page state under the freeze lock when a fault finds its range gone, so that a late write is not taken for a frozen page's and a read is not passed on past an arm.
The review found two windows in memory.c lazy_fault, where a fault holding frz_lock and lz_lock finds its address
neither armed nor in flight. It returned 0 there, and access_fault then asked for the page's state without a lock.
A write whose range a flight gave back while the write waited for frz_lock went on to thaw_fault, whose thaw released
every armed range of the 2 MiB region (lazy_drop_region_locked), each fetched under the lock. And after a release
that sets no hold (mem_lazy_skip's give-back of another buffer's range), the completion thread could arm the page
again between that 0 and access_fault's query, so a read saw an inaccessible page and went to the crash handler; that
one predates the ranges in flight.

Now a fault that finds a flight there waits for it as before. Otherwise it releases lz_lock, asks for the page's state
while it still holds frz_lock, under which no arm runs, and runs the access again when the page allows it: a read on
a readable page, a write on a read-write one. A write to a frozen page still goes on to thaw_fault, and an address
that is not ours to the crash handler.

The new late_write scenario of tests/lazy_fetch_check.py has a thread write a range whose release starts and ends
while the write waits for frz_lock: the write runs again and the region's other range stays armed. With the old
return of 0 it fails, the other range released. The read's window has no check: the arm would have to land between
two instructions. Not reproduced in the game before the change; lazy_fetch_check passed in all its scenarios.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
da60e0d327 Make wb_lazy_unlocked = 0 bring back 0.10.1's locking in evictions and releases as well, which the review found it did not.
With the setting at 0, frz_lock was held across every fault's and release's fetch again and mem_thaw released under the
lock, but two parts of the new code stayed. vk_memory.c mirror_evict and hot_check still fetched their victims'
armed ranges before taking pool_lock again, and a release still put up to 16 copies in one submission. Those are the
parts that run only with a device and have no check without one, so a device loss caused by them was the one case the
switch could not take back. Now vk_memory.c reads the setting too (evict_unlocked, once on the replay thread, where
both run): at 0 neither drops pool_lock, and the release under it fetches everything, victim by victim, as before. And
memory.c lazy_take_batch_locked takes one segment at a time at 0, so each is fetched on its own.

What stays at 0 is the table of ranges in flight: a touch of a range being fetched waits for that fetch and runs
again (a write no longer goes on to thaw_fault and releases its region), the hold is set as the range goes out, and
mem_lazy_touches counts the range. memory.c "Ranges in flight", nx2navi.ini.example and docs/building.md say so.

tests/lazy_fetch_check.py gains locked_release, the release scenario at 0: three fetches of one copy each, where the
batched release makes two of up to two copies. It fails with the cap taken out. The eviction's side has a compile
check only. Not reproduced before the change; lazy_fetch_check passed in all its scenarios.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
2e077edc9d Fetch the lazy write-back's ranges without the freeze lock, so a touch of GPU-written memory holds up only the threads that touch that range, with wb_lazy_unlocked = 0 to go back.
A CPU touch of an armed range ran its fetch (vk_memory.c lazy_fetch: a GPU copy after everything handed over, and a wait of up to 10 s) inside the fault handler under frz_lock. Every other fault, thaw, freeze and arm waited for the whole copy meanwhile: the replay's copies of guest memory (mem_copy_readable, vk_memory.c mirror_capture) and its freeze pass, the completion thread's arms, file reads and pool finalizes. A write that lost the race to the fetch went to thaw_fault after it and released every other armed range of its 2 MiB region, a second fetch. mirror_evict and hot_check fetched their victims' ranges under the exclusive pool_lock, which the completion thread's arms and every pool lookup wait for. This is AUDIT F-4-05 (C-12); the design's probe of the unchanged code had an unrelated mem_copy_readable, arm and freeze each wait 299 ms behind a 300 ms fetch, and the losing write 378.8 ms with two fetches.

A range now leaves the armed list into a table of ranges in flight (memory.c "Ranges in flight", lz_fly) in the same lz_lock hold, and stays there, its pages inaccessible, until its bytes are down and its pages read-write again (lazy_fly_run), without frz_lock. A fault on a range in flight waits for that entry alone on lz_lock's condition and runs the access again, the losing write included; a thaw, a mapping change, a release of its buffer, an arm, a freeze and mem_lazy_touches see it. Every store to lzr is made under lz_lock from the list, the entries and the region an arm is protecting (lz_arm_r, which replaces the previous commit's restore). An entry starts under frz_lock and lz_lock and ends under lz_lock alone, so a holder of frz_lock may wait for one: its fetch takes only the queue lock, which no frz_lock taker holds. mem_thaw releases what is armed in its regions before it takes frz_lock (lazy_release_regions), so file reads and pool finalizes fetch without it; a mapping change's thaw still fetches under it, as its page calls follow. The holds a fault sets (lz_hot, lz_hots) are now set as the range goes into flight, so an arm meanwhile finds them. The fetch takes several copies in one submission and one wait (MemLazyCopy), and the back end has four slots of command pool, command buffer and fence, so fetches of different ranges run at once; a release takes up to 16 armed ranges out together.

mirror_evict and hot_check now fetch what only their victims' mirrors hold before taking pool_lock exclusively, and then, under the lock, release only what was armed since (usually nothing) before the mirror leaves the list and the device address table names the imported buffer again. The design moved the release after the lock instead, which would have had the table change before the fetch: a shader of a submission still running reads through the table as it changes, and the fetch waits for every submission handed over. A victim that vk_pool_fini took off the list meanwhile is left to it; hot_check copies its log line's fields under the lock, as its records may be freed once the lock drops.

The new setting wb_lazy_unlocked (NX2NAVI_WB_LAZY_UNLOCKED, default 1, restart) set to 0 keeps frz_lock from the take-out to the end of each flight and leaves mem_thaw's releases to the thaw, as in 0.10.1; the table, the slots and the eviction order stay. The "mem: lazy write-back" line counts segments fetched; a second line gives the touches that waited for a range in flight, the longest such wait and the most ranges in flight at once; "mem: sizes" counts fetches that found every slot busy. The frz_lock and lz_lock comments say what holds what now (B-12). mem_lazy_touches has callers outside memory.c (camera_speed.c, nvn_api_cmd.c ctl_fingerprint), which now skip a range in flight instead of faulting on it and waiting.

Not reproduced in the game before the change. tests\lazy_fetch_check.py now runs 20 scenarios as their own processes: arm races, no stall of freezes, copies, arms, thaws and releases of other ranges during a fetch, joins with frz_lock held elsewhere, a losing write that keeps the region's other range armed, two fetches at once, a thaw, a mapping change and a release meeting a range in flight, an arm refused over one, the foreign give-back, the chunk mode, the old wait with wb_lazy_unlocked = 0 (joins and losing writes as well), the report lines, 80 touching threads past the 64 entries, and a freeze refused; all pass, in six runs. Taking the join before frz_lock out fails the join scenario. cfg_rows_check reports 0 problems; every runtime file compiles, and memory.c and vk_memory.c add no -Wall -Wextra warning; the test matrix passes 75 of 77, installer_test and worktree_test failing for the environment as before. vk_memory.c's slots and the eviction's order have no check without a device.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
f586c291fa Give the longest lazy fetch and the longest wait of a fault for the freeze lock on the "mem: lazy write-back" line.
A fetch of lazily written pages runs inside the fault handler under frz_lock (memory.c lazy_gone_locked and lazy_piece_fetch_locked call vk_memory.c lazy_fetch, a GPU copy after everything handed over and a wait for it), so every other fault, thaw, freeze and arm waits for the whole copy. The line gave only the fetches' sum: the 0.10.0 session's 848 fetches took 415.8 ms in all, and how long one took, or how long another thread waited behind it, was not known. Both fetch sites now keep the longest fetch, and lazy_fault keeps the longest time a fault waited to take frz_lock; the report line prints them as "the longest X ms, faults waited up to Y ms for the freeze lock" and starts both over. No tool parses the line.

tests\lazy_fetch_check.py's new report scenario fetches for 30 and 120 ms and has a third fault wait 60 ms for frz_lock: the line reads 120.29 ms and 60.03 ms, and the next one 0.00 for both. arm_race, arm_race_slow, wb_lazy_check and heap_lazy_copy_check passed; memory.c compiles without a warning under -Wall -Wextra.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
ef0673fba0 Mark a region for the lazy write-back before its pages become inaccessible, so that a read during an arm is not passed to the crash handler.
mem_lazy_arm made a range's pages inaccessible (lazy_protect_locked: the protect, then the back end's clean test) and set the region's lzr mark only after the segment was in the list. lazy_fault returns 0 at once for an unmarked region, without taking frz_lock, so a CPU read in that window found the page inaccessible in access_fault and went to the crash handler as an access violation. A write survived only because thaw_fault waits for frz_lock. Arms come from the completion thread (vk_memory.c vk_gpu_written_back) when a submission's fence completes, which is when the game's own fence wait returns and it reads the results; the window lasts as long as the protect and the clean test, which can take tens of milliseconds while the driver pins memory.

The mark is now set before the protect, after the cheap refusals, and put back when the protect or the clean test refuses the arm. A fault from the protect on finds the mark, waits for frz_lock and then finds the segment, or the pages given back. Every store to lzr is made under frz_lock, which the arm holds, so putting it back is exact.

Not reproduced in the game before the change. The new tests\lazy_fetch_check.py reproduces it with the real Windows fault handler: a thread reads a page while it is armed and given back 20,000 times (arm_race: an access violation in 5 of 6 runs before the change), and once with a clean test of 100 ms (arm_race_slow: an access violation in every run). Both pass in every run since. wb_lazy_check, heap_lazy_copy_check, freeze_hold_check and mem_growth_check passed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
5ee960f811 Add NX2NAVI_GEN_SEED, a test switch that starts the page write generations at a given value, so a play-test can cross 2^32 in seconds instead of 15 to 66 hours.
Whether the 64-bit generations hold past 2^32 in the game, with every texture check, mirror sync and shadow copy of a
real scene, could not be seen in play: reaching 2^32 takes 15 to 66 hours. gen_wrap_check shows it for the generation
code alone. NX2NAVI_GEN_SEED=<n> (nx2navi.ini gen_seed, diagnostics, default 0) makes the run's first generation
n + 1: with 4294966000 a run passes 2^32 within its first seconds, and with 1099511627000 the low 32 bits pass
through 0 again.

The first gen_next, before anything is stamped, reads the value (gen_seed). It is parsed with strtoull, decimal or
0x hexadecimal: strtoul stops at 2^32 - 1 where unsigned long is 32 bits, as on Windows, and would turn a seed past
32 bits into 4294967295 without a word. It is capped at 2^62, far from the 64-bit end, and only ever raises the
counter, since a lower value would leave stamps newer than it that read as written at every check. A value that is
no number, or negative, is logged and ignored; a seed in use is logged as "mem: page generations start at <n>
(NX2NAVI_GEN_SEED, a test switch)". Without the switch the only cost is one compare a generation that is true once.

cfg.c gets its metas row (0, 0..4611686018427387904, read at start), its known[] name and its sections[] entry, each
on a line of its own next to the other memory test switches (STAGING_FAIL_AT, VRAM_BALLAST_MB); nx2navi.ini.example
gets a commented gen_seed entry under [diagnostics]. The settings menu does not show it.

tests/gen_wrap_check runs the generation code a second way: the counter left at 0 and the seed taken from
NX2NAVI_GEN_SEED, for 4294967288, 0xfffffffff8, 0x7ffffffffffffff8 (capped), 12abc, -8, 0 and unset, each with the
start and the log line expected of it, then the same compares across the boundary. With strtoul in place of strtoull
two of those runs fail. Checks that passed: gen_wrap_check, gen_width_check, cfg_rows_check (0 problems),
pass_cc_check on vk_memory.c and cfg.c, no new -Wall -Wextra warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
028a6ce128 Count the page generation tables in trace=perf's "mem: sizes" line, so a play-test shows what the 64-bit generations cost in host memory.
The tables of page write generations take 8 bytes a page since the generations became 64-bit: each 256 MiB chunk of
guest address space stamped is 512 KiB a table instead of 256 KiB, and reg_max and creg_max are 2 MiB each. The
estimate for an outdoor area was 27 MiB of chunks at most, with less resident, but nothing in a log said how many
chunks a session allocates. slot_in now counts them (under gen_lock, an atomic add read without the lock), and
vk_mem_sizes adds "page generation table chunks N, X MiB" after "page generations handed out": the chunks of both
tables and the two per-region arrays, as allocated (only the pages that hold stamps become resident).

Only trace=perf logs the line; slot_in allocates a chunk a few dozen times a session. Checks that passed:
pass_cc_check on vk_memory.c, no new -Wall -Wextra warnings, gen_wrap_check, gen_width_check.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
6b95c867c8 Make the page write generations 64-bit, so a long session no longer wraps them and every texture, mirror and shadow check after the wrap answers wrong.
vk_memory.c handed out page write generations from a 32-bit counter (gen_now) and every consumer compared them as
32-bit values, with 0 meaning "never stamped". At 124 to 336 generations a frame (a 0.10.0 log with trace = perf) the
counter wrapped after about 66 h of play at 60 fps, 27.6 h at 144 and 15 to 17 h at 240. Past the wrap a stamp made
before it compared newer than every later check and one made after it older: textures and VRAM mirrors missed the
CPU's writes (stale textures, and the lazy fetch copying older mirror bytes over a CPU write), pages nothing wrote read
as written at every check (a hash or upload every time, and every such page of a mirrored pool copied up at each
sync), and the generation handed out at the wrap was 0. At 64 bits nothing wraps: 2^64 lasts millions of years at
336 a frame and 240 fps. Below 2^32 every value and every compare is the same as before.

What changed: gen_now, both tables (gen_chunk, cgen_chunk), reg_max and creg_max, PoolRec's synced_gen, mgen and sgen,
dirty_gen and DIRTY_UNSYNCED, VSample.mgen, every local and parameter that holds a generation in vk_memory.c,
vk_texture.c (Tex.wgen and gpu_gen, PEnt.since, storage_rewritten and its callers, sweep_hard's write-back check
included) and tex_probe.c, and vk.h's vk_mem_changed, vk_mem_changed_snap and vk_mem_unchanged(_why)_locked. The verify
mode's mismatch line prints the generations with %llu. The readers that take no gen_lock (lazy_clean, cpu_wrote_since,
verify_cb, vk_mem_sizes) stay lock-free at 64 bits: the slots, mgen, gen_now and a new chunk pointer are now stored
atomically (relaxed, or release where a reader acquires), the same instructions on x86-64, and a static assertion
holds that 64-bit atomics are lock-free. No lock, wait or thread changes.

gen_next, which hands out every generation, no longer warns at 0xC0000000 and up or at the wrap, which cannot come;
it logs one line when the count reaches 2^32, where earlier versions ran out, with the frame, the hours and the rate.

Cost: the tables take 8 bytes a page instead of 4 (512 KiB more per 256 MiB of guest address space stamped, both
tables together, allocated as before when a chunk is first stamped), reg_max and creg_max 2 MiB each instead of 1,
Tex and PoolRec 8 bytes more, PEnt 8 more. The page walks read 8 bytes a slot; a micro-benchmark of the walk measured no
difference.

Not reproduced in the game before the change: the wrong answers were reproduced by compiling vk_memory.c's generation
code with the counter seeded below 2^32. tests/gen_wrap_check now runs that code, with the readers outside its
section, across 2^32, 2^40 and 2^63 and checks that the compares answer as below the boundary; on the code before the
change 19 of its checks fail at 2^32 and the other two seeds do not fit the counter. The new tests/gen_width_check
compiles vk_memory.c, vk_texture.c and tex_probe.c with -Wshorten-64-to-32, -Wformat and -Wincompatible-pointer-types
as errors, so a 32-bit local left behind is found (shown by reverting single locals, fields and formats). Checks that
passed: gen_wrap_check, gen_width_check, pass_cc_check of every runtime file, no new -Wall -Wextra warnings in the
three files, tex_release_check, tex_compact_check, wb_lazy_check, gpa_check, mem_growth_check, mem_budget_check.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
879c76f91a Name every caller of reflex_state in nx2_sl.cpp's comment on why its calls never overlap.
The review found the comment incomplete: it named nx2sl_available on the device's thread and nx2sl_reflex2 under the
runtime's present lock, but nx2sl_reflex2 also runs on the device's thread with no lock (vk_dlssg.c
vk_dlssg_device_ready, reflex_apply(0)), and an older runtime reaches it through nx2sl_reflex. The comment now says
the runtime calls them on the device's thread before any present or under its present lock, never two at once; the
conclusion is unchanged. Comment only: the DLL builds with MSVC against the pinned Streamline SDK and passes the
export check.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
bb0317ac08 Log Reflex's frame limit as each of the first 16 swapchains hands it to Streamline again.
The review found that a play-test item expected nx2_sl.cpp nx2sl_reflex2's line after each resize, fullscreen toggle
or HDR change, but that export logs only its first 8 calls, which the device, the first swapchain, the first
hand-over and a movie use up, so a later rebuild usually logged nothing and the tester would read that as a failure.
vk_dlssg.c vk_dlssg_create_swapchain now logs, for each of the first 16 swapchains with a DLL that has nx2sl_reflex2,
the limit it sent and the one Streamline took ("dlssg: swapchain N: Reflex's mode and frame limit sent again"), on
the present thread under the present and queue locks where reflex_apply already logs a refused limit. The compile
check and -Wall on vk_dlssg.c (no new warnings) pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
46bf4823e6 Count Reflex's sleep at the frame start as the pacer's wait only while Reflex holds a frame limit.
The review found that vk_dlssg.c vk_dlssg_frame_start timed every held sleep as PERF_PACE, though the sleep takes the
pacer's place only while Streamline holds a frame limit. With reflex_limit 0, a DLL older than nx2sl_reflex2, during a
movie, after a refused limit or once the frames showed that Reflex does not hold the limit, the sleep is the driver's
wait for the GPU while the pacer waits as well, so GPU-bound frames read "bound: pace" in the perf windows (the
misattribution P-7 and F-7-04 describe), and reflex_limit 0 did not give back the old attribution. The sleep is now
timed only when it holds the frame (as before) and limit_set is not 0; the statistics line's held frame starts and
their time are counted as they were. Not reproduced before the change: the compile check and -Wall on vk_dlssg.c (no
new warnings), dlssg_pace_check and the test matrix pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:50 -05:00
chrg6
e0843439a7 Add the reflex_limit setting (NX2NAVI_REFLEX_LIMIT, default 1, live), which gives the frame limit back to the game's pacer with DLSS Frame Generation.
Reflex's hold of the frame limit with DLSS-G is unmeasured here (no RTX 40 or 50), so a player needs a way back to
the pacing of 0.10.1 without an older DLL. vk_dlssg.c limit_allowed reads the setting through cfg_bool, again
whenever a setting changed (cfg_version), as vk_lowlat.c reads low_latency; with 0 the presenting guest thread asks
Reflex for no limit, the present thread hands Streamline 0, and the "frame rate:" line names NX2NAVI_REFLEX_LIMIT=0 as
the reason the game's own limiter holds it. The row sits beside FG_MULT in metas[], beside REFLEX in known[] and
beside LOW_LATENCY in sections[] (display), and nx2navi.ini.example explains it under framegen_multiplier.
cfg_rows_check reports 0 problems; vk_dlssg.c and cfg.c compile clean with the dev build's flags.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:49 -05:00
chrg6
dc3f424fde Let Reflex hold the frame limit with DLSS Frame Generation, and the game's pacer keep only its ceiling while it does.
With DLSS-G, Reflex's sleep held each frame's start with no limit while the game's pacer still held the limit on the
present thread and at the window image's acquire (A-16, P-16): the pacer's wait came after the picture was drawn, so
the pad was read up to a period before the picture left, on top of Reflex's own wait for the GPU.
vk_dlssg.c vk_dlssg_pace, on the presenting guest thread in nvnQueuePresentTexture, asks dlssg_pace.h which limit Reflex
should hold and stores it; vk_dlssg_frame on the present thread hands a changed limit to Streamline before its present
(reflex_apply, under the present lock, which keeps Streamline's options calls on one thread at a time), as does each
new swapchain, which has neither mode nor limit. GfxBackend.frame_paced now returns the period the pacer still holds:
vk_backend.c vk_be_frame_paced gives the game's ceiling (frame_step_ceiling_ns, 240) while Reflex holds the limit,
0 where a driver's low latency holds it (vk_lowlat_pace, unchanged), else the period; during a movie the back ends get
0 and the pacer keeps the movie's period and catch-up slots as before. The pacer takes the limit back before the first
host present, after 200 ms without one, with a DLL older than nx2sl_reflex2 (one log line at start), for a limit
Streamline refused, and for the run when the frames show Reflex does not hold it; the "frame rate:" lines name who
holds it, the "dlssg: present N" line counts the frame starts Reflex held, and Reflex's sleep now counts as pacing
(perf_end_any, on whichever thread acquires). No wait is added on a drawing thread: the limiter's wait moves from the
acquire's sleep_until and the present thread's slot to Reflex's sleep a few lines later in the same acquire, and the
present thread's options call returns at once. docs/architecture.md names Reflex as a limiter. Not reproduced before
the change (no RTX 40 or 50 here); vk_dlssg.c, vk_backend.c and nvn_api_obj.c compile clean with the dev build's
flags, and dlssg_pace_check, fg_reset_check and frame_step_check pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:49 -05:00
chrg6
2bf32649af Name the game's frame-rate ceiling as a period of its own, frame_step_ceiling_ns, for a pacer that hands the limit over.
When Reflex holds the frame limit with DLSS Frame Generation, the game's pacer is to keep only the rate the game's
code survives (240 a second; the quick menu read address 0 at 640 to 700), so that a limiter that lets go, or one
that counts other frames than the game's, cannot run the game past it. frame_step_ceiling_ns gives that period
from LIMIT_FREE_MAX_HZ, or LIMIT_MAX_HZ for a build with the fps60 or realtime patch set; frame_step_cap_ns is left
as it is. tests/frame_step_check.c checks 4166666 ns for a game stepping by the clock; frame_step_check passes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
2d6f9f07b2 Decide who holds the frame limit with DLSS Frame Generation in a header of its own, runtime/dlssg_pace.h, with a check.
Reflex can hold the frame limit where each frame starts (nx2sl_reflex2), but the driver only takes a limit through
Streamline, the SDK does not say whether its minimum interval counts the frames the game draws or every frame on
screen, and nothing tells the game that the limit holds. dlssg_pace_step decides from what the presenting guest
thread sees alone: the limit Reflex should be given (the drawn period, 0 during a movie, before the first host
present, after 200 ms without one, or when not allowed), and whether Reflex holds this frame's limit, which it does
only once Streamline took exactly that limit. Over windows of 60 presents Reflex held to one period and multiplier,
over half sooner than three quarters of the period means nothing holds the game and the pacer takes the limit for
the run; over half the multiplier periods apart means the driver counted the frames on screen, and Reflex is given
the period over the multiplier from then on, or the pacer the limit if it happens again. Being pure, the decision is
checked without a GPU: tests/dlssg_pace_check.py builds it with undefined-behaviour traps and drives it through
each case; four mutations of the header (the window's majority, the idle bound, a host present stamped after the
clock read, the applied-limit test) each fail it. docs/building.md "Checks" names it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
660c42f6d2 Give nx2navi_dlssg.dll an export that hands Reflex a frame limit along with its mode.
With DLSS Frame Generation, nx2sl_reflex set only Reflex's mode, so Reflex's sleep at the frame's start held no
limit and the game's own pacer held the finished picture until its turn, a second wait after the frame was drawn
(A-16, P-16). nx2sl_reflex2(mode, frame_limit_us) fills sl::ReflexOptions::frameLimitUs as well, which the driver's
sleep then holds as its minimum interval. Streamline drops both with the swapchain and a call without the limit
clears it, so every call carries both. It returns 1 only when Streamline took them and, for a limit, did not report
Reflex's sleep as unavailable on this GPU and driver (slReflexGetState, asked from nx2sl_available on the device's
thread and from nx2sl_reflex2 under the runtime's present lock, never at once). nx2sl_reflex stays exported for older
runtimes and sets the mode alone; NX2SL_VERSION stays 3, since the runtime refuses any other version and the new
export is optional. Built with tools/build_dlssg.py against the pinned Streamline 2.14.16 SDK: clean at /W3, 15
exports including nx2sl_reflex2.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
af2ecc3c03 Correct the comments on RivaTuner Statistics Server's pacing beside the ceiling, the state file's writers and the crash path.
The review found three comments that claimed more than the code does. rtss_decide.h rtss_pace_period and frame_rate.c
(above frame_step_cap_ns) said the frames RTSS lets through always come further apart than the 240 ceiling, so the two
limiters never stack: at a limit of 240, or unlimited, which RTSS gets as 240, the ceiling is RTSS's own period, and
both may hold a frame. rtss_limit.c state_put said it runs on the worker thread only, but rtss_limit_restore's put back
calls it on the exiting thread, under the lock. And rtss_limit.c's header said a crash and a fatal error make no call
into RTSS's library, which holds only while the state file has the run's line. The comments now say so, and the header
names the exception at 240. (The comment above rtss_decide was rewrapped and lost its reference to rtss_fresh with the
unproven-hold rule.)

No behaviour changes. Checks that passed: tests/rtss_limit_check.py, tests/frame_step_check.py, pass_cc_check.py --wall
on rtss_limit.c and frame_rate.c (only frame_rate.c's existing game_interval warning), tools/check_content.py --staged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
9f45d7448f Leave rtss-profile.txt alone when it is there but does not read, so the lines of other executables in the folder stay.
The review found that rtss_limit.c state_put took a state file plat_file_read could not read for a missing one: a file
another program held open (a scan) was rewritten with this profile's line alone, or deleted when the run removed its
line, and the lines of other executables that share the folder (dev or probe builds) were lost. state_put now returns
0 when the read gave nothing but plat_file_size finds the file with bytes in it; the run then goes on as with a file
that cannot be written (one log line, and a crash puts the profile back as an orderly exit does). An empty file is
still rewritten. tests/rtss_worker_check.c state_locked checks a file holding another executable's line that does not
read: the run limits as before, logs it, and the file stays as it was through the exit.

Not reproduced before the change. Checks that passed: tests/rtss_limit_check.py, pass_cc_check.py --wall on
rtss_limit.c, tools/check_content.py --staged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
da822abec3 Keep the state file's line ahead of RivaTuner Statistics Server's profile, and keep what a run owes the profile after a put back that did not land.
The review found three ways rtss-profile.txt could fall behind the profile. Every write after a run's first one saved
the profile before the line, so a kill between the two left a line whose value the profile no longer held, and the
next start took the game's limit for the player's (rtss_restore_plan LEAVE). A put back whose write did not land kept
the line, but rtss_put_back_done cleared S.changed, so a later limit took the run's first-write path again and read the
game's own limit from the profile as the value to put back, overwriting the true one in the line. And put_back removed
the line after DeleteProfile without checking that the profile went away.

The line now carries the value the run wrote before the newest one (rtss_decide.h RtssSaved prev_limit and prev_den,
rtss1 with two more fields; the file came with this branch and was never released). rtss_limit.c apply writes the new
value into the line before the profile (record_before) and drops the previous one once the profile holds the new one,
or returns to it when RTSS refused it (record_after), and rtss_restore_plan takes a profile holding either value for
the game's. A new owed flag keeps had_profile, old_limit and old_denominator from the run's first write until a put
back lands: a later first write keeps them and the line's values, a failed first write leaves the line alone, and
rtss_limit_restore tries the put back again at the exit. put_back and restore_left check that a deleted profile is
gone before they remove the line, and restore_left's lines name the value the profile held. tests/rtss_limit_check.c
covers the new fields and the plan for either value; tests/rtss_worker_check.c checks the line before and after a
write (state_record), a run killed inside the save of its second limit and the next start removing the profile
(kill_restart_mid_write), a put back that did not land followed by a limit and the exit (put_back_owed), and a removal
that did not land (put_back_undeleted). Lines over 120 columns from the earlier fixes are wrapped.

Not reproduced before the change. Checks that passed: tests/rtss_limit_check.py (44 of 44), tests/exit_order_check.py,
tests/fatal_exit_check.py, pass_cc_check.py --wall on rtss_limit.c, tools/check_content.py --staged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
1744681f1a Give up on RivaTuner Statistics Server for the run when it hooks the game with the limit for 30 s and never shows it holds it.
The review found that nothing ended an RTSS hold that was never proven: when RTSS listed the game with the limit in
its profile but one sign never came (presents of another API value than Vulkan's, the update flag never cleared, a
period that never moved), the profile kept the limit and the game's pacer held the frames too for the whole run, two
limiters where the pacing rules allow one, with only a reason line in the log and nothing in the menu. Before the
hold had to be proven, RTSS alone held the game once it listed it. rtss_decide.h rtss_decide now counts the polls in a
row where RTSS lists the game, the profile holds a limit of this run (not "none"), a picture went out within 200 ms
and the hold is not shown; at RTSS_UNPROVEN_POLLS (60, 30 s) it puts the profile back and RTSS is out for the run
(RTSS_FAIL_UNPROVEN, "RivaTuner Statistics Server never showed it holds the game"). A minimised window, RTSS not
hooked, a want of no limit, a kept-out want or a poll that shows the hold start the count again. rtss_limit.c logs the
give-up, rtss_limit_state returns 3 for it, so the menu says RivaTuner does not limit the game and the game does, and
rtss_out_of_limit counts it, so a limit set in RTSS itself is reported as a second limiter. The comment above
rtss_decide, rewritten for the rule, is wrapped at 120 columns and no longer cites rtss_fresh, which does not exist.
tests/rtss_limit_check.c checks each sign that never comes (given up at the 60th poll, once, for the run), the polls
that do not count and a shown hold restarting the count; tests/rtss_worker_check.c unproven_run checks it against the
real worker. docs/architecture.md, docs/building.md and nx2navi.ini.example describe the rule and the remaining window
of two limiters before it.

This changes when the game takes the limit back from RTSS, beyond the approved rules: an RTSS whose hold never shows
in its shared memory (a version or setting that differs) is now left out after 30 s and the game's limiter holds
alone, as with rtss = 0, where it held the frames beside the game's limiter before. Not reproduced before the change.
Checks that passed: tests/rtss_limit_check.py (41 of 41), pass_cc_check.py --wall on rtss_limit.c,
tools/check_content.py --staged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
a7b4f0c7cb Put back a profile a run that did not close left in RivaTuner Statistics Server, also at a start that keeps RTSS out.
The review found that the worker read rtss-profile.txt only on its first poll with the hooks library, which it loads
only for a limit, so a start with frame generation on, with the driver's low latency on from the first present, or
with NX2NAVI_RTSS=0 never put back a killed run's limit: RTSS then went on limiting the game at that value (counting
the generated frames too under frame generation), and the log and the menu reported it as the player's own limit
beside the game's limiter. rtss_limit.c worker now looks at the state file on its first poll that finds RTSS running
(left_start): a limit puts the profile back with the library it loads anyway, as before; frame generation and low
latency load the library for that one restore (restore_left, which now takes the line found); NX2NAVI_RTSS=0 keeps the
thread read-only and logs that the file holds the game's limit from a run that did not close and that a start with
RTSS let in puts it back. A restore the library could not be loaded for waits for the first poll with the library.
tests/rtss_worker_check.c adds kill_restart_framegen (the next start asks frame generation first: the profile is
removed, no limit written, the line gone) and kill_restart_setting (NX2NAVI_RTSS=0: no library loaded, the profile and
the line stay, the log says so). The rtss_limit.c header, docs/architecture.md, docs/building.md and
nx2navi.ini.example say so, and the ini text no longer says the game alone paces movies under RTSS.

Not reproduced before the change. Checks that passed: tests/rtss_limit_check.py (40 of 40),
pass_cc_check.py --wall on rtss_limit.c, tools/check_content.py --staged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
ad84e0a1c0 Say what RivaTuner Statistics Server does with a movie at the default limit of 60, where the profile keeps the limit.
The review found that rtss_decide.h rtss_target writes no limit only while a movie runs faster than the play limit, so
at the default limit of 60 a 60 fps movie leaves 60 in RTSS's profile and RTSS holds the movie beside the game's pacer.
Its comment said RTSS at the movie's rate would space out the frames that catch up after a late frame, and README.md
("one limiter at a time") and docs/architecture.md read as if every movie under RTSS were the pacer's alone. The
approved rule writes nothing to the profile at movies when the limit is 60 or more, so the code stays: rtss_target's
and rtss_pace_period's comments now say that at the movie's own rate RTSS keeps holding it and that whether its
limiter lets the frames that catch up through is not measured, README.md drops "one limiter at a time", and
docs/architecture.md lists the movie at 60 among the exceptions to one limiter. tests/rtss_limit_check.c pins a movie
at 60 under a limit of 60: no write, the profile keeps 60, held again after it.

No behaviour changes. Checks that passed: tests/rtss_limit_check.py, tools/check_content.py --staged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
80b96a748d Describe RivaTuner Statistics Server's hand-off, the pacer's fallbacks and rtss-profile.txt as they behave now.
docs/architecture.md said the pacer holds nothing while RTSS holds the limit, through movies and minimised windows,
that the pacer stands aside once RTSS lists the process, that RTSS fails at 1.2 times the limit over six polls, and
that a killed run leaves its limit for the next run to take as the player's. The one-limiter and RTSS paragraphs now
name the pacer's fallbacks (before the first host present, a movie, 200 ms with no host present, a limiter seen
letting frames through, any failure) and its 240 ceiling beside RTSS, the settle rule, a movie's profile, the proof
RTSS must give before the pacer stands aside, the 1.05 rule over three new periods, the rounding down, and the state
file with rtss_limit_crash. nx2navi.ini.example's rtss lines, README.md's frame generation item, docs/building.md's
rtss_limit_check and frame_step_check paragraphs and rtss_limit.c's header follow, and the comments at main.c
crash_handler's and dispatch.c fatal_end's rtss_limit_crash say when RTSS's library is still called there (no line in
the state file, never on RTSS's own thread). No code changes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
60694532a1 Round a fractional limit down for a RivaTuner Statistics Server without the denominator field, so "display" never holds the game above the refresh.
apply() rounded the millihertz to the nearest whole frame where RTSS has no FramerateLimitDenominator, so "display" on
a 59.94 Hz monitor wrote 60 and on a 143.856 Hz one 144: above the refresh, one frame more than the monitor shows
every 17 s and every 7 s. Only "display" asks a fractional rate under RTSS (fixed limits and movie rates are whole,
frame generation keeps RTSS out), so the new rtss_decide.h rtss_limit_value rounds every fraction down when the
denominator is missing (59/1, 143/1), and keeps a whole rate over 1 and a fraction over 1000 where RTSS knows the
denominator. VRR and present-wait handling stay to be measured.

Not reproduced before the change. rtss_limit_check (T15: rtss_limit_value at 59940, 60000, 143856, 0 and 240000 with
and without the denominator) and rtss_worker_check (round_logged_whole now 59/1 and 143/1, round_logged) pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
4d4b97b043 Let a want reach RivaTuner Statistics Server's profile only once it stayed two polls, and write nothing before the first host present.
A want that flips between a limit and kept-out created and removed RTSS's profile at each flip: with a driver's low
latency vk_lowlat_on() is 0 before the first swapchain and during every swapchain rebuild, so a warm start without the
boot screen wrote the profile and removed it a poll later, and a poll that fell into a rebuild did the same (lane A's
probe: 6 creations and 5 removals for a want flipping each poll, 1 and 1 at start before RTSS hooked the game).
rtss_decide now counts the polls in a row a want stayed: a limit goes into the profile and a kept-out want puts it
back only after RTSS_SETTLE_POLLS (2), while standing aside still ends at once (rtss_limit_active compares the want),
and nothing is written before the first host present (RTSS_WHY_NO_PRESENT). A limit change reaches the profile up to a
second later; the game's limiter holds the new limit meanwhile. The created and removed lines were numbered before.

Not reproduced before the change. rtss_limit_check (T8: kept out for one poll and for two, a second reason, a limit
again; T9: 20 flips after a written limit give no write or put back, 20 from the start none, flips every two polls
act; no write while no host present) and rtss_worker_check (churn_run and every scenario that now settles) pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:26 -05:00
chrg6
f7899e5443 Let the game's limiter stand aside only once RivaTuner Statistics Server is seen holding the game, and give up on RTSS at 1.05 times the limit over three fresh periods.
The worker stood aside as soon as RTSS listed the process and the profile read back, before RTSS had seen a Vulkan
present (with InjectionDelay=15000 it lists the game some seconds after start) or read the new profile. Its frame
count check took the newest period whatever its age: a period frozen at 144 fps after the limit went to 60 and the
window was minimised failed RTSS for the run, and 1.15 times the limit was never caught (lane A's probe: listed_no_api,
stale_fps, over115). rtss_decide now parses the entry's flags and stands aside only when the profile held the target
at the poll before, RTSS lists the game with the Vulkan API value (flags & 0xFFFF == 0xA), RTSS_UPDATE_ASKED is clear,
a new period was seen within max(2.5 s, 2.5 times its length), no movie plays and a host present went out within
200 ms. Only a new period after a poll that stood aside is checked: three in a row over 1.05 times the limit put the
profile back for the run, and a stale period counts neither way. The worker reads the poll time and the newest host
present (gfx_window_hosted_ns), and logs each change of why the game's limiter paces, up to 100 lines.

Not reproduced before the change. rtss_limit_check (the flags at +264; T2 injection delay through not listed, no API,
the update flag, no period and holding; T3 quit; T6 the frame count at 144, 69, 63 fps, a frozen period and a newer
want; the idle case) and rtss_worker_check (over_run, over_slack_run, new api_run and idle_run, and every scenario that
now holds from the poll after the write) pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:25 -05:00
chrg6
04f4d44747 Keep RivaTuner Statistics Server's profile as it was in rtss-profile.txt, so a start after a kill or crash puts it back, and leave it to that start at a crash.
An exit that skipped rtss_limit_restore (Task Manager, a hang, lost_wait_exit, a kill during the pipeline save) left
the game's limit in Profiles\<exe>.cfg, and the next run's first apply took it for the player's own: that run then
put the stale limit back at its close, and a profile the game had made stayed for good (lane A's probe: kill1, then
kill2 'held a limit of 60/1'). The crash handler and the fatal error's end also called into RTSSHooks64.dll from a
process that had failed (P-13).

rtss_limit.c now keeps one line per profile in rtss-profile.txt next to the executable (rtss_decide.h RtssSaved:
whether the profile existed, its old limit and denominator, the value the run wrote), written atomically before the
run's first change of the profile, updated with each value written, and removed once the profile is back (put_back;
also when the first write did not stick). On the first poll with the hooks library a start that finds a line of its
profile puts the profile back first (restore_left, rtss_restore_plan): the old limit, or no profile, unless the player
changed or removed it since, which the written value tells. main.c crash_handler and dispatch.c fatal_end call the
new rtss_limit_crash, which closes the run and leaves the profile to that start with no call into RTSS's library; only
when the file could not be written does it put the profile back as before, and never from the worker's own thread. A
file that cannot be written leaves the run as it was, with one line in the log. Profiles frozen by earlier versions
cannot be told from the player's own.

Not reproduced before the change. rtss_limit_check (T12: the line's format and parse, the file's rewrite, each plan;
the static check that both crash paths go through rtss_limit_crash), rtss_worker_check (state_record, state_player,
state_unwritable, crash_recorded, and kill_restart with its player and edited variants, each a killed run and the next
start in processes of their own), fatal_exit_check and exit_order_check pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:25 -05:00
chrg6
208a10718d Let the game's pacer hold a movie while RivaTuner Statistics Server holds the limit, and keep the play limit in RTSS's profile through it.
During a movie frame_step_cap_ns wanted the movie's rate and returned 0 once RTSS held it, so pace() kept none of the
movie's slots while ocarina_step still gave one movie frame per update: each late frame put the picture behind the
sound for good, the fall-behind d3e50328 fixed for the game's pacer. Each movie start and end also rewrote RTSS's
profile (two saves and two UpdateProfiles per movie in lane A's probe). Now frame_step_cap_ns wants the play limit
through the movie and tells RTSS of the movie (rtss_limit_movie); rtss_pace_period keeps the movie's period from its
first frame, and the worker leaves the movie to the pacer (RTSS_WHY_MOVIE) with the profile as it is. Under a limit
below the movie's rate (30, 40, 45) the profile has no limit for the movie's length (rtss_decide.h rtss_target), since
RTSS at the movie's rate would space out the frames the pacer sends at once to catch up; that is still two writes per
movie there, and none at 60 or more.

Not reproduced before the change. rtss_limit_check (the movie case at 144 and 40, rtss_target, rtss_pace_period),
rtss_worker_check (movie_run, new movie_low_run) and frame_step_check (the want and period through a movie under RTSS)
pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:25 -05:00
chrg6
27a034e946 Keep the 240 ceiling while RivaTuner Statistics Server holds the limit, and take the limit back after 200 ms with no host present.
frame_step_cap_ns returned 0 once RTSS held the limit, so pace() set each slot to now and the acquire never waited.
RTSS holds frames inside the host present, and a minimised window (gfx_present.c present_image skips the present) or
a frame hold gives it none to hold: the game then ran with no limiter but pace_smooth's floor, which follows the
intervals down, and the quick menu is where 640-700 fps read address 0 (frame_rate.c LIMIT_FREE_MAX_HZ).
gfx_present.c flip_locked now stores the time of each host present that took its picture (gfx_window_hosted_ns), and
rtss_decide.h rtss_pace_period gives the pacer the limit's period before the first host present and once none went
out for 200 ms (as vk_lowlat.c HOLD_IDLE_NS), and the 240 ceiling's period, never 0, while RTSS holds. The frames RTSS
lets through come further apart than the ceiling, so pace() waits for none of them; no new wait site.

Not reproduced before the change. frame_step_check (new cases: holding with a host present 10 and 199 ms ago, 250 ms
ago, none yet, not holding, 144 minimised, unlimited) and rtss_limit_check (rtss_pace_period) pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 20:12:25 -05:00
chrg6
d9940113ef Give a network task's system event back to the event table before its record is freed. The table counts entries in host memory as live, so each task the game destroyed kept a slot for the rest of the run, and enough of them would fill it. Not reproduced before the change; system_event_check passed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:46:50 -05:00
chrg6
52d996d4aa Count the low-latency driver's sleep as pacing on whichever thread acquires, as the other acquire waits are counted since the pacing timing changed: vk_lowlat.c frame_begin runs on the acquiring thread, which need not be the one that presents, and perf_end dropped a wait made there.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:46:49 -05:00
chrg6
359591c237 Note the module list from each image's header and GetModuleFileNameA, which takes 0.3 ms where it took 25 ms.
The review of the ExitProcess note found that platform/win32.c plat_exit_note_modules, which runs on the window thread
at every focus loss, was quadratic in the number of modules: GetModuleInformation and GetModuleBaseNameA each walk the
loader's module list through ReadProcessMemory. Its probe measured 27.5 ms a note with 292 modules.

Each module's size now comes from SizeOfImage in its own image header and its name from GetModuleFileNameA, the part
after the last backslash. The header is read through ReadProcessMemory into a 1 KiB buffer rather than in place: a
module unloaded since EnumProcessModules then fails the read and is left out, where a plain read would fault. The
table's contents are the same.

Measured with a probe beside this change (283 modules loaded, scratchpad rv_fxexit_fix\modtime_rpm): 25.3 ms a note
before, 0.27 ms after, and both ways note the same 283 bases, sizes and names. tests/exit_tls_check.py, whose note
names the program's own module from that table, and the compile check of win32.c pass, with no new -Wall warnings.
The review's optional second note after the device is made is left out: vk_device.c create_device is another
branch's.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:24 -05:00
chrg6
e4f10a5a02 Log the hang watch's list of guest thread waits after the kernel lock is released, not while it is held.
The review of the hang watch found that kernel.c kern_wait_report called rt_log for up to 16 lines while it held the
kernel lock G. When the log queue is full and the log thread is stuck holding its write lock, one of the stalls the
watch reports, the put that drains the queue waits there with G held, which turns a log stall into a freeze of every
guest thread. wait_watch (NX2NAVI_WAIT_WATCH_S, off by default) had the same pattern.

wait_pass now formats its lines into a buffer of the caller's (WAIT_LINE_CAP bytes a line, room for the 16 or 24 lines
asked for), and kern_wait_report and wait_watch log them with the new wait_lines_log after unlock. A line that finds
no room ends the pass unmarked, so wait_watch logs it at its next pass. The lines read as before.

Not reproduced before the change (it needs a stalled log thread beside a report). tests/hang_watch_check.py checks
statically that wait_pass logs nothing and that both callers log only after unlock; the check fails on the code
before this change and passes after it, and the program's cases pass. The compile check of kernel.c passes with no
-Wall warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:24 -05:00
chrg6
cd3f015732 Log the guest's ExitProcess line after the exit guard, and bring the comments on gfx_window_closed's return up to date.
The review of the exit sequence found that svc.c case 0x07 logged "guest ExitProcess at ...; exiting" before
rt_exit_begin. When a close had already begun, the guest thread that the guard then parks still claimed to exit, and
past the log cap the line fell outside the end's reserve, which rt_exit_begin opens. It also found that host.h and
gfx_wsys.h still said gfx_window_closed does not return, though it returns for its own thread coming back, and that
rt_exit_begin's comment spoke of the window thread handling a message inside the sequence, which since the close
runs on a thread of its own happens only in the fallback.

svc.c logs the line after rt_exit_begin returns 1. The two header comments say gfx_window_closed returns only for its
own thread coming back (gfx_present.c rt_exit_begin), and rt_exit_begin's comment says the window thread runs the
sequence itself only when gfx_window_win.c exit_start could make no thread.

tests/exit_order_check.py checks statically that 0x07's rt_log comes after rt_exit_begin; it fails on the code
before this change and passes after it. exit_order_check and the compile check of every runtime file (a header
changed) pass, with no new -Wall warnings in svc.c and gfx_present.c.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:24 -05:00
chrg6
66c607414b End a hung fatal error's report with its own code when a thread waiting beside it gives up, not with the crash code.
The review of the shared failure guard found that a thread parked in dispatch.c rt_ending_enter ended the process
with 71 after 30 s even when the first report was a FATAL (70, or 72 for a lost device), so a fatal error whose report
hung read as a crash to the launcher.

rt_ending_enter now takes the code its caller ends with: fatal_end passes its own, main.c crash_handler 71. The first
caller stores it next to ending_tid, and a waiting thread that gives up after 30 s ends the process with it. Before
the store the value is 71, as before.

Reproduced before the change with a one-off build of tests/fatal_exit_check.c given a fatal note that never returns
and a fault on another thread: the process ended with 71 after 30.5 s; with this change it ends with 70 after 30.5 s.
That case takes 30 s, so the matrix gets a static check instead: tests/fatal_exit_check.py holds rt_ending_enter to
storing the first caller's code and ending with it, fatal_end to passing its code, and crash_handler to its first
line `if (!rt_ending_enter(71)) plat_exit(71);`. fatal_exit_check and the compile check of every runtime file (a
header changed) pass, with no new -Wall warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:24 -05:00
chrg6
e6a6d6a809 Skip RivaTuner's restore at a fatal error on the RTSS thread too, not only at a crash there.
The review of the RTSS thread change found that only main.c crash_handler skipped rtss_limit_restore on the "rtss"
thread. An rt_fatal on that thread while it holds rtss_limit.c's lock still called the restore from dispatch.c
fatal_end, whose trylocks then spent a second for nothing before the end went on. A-14 names fatal_end.

fatal_end now reads plat_thread_name and calls rtss_limit_restore only when the thread is not "rtss", as
crash_handler does.

Not reproduced in the game before the change. tests/fatal_exit_check.c's stand-in restore now says it ran, and a new
fatal_rtss mode calls rt_fatal on a thread spawned with the name "rtss": on the code before this change the restore
ran there (the check fails), now it does not, and the process still ends with 70 after the rumble stop. The other
cases check that the restore runs from any other thread. fatal_exit_check and the compile check of dispatch.c pass,
with no -Wall warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:24 -05:00
chrg6
78e9d23211 Write a fatal error's frame-time summary after RivaTuner's restore and the rumble stop, not before them.
The review of the fatal summary change found that dispatch.c fatal_end called perf_summary_flush_try (ft_stop waits up
to 50 ms, then the summary and the stutter map are formatted) before rtss_limit_restore and rt_rumble_stop_bounded.
The lane's order for the failed ends, and A-6's for the orderly ones (gfx_present.c rt_exit_steps), puts what a kill
of the process must not lose first: RTSS's profile and the controller's rumble.

perf_summary_flush_try now runs after the restore and the bounded rumble stop, just before the stdio flushes and
plat_terminate. The summary's lines still reach the log: rt_log_sync ran at the start of fatal_end, so every later
line is written as it is made.

Not reproduced before the change (an order of steps). tests/fatal_exit_check.py checks statically that fatal_end
calls perf_summary_flush_try after rt_rumble_stop_bounded; the check fails on the code before this change and passes
after it, and the program's cases (fatal, fatal_rumble, crash_crash, fatal_crash, crash_fatal) pass. The compile
check of dispatch.c passes with no -Wall warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:24 -05:00
chrg6
313ef09081 Open the log's end reserve only when the process is ending, not at a bug capture's log sync in play.
The review of the log tail change found that main.c log_sink opened the 4 MiB reserve past the cap on g_log_final as
well as on g_log_ending. bugcap.c write_log_tail calls rt_log_sync in play, which sets g_log_final, so after a bug
capture in a trace = perf session past the cap, ordinary play lines filled the reserve (about 6 minutes at 12 KB/s),
the end's lines were dropped again, and the "end of the run follows" line came mid-run.

The reserve is now keyed on g_log_ending alone, as E-10's design had it. The ends that relied on rt_log_sync now call
rt_log_ending first: dispatch.c fatal_end and main.c crash_handler right after their guard, vk_device.c
lost_wait_exit's watchdog before its sync, and plat_exit through the new main.c log_sync_ending (plat_exit_hook in a
game run), so every path that ends through plat_exit marks the end too. rt_exit_begin and perf_summary_flush call it
as before.

Not reproduced in the game before the change. tests/log_cap_check.py gains a capture case (rt_log_sync in play, 1000
more play lines, then the end): on the code before this change 58 play lines took the reserve, the end line came at
the sync and no end line reached the file; now no play line passes the cap and the end's lines fill the reserve. A
fatal case runs rt_log_ending and then rt_log_sync, and static checks hold crash_handler, log_sync_ending, fatal_end
and lost_wait_exit to calling rt_log_ending before rt_log_sync. log_cap_check, fatal_exit_check (with a stand-in for
rt_log_ending) and the compile checks of main.c, dispatch.c and vk_device.c pass, with no new -Wall warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:23 -05:00
chrg6
b8b57cfd53 Take the file lock at the end of a save file call only when a CloseFile is pending, not at the end of every call.
The review of the save file use count found that fs_host.c file_release took `lock` exclusive whenever users dropped
to 0, which is the end of every single read_file, write_file, file_size, set_file_size and flush_file, not only when
a CloseFile had marked the slot. The exclusive acquisition waits for every shared holder, such as the fsync loop of a
non-save Commit (commit, FlushFileBuffers under the shared lock), and while it waits it holds up new file_of callers:
a new wait on the guest threads that the change had said it did not add.

file_release now takes the lock only when it finds closing set after its decrement. The decrement and the load of
closing, and close_file's store of closing and load of users, are sequentially consistent pairs, so either the last
call out sees the mark or close_file sees users at 0 and closes the file itself; the check under the lock still keeps
the close to one. closing is an int32_t read and stored with atomics; a slot is emptied by the new file_clear, which
stores users and closing as atomics because the call that has just left the slot may read closing without the lock,
and open_file fills only h, mode and save.

Not reproduced in the game before the change. tests/fs_file_use_check.py gains a case: 1000 reads with no close take
the exclusive lock 0 times, and a read ends while another thread holds the lock shared. It failed on the code before
this change (1000 exclusive takes, the read waited) and passes after it, as do its close-during-a-read case and its
20000 closes and opens beside a reader (six runs, no foreign bytes, no double close). The compile check of fs_host.c
passes with no -Wall warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:23 -05:00
chrg6
a135c19450 Ignore the focus loss that the close's own hide of the window sends, so the exit summary keeps the session's frames.
The review of the window close change found that hiding the active window in gfx_window_win.c exit_start makes
Windows send WM_ACTIVATEAPP(FALSE), which wnd_proc passed on to gfx_window_focus_changed(0). In play, perf.c
perf_focus_lost then asked for the session's reports, and the present thread, which runs until rt_exit_steps reaches
prof_write, wrote "frametime: at focus loss" for the whole session at its next present, with the stutter map and the
memory ranking, and started a new period. "frametime: exit" then covered only the few frames since the hide, or was
missing when none came, so a normal close no longer left the exit summary the play-test's "Normal close" item reads.

wnd_proc now returns at a focus loss once g_exit_started is set. exit_start sets it on the window thread before
ShowWindow, and the message comes on the same thread inside that call, so the check has no race. A gained focus is
still passed on.

Not reproduced before the change (it needs a play run). tests/exit_order_check.py checks statically that
WM_ACTIVATEAPP returns before gfx_window_focus_changed for a loss once g_exit_started is set and that exit_start sets
the flag before the hide; the check fails on the code before this change and passes after it. The compile check of
gfx_window_win.c passes with no new -Wall warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:23 -05:00
chrg6
b23acff301 Leave a line in the log when a module ends the process with ExitProcess, and write the gathered lines at a focus loss.
Two runs of 2026-11-13 (the 0.10.0 fresh install's nx2navi.1.log, 7,974,927 bytes, and the 0.9.2 test's
nx2navi.3.log) ended seconds after "window lost the focus" with exit code 0, no WER event and no line of their own:
an ExitProcess or exit() called by a module in the process and a kill from outside could not be told apart, and
both lose the up to 250 ms of lines the log thread was gathering. A direct ExitProcess skips perf.c's atexit sync.

(a) ExitProcess stops every other thread and then calls each module's detach, the executable's TLS callbacks
included; TerminateProcess calls none. win32.c exit_note_tls, in the .CRT$XLB list beside _tls_used, writes on
DLL_PROCESS_DETACH what the log queue holds (logq.c logq_peek: the bytes as they stand, no lock) and one line,
"exit: the process is ending through ExitProcess, not one of the game's own ends", with the return addresses that
led there. It runs under the loader lock with the other threads gone, so it uses no C runtime (hex written by hand)
and does not ask the loader: the addresses are named from a module list noted at start and at each focus loss
(plat_exit_note_modules: EnumProcessModules, two tables and a writers' lock the callback never takes), and
RtlCaptureStackBackTrace and WriteFile take no lock of the process. plat_exit, plat_terminate and plat_reexec set
the orderly flag (plat_exit_orderly), and main does once main_run has returned, so a tool mode's exit() leaves no
line; a library's exit() during a game run does, naming ucrtbase and its caller. main.c log_open sets it up for a
game run that writes a log file. Linux has empty stand-ins.

(b) gfx_window_focus_changed calls the new main.c rt_log_focus_lost at a focus loss: logq_nudge makes the log thread
write what it gathers at once (a flag log_main's gathering loop checks) and the module list is noted again.

Not reproduced before the change (the two ends are unexplained). The new tests/exit_tls_check.py links win32.c and
logq.c with a queued line still being gathered: ExitProcess from the program and exit() write the line and the note
naming exit_tls_check.exe; TerminateProcess, plat_exit and a return after plat_exit_orderly write no note.
tests/logq_check.c gains a nudge case, which fails with the flag left unset (252 ms) and passes with it.
thread_end_check, fatal_exit_check, crash_stack_check, exit_order_check, save_share_check and the compile checks of
every runtime file pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:23 -05:00
chrg6
dd78132e6d Tell a crash on the RTSS thread as the game's own call of RivaTuner's library, and skip the restore there.
main.c foreign_hook matches "hook", so a crash in RTSSHooks64.dll got the line that the library is not part of the
game and that a recording, overlay or capture program put it there, also when the game's own thread that sets the
frame limit (rtss_limit.c's "rtss" worker) had loaded and called it. The one RTSS crash on record (issue #231)
faulted in ntdll on that thread, where no note applied. And the crash handler then called rtss_limit_restore from
that thread, whose own lock it holds (SRW locks are not recursive), so the restore's trylocks spent a second.

The platform layer keeps the name a thread was made with in the thread itself (plat_thread_name, copied at its
start: a detached thread's struct may go before it ends). main.c crash_module_note picks the note: on the "rtss"
thread, whatever the module, the line says the fault is in the game's calls of RTSS's library or around them and
that rtss = 0 keeps the game from calling it; on other threads RTSSHooks64.dll stays foreign, as it is RTSS's
overlay there. The restore is skipped on the rtss thread. rtss_limit.c is unchanged: the thread is known by the name
it is spawned with. Crash text only, besides the skipped restore.

Not reproduced before the change. tests/foreign_hook_check.py checks the note's choice (RTSSHooks64.dll and ntdll on
the RTSS thread: the RTSS note; RTSSHooks64.dll elsewhere: foreign; a driver: none), and tests/crash_stack_check.py
that the crash report reads the faulting thread's name. They, thread_end_check, fatal_exit_check and the compile
checks of every runtime file pass.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-13 18:43:23 -05:00