Beanstalk Longshot test farm
Menu

Report rp_7b710fe891ba837a pass

#1359 perf(vk): a freed sparse mirror's buffer is kept for the next mirror of its size, not destroyed (RX_VK_SPARSE_SPARE_MB)

linux on lick/agent-b0d3c1,

This report disagrees with its peers.

zero-spares field.stutter_count: +100.0% against peers' +0.0%
zero-spares market.hitch_count: +0.0% against peers' -33.3%
zero-spares market.stutter_count: +40.0% against peers' -12.5%

Machine

Commit
3f72adcc48956b74dacfb826aa44597185ae16f3
Baseline
544229a260c5f5008f241b9cf48ae6b01c1602b9
App
nightly-156-4ec688b0, attempt 1
Game build
e0f3ccfa
os
linux
cpu
AMD Ryzen 9 3900X 12-Core Processor
gpu
NVIDIA GeForce RTX 3080 Ti
tags
linuxnvidianvidia-proprietarydgpu
driver
620.70.6
ram_gb
48
vram_mb
12,288
gpu_vendor
nvidia
os_version
Arch Linux (kernel 7.2.14)

Builds

main (baseline)

790.4 s warm

from 544229a260c5

lift
0.2 s
read
3.5 s
finish
22.3 s
compile
29.8 s
shaders
726.9 s

PR

87.5 s warm

Program cache: warm

Scenarios

default-spares

ok
Metric (median)mainPRΔ
exit_code00±0%
load_time_s20.2820.75+2.3%
load_time_max_s46.9339.24−16.4%
field.hitch_count20−100%
log_forbidden_hits00±0%
market.hitch_count21−50%
field.stutter_count1312−7.7%
log_expected_missing00±0%
market.stutter_count32−33.3%
field.frametime_max_ms55.4147.86−13.6%
field.frametime_p50_ms16.7216.83+0.7%
field.frametime_p99_ms33.7433.83+0.3%
market.frametime_max_ms80.4171.13−11.5%
market.frametime_p50_ms20.0918.85−6.2%
market.frametime_p99_ms36.3633.4−8.1%
field.frametime_1pct_low_ms42.0138.74−7.8%
market.frametime_1pct_low_ms47.138.39−18.5%
screenshot_diff—0—

Pictures against main: field diff 0, market diff 0

Runs (2)
Metricmain 0pr 1
hangnono
crashednono
exit_code00
load_time_s20.2820.75
load_time_max_s46.9339.24
field.hitch_count20
log_forbidden_hits00
market.hitch_count21
field.stutter_count1312
log_expected_missing00
market.stutter_count32
field.frametime_max_ms55.4147.86
field.frametime_p50_ms16.7216.83
field.frametime_p99_ms33.7433.83
market.frametime_max_ms80.4171.13
market.frametime_p50_ms20.0918.85
market.frametime_p99_ms36.3633.4
field.frametime_1pct_low_ms42.0138.74
market.frametime_1pct_low_ms47.138.39
screenshot_diff—0
loadsMarket_Child 46.93 sHyruleField_Child 6.09 sMarket_Child 7.81 sMarket_Child 39.24 sHyruleField_Child 15.13 sMarket_Child 7.89 s

zero-spares RX_VK_SPARSE_SPARE_MB=0 (PR only)

ok
Metric (median)mainPRΔ
exit_code00±0%
load_time_s7.17.07−0.4%
load_time_max_s8.468.44−0.2%
field.hitch_count01—
log_forbidden_hits00±0%
market.hitch_count33±0%
field.stutter_count36+100%
log_expected_missing00±0%
market.stutter_count57+40%
field.frametime_max_ms39.9952.56+31.4%
field.frametime_p50_ms16.7216.74+0.1%
field.frametime_p99_ms29.1428.13−3.5%
market.frametime_max_ms72.8673.22+0.5%
market.frametime_p50_ms18.9118.83−0.4%
market.frametime_p99_ms35.0735.13+0.2%
field.frametime_1pct_low_ms32.7236.09+10.3%
market.frametime_1pct_low_ms45.848.49+5.9%
screenshot_diff—0—

Pictures against main: field diff 0, market diff 0

Runs (2)
Metricmain 0pr 1
hangnono
crashednono
exit_code00
load_time_s7.17.07
load_time_max_s8.468.44
field.hitch_count01
log_forbidden_hits00
market.hitch_count33
field.stutter_count36
log_expected_missing00
market.stutter_count57
field.frametime_max_ms39.9952.56
field.frametime_p50_ms16.7216.74
field.frametime_p99_ms29.1428.13
market.frametime_max_ms72.8673.22
market.frametime_p50_ms18.9118.83
market.frametime_p99_ms35.0735.13
field.frametime_1pct_low_ms32.7236.09
market.frametime_1pct_low_ms45.848.49
screenshot_diff—0
loadsMarket_Child 8.46 sHyruleField_Child 4.95 sMarket_Child 7.88 sMarket_Child 8.44 sHyruleField_Child 4.94 sMarket_Child 7.84 s

Rules

ScenarioCellRuleLimitmainPRΔ %Status
default-sparescrashFalse00—pass
default-sparesscreenshot.max_diff0.05—0—pass
default-sparesload_time_max_s.max_delta_pct1546.9339.24−16.4%pass
default-spareslog_forbidden_hits.max000—pass
default-spareslog_expected_missing.max000—pass
zero-sparesRX_VK_SPARSE_SPARE_MB=0 (PR only)crashFalse00—pass
zero-sparesRX_VK_SPARSE_SPARE_MB=0 (PR only)screenshot.max_diff0.05—0—pass
zero-sparesRX_VK_SPARSE_SPARE_MB=0 (PR only)load_time_max_s.max_delta_pct158.468.44−0.2%pass
zero-sparesRX_VK_SPARSE_SPARE_MB=0 (PR only)log_forbidden_hits.max000—pass
zero-sparesRX_VK_SPARSE_SPARE_MB=0 (PR only)log_expected_missing.max000—pass

Log (PR side, end)

[hos host] vk:   read live, by pool: 0x1042800000 (4 MB, captured, state 2): 1;
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: replay reads of guest memory as submitted: 2080000 (1205008 pages from the captures)
[hos sead::MainThread] vk: readbacks of GPU results: 1278 of 3000 waited for their write-back, 0.06 ms each; 0 looked at the stopped queues first, 1 waited for their submission's last part (guest memory read in place after theirs), 0 waited for their whole submission (their pages written in place, no mirror)
[hos rmbl::VibrationThread] hid: 16384 vibration values, strongest 0.00 since the last count
[hos host] vk: a frame: 381 draws kept their neighbour's descriptors (0 of them through other slots' registrations)
[hos host] vk: a frame: 0.0 draws left out (not ready), 0.00 waits for shaders or pipelines (0.0 ms); translation 0 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 12.74, present 0.35, draw 7.20, descriptors 2.52, texture lookups 0.25, uploads 0.13, pipelines 0.33, programs 1.05, commands 2.80, dispatch 0.66, memory sync 0.82, queue submit 0.39, slot wait 0.03
[hos host] vk: a frame: mirrors 294 (1111 MB): 8.13 MB synced in 1253.9 copies, 18.31 MB written back
[hos host] vk: 4920 presents, 53.6 fps; a frame: 2405 draws (0 skipped), 126 dispatches, 158 passes, 146 barriers, 6 submits, 0.5 uploads (0.4 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 32554 pipelines (5614877 ms), 0 programs (0 failed), 1967 textures (3145 MB)
[hos host] vk: host time in GPU resource waits: 0.65 ms a frame (fences 0.02, timelines 0.64, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 515.0 full barriers recorded, 217.5 left out (nothing recorded since the one before), 0.0 pass starts ordered in their fragment stages only (RX_VK_PASS_BARRIER=game)
[hos host] vk: a frame: the game's barriers by their NVN flags: 0x1: 2.0; 0x8: 3.0; 0x9: 1.0; 0x11: 38.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 8.3 ms a frame (the longest submission 1.8 ms); 0.5 textures made a frame
[hos host] vk: a frame: 0.44 MB unswizzled on the CPU in 0.03 ms
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 427 dirty pages from their submits' captures, 39025 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2774 runs, 348.4 MB compared; 36 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9222 in captures, 5 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 2100000 (1218151 pages from the captures)
[hos host] vk: program picks: 3499477 kept (1471526 by the last picks' textures, a uniform buffer bound since), 2535 new, 1097987 with other textures (of 4600000)
[hos host] vk: replay reads of guest memory as submitted: 2120000 (1231381 pages from the captures)
[hos host] vk:   (its written part: 1 of 1 KB readable, 0 bytes not zero; levels L0)
[hos host] vk: texture 0x111516ac00 1775x1175x1 x8 levels 1 NVN 0x6 (32928 KB of storage) re-uploaded (its memory was written; flags 0x4): what the CPU wrote, 1 KB of it
[hos host] vk: sparse mirrors: 80, 1111 MB with memory of 1520 MB (408 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 884 binds (126.3 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 433 dirty pages from their submits' captures, 40169 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2768 runs, 347.1 MB compared; 37 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9667 in captures, 8 only in the collection, 17 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk:   (its written part: 1 of 1 KB readable, 0 bytes not zero; levels L0)
[hos host] vk: texture 0x111516ac00 1775x1175x1 x8 levels 1 NVN 0x6 (32928 KB of storage) re-uploaded (its memory was written; flags 0x4): what the CPU wrote, 1 KB of it
[hos Presentation Thread] vk: capture: a submit 0.25 ms (lock 0.02, scan 0.05 over 686 ranges, 3678 MB), 580 pages written, 580 copied (2.3 MB)
[hos host] vk: a capture taken (ms): all 0.102; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.097, bytes copied 0.000; 25 kept (170 MB), 11 spare buffers, 14780 texture pages copied (57 MB)
[hos host] vk: replay reads of guest memory as submitted: 2140000 (1244502 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 420, 2060.2 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 499 dirty pages from their submits' captures, 40280 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2895 runs, 364.3 MB compared; 35 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9168 in captures, 5 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos sead::MainThread] nvn: GPU waits in 5 s: 15 sync waits (26 ms), 0 polls timed out (0 ms), 0 queue finishes (0 ms)
[hos host] vk: readback pages refreshed from guest memory: 1508928 pages so far (0 held: may be written ahead in their submission)
[hos host] vk: replay reads of guest memory as submitted: 2160000 (1257656 pages from the captures)
[hos host] vk: sparse mirrors: 81, 1114 MB with memory of 1528 MB (413 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 885 binds (126.4 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 372 dirty pages from their submits' captures, 39164 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2755 runs, 348.1 MB compared; 37 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9172 in captures, 7 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos ModuleSystemWorker3] nvn: CommandBufferSignalEvent #50000: event 0x100d6c8004 mode 0 value 49999 (lr 0x80291504)
[hos host] vk: replay reads of guest memory as submitted: 2180000 (1270801 pages from the captures)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 108099 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.09, writes collected 0.07
[hos host] vk: replay reads of guest memory as submitted: 2200000 (1283878 pages from the captures)
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 641 KB; 0 made on demand (0 MB kept now); c1 stages: at most 255 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: write-backs left out (600 submits): 420, 2060.2 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 383 dirty pages from their submits' captures, 40885 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2754 runs, 347.1 MB compared; 36 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9677 in captures, 5 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos sead::MainThread] nvn: EventSignal (CPU) #10000: event 0x1006e38004 mode 0 value 1 (lr 0x8348c78c)
[hos host] vk: texture storage read as submitted (6001 times; 179370 pages from the copies, 0 of them rewritten in guest memory since their submit; 12221 pages kept, 411 spans, 1246 MB)
[hos host] vk: texture 0x1132655200 256x256x1 x1 levels 9 NVN 0x49 -> vk 139 (2000 textures, 3168 MB)
[hos host] vk: texture 0x10aecc4c00 768x768x1 x1 levels 10 NVN 0x8c -> vk 137 (ASTC decoded) (2000 textures, 3156 MB)
[hos host] vk: replay reads of guest memory as submitted: 2220000 (1297033 pages from the captures)
[hos host] vk: program picks: 3649328 kept (1523873 by the last picks' textures, a uniform buffer bound since), 2535 new, 1148136 with other textures (of 4800000)
[hos host] vk: sparse mirrors: 81, 1114 MB with memory of 1528 MB (413 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 885 binds (126.4 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 624 dirty pages from their submits' captures, 39276 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2892 runs, 364.3 MB compared; 36 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9200 in captures, 7 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos sead::MainThread] nvn: 750000 recordings made, 65469 live copies (22.7 MB), 569 texture copies; 5 command buffers' segments (4.2 MB)
[hos host] vk:   (its written part: 1 of 1 KB readable, 968 bytes not zero; levels L0)
[hos host] vk: texture 0x112a8e5c00 1144x3056x1 x3 levels 1 NVN 0x50 (10368 KB of storage) re-uploaded (its memory was written; flags 0): what the CPU wrote, 1 KB of it
[hos host] vk:   (its written part: 1 of 1 KB readable, 968 bytes not zero; levels L0)
[hos host] vk: texture 0x112a8e5c00 1144x3056x1 x3 levels 1 NVN 0x50 (10368 KB of storage) re-uploaded (its memory was written; flags 0): what the CPU wrote, 1 KB of it
[hos host] vk: replay reads of guest memory as submitted: 2240000 (1310159 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 498 dirty pages from their submits' captures, 39121 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2753 runs, 348.1 MB compared; 37 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9187 in captures, 8 only in the collection, 17 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: a frame: 374 draws kept their neighbour's descriptors (0 of them through other slots' registrations)
[hos host] vk: a frame: 0.0 draws left out (not ready), 0.00 waits for shaders or pipelines (0.0 ms); translation 0 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 13.19, present 0.39, draw 7.46, descriptors 2.57, texture lookups 0.19, uploads 0.08, pipelines 0.33, programs 1.12, commands 2.91, dispatch 0.69, memory sync 0.87, queue submit 0.38, slot wait 0.03
[hos host] vk: a frame: mirrors 295 (1114 MB): 8.16 MB synced in 1252.3 copies, 18.27 MB written back
[hos host] vk: 5040 presents, 54.3 fps; a frame: 2399 draws (0 skipped), 125 dispatches, 159 passes, 146 barriers, 6 submits, 1.4 uploads (0.1 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 32554 pipelines (5614877 ms), 0 programs (0 failed), 2041 textures (3166 MB)
[hos host] vk: host time in GPU resource waits: 0.81 ms a frame (fences 0.02, timelines 0.79, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 515.3 full barriers recorded, 217.6 left out (nothing recorded since the one before), 0.0 pass starts ordered in their fragment stages only (RX_VK_PASS_BARRIER=game)
[hos host] vk: a frame: the game's barriers by their NVN flags: 0x1: 2.0; 0x8: 3.0; 0x9: 1.0; 0x11: 38.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 8.3 ms a frame (the longest submission 1.8 ms); 1.2 textures made a frame
[hos host] vk: a frame: 0.11 MB unswizzled on the CPU in 0.01 ms
[hos host] vk: replay reads of guest memory as submitted: 2260000 (1323315 pages from the captures)
[hos sead::MainThread] vk: capture: a submit 0.22 ms (lock 0.02, scan 0.05 over 684 ranges, 3674 MB), 506 pages written, 506 copied (2.0 MB)
[hos host] vk: a capture taken (ms): all 0.062; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.056, bytes copied 0.000; 25 kept (172 MB), 12 spare buffers, 9406 texture pages copied (36 MB)
[hos ModuleSystemWorker1] nvn: DrawMeshTasksIndirect: 120000 so far (x1 0x1179846000 x2 0 x3 0x1179846000 x4 0x654040)
[hos host] vk: sparse mirrors: 81, 1114 MB with memory of 1528 MB (413 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 885 binds (126.4 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 421 dirty pages from their submits' captures, 40824 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2752 runs, 346.9 MB compared; 36 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9653 in captures, 5 only in the collection, 17 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: replay reads of guest memory as submitted: 2280000 (1336436 pages from the captures)
[hos host] vk: readback pages refreshed from guest memory: 1581433 pages so far (0 held: may be written ahead in their submission)
[hos ModuleSystemWorker5] vk: readbacks of GPU results: 1288 of 3000 waited for their write-back, 0.08 ms each; 0 looked at the stopped queues first, 0 waited for their submission's last part (guest memory read in place after theirs), 0 waited for their whole submission (their pages written in place, no mirror)
[hos host] vk: replay reads of guest memory as submitted: 2300000 (1349507 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 420, 2060.2 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 418 dirty pages from their submits' captures, 40475 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2869 runs, 365.2 MB compared; 37 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9050 in captures, 4 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 49642 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos sead::MainThread] nvn: CommandBufferWaitEvent #75000: event 0x100d6c8030 mode 0 value 51948 (now 51929) (lr 0x806614d8)
[hos host] vk: replay reads of guest memory as submitted: 2320000 (1362666 pages from the captures)
[hos host] vk:   (its written part: 1 of 1 KB readable, 927 bytes not zero; levels L0)
[hos host] vk: texture 0x11309bac00 2240x3032x1 x1 levels 1 NVN 0x4b (6650 KB of storage) re-uploaded (its memory was written; flags 0): what the CPU wrote, 1 KB of it
[hos host] vk:   (its written part: 1 of 1 KB readable, 927 bytes not zero; levels L0)
[hos host] vk: texture 0x11309bac00 2240x3032x1 x1 levels 1 NVN 0x4b (6650 KB of storage) re-uploaded (its memory was written; flags 0): what the CPU wrote, 1 KB of it
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.05, writes collected 0.07
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 641 KB; 0 made on demand (0 MB kept now); c1 stages: at most 255 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: sparse mirrors: 79, 1100 MB with memory of 1508 MB (407 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 885 binds (126.4 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 454 dirty pages from their submits' captures, 39749 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2719 runs, 346.6 MB compared; 38 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9524 in captures, 5 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: program picks: 3799517 kept (1576200 by the last picks' textures, a uniform buffer bound since), 2535 new, 1197947 with other textures (of 5000000)
[hos host] vk: replay reads of guest memory as submitted: 2340000 (1375798 pages from the captures)
[hos Presentation Thread] vi: 5108 frames presented (54.2 fps over the last 5.0 s, interval 1)
[hos Presentation Thread] vi: present intervals: <15.5 ms 0, 15.5-17.8 115, 17.8-22 147, 22-30 9, 30-40 1, over 40 0
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 305 dirty pages from their submits' captures, 39155 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2730 runs, 345.8 MB compared; 37 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9078 in captures, 6 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 2360000 (1388898 pages from the captures)
[hos NavMeshCAStepDepWorker1] vcores: 167406 takes, 43977 waits (mean 115 us, longest 43.1 ms, 0 over HOS_VCORE_LOG_MS), 21763 kicks, 21797 yields, 0 starved grants, 0 escapes in 5 s
[hos host] vk: replay reads of guest memory as submitted: 2380000 (1401955 pages from the captures)
[hos host] vk: sparse mirrors: 79, 1100 MB with memory of 1508 MB (407 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 885 binds (126.4 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 258 dirty pages from their submits' captures, 39060 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2763 runs, 348.7 MB compared; 30 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9085 in captures, 5 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos Presentation Thread] vk: capture: a submit 0.20 ms (lock 0.02, scan 0.05 over 684 ranges, 3670 MB), 468 pages written, 468 copied (1.8 MB)
[hos host] vk: a capture taken (ms): all 0.045; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.041, bytes copied 0.000; 25 kept (138 MB), 12 spare buffers, 7812 texture pages copied (30 MB)
[hos host] vk: recorder: 8904292 packets run in 10.0 s (busy 2.05 s, spun idle 4.43 s, slept 4460 times, 1982 of them under 200 us), 0 calls made directly (no room); the replay waited for it 38586 times, 184.00 ms in all, the longest 1.29 ms; vkEndCommandBuffer 16109/110.43 ms (max 1.29), vkResetDescriptorPool 16108/37.18 ms (max 0.23), replay_mu unlock 6366/36.36 ms (max 0.38), vkDestroyImageView 2/0.04 ms (max 0.02), vkQueueBindSparse 1/0.01 ms (max 0.01)
[hos host] vk: replay reads of guest memory as submitted: 2400000 (1415151 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 420, 2060.2 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 314 dirty pages from their submits' captures, 40289 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2854 runs, 363.0 MB compared; 31 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9095 in captures, 5 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: readback pages refreshed from guest memory: 1653936 pages so far (0 held: may be written ahead in their submission)
[hos host] vk: a frame: 384 draws kept their neighbour's descriptors (0 of them through other slots' registrations)
[hos host] vk: a frame: 0.0 draws left out (not ready), 0.00 waits for shaders or pipelines (0.0 ms); translation 0 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 12.84, present 0.38, draw 7.38, descriptors 2.53, texture lookups 0.14, uploads 0.01, pipelines 0.33, programs 1.12, commands 2.86, dispatch 0.67, memory sync 0.72, queue submit 0.38, slot wait 0.03
[hos host] vk: a frame: mirrors 295 (1097 MB): 8.14 MB synced in 1248.0 copies, 18.25 MB written back
[hos host] vk: 5160 presents, 55.0 fps; a frame: 2407 draws (0 skipped), 124 dispatches, 159 passes, 146 barriers, 6 submits, 0.4 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 32554 pipelines (5614877 ms), 0 programs (0 failed), 2056 textures (3168 MB)
[hos host] vk: host time in GPU resource waits: 0.58 ms a frame (fences 0.02, timelines 0.56, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 513.6 full barriers recorded, 217.6 left out (nothing recorded since the one before), 0.0 pass starts ordered in their fragment stages only (RX_VK_PASS_BARRIER=game)
[hos host] vk: a frame: the game's barriers by their NVN flags: 0x1: 2.0; 0x8: 3.0; 0x9: 1.0; 0x11: 38.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 8.2 ms a frame (the longest submission 1.9 ms); 0.3 textures made a frame
[hos host] vk: a frame: 0.01 MB unswizzled on the CPU in 0.00 ms
[hos host] vk: replay reads of guest memory as submitted: 2420000 (1428339 pages from the captures)
[hos host] vk: sparse mirrors: 79, 1106 MB with memory of 1516 MB (409 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 886 binds (126.5 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 435 dirty pages from their submits' captures, 39311 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2749 runs, 347.2 MB compared; 33 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9602 in captures, 2 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 64347 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: program picks: 3949735 kept (1628539 by the last picks' textures, a uniform buffer bound since), 2539 new, 1247725 with other textures (of 5200000)
[hos host] vk: replay reads of guest memory as submitted: 2440000 (1441444 pages from the captures)
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 641 KB; 0 made on demand (0 MB kept now); c1 stages: at most 255 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 430 dirty pages from their submits' captures, 40245 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2734 runs, 345.5 MB compared; 33 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9149 in captures, 6 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 2460000 (1454588 pages from the captures)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.06, writes collected 0.07
[hos host] vk: replay reads of guest memory as submitted: 2480000 (1467752 pages from the captures)
[hos host] vk: sparse mirrors: 79, 1106 MB with memory of 1516 MB (409 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 886 binds (126.5 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 420, 2060.2 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 448 dirty pages from their submits' captures, 40420 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2852 runs, 362.3 MB compared; 28 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9123 in captures, 4 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: render targets drawn over from another start, let go: 1187 so far (2714 MB), 1052 of them (2388 MB) between the passes
[hos host] vk: replay reads of guest memory as submitted: 2500000 (1480893 pages from the captures)
[hos sead::MainThread] vk: readbacks of GPU results: 1316 of 3000 waited for their write-back, 0.06 ms each; 0 looked at the stopped queues first, 1 waited for their submission's last part (guest memory read in place after theirs), 0 waited for their whole submission (their pages written in place, no mirror)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 481 dirty pages from their submits' captures, 39042 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2718 runs, 345.2 MB compared; 31 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9083 in captures, 6 only in the collection, 17 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 2520000 (1494023 pages from the captures)
[hos Presentation Thread] vk: capture: a submit 0.21 ms (lock 0.02, scan 0.05 over 685 ranges, 3672 MB), 506 pages written, 506 copied (2.0 MB)
[hos host] vk: a capture taken (ms): all 0.056; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.052, bytes copied 0.000; 25 kept (154 MB), 11 spare buffers, 5266 texture pages copied (20 MB)
[hos ModuleSystemWorker3] nvn: CommandBufferSignalEvent #55000: event 0x100d6c8014 mode 0 value 55000 (lr 0x80660bcc)
[hos host] vk: sparse mirrors: 79, 1106 MB with memory of 1516 MB (409 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 886 binds (126.5 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 385 dirty pages from their submits' captures, 39001 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2714 runs, 345.1 MB compared; 35 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9083 in captures, 5 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: replay reads of guest memory as submitted: 2540000 (1507112 pages from the captures)
[hos host] vk: readback pages refreshed from guest memory: 1726639 pages so far (0 held: may be written ahead in their submission)
[hos host] vk: program picks: 4099998 kept (1680931 by the last picks' textures, a uniform buffer bound since), 2540 new, 1297461 with other textures (of 5400000)
[hos host] vk: replay reads of guest memory as submitted: 2560000 (1520282 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 443 dirty pages from their submits' captures, 39932 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2706 runs, 345.9 MB compared; 37 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9553 in captures, 6 only in the collection, 20 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 60632 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: a frame: 375 draws kept their neighbour's descriptors (0 of them through other slots' registrations)
[hos host] vk: a frame: 0.0 draws left out (not ready), 0.00 waits for shaders or pipelines (0.0 ms); translation 0 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 13.10, present 0.41, draw 7.41, descriptors 2.50, texture lookups 0.08, uploads 0.01, pipelines 0.34, programs 1.12, commands 2.92, dispatch 0.68, memory sync 0.75, queue submit 0.40, slot wait 0.03
[hos host] vk: a frame: mirrors 297 (1109 MB): 8.23 MB synced in 1250.2 copies, 18.17 MB written back
[hos host] vk: 5280 presents, 54.4 fps; a frame: 2422 draws (0 skipped), 123 dispatches, 160 passes, 146 barriers, 6 submits, 0.2 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 32554 pipelines (5614877 ms), 0 programs (0 failed), 2069 textures (3171 MB)
[hos host] vk: host time in GPU resource waits: 0.63 ms a frame (fences 0.02, timelines 0.61, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 514.2 full barriers recorded, 217.9 left out (nothing recorded since the one before), 0.0 pass starts ordered in their fragment stages only (RX_VK_PASS_BARRIER=game)
[hos host] vk: a frame: the game's barriers by their NVN flags: 0x1: 2.0; 0x8: 3.0; 0x9: 1.0; 0x11: 38.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 8.2 ms a frame (the longest submission 1.7 ms); 0.1 textures made a frame
[hos host] vk: a frame: 0.02 MB unswizzled on the CPU in 0.00 ms
[hos ModuleSystemWorker3] nvn: CommandBufferWaitEvent #80000: event 0x100d6c8030 mode 0 value 55748 (now 55729) (lr 0x806614d8)
[hos host] vk: replay reads of guest memory as submitted: 2580000 (1533471 pages from the captures)
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 641 KB; 0 made on demand (0 MB kept now); c1 stages: at most 255 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: sparse mirrors: 80, 1109 MB with memory of 1520 MB (410 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 887 binds (126.6 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 420, 2060.2 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 300 dirty pages from their submits' captures, 39984 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2803 runs, 361.6 MB compared; 32 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 8990 in captures, 6 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.06, writes collected 0.07
[hos host] vk: replay reads of guest memory as submitted: 2600000 (1546555 pages from the captures)
[hos sead::MainThread] nvn: GPU waits in 5 s: 23 sync waits (25 ms), 0 polls timed out (0 ms), 0 queue finishes (0 ms)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 421 dirty pages from their submits' captures, 39452 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2739 runs, 345.2 MB compared; 34 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9040 in captures, 5 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 2620000 (1559721 pages from the captures)
[hos ModuleSystemWorker1] nvn: 800000 recordings made, 66026 live copies (22.7 MB), 569 texture copies; 6 command buffers' segments (4.4 MB)
[hos host] vk: sparse mirrors: 80, 1109 MB with memory of 1520 MB (410 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 887 binds (126.6 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 559 dirty pages from their submits' captures, 38914 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2700 runs, 344.9 MB compared; 32 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9040 in captures, 3 only in the collection, 17 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: replay reads of guest memory as submitted: 2640000 (1572918 pages from the captures)
[hos host] vk: texture slots registered again at another size: 200000 (of 301392 registrations; slot 16307 480x270 -> 1920x1080)
[hos host] vk:   (its written part: 2 of 2 KB readable, 1189 bytes not zero; levels L0)
[hos host] vk: texture 0x10baa31e00 3056x1840x1 x1 levels 1 NVN 0x4b (5539 KB of storage) re-uploaded (its memory was written; flags 0): what the CPU wrote, 2 KB of it
[hos Presentation Thread] vk: capture: a submit 0.23 ms (lock 0.02, scan 0.05 over 684 ranges, 3676 MB), 525 pages written, 525 copied (2.1 MB)
[hos host] vk: a capture taken (ms): all 0.070; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.065, bytes copied 0.000; 25 kept (138 MB), 11 spare buffers, 11969 texture pages copied (46 MB)
[hos host] vk:   (its written part: 2 of 2 KB readable, 1189 bytes not zero; levels L0)
[hos host] vk: texture 0x10baa31e00 3056x1840x1 x1 levels 1 NVN 0x4b (5539 KB of storage) re-uploaded (its memory was written; flags 0): what the CPU wrote, 2 KB of it
[hos host] vk: program picks: 4250198 kept (1733089 by the last picks' textures, a uniform buffer bound since), 2541 new, 1347260 with other textures (of 5600000)
[hos host] vk: replay reads of guest memory as submitted: 2660000 (1586034 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 597 dirty pages from their submits' captures, 38721 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2706 runs, 345.9 MB compared; 32 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9037 in captures, 4 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: readback pages refreshed from guest memory: 1799143 pages so far (0 held: may be written ahead in their submission)
[hos host] vk: replay reads of guest memory as submitted: 2680000 (1599168 pages from the captures)
[hos host] vk: sparse mirrors: 82, 1122 MB with memory of 1540 MB (417 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 889 binds (126.9 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 507 dirty pages from their submits' captures, 38992 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2706 runs, 344.9 MB compared; 31 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9056 in captures, 7 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 80972 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: 2 textures retired (2084 left, 3231 MB): budget 2, drawn over 0, dead targets 0
[hos host] vk: replay reads of guest memory as submitted: 2700000 (1612254 pages from the captures)
[hos ModuleSystemWorker3] vk: readbacks of GPU results: 1275 of 3000 waited for their write-back, 0.07 ms each; 0 looked at the stopped queues first, 7 waited for their submission's last part (guest memory read in place after theirs), 0 waited for their whole submission (their pages written in place, no mirror)
[hos Presentation Thread] vi: 5379 frames presented (54.1 fps over the last 5.0 s, interval 1)
[hos Presentation Thread] vi: present intervals: <15.5 ms 0, 15.5-17.8 88, 17.8-22 181, 22-30 2, 30-40 0, over 40 0
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 641 KB; 0 made on demand (0 MB kept now); c1 stages: at most 255 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 366 dirty pages from their submits' captures, 39056 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2712 runs, 345.5 MB compared; 24 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9026 in captures, 5 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 2720000 (1625478 pages from the captures)
[hos NavMeshCAStepDepWorker1] vcores: 166052 takes, 43088 waits (mean 118 us, longest 36.9 ms, 0 over HOS_VCORE_LOG_MS), 21355 kicks, 21333 yields, 0 starved grants, 0 escapes in 5 s
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.08, writes collected 0.07
[hos host] vk: memory: working set 20485 MB (peak 20485), private 11424 MB, staging 208 MB in 13 arenas, 32 made so far (13 ms)
[hos host] vk: memory heaps: heap 0 (device): 186 blocks 5682 MB, 20255 allocations 5046 MB; used 6708 of 8848 MB; heap 1 (system): 16 blocks 896 MB, 40 allocations 818 MB; used 3124 of 36053 MB;
[hos host] vk: image blocks: 1 blocks 32 MB, 5 images 16 MB; 1 free ranges, the largest 15 MB
[hos host] vk: image blocks (render targets): 16 blocks 512 MB, 131 images 210 MB; 44 free ranges, the largest 32 MB
[hos host] vk: image blocks (textures, compacted): 78 blocks 2496 MB, 1948 images 2422 MB; 1029 free ranges, the largest 16 MB
[hos host] vk: render target images: 506 reused from the pool, 823 made new, 688 destroyed from it; it holds 0 (0 MB)
[hos host] vk: VRAM: 6708 MB used of 8848 MB allowed (635 MB of it free in the blocks); textures 3231 MB (budget 6775 MB), 2086 textures
[hos host] vk: textures: decoded ASTC 1693 MB, rendered 844 MB (69 drawn in one frame, kept: 40518 KB), other 506 MB; used in the last 8 presents 2187 MB
[hos host] vk: decoded ASTC used again: 830 textures (1654 MB) copied from ones the game finalized, not decoded; 830 of those retired, 52 listed; 1659 MB of blocks hashed in 217 ms
[hos host] vk:   288 MB: 0x1147c16000 12288x12288x1 x1 levels 1 NVN 0x32 rendered, last used 1 presents ago
[hos host] vk:   64 MB: 0x113755c000 4096x4096x1 x1 levels 1 NVN 0x34 rendered, last used 1 presents ago
[hos host] vk:   64 MB: 0x113cba9000 512x512x64 x1 levels 1 NVN 0x3d rendered, last used 1 presents ago
[hos host] vk:   54 MB: 0x10002d97a00 3072x3072x1 x3 levels 1 NVN 0x32 rendered, last used 858 presents ago
[hos host] vk:   54 MB: 0x100030a4200 3072x3072x1 x3 levels 1 NVN 0x32 rendered, last used 1 presents ago
[hos host] vk:   37 MB: 0x111b417400 1800x1380x1 x8 levels 1 NVN 0x6, last used 1225 presents ago
[hos host] vk:   32 MB: 0x1159c17000 4096x4096x1 x1 levels 1 NVN 0x32 rendered, last used 1 presents ago
[hos host] vk:   31 MB: 0x111516ac00 1775x1175x1 x8 levels 1 NVN 0x6, last used 1 presents ago
[hos host] vk: finalized textures the GPU drew once, kept for being made again: 55 (9369 KB): 140x42x1 NVN 0x25: 1 (22 KB); 204x106x1 NVN 0x1: 1 (21 KB); 260x260x1 NVN 0xd: 53 (9325 KB);
[hos Presentation Thread] vk: presents after the CPU's signal (600): 177 waited for it, 0.48 ms each, 0 let go after RX_VK_FRAME_GATE_MS
[hos host] vk: malloc: 12692 MB in arenas (10580 MB in use, 2112 MB free, 10 MB of it at the tops), 37 MB mapped
[hos host] vk: status: 600 frames, the slowest 38.3 ms; VRAM 6.6 of 8.6 GB
[hos host] vk: a frame: 375 draws kept their neighbour's descriptors (0 of them through other slots' registrations)
[hos host] vk: a frame: 0.0 draws left out (not ready), 0.00 waits for shaders or pipelines (0.0 ms); translation 0 ms in all, 1 jobs queued
[hos host] vk: a frame's replay time (ms): submit 13.40, present 0.41, draw 7.66, descriptors 2.64, texture lookups 0.19, uploads 0.09, pipelines 0.34, programs 1.13, commands 3.01, dispatch 0.67, memory sync 0.87, queue submit 0.38, slot wait 0.03
[hos host] vk: a frame: mirrors 299 (1122 MB): 8.23 MB synced in 1250.1 copies, 18.19 MB written back
[hos host] vk: 5400 presents, 52.6 fps; a frame: 2432 draws (0 skipped), 122 dispatches, 160 passes, 146 barriers, 6 submits, 0.2 uploads (0.4 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 32554 pipelines (5614877 ms), 0 programs (0 failed), 2086 textures (3231 MB)
[hos host] vk: host time in GPU resource waits: 0.61 ms a frame (fences 0.02, timelines 0.60, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 513.4 full barriers recorded, 217.5 left out (nothing recorded since the one before), 0.0 pass starts ordered in their fragment stages only (RX_VK_PASS_BARRIER=game)
[hos host] vk: a frame: the game's barriers by their NVN flags: 0x1: 2.0; 0x8: 3.0; 0x9: 1.0; 0x11: 38.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 8.2 ms a frame (the longest submission 1.7 ms); 0.2 textures made a frame
[hos host] vk: a frame: 0.43 MB unswizzled on the CPU in 0.04 ms
[hos host] vk: the render thread's whole frames (600 presents): 0 jobs went on before their present after 1000 ms, 0 times a game's thread waited for the work handed over, 0 times for a fence of the frame in the making
[hos host] vk: the render thread (600 presents, 18.53 ms a frame): its jobs 13.81 ms a frame (submits 9.71, presents 0.40, others 3.70), idle 4.73; the game waited for room at 0 presents, 0.00 ms each (0.00 a frame; 168.88 jobs queued at a present); 0 catch-ups (a finish, a texture read back, a write-back not recorded yet: 0), 0.00 ms each; 0 presents held for the CPU's signal, 0.00 ms each (0 let go by RX_VK_HOLD_MS); 182 captures at a CPU signal; the submits' waits read in 0.053 ms a frame; readbacks: 3000 next to a texture's storage (no catch-up), 0 in one, 0 read exactly (after the work handed over: among the catch-ups); a present made 13.86 ms after its hand-over (24.88 at most)
[hos host] vk: sparse mirrors: 82, 1122 MB with memory of 1540 MB (417 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 889 binds (126.9 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 401 dirty pages from their submits' captures, 39202 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2736 runs, 347.8 MB compared; 39 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 8838 in captures, 96 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: malloc_trim: 2102 MB free in the arenas now, the process's private memory 11424 -> 11249 MB (52 ms)
[hos host] vk: replay reads of guest memory as submitted: 2740000 (1638640 pages from the captures)
[hos host] vk: program picks: 4400286 kept (1786113 by the last picks' textures, a uniform buffer bound since), 2541 new, 1397172 with other textures (of 5800000)
[hos host] vk: replay reads of guest memory as submitted: 2760000 (1651762 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 454 dirty pages from their submits' captures, 39173 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2750 runs, 348.6 MB compared; 27 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9062 in captures, 5 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos Presentation Thread] vk: capture: a submit 0.23 ms (lock 0.03, scan 0.05 over 685 ranges, 3704 MB), 508 pages written, 508 copied (2.0 MB)
[hos host] vk: a capture taken (ms): all 0.066; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.061, bytes copied 0.000; 25 kept (154 MB), 9 spare buffers, 14644 texture pages copied (57 MB)
[hos host] vk: replay reads of guest memory as submitted: 2780000 (1664940 pages from the captures)
[hos host] vk: texture #1821 0x1001e87e800 1920x1080x1 x1 NVN 0x3f the GPU drew goes with no copy of its texels (live, unused for 3647 frames, over the budget): made again over this storage, it starts empty
[hos host] vk: 1 textures retired (2089 left, 3224 MB): budget 1, drawn over 0, dead targets 0; GPU-drawn with no copy kept: 1 live, 0 finalized, 7 MB
[hos host] vk: sparse mirrors: 82, 1122 MB with memory of 1540 MB (417 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 889 binds (126.9 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 398 dirty pages from their submits' captures, 39042 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2754 runs, 350.9 MB compared; 29 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9001 in captures, 7 only in the collection, 17 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: readback pages refreshed from guest memory: 1871648 pages so far (0 held: may be written ahead in their submission)
[hos host] vk: replay reads of guest memory as submitted: 2800000 (1678081 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 382 dirty pages from their submits' captures, 39004 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2739 runs, 348.9 MB compared; 27 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9024 in captures, 6 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 61746 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: replay reads of guest memory as submitted: 2820000 (1691216 pages from the captures)
[hos host] vk: queues ordered by their events: 30000 submissions stopped at a wait, 29983 went on once it was reached, 14 waits gone past; made to go by presents 0, releases 0, waits for the replay 0, 50 ms 1 (stopped 50.06 ms at most); waits met 84725, not met 0; captures at a CPU event's wait 1606, 0 of them at a wait met on arrival
[hos host] vk: replay reads of guest memory as submitted: 2840000 (1704304 pages from the captures)
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 641 KB; 0 made on demand (0 MB kept now); c1 stages: at most 255 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: sparse mirrors: 81, 1119 MB with memory of 1532 MB (412 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 889 binds (126.9 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 471 dirty pages from their submits' captures, 38866 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2759 runs, 351.2 MB compared; 29 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9033 in captures, 6 only in the collection, 18 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos ModuleSystemWorker3] nvn: CommandBufferWaitEvent #85000: event 0x100d6c8030 mode 0 value 59548 (now 59529) (lr 0x806614d8)
[hos host] vk: replay reads of guest memory as submitted: 2860000 (1717456 pages from the captures)
[hos host] vk: program picks: 4550939 kept (1839580 by the last picks' textures, a uniform buffer bound since), 2541 new, 1446519 with other textures (of 6000000)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.06, writes collected 0.07
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 431 dirty pages from their submits' captures, 38932 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2741 runs, 348.9 MB compared; 27 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9040 in captures, 5 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos ModuleSystemWorker1] nvn: CommandBufferSignalEvent #60000: event 0x100d6c8020 mode 0 value 60000 (lr 0x80664f14)
[hos host] vk: replay reads of guest memory as submitted: 2880000 (1730659 pages from the captures)
[hos host] vk: sparse mirrors: 81, 1119 MB with memory of 1532 MB (412 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 889 binds (126.9 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 417 dirty pages from their submits' captures, 39231 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2750 runs, 349.2 MB compared; 28 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9041 in captures, 6 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: a frame: 375 draws kept their neighbour's descriptors (0 of them through other slots' registrations)
[hos host] vk: a frame: 0.0 draws left out (not ready), 0.00 waits for shaders or pipelines (0.0 ms); translation 0 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 13.08, present 0.38, draw 7.44, descriptors 2.51, texture lookups 0.15, uploads 0.02, pipelines 0.34, programs 1.10, commands 2.96, dispatch 0.68, memory sync 0.75, queue submit 0.39, slot wait 0.03
[hos host] vk: a frame: mirrors 299 (1119 MB): 8.11 MB synced in 1252.8 copies, 18.40 MB written back
[hos host] vk: 5520 presents, 53.1 fps; a frame: 2430 draws (0 skipped), 122 dispatches, 159 passes, 146 barriers, 6 submits, 0.4 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 32554 pipelines (5614877 ms), 0 programs (0 failed), 2106 textures (3228 MB)
[hos host] vk: host time in GPU resource waits: 0.53 ms a frame (fences 0.02, timelines 0.51, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 512.4 full barriers recorded, 216.2 left out (nothing recorded since the one before), 0.0 pass starts ordered in their fragment stages only (RX_VK_PASS_BARRIER=game)
[hos host] vk: a frame: the game's barriers by their NVN flags: 0x1: 2.0; 0x8: 3.0; 0x9: 1.0; 0x11: 38.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 8.2 ms a frame (the longest submission 1.9 ms); 0.4 textures made a frame
[hos host] vk: a frame: 0.03 MB unswizzled on the CPU in 0.00 ms
[hos host] vk: replay reads of guest memory as submitted: 2900000 (1743833 pages from the captures)
[hos sead::MainThread] vk: capture: a submit 0.22 ms (lock 0.02, scan 0.05 over 686 ranges, 3712 MB), 505 pages written, 505 copied (2.0 MB)
[hos host] vk: a capture taken (ms): all 0.060; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.056, bytes copied 0.000; 25 kept (154 MB), 7 spare buffers, 8513 texture pages copied (33 MB)
[hos ModuleSystemWorker3] vk: readbacks of GPU results: 1274 of 3000 waited for their write-back, 0.06 ms each; 0 looked at the stopped queues first, 0 waited for their submission's last part (guest memory read in place after theirs), 0 waited for their whole submission (their pages written in place, no mirror)
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 420 dirty pages from their submits' captures, 39103 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2744 runs, 349.9 MB compared; 29 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9010 in captures, 6 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 2920000 (1756931 pages from the captures)
[hos host] vk: readback pages refreshed from guest memory: 1944154 pages so far (0 held: may be written ahead in their submission)
[hos host] vk: replay reads of guest memory as submitted: 2940000 (1770168 pages from the captures)
[hos host] control: quit asked by a client
[hos host] vk: page table census (close): no pool outside its windows, 64-72 GB and 508-512 GB (4252 pools in the table now)
[hos host] vk: sparse mirrors: 79, 1106 MB with memory of 1512 MB (405 MB spared); blocks given memory later 12430, given back 19728; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 426; 889 binds (126.9 ms); 0 made whole for want of memory
[hos host] vk: write-backs left out (600 submits): 399, 1957.1 MB; 31 bindings followed, 43 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 1686 so far
[hos host] vk: mirror syncs (600 submits): 335 dirty pages from their submits' captures, 39264 from earlier ones', 0 from guest memory as it was at the replay; 25 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: write-back diff (600 submits): 2732 runs, 348.2 MB compared; 28 pages read into the shadows; 308 MB of shadows; pages the shadows knew that the CPU wrote: 9005 in captures, 9 only in the collection, 19 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 156 live (2208 MB), 0 retired still imported (0 MB), 77 retired given back (1716 MB)
[hos host] vk: pipeline cache: play file 11 saved (2639 KB): 8 pipelines, 8 from the driver's cache, 0 compiled; 0 more from the start's cache
[hos host] control: exiting
[hos host] stop: close work done (the game's saves are not waited for)