Beanstalk Longshot test farm
Menu

Report rp_8fdadaaac118fe30 fail

#1702 perf(vk): the presenter thread on by default with NVIDIA on Linux and AMD on Windows, with NVIDIA Reflex; after #1121

linux+amd on Alberania/agent-af6cf3,

This report disagrees with its peers.

default-presenter kokiri-after-motion.draws_mean: -10.6% against peers' +23.0%
default-presenter kokiri-after-motion.fps_mean: +30.7% against peers' +3.2%
default-presenter kokiri-after-motion.frametime_1pct_low_ms: -86.9% against peers' -11.0%
default-presenter kokiri-after-motion.frametime_max_ms: -92.9% against peers' -4.7%
default-presenter kokiri-after-motion.hitch_count: +33.3% against peers' -5.9%
default-presenter kokiri-standing.fps_mean: +41.0% against peers' +5.5%
default-presenter kokiri-standing.frametime_1pct_low_ms: -91.7% against peers' -27.4%
default-presenter kokiri-standing.frametime_max_ms: -95.4% against peers' -30.0%
default-presenter kokiri-standing.stutter_count: -28.6% against peers' -100.0%
default-presenter market-after-motion.draws_mean: -32.3% against peers' +2.7%
default-presenter market-after-motion.fps_mean: +44.5% against peers' +3.0%
default-presenter market-after-motion.frametime_1pct_low_ms: -93.8% against peers' -29.3%
default-presenter market-after-motion.frametime_max_ms: -94.5% against peers' -35.6%
default-presenter market-after-motion.stutter_count: -40.0% against peers' -100.0%
default-presenter market-standing.fps_mean: +688.9% against peers' +0.0%
default-presenter market-standing.frametime_1pct_low_ms: -93.8% against peers' +1.8%
default-presenter market-standing.frametime_max_ms: -93.6% against peers' +6.5%
default-presenter market-standing.frametime_p99_ms: -93.9% against peers' -8.8%
default-presenter market-standing.hitch_count: +148.0% against peers' +4.2%
default-presenter run_frametime_max_ms: -72.9% against peers' +234.5%
default-presenter run_hitch_count: -41.9% against peers' +42.7%
forced-presenter kokiri-after-motion.draws_mean: -10.4% against peers' -39.6%
forced-presenter kokiri-after-motion.fps_mean: +29.7% against peers' -3.2%
forced-presenter kokiri-after-motion.frametime_1pct_low_ms: -88.6% against peers' +12.5%
forced-presenter kokiri-after-motion.frametime_max_ms: -93.8% against peers' +10.5%
forced-presenter kokiri-after-motion.hitch_count: +33.3% against peers' +4.2%
forced-presenter kokiri-standing.draws_mean: +6.4% against peers' +39.9%
forced-presenter kokiri-standing.fps_mean: +30.2% against peers' -10.1%
forced-presenter kokiri-standing.frametime_1pct_low_ms: -84.7% against peers' +70.0%
forced-presenter kokiri-standing.frametime_max_ms: -91.1% against peers' +79.4%
forced-presenter kokiri-standing.frametime_p99_ms: -14.5% against peers' +50.2%
forced-presenter kokiri-standing.hitch_count: -50.0% against peers' +611.8%
forced-presenter run_frametime_max_ms: -77.3% against peers' -10.0%

Machine

Commit
3569b8da6a4f90bc036bbb40cb52993b865b0a39
Baseline
09cbdd979eacfbd0c83e27fad7e50380985af02d
App
nightly-177-99542c21, attempt 1
Game build
e0f3ccfa
os
linux
cpu
AMD Ryzen 7 5800X3D 8-Core Processor
gpu
AMD Radeon RX 7900 XTX (RADV NAVI31)
tags
linuxamdradvdgpu
driver
26.2.7
ram_gb
64
vram_mb
24,560
gpu_vendor
amd
os_version
Bazzite (kernel 7.3.4)

Builds

main (baseline)

117.4 s warm

from 09cbdd979eac

lift
0.8 s
read
4.7 s
finish
16.2 s
compile
26.3 s
shaders
56.5 s

PR

80.5 s warm

Program cache: warm

Scenarios

default-presenter

error

the game refused a step: refused: the game refused the teleport: an event is playing (a cutscene, a map change's)

Metric (median)mainPRΔ
exit_code00±0%
run_hitch_count64.537.5−41.9%
log_forbidden_hits00±0%
log_expected_missing00±0%
run_frametime_max_ms1,042.56282.06−72.9%
kokiri-standing.fps_mean71100.08+41%
market-standing.fps_mean9.0771.55+688.9%
kokiri-standing.draws_mean668.15708.25+6%
market-standing.draws_mean2,322.652,467.2+6.2%
kokiri-standing.hitch_count20.5−75%
market-standing.hitch_count12.531+148%
kokiri-after-motion.fps_mean76.1699.55+30.7%
market-after-motion.fps_mean63.6291.91+44.5%
kokiri-standing.stutter_count3.52.5−28.6%
market-standing.stutter_count14.534+134.5%
kokiri-after-motion.draws_mean811.45725.65−10.6%
market-after-motion.draws_mean1,329.05899.8−32.3%
kokiri-after-motion.hitch_count1.52+33.3%
market-after-motion.hitch_count2.52−20%
kokiri-standing.frametime_max_ms1,005.5746.24−95.4%
kokiri-standing.frametime_p50_ms12.5810.5−16.6%
kokiri-standing.frametime_p99_ms15.5812.86−17.5%
market-standing.frametime_max_ms1,011.3864.34−93.6%
market-standing.frametime_p50_ms14.6912.35−16%
market-standing.frametime_p99_ms1,010.3361.66−93.9%
kokiri-after-motion.stutter_count3.55+42.9%
market-after-motion.stutter_count53−40%
kokiri-after-motion.frametime_max_ms999.1370.78−92.9%
kokiri-after-motion.frametime_p50_ms11.379.73−14.4%
kokiri-after-motion.frametime_p99_ms14.3413.95−2.7%
market-after-motion.frametime_max_ms1,006.1755.69−94.5%
market-after-motion.frametime_p50_ms11.5310.67−7.5%
market-after-motion.frametime_p99_ms18.1814.95−17.7%
kokiri-standing.frametime_1pct_low_ms198.0216.52−91.7%
market-standing.frametime_1pct_low_ms1,010.9562.4−93.8%
kokiri-after-motion.frametime_1pct_low_ms201.7526.51−86.9%
market-after-motion.frametime_1pct_low_ms408.1625.13−93.8%
screenshot_diff—0.87—

Pictures against main: kokiri-standing 5 s diff 0.7017 (main against itself 0.0294), kokiri-standing 9 s diff 0.7119 (main against itself 0.0355), kokiri-after-motion 5 s diff 0.8731 (main against itself 0.3369), kokiri-after-motion 9 s diff 0.969 (main against itself 0.3743), market-standing 5 s diff 0.0868 (main against itself 0.0215), market-standing 9 s diff 0.0823 (main against itself 0.0249), market-after-motion 5 s diff 0.7533 (main against itself 0.7316), market-after-motion 9 s diff 0.7516 (main against itself 0.7232), kokiri-standing diff 0.6936 (main against itself 0.0227), kokiri-after-motion diff 0.984 (main against itself 0.3815), market-standing diff 0.0934 (main against itself 0.0276), market-after-motion diff 0.72 (main against itself 0.6827)

Runs (4)
Metricmain 0pr 1pr 2main 3
hangnononono
crashednononono
exit_code00—0
run_hitch_count6366966
log_forbidden_hits0000
log_expected_missing0000
run_frametime_max_ms1,046.78383.7180.411,038.35
kokiri-standing.fps_mean70.5384.81115.3571.46
market-standing.fps_mean7.1871.55—10.96
kokiri-standing.draws_mean661.6675.1741.4674.7
market-standing.draws_mean2,305.42,467.2—2,339.9
kokiri-standing.hitch_count2012
market-standing.hitch_count1331—12
kokiri-after-motion.fps_mean80.3891.81107.2871.94
market-after-motion.fps_mean68.2791.91—58.96
kokiri-standing.stutter_count3234
market-standing.stutter_count1534—14
kokiri-after-motion.draws_mean810.4823.2628.1812.5
market-after-motion.draws_mean1,300.2899.8—1,357.9
kokiri-after-motion.hitch_count1222
market-after-motion.hitch_count22—3
kokiri-standing.frametime_max_ms1,004.5530.1962.281,006.59
kokiri-standing.frametime_p50_ms12.5512.378.6212.61
kokiri-standing.frametime_p99_ms15.9114.4211.315.26
market-standing.frametime_max_ms1,010.8264.34—1,011.94
market-standing.frametime_p50_ms14.9212.35—14.45
market-standing.frametime_p99_ms1,010.4961.66—1,010.17
kokiri-after-motion.stutter_count3464
market-after-motion.stutter_count53—5
kokiri-after-motion.frametime_max_ms998.0258.3183.241,000.24
kokiri-after-motion.frametime_p50_ms11.3910.518.9511.34
kokiri-after-motion.frametime_p99_ms14.3513.8414.0514.32
market-after-motion.frametime_max_ms1,003.7755.69—1,008.56
market-after-motion.frametime_p50_ms11.2610.67—11.8
market-after-motion.frametime_p99_ms16.9114.95—19.44
kokiri-standing.frametime_1pct_low_ms198.1317.1615.88197.92
market-standing.frametime_1pct_low_ms1,010.6962.4—1,011.22
kokiri-after-motion.frametime_1pct_low_ms140.0524.6128.41263.45
market-after-motion.frametime_1pct_low_ms302.9525.13—513.37
screenshot_diff—0.750.98—

forced-presenter RX_VK_PRESENT_THREAD=1 (PR only)

error

the game refused a step: refused: the game refused the teleport: an event is playing (a cutscene, a map change's)

Metric (median)mainPRΔ
exit_code00±0%
run_hitch_count6335.5−43.7%
log_forbidden_hits00±0%
log_expected_missing00±0%
run_frametime_max_ms1,044.77237.32−77.3%
kokiri-standing.fps_mean76.599.58+30.2%
market-standing.fps_mean10.4672.21+590%
kokiri-standing.draws_mean675.95719.1+6.4%
market-standing.draws_mean2,221.752,486.8+11.9%
kokiri-standing.hitch_count10.5−50%
market-standing.hitch_count1228+133.3%
kokiri-after-motion.fps_mean76.899.64+29.7%
market-after-motion.fps_mean74.1989.89+21.2%
kokiri-standing.stutter_count1.52.5+66.7%
market-standing.stutter_count1330+130.8%
kokiri-after-motion.draws_mean804.6720.65−10.4%
market-after-motion.draws_mean1,169.35904.3−22.7%
kokiri-after-motion.hitch_count1.52+33.3%
market-after-motion.hitch_count1.50−100%
kokiri-standing.frametime_max_ms515.5345.76−91.1%
kokiri-standing.frametime_p50_ms12.7210.47−17.7%
kokiri-standing.frametime_p99_ms15.313.09−14.5%
market-standing.frametime_max_ms1,010.7564.3−93.6%
market-standing.frametime_p50_ms14.7412.33−16.3%
market-standing.frametime_p99_ms1,010.0661.34−93.9%
kokiri-after-motion.stutter_count3.55+42.9%
market-after-motion.stutter_count42−50%
kokiri-after-motion.frametime_max_ms1,002.7462.37−93.8%
kokiri-after-motion.frametime_p50_ms11.629.73−16.3%
kokiri-after-motion.frametime_p99_ms14.7214−4.9%
market-after-motion.frametime_max_ms1,002.5129.41−97.1%
market-after-motion.frametime_p50_ms11.3110.94−3.2%
market-after-motion.frametime_p99_ms18.7214.77−21.1%
kokiri-standing.frametime_1pct_low_ms107.6716.43−84.7%
market-standing.frametime_1pct_low_ms1,010.4961.84−93.9%
kokiri-after-motion.frametime_1pct_low_ms220.9725.2−88.6%
market-after-motion.frametime_1pct_low_ms223.4617.89−92%
screenshot_diff—0.92—

Pictures against main: kokiri-standing 5 s diff 0.7049 (main against itself 0.0199), kokiri-standing 9 s diff 0.7163 (main against itself 0.0607), kokiri-after-motion 5 s diff 0.8335 (main against itself 0.3065), kokiri-after-motion 9 s diff 0.9712 (main against itself 0.3013), market-standing 5 s diff 0.1632 (main against itself 0.0844), market-standing 9 s diff 0.2228 (main against itself 0.0847), market-after-motion 5 s diff 0.8372 (main against itself 0.9239), market-after-motion 9 s diff 0.8418 (main against itself 0.9215), kokiri-standing diff 0.6916 (main against itself 0.0991), kokiri-after-motion diff 0.9884 (main against itself 0.3112), market-standing diff 0.1935 (main against itself 0.0756), market-after-motion diff 0.8296 (main against itself 0.9243)

Runs (4)
Metricmain 0pr 1pr 2main 3
hangnononono
crashednononono
exit_code0—00
run_hitch_count62145764
log_forbidden_hits0000
log_expected_missing0000
run_frametime_max_ms1,044.69103.81370.821,044.86
kokiri-standing.fps_mean81.41115.1384.0271.59
market-standing.fps_mean9.11—72.2111.82
kokiri-standing.draws_mean675.3759.1679.1676.6
market-standing.draws_mean2,291.3—2,486.82,152.2
kokiri-standing.hitch_count0102
market-standing.hitch_count12—2812
kokiri-after-motion.fps_mean77.35107.8291.4676.24
market-after-motion.fps_mean77.85—89.8970.53
kokiri-standing.stutter_count1412
market-standing.stutter_count12—3014
kokiri-after-motion.draws_mean809.3621.9819.4799.9
market-after-motion.draws_mean991.1—904.31,347.6
kokiri-after-motion.hitch_count2221
market-after-motion.hitch_count1—02
kokiri-standing.frametime_max_ms26.0563.7327.781,005.01
kokiri-standing.frametime_p50_ms12.738.5212.4212.71
kokiri-standing.frametime_p99_ms15.411.4914.6815.2
market-standing.frametime_max_ms1,010.78—64.31,010.72
market-standing.frametime_p50_ms15.16—12.3314.32
market-standing.frametime_p99_ms1,010.2—61.341,009.93
kokiri-after-motion.stutter_count4643
market-after-motion.stutter_count3—25
kokiri-after-motion.frametime_max_ms1,002.765.2859.471,002.79
kokiri-after-motion.frametime_p50_ms11.428.9110.5511.83
kokiri-after-motion.frametime_p99_ms14.414.1913.8115.04
market-after-motion.frametime_max_ms1,001.78—29.411,003.23
market-after-motion.frametime_p50_ms11.32—10.9411.3
market-after-motion.frametime_p99_ms17.68—14.7719.76
kokiri-standing.frametime_1pct_low_ms17.8916.4516.42197.44
market-standing.frametime_1pct_low_ms1,010.55—61.841,010.42
kokiri-after-motion.frametime_1pct_low_ms299.826.2624.14142.14
market-after-motion.frametime_1pct_low_ms143.78—17.89303.15
screenshot_diff—0.990.84—

Rules

ScenarioCellRuleLimitmainPRΔ %Status
default-presentercrashFalse00—pass
default-presenterkokiri-standing.fps_mean.min_delta_pct1571100.08+41%pass
default-presenterkokiri-after-motion.fps_mean.min_delta_pct1576.1699.55+30.7%pass
default-presentermarket-standing.fps_mean.min_delta_pct159.0771.55+688.9%pass
default-presentermarket-after-motion.fps_mean.min_delta_pct1563.6291.91+44.5%pass
default-presenterscreenshot.max_diff0.40.730.98—fail
default-presenterkokiri-standing.frametime_p99_ms.max_delta_pct2015.5812.86−17.5%pass
default-presenterkokiri-after-motion.frametime_p99_ms.max_delta_pct2014.3413.95−2.7%pass
default-presentermarket-standing.frametime_p99_ms.max_delta_pct201,010.3361.66−93.9%pass
default-presentermarket-after-motion.frametime_p99_ms.max_delta_pct2018.1814.95−17.7%pass
default-presenterlog_forbidden_hits.max000—pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)crashFalse00—pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)kokiri-standing.fps_mean.min_delta_pct1576.599.58+30.2%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)kokiri-after-motion.fps_mean.min_delta_pct1576.899.64+29.7%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)market-standing.fps_mean.min_delta_pct1510.4672.21+590%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)market-after-motion.fps_mean.min_delta_pct1574.1989.89+21.2%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)screenshot.max_diff0.40.920.99—fail
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)kokiri-standing.frametime_p99_ms.max_delta_pct2015.313.09−14.5%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)kokiri-after-motion.frametime_p99_ms.max_delta_pct2014.7214−4.9%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)market-standing.frametime_p99_ms.max_delta_pct201,010.0661.34−93.9%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)market-after-motion.frametime_p99_ms.max_delta_pct2018.7214.77−21.1%pass
forced-presenterRX_VK_PRESENT_THREAD=1 (PR only)log_forbidden_hits.max000—pass

Pictures of the failure

This report had 2 pictures of its failure (the last frame of a stuck step, or main's and the PR's picture where they differed too much). The farm keeps those for a day: they were deleted.

Log (PR side, end)

[hos host] vk: a frame: the game's barriers by their NVN flags: 0x1: 2.0; 0x8: 3.0; 0x9: 1.0; 0x11: 37.1; 0x12: 77.7; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.7 ms a frame (the longest submission 4.1 ms); 1.9 textures made a frame
[hos host] vk: a frame: 0.03 MB unswizzled on the CPU in 0.00 ms
[hos host] vk: the render thread (600 presents, 13.32 ms a frame): its jobs 8.23 ms a frame (submits 3.33, presents 0.20, others 4.70), idle 5.09; the game waited for room at 1 presents, 13.90 ms each (0.02 a frame; 2.65 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); 387 captures at a CPU signal; the submits' waits read in 0.047 ms a frame; readbacks: 2985 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 7.50 ms after its hand-over (27.20 at most)
[hos host] vk: frame overlap (600 presents): the GPU between recordings, ms a frame, idle + in the next one's uploads: up to a frame's first recording with a draw 2.04 + 0.20 (the window's copy is in the idle), after it 3.27 + 0.63; after a frame's last recording that reads or writes the game's memory come 1.00 recordings (0.01 ms of GPU time), after its last that reads it 1.00 (0.01 ms)
[hos host] vk: frame ahead (600 presents): 10.50 parts a frame recorded in the signal's job, 80.2% of them took nothing that came with the signal; of the others, 2.08 a frame took pages of the signal's capture into a mirror (73.2 pages a frame) and 0.01 pages read as guest memory had them (2.2 pages a frame); a frame's first part took nothing in 38 of 382 frames, its first three in 38
[hos host] vk: frame ahead, what came with the signal: 387 signals with a capture, each of 131.6 pages in 11.1 pools; between a frame's last submit and that signal 0.00 texture and sampler registrations, 0.00 releases and 0.00 other CPU signals a signal (0 of them with any)
[hos host] vk: frame overlap, by place: idle + uploads before a frame's 2nd to 12th recording with a draw, then all later ones: 0.57+0.06, 0.38+0.08, 0.25+0.04, 0.34+0.04, 0.43+0.07, 0.23+0.08, 0.22+0.04, 0.17+0.13, 0.09+0.03, 0.08+0.02, 0.04+0.01, 0.48+0.04
[hos host] vk: replay reads of guest memory as submitted: 4180000 (2454133 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 605 dirty pages from their submits' captures, 56281 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: mirror syncs' pass (600 submits): 1102 dispatches, 44795 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 6323 runs, 922.3 MB compared; 107 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 19908 in captures, 1 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: late pages found at the replay (600 takes): 0 pages in 0 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos host] vk: replay reads of guest memory as submitted: 4200000 (2469048 pages from the captures)
[hos ModuleSystemWorker1] vk: readbacks of GPU results: 1166 of 3000 waited for their write-back, 0.25 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 ModuleSystemWorker1] vk: invalidates since the last readbacks line: 8781 kept their virtual core (nothing there to wait for), 2995 gave it up (counted by 256 a thread; RX_VK_INVALIDATE_QUICK=0: none kept)
[hos SceneLowPriority] vcores: 138413 takes, 27547 waits (mean 84 us, longest 11.7 ms, 0 over HOS_VCORE_LOG_MS), 12428 kicks, 12384 yields, 0 starved grants, 0 escapes in 5 s
[hos host] vk: replay reads of guest memory as submitted: 4220000 (2483985 pages from the captures)
[hos host] vk: 1 textures retired (3361 left, 6277 MB): budget 0, drawn over 1, dead targets 0
[hos host] vk: readback pages refreshed from guest memory: 1811884 pages so far (0 held: may be written ahead in their submission)
[hos sead::MainThread] vk: capture: a submit 0.10 ms (lock 0.00, scan 0.04 over 1077 ranges, 2952 MB), 191 pages written, 191 copied (0.7 MB)
[hos host] vk: a capture taken (ms): all 0.006; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.003, bytes copied 0.000; 25 kept (124 MB), 10 spare buffers, 2318 texture pages copied (9 MB)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 1930 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: recorder: 9082920 packets run in 10.0 s (busy 1.73 s, spun idle 4.15 s, slept 4057 times, 1563 of them under 200 us), 0 calls made directly (no room); the replay waited for it 35445 times, 100.05 ms in all, the longest 0.57 ms; vkEndCommandBuffer 12805/41.57 ms (max 0.39), replay_mu unlock 9683/39.17 ms (max 0.57), vkResetDescriptorPool 12762/18.17 ms (max 0.02), vmaInvalidateAllocation 132/0.99 ms (max 0.13), vkQueueSubmit2 42/0.07 ms (max 0.01), vkQueueBindSparse 21/0.07 ms (max 0.01)
[hos host] vk: replay reads of guest memory as submitted: 4240000 (2498884 pages from the captures)
[hos host] vk: sparse mirrors: 104, 993 MB with memory of 1520 MB (526 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 366 binds (104.0 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 94 kept as spares, 0 destroyed; 8 spares now, 312 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 775 dirty pages from their submits' captures, 56166 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: mirror syncs' pass (600 submits): 1094 dispatches, 44778 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 6317 runs, 922.3 MB compared; 97 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20098 in captures, 0 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos host] vk: replay reads of guest memory as submitted: 4260000 (2513696 pages from the captures)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos host] vk: replay reads of guest memory as submitted: 4280000 (2528568 pages from the captures)
[hos host] vk: a frame: 142 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 54 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 4.58, present 0.15, draw 2.52, descriptors 1.02, texture lookups 0.04, uploads 0.00, pipelines 0.12, programs 0.42, commands 0.77, dispatch 0.49, memory sync 0.25, queue submit 0.77, slot wait 0.04
[hos host] vk: a frame: mirrors 198 (993 MB): 4.26 MB synced in 859.3 copies, 17.73 MB written back
[hos host] vk: 19920 presents, 90.1 fps; a frame: 908 draws (0 skipped), 96 dispatches, 145 passes, 146 barriers, 6 submits, 0.0 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 43659 pipelines (1179 ms), 0 programs (10 failed), 3361 textures (6277 MB)
[hos host] vk: host time in GPU resource waits: 2.34 ms a frame (fences 0.00, timelines 2.34, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 429.4 full barriers recorded, 227.0 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: 37.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.5 ms a frame (the longest submission 4.0 ms); 0.0 textures made a frame
[hos host] vk: a frame: 0.00 MB unswizzled on the CPU in 0.00 ms
[hos host] vk: write-backs left out (600 submits): 1113, 5459.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 566 dirty pages from their submits' captures, 56779 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: mirror syncs' pass (600 submits): 1103 dispatches, 45276 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 6344 runs, 937.9 MB compared; 99 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20450 in captures, 0 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: late pages found at the replay (600 takes): 0 pages in 0 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos host] vk: replay reads of guest memory as submitted: 4300000 (2543510 pages from the captures)
[hos Presentation Thread] vk: capture: a submit 0.09 ms (lock 0.00, scan 0.04 over 1077 ranges, 2946 MB), 191 pages written, 191 copied (0.7 MB)
[hos host] vk: a capture taken (ms): all 0.006; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.003, bytes copied 0.000; 25 kept (126 MB), 11 spare buffers, 2318 texture pages copied (9 MB)
[hos host] vk: replay reads of guest memory as submitted: 4320000 (2558565 pages from the captures)
[hos host] vk: mirror of pool 0x7f9a000000 (4 MB) given back: idle, 6039 submissions since its sync
[hos sead::MainThread] nvn: GPU waits in 5 s: 142 sync waits (143 ms), 0 polls timed out (0 ms), 0 queue finishes (0 ms)
[hos ModuleSystemWorker1] nvn: CommandBufferWaitEvent #190000: event 0x100d6c8008 mode 0 value 114392 (now 114373) (lr 0x8027944c)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 3362 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: replay reads of guest memory as submitted: 4340000 (2573568 pages from the captures)
[hos host] vk: mirror of pool 0x1055400000 (4 MB) given back: idle, 6024 submissions since its sync
[hos host] vk: mirror of pool 0x7f73800000 (4 MB) given back: idle, 6024 submissions since its sync
[hos host] vk: 97 mirrors given back (857 MB in all); 989 MB of mirrors now, 56 MB of them idle and kept for what the GPU wrote there alone
[hos ModuleSystemWorker1] vk: readbacks of GPU results: 1211 of 3000 waited for their write-back, 0.28 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 ModuleSystemWorker1] vk: invalidates since the last readbacks line: 8557 kept their virtual core (nothing there to wait for), 2963 gave it up (counted by 256 a thread; RX_VK_INVALIDATE_QUICK=0: none kept)
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 293 KB; 0 made on demand (0 MB kept now); c1 stages: at most 105 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: sparse mirrors: 101, 989 MB with memory of 1508 MB (518 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 368 binds (104.1 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 97 kept as spares, 0 destroyed; 11 spares now, 324 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 721 dirty pages from their submits' captures, 56075 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: mirror syncs' pass (600 submits): 1089 dispatches, 44704 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5957 runs, 901.9 MB compared; 96 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20080 in captures, 0 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos host] vk: program picks: 5295248 kept (1884746 by the last picks' textures, a uniform buffer bound since), 72666 new, 2432085 with other textures (of 7800000)
[hos ModuleSystemWorker1] nvn: CommandBufferSignalEvent #115000: event 0x100d6c8010 mode 0 value 115002 (lr 0x80660bcc)
[hos host] vk: replay reads of guest memory as submitted: 4360000 (2588631 pages from the captures)
[hos sead::MainThread] nvn: QueueSubmitCommands #120000: queue 0x10100d6bf020 (its #19999), 1 handles (lr 0x80666be0)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos sead::MainThread] nvn: EventSignal (CPU) #40000: event 0x1006e38004 mode 0 value 1 (lr 0x8348c78c)
[hos ModuleSystemWorker1] nvn: 1150000 recordings made, 67005 live copies (19.7 MB), 982 texture copies; 4 command buffers' segments (4.2 MB)
[hos host] vk: replay reads of guest memory as submitted: 4380000 (2603731 pages from the captures)
[hos ModuleSystemWorker2] nvn: DrawMeshTasksIndirect: 220000 so far (x1 0x117a4f2030 x2 0 x3 0x117a4f2000 x4 0xf33080)
[hos host] vk: write-backs left out (600 submits): 1113, 5459.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 464 dirty pages from their submits' captures, 57172 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: mirror syncs' pass (600 submits): 1116 dispatches, 45459 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 6080 runs, 947.9 MB compared; 100 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20400 in captures, 0 only in the collection, 53 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: late pages found at the replay (600 takes): 0 pages in 0 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos host] vk: replay reads of guest memory as submitted: 4400000 (2618576 pages from the captures)
[hos host] vk: a frame: 131 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 54 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 4.53, present 0.14, draw 2.60, descriptors 1.06, texture lookups 0.05, uploads 0.00, pipelines 0.12, programs 0.44, commands 0.79, dispatch 0.44, memory sync 0.26, queue submit 0.80, slot wait 0.04
[hos host] vk: a frame: mirrors 198 (989 MB): 4.26 MB synced in 858.5 copies, 17.61 MB written back
[hos host] vk: 20040 presents, 89.5 fps; a frame: 897 draws (0 skipped), 90 dispatches, 145 passes, 146 barriers, 6 submits, 0.0 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 43659 pipelines (1179 ms), 0 programs (10 failed), 3362 textures (6278 MB)
[hos host] vk: host time in GPU resource waits: 2.35 ms a frame (fences 0.00, timelines 2.35, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 423.2 full barriers recorded, 227.1 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: 37.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.5 ms a frame (the longest submission 4.0 ms); 0.0 textures made a frame
[hos host] vk: a frame: 0.00 MB unswizzled on the CPU in 0.00 ms
[hos Presentation Thread] vk: capture: a submit 0.10 ms (lock 0.00, scan 0.04 over 1078 ranges, 2948 MB), 195 pages written, 195 copied (0.8 MB)
[hos host] vk: a capture taken (ms): all 0.007; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.004, bytes copied 0.000; 25 kept (126 MB), 10 spare buffers, 2318 texture pages copied (9 MB)
[hos host] vk: mirror of pool 0x7fa2000000 (32 MB) given back: idle, 6036 submissions since its sync
[hos host] vk: replay reads of guest memory as submitted: 4420000 (2633779 pages from the captures)
[hos host] vk: mirror of pool 0x7fa4000000 (12 MB) given back: idle, 6061 submissions since its sync
[hos host] vk: mirror of pool 0x7f7f000000 (16 MB) given back: idle, 6047 submissions since its sync
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 4256 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: replay reads of guest memory as submitted: 4440000 (2648838 pages from the captures)
[hos host] vk: sparse mirrors: 98, 944 MB with memory of 1448 MB (503 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 371 binds (105.3 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 100 kept as spares, 0 destroyed; 14 spares now, 384 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 486 dirty pages from their submits' captures, 56213 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: mirror syncs' pass (600 submits): 1092 dispatches, 44658 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5796 runs, 910.9 MB compared; 103 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20063 in captures, 0 only in the collection, 53 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos host] vk: mirror of pool 0x7fa1000000 (8 MB) given back: idle, 6041 submissions since its sync
[hos Presentation Thread] vi: 20097 frames presented (89.4 fps over the last 5.0 s, interval 1)
[hos Presentation Thread] vi: present intervals: <15.5 ms 445, 15.5-17.8 3, 17.8-22 0, 22-30 0, 30-40 0, over 40 0
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos host] vk: replay reads of guest memory as submitted: 4460000 (2664046 pages from the captures)
[hos host] vk: mirror of pool 0x7fb5c00000 (4 MB) given back: idle, 6051 submissions since its sync
[hos host] vk: mirror of pool 0x7fa7400000 (12 MB) given back: idle, 6027 submissions since its sync
[hos host] vk: mirror of pool 0x7f72400000 (4 MB) given back: idle, 6026 submissions since its sync
[hos host] vk: mirror of pool 0x7f73000000 (4 MB) given back: idle, 6025 submissions since its sync
[hos host] vk: mirror of pool 0x1041800000 (4 MB) given back: idle, 6015 submissions since its sync
[hos host] vk: mirror of pool 0x7f9f800000 (8 MB) given back: idle, 6013 submissions since its sync
[hos host] vk: mirror of pool 0x7f9e800000 (4 MB) given back: idle, 6010 submissions since its sync
[hos host] vk: mirror of pool 0x7f86000000 (8 MB) given back: idle, 6010 submissions since its sync
[hos host] vk: mirror of pool 0x7f9c000000 (8 MB) given back: idle, 6062 submissions since its sync
[hos host] vk: mirror of pool 0x7f9b400000 (12 MB) given back: idle, 6049 submissions since its sync
[hos host] vk: mirror of pool 0x7fb3400000 (8 MB) given back: idle, 6031 submissions since its sync
[hos host] vk: 112 mirrors given back (940 MB in all); 906 MB of mirrors now, 56 MB of them idle and kept for what the GPU wrote there alone
[hos host] vk: replay reads of guest memory as submitted: 4480000 (2679116 pages from the captures)
[hos host] vk: mirror of pool 0x7fa1800000 (4 MB) given back: idle, 6053 submissions since its sync
[hos host] vk: mirror of pool 0x1043400000 (4 MB) given back: idle, 6051 submissions since its sync
[hos ModuleSystemWorker1] vk: readbacks of GPU results: 1202 of 3000 waited for their write-back, 0.27 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 ModuleSystemWorker1] vk: invalidates since the last readbacks line: 8700 kept their virtual core (nothing there to wait for), 3076 gave it up (counted by 256 a thread; RX_VK_INVALIDATE_QUICK=0: none kept)
[hos host] vk: readback pages refreshed from guest memory: 1888546 pages so far (0 held: may be written ahead in their submission)
[hos host] vk: mirror of pool 0x7f9e000000 (8 MB) given back: idle, 6015 submissions since its sync
[hos sead::MainThread] vk: capture: a submit 0.09 ms (lock 0.00, scan 0.04 over 1079 ranges, 2950 MB), 193 pages written, 193 copied (0.8 MB)
[hos host] vk: a capture taken (ms): all 0.007; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.004, bytes copied 0.000; 25 kept (154 MB), 11 spare buffers, 2652 texture pages copied (10 MB)
[hos host] vk: mirror of pool 0x7f69c00000 (4 MB) given back: idle, 6041 submissions since its sync
[hos host] vk: late pages found at the replay (600 takes): 0 pages in 0 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos host] vk: replay reads of guest memory as submitted: 4500000 (2694327 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 1113, 5459.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 586 dirty pages from their submits' captures, 57335 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: mirror syncs' pass (600 submits): 1115 dispatches, 45552 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5865 runs, 926.8 MB compared; 99 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20484 in captures, 0 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: mirror of pool 0x7fe1400000 (8 MB) given back: idle, 6055 submissions since its sync
[hos host] vk: mirror of pool 0x105a800000 (4 MB) given back: idle, 6043 submissions since its sync
[hos host] vk: mirror of pool 0x7fe0c00000 (4 MB) given back: idle, 6043 submissions since its sync
[hos host] vk: mirror of pool 0x7f9f000000 (4 MB) given back: idle, 6043 submissions since its sync
[hos host] vk: mirror of pool 0x7f9ec00000 (4 MB) given back: idle, 6043 submissions since its sync
[hos host] vk: mirror of pool 0x7fc0400000 (20 MB) given back: idle, 6030 submissions since its sync
[hos host] vk: mirror of pool 0x105cc00000 (4 MB) given back: idle, 6053 submissions since its sync
[hos host] vk: mirror of pool 0x7fe2400000 (8 MB) given back: idle, 6042 submissions since its sync
[hos host] vk: mirror of pool 0x7fa0400000 (8 MB) given back: idle, 6006 submissions since its sync
[hos host] vk: mirror of pool 0x7f58000000 (4 MB) given back: idle, 6058 submissions since its sync
[hos host] vk: replay reads of guest memory as submitted: 4520000 (2709475 pages from the captures)
[hos host] vk: a frame: 131 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 54 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 4.50, present 0.15, draw 2.58, descriptors 1.07, texture lookups 0.05, uploads 0.00, pipelines 0.12, programs 0.42, commands 0.78, dispatch 0.43, memory sync 0.25, queue submit 0.78, slot wait 0.04
[hos host] vk: a frame: mirrors 198 (865 MB): 4.26 MB synced in 859.0 copies, 17.51 MB written back
[hos host] vk: 20160 presents, 90.5 fps; a frame: 899 draws (0 skipped), 86 dispatches, 144 passes, 145 barriers, 6 submits, 0.1 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 43659 pipelines (1179 ms), 0 programs (10 failed), 3367 textures (6279 MB)
[hos host] vk: host time in GPU resource waits: 2.28 ms a frame (fences 0.00, timelines 2.28, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 419.3 full barriers recorded, 227.0 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: 37.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.4 ms a frame (the longest submission 3.8 ms); 0.1 textures made a frame
[hos host] vk: a frame: 0.01 MB unswizzled on the CPU in 0.00 ms
[hos ModuleSystemWorker3] nvn: CommandBufferWaitEvent #195000: event 0x100d6c8008 mode 0 value 118192 (now 118173) (lr 0x8027944c)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 4563 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: replay reads of guest memory as submitted: 4540000 (2724643 pages from the captures)
[hos host] vk: slow present: 17.0 ms:
[hos host] vk: present 20189 written to <data>/jobs/jb_9b21cc79f76459e4/runs/forced-presenter-g0-02-pr/shots/forced-presenter-c0-pr-02-market-after-motion_5s.png (the control bridge)
[hos host] vk: mirror of pool 0x7f72800000 (4 MB) given back: idle, 6035 submissions since its sync
[hos host] vk: sparse mirrors: 71, 864 MB with memory of 1272 MB (407 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 381 binds (107.5 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 126 kept as spares, 0 destroyed; 40 spares now, 556 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 630 dirty pages from their submits' captures, 56339 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: mirror syncs' pass (600 submits): 1103 dispatches, 44814 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5757 runs, 911.3 MB compared; 113 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 19784 in captures, 0 only in the collection, 53 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos host] vk: render targets drawn over from another start, let go: 1465 so far (4729 MB), 1197 of them (3791 MB) between the passes
[hos ModuleSystemWorker1] barrier: 313080 looks: 67320 found the masks moved already, 171289 waited until they moved, 74471 for the whole 1 ms (a wait 402 us on average); 137888 wakes
[hos host] vk: mirror of pool 0x105bc00000 (4 MB) given back: idle, 6049 submissions since its sync
[hos host] vk: mirror of pool 0x1057400000 (4 MB) given back: idle, 6001 submissions since its sync
[hos host] vk: mirror of pool 0x105b400000 (4 MB) given back: idle, 6001 submissions since its sync
[hos host] vk: mirror of pool 0x105c400000 (4 MB) given back: idle, 6001 submissions since its sync
[hos host] vk: 131 mirrors given back (983 MB in all); 864 MB of mirrors now, 56 MB of them idle and kept for what the GPU wrote there alone
[hos host] vk: program picks: 5414123 kept (1903596 by the last picks' textures, a uniform buffer bound since), 72666 new, 2513210 with other textures (of 8000000)
[hos host] vk: replay reads of guest memory as submitted: 4560000 (2739782 pages from the captures)
[hos host] vk: mirror of pool 0x105c000000 (4 MB) given back: idle, 6032 submissions since its sync
[hos host] vk: mirror of pool 0x1057000000 (4 MB) given back: idle, 6039 submissions since its sync
[hos host] vk: replay reads of guest memory as submitted: 4580000 (2754965 pages from the captures)
[hos host] vk: queues ordered by their events: 90000 submissions stopped at a wait, 89319 went on once it was reached, 680 waits gone past; made to go by presents 0, releases 0, waits for the replay 0, 50 ms 50 (stopped 50.06 ms at most); waits met 195792, not met 0; captures at a CPU event's wait 7031, 0 of them at a wait met on arrival
[hos Presentation Thread] vk: capture: a submit 0.09 ms (lock 0.00, scan 0.04 over 1080 ranges, 2952 MB), 188 pages written, 188 copied (0.7 MB)
[hos host] vk: a capture taken (ms): all 0.006; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.003, bytes copied 0.000; 25 kept (146 MB), 9 spare buffers, 2652 texture pages copied (10 MB)
[hos host] vk: mirror of pool 0x7f69800000 (4 MB) given back: idle, 6040 submissions since its sync
[hos host] vk: mirror of pool 0x7f6ec00000 (20 MB) given back: idle, 6023 submissions since its sync
[hos host] vk: late pages found at the replay (600 takes): 46 pages in 1 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos host] vk: replay reads of guest memory as submitted: 4600000 (2770125 pages from the captures)
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 293 KB; 0 made on demand (0 MB kept now); c1 stages: at most 105 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: write-backs left out (600 submits): 1113, 5459.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 550 dirty pages from their submits' captures, 56862 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: mirror syncs' pass (600 submits): 1102 dispatches, 45299 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5891 runs, 927.8 MB compared; 110 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20470 in captures, 0 only in the collection, 53 only by the replay itself
[hos host] vk:   read live, by pool:
[hos sead::MainThread] nvn: CommandBufferSignalEvent #120000: event 0x100d6c8014 mode 0 value 120000 (lr 0x80660bcc)
[hos host] vk: replay reads of guest memory as submitted: 4620000 (2785321 pages from the captures)
[hos host] vk: mirror of pool 0x7f83000000 (12 MB) given back: idle, 6049 submissions since its sync
[hos sead::MainThread] vk: readbacks of GPU results: 1219 of 3000 waited for their write-back, 0.26 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 sead::MainThread] vk: invalidates since the last readbacks line: 8499 kept their virtual core (nothing there to wait for), 3021 gave it up (counted by 256 a thread; RX_VK_INVALIDATE_QUICK=0: none kept)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 2417 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: a frame: 129 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 54 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 4.52, present 0.27, draw 2.59, descriptors 1.06, texture lookups 0.04, uploads 0.00, pipelines 0.12, programs 0.43, commands 0.78, dispatch 0.42, memory sync 0.25, queue submit 0.81, slot wait 0.04
[hos host] vk: a frame: mirrors 198 (837 MB): 4.25 MB synced in 858.0 copies, 17.51 MB written back
[hos host] vk: 20280 presents, 90.3 fps; a frame: 897 draws (0 skipped), 86 dispatches, 145 passes, 145 barriers, 6 submits, 0.0 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 43659 pipelines (1179 ms), 0 programs (10 failed), 3369 textures (6280 MB)
[hos host] vk: host time in GPU resource waits: 2.16 ms a frame (fences 0.00, timelines 2.16, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 418.8 full barriers recorded, 227.0 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: 37.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.4 ms a frame (the longest submission 4.1 ms); 0.0 textures made a frame
[hos host] vk: a frame: 0.00 MB unswizzled on the CPU in 0.00 ms
[hos host] vk: replay reads of guest memory as submitted: 4640000 (2800460 pages from the captures)
[hos host] vk: sparse mirrors: 62, 837 MB with memory of 1212 MB (374 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 388 binds (108.5 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 136 kept as spares, 0 destroyed; 50 spares now, 620 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 378 dirty pages from their submits' captures, 56180 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: mirror syncs' pass (600 submits): 1090 dispatches, 44572 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5746 runs, 909.3 MB compared; 96 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 19876 in captures, 0 only in the collection, 51 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos ModuleSystemWorker1] vcores: 152979 takes, 30257 waits (mean 68 us, longest 11.0 ms, 0 over HOS_VCORE_LOG_MS), 14145 kicks, 13939 yields, 0 starved grants, 0 escapes in 5 s
[hos host] vk: replay reads of guest memory as submitted: 4660000 (2815742 pages from the captures)
[hos sead::MainThread] vk: capture: a submit 0.09 ms (lock 0.00, scan 0.04 over 1080 ranges, 2954 MB), 190 pages written, 190 copied (0.7 MB)
[hos host] vk: a capture taken (ms): all 0.007; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.005, bytes copied 0.000; 25 kept (124 MB), 11 spare buffers, 2652 texture pages copied (10 MB)
[hos host] vk: replay reads of guest memory as submitted: 4680000 (2830959 pages from the captures)
[hos host] vk: late pages found at the replay (600 takes): 0 pages in 0 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos host] vk: replay reads of guest memory as submitted: 4700000 (2846210 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 572 dirty pages from their submits' captures, 56148 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: mirror syncs' pass (600 submits): 1094 dispatches, 44649 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5767 runs, 911.3 MB compared; 108 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20132 in captures, 3 only in the collection, 53 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 4720000 (2861437 pages from the captures)
[hos sead::MainThread] nvn: CommandBufferWaitEvent #200000: event 0x100d6c8008 mode 0 value 121992 (now 121973) (lr 0x8027944c)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 4257 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: replay reads of guest memory as submitted: 4740000 (2876774 pages from the captures)
[hos host] vk: readback pages refreshed from guest memory: 1965332 pages so far (0 held: may be written ahead in their submission)
[hos ModuleSystemWorker5] nvn: 1200000 recordings made, 66959 live copies (19.7 MB), 982 texture copies; 5 command buffers' segments (4.2 MB)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos host] vk: memory: working set 9498 MB (peak 12080), private 3065 MB, staging 32 MB in 2 arenas, 37 made so far (1 ms)
[hos host] vk: memory heaps: heap 0 (system): 19 blocks 984 MB, 39 allocations 872 MB; used 2820 of 31846 MB; heap 1 (device): 252 blocks 8243 MB, 17045 allocations 7712 MB; used 8902 of 21493 MB;
[hos host] vk: image blocks: 1 blocks 32 MB, 6 images 31 MB; 2 free ranges, the largest 0 MB
[hos host] vk: image blocks (render targets): 16 blocks 512 MB, 106 images 245 MB; 68 free ranges, the largest 30 MB
[hos host] vk: image blocks (textures, compacted): 142 blocks 4544 MB, 3249 images 4498 MB; 712 free ranges, the largest 5 MB
[hos host] vk: render target images: 868 reused from the pool, 800 made new, 691 destroyed from it; it holds 0 (0 MB)
[hos host] vk: VRAM: 8902 MB used of 21493 MB allowed (530 MB of it free in the blocks); textures 6280 MB (budget 11000 MB), 3371 textures
[hos host] vk: budget: VRAM 8243 of 21493 MB (8902 MB by the driver, 658 MB unaccounted): textures 6280, mirrors 837, blocks free 530, shadows 288, other buffers 227, other images 46, staging 32; RAM 2754 MB (the process's private memory 3065 MB): imports 1812, capture spares 752, captures 160, capture table 30, others 0
[hos host] vk: textures: decoded ASTC 4088 MB, rendered 860 MB (51 drawn in one frame, kept: 34055 KB), other 713 MB; used in the last 8 presents 2129 MB
[hos host] vk: decoded ASTC used again: 149 textures (160 MB) copied from ones the game finalized, not decoded; 149 of those retired, 370 listed; 865 MB of blocks hashed in 80 ms
[hos host] vk:   288 MB: 0x1147c16000 12288x12288x1 x1 levels 1 NVN 0x32 rendered, last used 1 presents ago
[hos host] vk:   214 MB: 0x10a5b22c00 1365x1365x1 x90 levels 11 NVN 0x80 ASTC, last used 4744 presents ago
[hos host] vk:   214 MB: 0x10a22aa600 1365x1365x1 x90 levels 11 NVN 0x8e ASTC, last used 4744 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: 0x113755c000 4096x4096x1 x1 levels 1 NVN 0xa, last used 7882 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: 0x100030a4200 3072x3072x1 x3 levels 1 NVN 0x32 rendered, last used 1 presents ago
[hos host] vk:   54 MB: 0x10002d97a00 3072x3072x1 x3 levels 1 NVN 0x32 rendered, last used 667 presents ago
[hos host] vk: finalized textures the GPU drew once, kept for being made again: 56 (12391 KB): 8x8x1 NVN 0x3f: 1 (2 KB); 50x57x1 NVN 0x25: 1 (11 KB); 62x41x1 NVN 0x25: 1 (9 KB); 64x62x1 NVN 0x25: 1 (15 KB); 73x33x1 NVN 0x25: 1 (9 KB); 74x32x1 NVN 0x25: 1 (9 KB); 90x34x1 NVN 0x25: 1 (11 KB); 114x121x1 NVN 0x1: 1 (13 KB); 126x105x1 NVN 0x1: 1 (12 KB); 128x126x1 NVN 0x1: 1 (15 KB); 137x97x1 NVN 0x1: 1 (12 KB); 138x96x1 NVN 0x1: 1 (12 KB); 154x98x1 NVN 0x1: 1 (14 KB); 165x112x1 NVN 0x1: 1 (18 KB); 198x40x1 NVN 0x25: 2 (61 KB); 242x40x1 NVN 0x25: 3 (113 KB);
[hos Presentation Thread] vk: presents after the CPU's signal (600): 226 waited for it, 0.49 ms each, 0 let go after RX_VK_FRAME_GATE_MS
[hos host] vk: replay reads of guest memory as submitted: 4760000 (2891901 pages from the captures)
[hos host] vk: status: 600 frames, the slowest 26.9 ms; VRAM 8.7 of 21.0 GB
[hos host] vk: program census (play): created 0 (0 with stages; 0 with no present for 100 ms before, 0 in a frame of 100 ms or more; 0.0 ms of the game's threads), freed 0 (0 never used); first uses 0: program ready 0, loaded 0, translated 0, failed 0; pipeline ready 0, from a cache 0, compiled 0, made 0, left out 0, none 0; 0 with no create call seen, 0 whose stages are not the create call's; after the create call: under 16.7 ms 0, under 100 ms 0, under 1 s 0, longer 0; needing work 0 (0 of them 100 ms or more after the create call, 0 created in a load or a long frame); in play: 23 program records loaded (5.9 ms), 1 new graphics pipelines, pipelines from the driver's cache 0, from the start's cache 0, compiled 1; 0 draws and dispatches left out, 0 waits (0.0 ms)
[hos host] vk: program loads at create (play): 5049 queued, 5049 loaded (median 0.14 ms, 99th percentile 1.29 ms, worst 193.21 ms, 1688 ms and 1046.5 MB read in all); 6969 programs (1465 of them compute variants without their code, 0 graphics programs no note names), 9153 graphics pipelines (0 not in a cache), 407 compute pipelines (250 not in a cache); a first use ran 0 itself and waited for 0 (0.0 ms); 1635 pipelines made at create taken by draws; 3199 create calls with nothing kept for their stages, 0 before the index; units 8248 (5049 ready, 0 loading), 14342 programs alive; 0 noted graphics pipeline probes deferred to first use
[hos host] vk: module identifiers (play): 658 known, 0 learned in this run; 407 compute pipelines made by them at a create call, 250 asked that way were in no cache (their code read, as before)
[hos host] vk: a frame: 132 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 54 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 4.53, present 0.14, draw 2.58, descriptors 1.06, texture lookups 0.05, uploads 0.00, pipelines 0.12, programs 0.43, commands 0.78, dispatch 0.45, memory sync 0.28, queue submit 0.79, slot wait 0.04
[hos host] vk: a frame: mirrors 198 (837 MB): 4.25 MB synced in 858.2 copies, 17.51 MB written back
[hos host] vk: 20400 presents, 90.6 fps; a frame: 901 draws (0 skipped), 86 dispatches, 145 passes, 145 barriers, 6 submits, 0.1 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 43659 pipelines (1179 ms), 0 programs (10 failed), 3371 textures (6280 MB)
[hos host] vk: host time in GPU resource waits: 2.15 ms a frame (fences 0.00, timelines 2.15, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 419.0 full barriers recorded, 227.0 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: 37.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.4 ms a frame (the longest submission 4.1 ms); 0.1 textures made a frame
[hos host] vk: a frame: 0.00 MB unswizzled on the CPU in 0.00 ms
[hos host] vk: the render thread (600 presents, 11.09 ms a frame): its jobs 5.26 ms a frame (submits 2.14, presents 0.17, others 2.95), idle 5.82; the game waited for room at 1 presents, 11.77 ms each (0.02 a frame; 3.25 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); 403 captures at a CPU signal; the submits' waits read in 0.028 ms a frame; readbacks: 2992 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 4.63 ms after its hand-over (21.75 at most)
[hos host] vk: frame overlap (600 presents): the GPU between recordings, ms a frame, idle + in the next one's uploads: up to a frame's first recording with a draw 2.47 + 0.19 (the window's copy is in the idle), after it 1.70 + 0.26; after a frame's last recording that reads or writes the game's memory come 1.01 recordings (0.01 ms of GPU time), after its last that reads it 1.01 (0.01 ms)
[hos host] vk: frame ahead (600 presents): 4.80 parts a frame recorded in the signal's job, 83.5% of them took nothing that came with the signal; of the others, 0.79 a frame took pages of the signal's capture into a mirror (13.8 pages a frame) and 0.00 pages read as guest memory had them (0.0 pages a frame); a frame's first part took nothing in 217 of 403 frames, its first three in 217
[hos host] vk: frame ahead, what came with the signal: 403 signals with a capture, each of 40.8 pages in 7.1 pools; between a frame's last submit and that signal 0.00 texture and sampler registrations, 0.00 releases and 0.00 other CPU signals a signal (0 of them with any)
[hos host] vk: frame overlap, by place: idle + uploads before a frame's 2nd to 12th recording with a draw, then all later ones: 0.66+0.09, 0.43+0.06, 0.20+0.07, 0.20+0.03, 0.21+0.01, 0.01+0.00, 0.00+0.00, 0.00+0.00, 0.00+0.00, 0.00+0.00, 0.00+0.00, 0.00+0.00
[hos host] vk: sparse mirrors: 62, 837 MB with memory of 1212 MB (374 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 388 binds (108.5 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 136 kept as spares, 0 destroyed; 50 spares now, 620 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 483 dirty pages from their submits' captures, 56553 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: mirror syncs' pass (600 submits): 1106 dispatches, 44913 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5763 runs, 910.4 MB compared; 98 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20497 in captures, 0 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos sead::MainThread] vk: capture: a submit 0.10 ms (lock 0.00, scan 0.04 over 1080 ranges, 2952 MB), 193 pages written, 193 copied (0.8 MB)
[hos host] vk: a capture taken (ms): all 0.007; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.004, bytes copied 0.000; 25 kept (124 MB), 10 spare buffers, 2795 texture pages copied (10 MB)
[hos host] vk: program picks: 5532847 kept (1922539 by the last picks' textures, a uniform buffer bound since), 72666 new, 2594486 with other textures (of 8200000)
[hos sead::MainThread] nvn: GPU waits in 5 s: 85 sync waits (70 ms), 0 polls timed out (0 ms), 0 queue finishes (0 ms)
[hos ModuleSystemWorker5] vk: readbacks of GPU results: 1194 of 3000 waited for their write-back, 0.27 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 ModuleSystemWorker5] vk: invalidates since the last readbacks line: 8814 kept their virtual core (nothing there to wait for), 2962 gave it up (counted by 256 a thread; RX_VK_INVALIDATE_QUICK=0: none kept)
[hos host] vk: replay reads of guest memory as submitted: 4780000 (2907173 pages from the captures)
[hos host] vk: replay reads of guest memory as submitted: 4800000 (2922286 pages from the captures)
[hos host] vk: late pages found at the replay (600 takes): 0 pages in 0 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 443 dirty pages from their submits' captures, 56459 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: mirror syncs' pass (600 submits): 1094 dispatches, 44776 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5776 runs, 910.4 MB compared; 98 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20024 in captures, 1 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: replay reads of guest memory as submitted: 4820000 (2937576 pages from the captures)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 6500 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos host] vk: replay reads of guest memory as submitted: 4840000 (2952675 pages from the captures)
[hos Presentation Thread] vk: capture: a submit 0.09 ms (lock 0.00, scan 0.04 over 1079 ranges, 2958 MB), 190 pages written, 190 copied (0.7 MB)
[hos host] vk: a capture taken (ms): all 0.006; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.004, bytes copied 0.000; 25 kept (146 MB), 10 spare buffers, 2795 texture pages copied (10 MB)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos host] vk: replay reads of guest memory as submitted: 4860000 (2967995 pages from the captures)
[hos host] vk: the recording streams: 48 of 2 MB, the most one held 293 KB; 0 made on demand (0 MB kept now); c1 stages: at most 105 a recording; 0 descriptor pools made when a recording's own ran out
[hos host] vk: sparse mirrors: 62, 837 MB with memory of 1212 MB (374 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 388 binds (108.5 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 136 kept as spares, 0 destroyed; 50 spares now, 620 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1113, 5459.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 458 dirty pages from their submits' captures, 56725 from earlier ones', 0 from guest memory as it was at the replay; 24 captures kept; 0 pages the GPU wrote copied over (0 live); 0 kept for the GPU (RX_VK_PROTECT_BARE)
[hos host] vk: mirror syncs' pass (600 submits): 1101 dispatches, 45175 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5863 runs, 927.8 MB compared; 102 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 19962 in captures, 1 only in the collection, 53 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos host] vk: replay reads of guest memory as submitted: 4880000 (2983204 pages from the captures)
[hos host] vk: a frame: 132 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 54 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 4.50, present 0.15, draw 2.60, descriptors 1.07, texture lookups 0.05, uploads 0.00, pipelines 0.12, programs 0.43, commands 0.79, dispatch 0.42, memory sync 0.25, queue submit 0.79, slot wait 0.04
[hos host] vk: a frame: mirrors 198 (837 MB): 4.26 MB synced in 859.3 copies, 17.51 MB written back
[hos host] vk: 20520 presents, 89.6 fps; a frame: 903 draws (0 skipped), 86 dispatches, 145 passes, 146 barriers, 6 submits, 0.1 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 43659 pipelines (1179 ms), 0 programs (10 failed), 3374 textures (6281 MB)
[hos host] vk: host time in GPU resource waits: 2.39 ms a frame (fences 0.00, timelines 2.39, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 419.2 full barriers recorded, 227.1 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: 37.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.6 ms a frame (the longest submission 4.1 ms); 0.1 textures made a frame
[hos host] vk: a frame: 0.00 MB unswizzled on the CPU in 0.00 ms
[hos ModuleSystemWorker1] nvn: CommandBufferSignalEvent #125000: event 0x100d6c8020 mode 0 value 125000 (lr 0x80667410)
[hos host] vk: replay reads of guest memory as submitted: 4900000 (2998461 pages from the captures)
[hos host] vk: late pages found at the replay (600 takes): 0 pages in 0 takes (RX_VK_LATE_AT_REPLAY=0: only the ones the capture found)
[hos ModuleSystemWorker2] nvn: DrawMeshTasksIndirect: 240000 so far (x1 0x117d058000 x2 0x2 x3 0x117d058000 x4 0x44b040)
[hos Presentation Thread] vi: 20549 frames presented (90.4 fps over the last 5.0 s, interval 1)
[hos Presentation Thread] vi: present intervals: <15.5 ms 444, 15.5-17.8 7, 17.8-22 1, 22-30 0, 30-40 0, over 40 0
[hos host] vk: slow present: 16.2 ms:
[hos host] vk: present 20550 written to <data>/jobs/jb_9b21cc79f76459e4/runs/forced-presenter-g0-02-pr/shots/forced-presenter-c0-pr-02-market-after-motion_9s.png (the control bridge)
[hos host] vk: replay reads of guest memory as submitted: 4920000 (3013577 pages from the captures)
[hos host] vk: write-backs left out (600 submits): 1113, 5459.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 689 dirty pages from their submits' captures, 57191 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: mirror syncs' pass (600 submits): 1113 dispatches, 45562 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5864 runs, 927.8 MB compared; 106 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20367 in captures, 0 only in the collection, 53 only by the replay itself
[hos host] vk:   read live, by pool:
[hos ModuleSystemWorker1] nvn: CommandBufferWaitEvent #205000: event 0x100d6c8008 mode 0 value 125792 (now 125773) (lr 0x8027944c)
[hos sead::MainThread] vk: readbacks of GPU results: 1174 of 3000 waited for their write-back, 0.28 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 sead::MainThread] vk: invalidates since the last readbacks line: 8518 kept their virtual core (nothing there to wait for), 3002 gave it up (counted by 256 a thread; RX_VK_INVALIDATE_QUICK=0: none kept)
[hos host] vk: replay reads of guest memory as submitted: 4940000 (3028798 pages from the captures)
[hos host] vk: shadow staging (600 takes): 0 shadow copies from staging, 4136 from the capture buffer; capture buffers: 600 host-cached, 0 not (write-combined: staged)
[hos Presentation Thread] vk: capture: a submit 0.09 ms (lock 0.00, scan 0.04 over 1079 ranges, 2958 MB), 191 pages written, 191 copied (0.7 MB)
[hos host] vk: a capture taken (ms): all 0.006; GPU wait for late pages 0.000 (0 of 600 takes), pools lock 0.000, retires and stamps 0.003, bytes copied 0.000; 25 kept (138 MB), 9 spare buffers, 2827 texture pages copied (11 MB)
[hos host] vk: a submit's memory work (ms): maps 0.00, capture taken 0.01, writes collected 0.03
[hos host] vk: replay reads of guest memory as submitted: 4960000 (3043929 pages from the captures)
[hos host] vk: sparse mirrors: 62, 837 MB with memory of 1212 MB (374 MB spared); blocks given memory later 2578, given back 6259; bindings served from guest memory meanwhile 0 (0 reaching blocks with memory too), from the mirror past every game pool 0; 388 binds (108.5 ms); 0 made whole for want of memory
[hos host] vk: sparse mirrors' buffers: 86 taken from the spares, 112 made new; 136 kept as spares, 0 destroyed; 50 spares now, 620 MB of address space (RX_VK_SPARSE_SPARE_MB=4096)
[hos host] vk: write-backs left out (600 submits): 1092, 5356.4 MB; 23 bindings followed, 34 ranges the CPU read, 0 resets; CPU reads or writes of one left out: 0 so far
[hos host] vk: mirror syncs (600 submits): 644 dirty pages from their submits' captures, 56451 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: mirror syncs' pass (600 submits): 1106 dispatches, 44906 chunks; 0 syncs as copies (no table room)
[hos host] vk: write-back diff (600 submits): 5763 runs, 910.3 MB compared; 102 pages read into the shadows; 288 MB of shadows; pages the shadows knew that the CPU wrote: 20129 in captures, 1 only in the collection, 52 only by the replay itself
[hos host] vk:   read live, by pool:
[hos host] vk: imports: 126 live (1812 MB), 0 retired still imported (0 MB), 21 retired given back (476 MB)
[hos host] vk: replay reads of guest memory as submitted: 4980000 (3059181 pages from the captures)
[hos host] vk: program picks: 5651749 kept (1941396 by the last picks' textures, a uniform buffer bound since), 72666 new, 2675584 with other textures (of 8400000)
[hos host] vk: replay reads of guest memory as submitted: 5000000 (3074348 pages from the captures)
[hos host] vk: a frame: 133 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 54 ms in all, 0 jobs queued
[hos host] vk: a frame's replay time (ms): submit 4.50, present 0.27, draw 2.56, descriptors 1.04, texture lookups 0.04, uploads 0.00, pipelines 0.12, programs 0.42, commands 0.78, dispatch 0.43, memory sync 0.25, queue submit 0.78, slot wait 0.04
[hos host] vk: a frame: mirrors 198 (837 MB): 4.26 MB synced in 858.7 copies, 17.51 MB written back
[hos host] vk: 20640 presents, 89.7 fps; a frame: 902 draws (0 skipped), 86 dispatches, 145 passes, 145 barriers, 6 submits, 0.0 uploads (0.0 MB), 82 ub updates (57 overlays), GPU-bound wait 0.0 ms (recording slots only); 43659 pipelines (1179 ms), 0 programs (10 failed), 3375 textures (6283 MB)
[hos host] vk: host time in GPU resource waits: 2.38 ms a frame (fences 0.00, timelines 2.38, idle 0.00, queries 0.00; sum across threads, not frame time; excludes pacing and driver submit/present blocking)
[hos host] vk: a frame: 419.1 full barriers recorded, 227.1 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: 37.0; 0x12: 76.5; 0x13: 2.0; 0x21: 4.0; 0x28: 16.0; 0x92: 3.0; 0x97: 1.0;
[hos host] vk: GPU time 6.5 ms a frame (the longest submission 4.0 ms); 0.0 textures made a frame
[hos host] vk: a frame: 0.02 MB unswizzled on the CPU in 0.00 ms
[hos host] vk: slow present: 16.3 ms:
[hos host] vk: present 20640 written to <data>/jobs/jb_9b21cc79f76459e4/runs/forced-presenter-g0-02-pr/shots/forced-presenter-c0-pr-02-market-after-motion.png (the control bridge)
[hos host] control: quit asked by a client
[hos host] vk: program census (close): created 0 (0 with stages; 0 with no present for 100 ms before, 0 in a frame of 100 ms or more; 0.0 ms of the game's threads), freed 0 (0 never used); first uses 0: program ready 0, loaded 0, translated 0, failed 0; pipeline ready 0, from a cache 0, compiled 0, made 0, left out 0, none 0; 0 with no create call seen, 0 whose stages are not the create call's; after the create call: under 16.7 ms 0, under 100 ms 0, under 1 s 0, longer 0; needing work 0 (0 of them 100 ms or more after the create call, 0 created in a load or a long frame); in play: 23 program records loaded (5.9 ms), 1 new graphics pipelines, pipelines from the driver's cache 0, from the start's cache 0, compiled 1; 0 draws and dispatches left out, 0 waits (0.0 ms)
[hos host] vk: program loads at create (close): 5049 queued, 5049 loaded (median 0.14 ms, 99th percentile 1.29 ms, worst 193.21 ms, 1688 ms and 1046.5 MB read in all); 6969 programs (1465 of them compute variants without their code, 0 graphics programs no note names), 9153 graphics pipelines (0 not in a cache), 407 compute pipelines (250 not in a cache); a first use ran 0 itself and waited for 0 (0.0 ms); 1635 pipelines made at create taken by draws; 3199 create calls with nothing kept for their stages, 0 before the index; units 8248 (5049 ready, 0 loading), 14342 programs alive; 0 noted graphics pipeline probes deferred to first use
[hos host] vk: module identifiers (close): 658 known, 0 learned in this run; 407 compute pipelines made by them at a create call, 250 asked that way were in no cache (their code read, as before)
[hos host] vk: page table census (close): no pool outside its windows, 64-72 GB and 508-512 GB (4311 pools in the table now)
[hos host] control: exiting
[hos host] stop: close work done (a file the game was writing waited for, at most 3 s)