Game: Periodic ~4.6 s stalls of 100-200 ms (wall-clock cadence, camera-movement dependent) in Castle Town market and Kokiri Forest: host blocked in present or texture lookups #1988

Open
opened 2026-11-21 02:51:33 +00:00 by Sylvohm · 1 comment

Longshot version

v3.0.0 stable (source bd6a0cdd); Nightly 649 (source 846e5399)

Channel

Release

What kind of problem

Low fps or stutter

What happened

Disclaimer

I’m not a developer, and I’ve been looking into my problem with Claude AI, who has also written the following.

Summary

With the camera moving, the game stalls for 100-200 ms (up to ~240 ms) about every 4.6 s in busy scenes. The cadence is wall-clock: the period is the same at a 30 fps cap and at ~52-60 fps, and with different graphics settings. The GPU is at ~35 % (~8 ms per frame), no shader is built, there are 0-1 uploads per frame, and the system measures clean.

RX_VK_SLOW_LOG=50 and HOS_PROFILE=1 place the lost time either in texture lookups (a single frame) or in present followed by texture lookups on the next frame. In both cases the game's main thread is blocked and no game thread is on the CPU. Nothing I changed (system, kernel, driver, desktop, V-Sync, FPS cap, ~12 Longshot switches) removes it. From outside I cannot tell what the game waits for, so I am looking for pointers: which call blocks, and what runs on a ~4.6 s cadence?

Environment

Game Longshot v3.0.0 stable (source bd6a0cdd); Nightly 649 (846e5399) also tested, same result
CPU / RAM i5-12400F (6 cores, 12 threads), 32 GB DDR4
GPU RTX 3060 12 GB, NVIDIA 595.99.02 and 615.71.09 (same result)
OS Linux Mint 22.3, X11 (Cinnamon; Openbox without compositor also tried)
Kernels 7.0.0-43 and 6.8.0-149 (same result)

Two configurations show the same stalls:

  • Config A (daily): DLSS Quality 67 %, 1920x1200, FIFO swapchain ("V-Sync Auto"), built-in frame generation Auto, Reflex, ~52 game fps. Earlier sessions at 2048x1152 without DLSS and with V-Sync off also showed it.
  • Config B (bench): handheld profile, 30 fps cap, no DLSS / FG, V-Sync Off (immediate), everything at minimum.

Another recomp (HeyListen) shows the same symptom on this machine. I have not checked whether it shares this renderer.

What I see

  • Moving the camera continuously (Kokiri Forest, Castle Town market): a stall of 100-200 ms about every 4.6 s.
  • Moving only the character: much rarer. Standing still in a calm area (Hyrule Field): none.
  • Standing still next to the market fountain: the stalls continue at about half the rate (~9.6 s; in one 30 s test, big stalls at 28.3, 37.9, 47.9 and 57.4 s). They disappear when I move away from that spot.

Evidence

1. Cadence

  • Config A, MangoHud: 6 runs of 55 s with continuous camera movement (3 with RX_VK_COMPACT=0, 3 without, same result).
Without With RX_VK_COMPACT=0
Stalls >= 50 ms per run 9, 12, 8 (mean 9.7) 8, 11, 10 (mean 9.7)
Stalls >= 100 ms per run 6, 9, 5 (mean 6.7) 4, 6, 6 (mean 5.3)

Of 52 intervals between stalls, 44 are 3.6-5.8 s (median 4.6 s, SD 0.46 s). The other 8 are 8.6-10.7 s, i.e. two beats with one below 50 ms. So it is not an exact timer.

  • Config B, RX_VK_SLOW_LOG run (61 s): 11 stalls >= 95 ms (100-195 ms), period fit 4.58 s (R = 0.91).
  • Earlier sessions: fitted 4.58 s and 4.70 s at a 30 fps cap, and median 4.64 s over 89 intervals with DLSS and frame generation.
  • The period is 4.6-4.7 s in both configs although the game frame rate differs (30 vs ~52). It looks time-based rather than frame-count-based (my inference).

2. Where the time goes

All stalls have 0-1 uploads, 0 new pipelines / programs, and 1600-2300 draws.

(a) One frame, submit dominated by texture lookups (normally ~1-2 ms per frame):

vk: slow submit: 164.1 ms: draw 1.1 texture lookups 162.5
vk: slow submit: 101.0 ms: ... texture lookups 99.4
vk: slow submit: 129.6 ms: texture lookups 128.1
vk: slow submit: 163.4 ms: texture lookups 162.4

(b) A slow present, then texture lookups on the next frame (4 of the 11 largest stalls):

vk: slow frame at present 2107: 156.3 ms; submit 22.2, draw 7.9, ... 0 uploads
vk: slow present: 105.8 ms:
vk: slow submit: 55.6 ms: texture lookups 54.5
vk: slow frame at present 2108: 69.7 ms; submit 67.5, ... 0 uploads

The same pattern appears at presents 1172 (frame 241.9 ms, submit 46.9, present 169.1), 1588 (104.6 / 15.5 / 83.8, next frame lookups 78.5) and 1833 (135.2 / 24.1 / 103.9, next frame lookups 37.5). My reading: one blockage of ~150-230 ms that lands in whichever host call is running.

HOS_PROFILE=1 (88 s, Castle Town market; report.md and the profile folder attached):

  • 5469 frames, 62.1 fps, p99 46.4 ms, p99.9 181 ms. 47 frames >= 50 ms (16 over 100 ms).
  • After the load-in (from 25 s), the 8 largest stalls (106-194 ms) are all in tex_lookup_ms (80-166 ms) or tex_frame_ms (168 ms, with wait_render_ms 164 ms). They have 0 or 1 uploads, so this is not upload volume.
  • During these stalls main_blocked_ms is 79 ms on average (12.6 ms in steady frames) while main_vcore_ms stays at 0.06 ms. The main thread is not waiting for a virtual core.
  • The sampler shows sead::MainThread waiting in libc.so.6 100 % of the time, and in most sweeps no game thread is on the CPU. The wait is in a lock or the driver. longshot_vk.so offsets are unresolved on my side, so I cannot name the call.
  • Other game log lines: slow upload 59.2 ms: ... 512x512 ... ASTC, 236 KB.

perf (other session): mkw_vk display blocked 122.7 ms at 159.14 s and 169.4 ms at 163.66 s (4.5 s apart).

3. Not the system

  • cyclictest (30 s): max latency 6-157 µs.
  • perf sched: no game thread waits more than ~5 ms for a CPU.
  • turbostat: SMI = 0, no thermal or power limit.
  • No swap-in or disk read at stall times.
  • During a stall CPU load and frequency drop: the game is waiting, not starved.
  • Other Vulkan games and vkcube run fine on the same machine.

4. Tried without effect

Test Result
V-Sync On/Off, 30 fps cap, minimum graphics, windowed, Openbox (no compositor), compositing on/off, FG on/off, DLSS on/off no change
Kernels 7.0.0-43 / 6.8.0-149, NVIDIA 595.99.02 / 615.71.09 no change
NVIDIA shader disk cache off, Wi-Fi/Bluetooth off, SDL audio drivers no change
Controller: wired, Bluetooth, and no controller at all no change
Monitoring tools stopped (nvidia-smi, pidstat, MangoHud) no change
RX_VK_PRESENT_THREAD=1, threaded rendering off no change
RX_VK_COMPACT=0, HOS_TEX_COMPACT_MS=off / 10000 no measurable change
RX_VK_UPLOAD_BUDGET_MS=10, RX_VK_ASTC_RGBA8=1 no change
HOS_CORES 8 / 10 / 12 no change
HOS_RESERVE_CORE=0 worse

The sparse-mirror destruction fix described in your docs is active here (0 destroyed).

5. Thread-level observations

  • Virtual-core waits ("vcores: X waited N ms for a core ... never asked") are frequent: 94 waits >= 50 ms in the 88 s profile, and in another run 128 in 86 s (65 >= 100 ms), with chains NavMeshCAStepDepWorker1 → tera::FileResourceMgr → PreActorModelCompose → Porter on core 1. Only 7 of the 94 overlap a stall, and none of the 11 largest stalls has one within ±4 log lines (14 % expected by chance). The waiting threads are loaders (tera::FileResourceMgr, PreActorModelCompose, GeneralFormatAsync_), blocked while sead::MainThread holds the core. I read this as a consequence of the main thread being blocked rather than the cause.
  • Replay thread (mkw_vk replay): scheduled ~3000 times per second (~100 per frame, a sched_yield spin around Recorder::wait_idle). In a 45 s bpftrace capture with stalls it showed no off-CPU wait >= 40 ms. A 40 s perf record shows 60 % on CPU, no continuous on-CPU stretch >= 80 ms, longest gap 85 ms. It appears to wait for work rather than cause the stall; the in-game overlay shows "Waiting on: Game's main thread" rising during the holes. This is hard to reconcile with tex_lookup_ms of 165 ms.
  • bpftrace on all threads, 58 s, no controller connected: of all ioctl / mmap / mprotect / munmap / mremap / madvise calls, only one lasted >= 40 ms (a 42 ms ioctl in mkw_vk replay). So there is no long NVIDIA ioctl or memory syscall during the stalls.
  • futex waits >= 60 ms in Presentation Th (60-101 ms) occur at seconds 3, 13, 18, 22, 27, 32 and 46 of the capture (period fit 4.76 s, R = 0.94, 1 s clock resolution). At 3, 22 and 46 s, sead::MainThread (52-63 ms) and mkw_vk recorder (41 ms) also wait at the same moment, so the whole pipeline stalls at once. The larger events fall on the same ~4.76 s grid every 4-5 cycles. The main replay thread never waits >= 40 ms in futex; the presentation thread is starved while the replay does not sleep.
  • Guest thread OptimizeChangeBanc: in a separate 37 s perf capture it was the only thread with CPU bursts >= 80 ms: 94 ms at t = 22.05 s and 203 ms at t = 26.51 s, 4.46 s apart (hot function f_805aeea0, 18 % of its samples). In the profile above it is active in 2 of the 8 largest stalls. Two events are too few to conclude.
  • Render-target churn is high in this scene: 7591 targets and 5271 MB "let go" in the first run, while the pool holds 7-12 MB of 768 MB.

6. Side note (not the same problem)

A performance power profile (powerprofilesctl set performance, min_perf_pct=100) cuts replay CPU per frame from 15.8 to 10.5 ms and 50-65 ms frames from ~85/min to ~35/min (30 fps cap). It does not remove the periodic 100-200 ms stalls.

Questions

  1. Is a ~4.6 s cadence (SD ≈ 0.5 s), tied to camera movement, known? Is there a periodic job in the game or the host whose cost grows with the scene (map placement optimisation, navmesh update, actor streaming, render-target / texture release, descriptor or sampler rebuild, a timer)?
  2. What does the main thread wait for during these stalls? The sampler only shows libc.so.6. Is there a switch, a log or a Tracy build that names the object it waits on? Could the log break down slow present (acquire vs queue present) and texture lookups (per-call time)?
  3. How can tex_lookup_ms reach 165 ms with no upload? Could present and texture lookups be waiting on the same lock shared with a game thread, or on the same driver operation?
  4. In the perf stacks the replay thread is often in Recorder::spin<…wait_idle…> / sched_yield before long gaps. What can keep it waiting 100+ ms there, and what is the recorder doing at that time?
  5. I tested Nightly 649 (846e5399) and the stalls are unchanged. Are #994 (render-target reuse), #838 (render thread leaving its core) and the occlusion virtual-core fix included in that nightly? If not, is there a branch or build with them that I can test? I can also run a debug build or any switch you suggest, and report.

Where in the game

Kokiri Forest, Town Market, while moving the camera

How often

Every time

Graphics card and driver version

RTX 3060 12 GB, NVIDIA 595.99.02 and 615.71.09 (same result)

CPU, RAM and operating system

i5-12400F (6 cores, 12 threads), 32 GB DDR4, Linux Mint 22.3, X11 (Cinnamon; Openbox without compositor also tried)

Changes from the defaults

Frame generation: Auto (Built-in FG) (default: Off)
Frame generation by: Built-in FG (default: FSR 3.1 FG)
Fullscreen: On (default: Off)
GPU device: 0: NVIDIA GeForce RTX 3060 (default: Auto)
Volume: 50% (default: 100%)
FPS counter: Dev (default: Off)
Menu size: Small (default: Medium)
Depends on this PC (the values now, not counted as changes):
Quality preset: Auto (this PC) (default depends on this PC)
Output resolution: 1920x1200 (native) (default depends on this PC)
Profile: Docked (default depends on this PC)
Upscaler: NVIDIA DLSS (default depends on this PC)
Render resolution: Quality 67% (1280x800) (default depends on this PC)
Sharpness: 40% (default depends on this PC)
Window scaling filter: Not needed (window = output) (default depends on this PC)
Game speed: Real time (default depends on this PC)
Frame pacing: On (default depends on this PC)
Switches set:
LONGSHOT_NO_UPDATE_CHECK=1
SDL_VIDEO_WAYLAND_WMCLASS=(value left out)
SDL_VIDEO_X11_WMCLASS=(value left out)

longshot_log.txt

No response

Screenshots or video

No response

Before you send it

  • The problem is still there on the newest release or nightly.
  • I searched the open issues for it.
  • I'm not attaching game files, keys, the game's executable, its .bin caches or saves.
### Longshot version v3.0.0 stable (source bd6a0cdd); Nightly 649 (source 846e5399) ### Channel Release ### What kind of problem Low fps or stutter ### What happened # Disclaimer I’m not a developer, and I’ve been looking into my problem with Claude AI, who has also written the following. ## Summary With the camera moving, the game stalls for 100-200 ms (up to ~240 ms) about every **4.6 s** in busy scenes. The cadence is **wall-clock**: the period is the same at a 30 fps cap and at ~52-60 fps, and with different graphics settings. The GPU is at ~35 % (~8 ms per frame), no shader is built, there are 0-1 uploads per frame, and the system measures clean. `RX_VK_SLOW_LOG=50` and `HOS_PROFILE=1` place the lost time either in `texture lookups` (a single frame) or in `present` followed by `texture lookups` on the next frame. In both cases the game's main thread is blocked and no game thread is on the CPU. Nothing I changed (system, kernel, driver, desktop, V-Sync, FPS cap, ~12 Longshot switches) removes it. From outside I cannot tell what the game waits for, so I am looking for pointers: which call blocks, and what runs on a ~4.6 s cadence? ## Environment | | | |---|---| | Game | Longshot v3.0.0 stable (source bd6a0cdd); Nightly 649 (846e5399) also tested, same result | | CPU / RAM | i5-12400F (6 cores, 12 threads), 32 GB DDR4 | | GPU | RTX 3060 12 GB, NVIDIA 595.99.02 and 615.71.09 (same result) | | OS | Linux Mint 22.3, X11 (Cinnamon; Openbox without compositor also tried) | | Kernels | 7.0.0-43 and 6.8.0-149 (same result) | Two configurations show the same stalls: - **Config A (daily)**: DLSS Quality 67 %, 1920x1200, FIFO swapchain ("V-Sync Auto"), built-in frame generation Auto, Reflex, ~52 game fps. Earlier sessions at 2048x1152 without DLSS and with V-Sync off also showed it. - **Config B (bench)**: handheld profile, 30 fps cap, no DLSS / FG, V-Sync Off (immediate), everything at minimum. Another recomp (HeyListen) shows the same symptom on this machine. I have not checked whether it shares this renderer. ## What I see - Moving the camera continuously (Kokiri Forest, Castle Town market): a stall of 100-200 ms about every 4.6 s. - Moving only the character: much rarer. Standing still in a calm area (Hyrule Field): none. - Standing still next to the market fountain: the stalls continue at about half the rate (~9.6 s; in one 30 s test, big stalls at 28.3, 37.9, 47.9 and 57.4 s). They disappear when I move away from that spot. ## Evidence ### 1. Cadence - **Config A, MangoHud**: 6 runs of 55 s with continuous camera movement (3 with `RX_VK_COMPACT=0`, 3 without, same result). | | Without | With `RX_VK_COMPACT=0` | |---|---|---| | Stalls >= 50 ms per run | 9, 12, 8 (mean 9.7) | 8, 11, 10 (mean 9.7) | | Stalls >= 100 ms per run | 6, 9, 5 (mean 6.7) | 4, 6, 6 (mean 5.3) | Of 52 intervals between stalls, 44 are 3.6-5.8 s (median 4.6 s, SD 0.46 s). The other 8 are 8.6-10.7 s, i.e. two beats with one below 50 ms. So it is not an exact timer. - **Config B, `RX_VK_SLOW_LOG` run (61 s)**: 11 stalls >= 95 ms (100-195 ms), period fit 4.58 s (R = 0.91). - **Earlier sessions**: fitted 4.58 s and 4.70 s at a 30 fps cap, and median 4.64 s over 89 intervals with DLSS and frame generation. - The period is 4.6-4.7 s in both configs although the game frame rate differs (30 vs ~52). It looks time-based rather than frame-count-based (my inference). ### 2. Where the time goes All stalls have 0-1 uploads, 0 new pipelines / programs, and 1600-2300 draws. **(a) One frame, submit dominated by `texture lookups`** (normally ~1-2 ms per frame): ``` vk: slow submit: 164.1 ms: draw 1.1 texture lookups 162.5 vk: slow submit: 101.0 ms: ... texture lookups 99.4 vk: slow submit: 129.6 ms: texture lookups 128.1 vk: slow submit: 163.4 ms: texture lookups 162.4 ``` **(b) A slow `present`, then `texture lookups` on the next frame** (4 of the 11 largest stalls): ``` vk: slow frame at present 2107: 156.3 ms; submit 22.2, draw 7.9, ... 0 uploads vk: slow present: 105.8 ms: vk: slow submit: 55.6 ms: texture lookups 54.5 vk: slow frame at present 2108: 69.7 ms; submit 67.5, ... 0 uploads ``` The same pattern appears at presents 1172 (frame 241.9 ms, submit 46.9, present 169.1), 1588 (104.6 / 15.5 / 83.8, next frame lookups 78.5) and 1833 (135.2 / 24.1 / 103.9, next frame lookups 37.5). My reading: one blockage of ~150-230 ms that lands in whichever host call is running. **`HOS_PROFILE=1`** (88 s, Castle Town market; `report.md` and the profile folder attached): - 5469 frames, 62.1 fps, p99 46.4 ms, p99.9 181 ms. 47 frames >= 50 ms (16 over 100 ms). - After the load-in (from 25 s), the 8 largest stalls (106-194 ms) are all in `tex_lookup_ms` (80-166 ms) or `tex_frame_ms` (168 ms, with `wait_render_ms` 164 ms). They have 0 or 1 uploads, so this is not upload volume. - During these stalls `main_blocked_ms` is 79 ms on average (12.6 ms in steady frames) while `main_vcore_ms` stays at 0.06 ms. The main thread is not waiting for a virtual core. - The sampler shows `sead::MainThread` waiting in `libc.so.6` 100 % of the time, and in most sweeps no game thread is on the CPU. The wait is in a lock or the driver. `longshot_vk.so` offsets are unresolved on my side, so I cannot name the call. - Other game log lines: `slow upload 59.2 ms: ... 512x512 ... ASTC, 236 KB`. **`perf`** (other session): `mkw_vk display` blocked 122.7 ms at 159.14 s and 169.4 ms at 163.66 s (4.5 s apart). ### 3. Not the system - `cyclictest` (30 s): max latency 6-157 µs. - `perf sched`: no game thread waits more than ~5 ms for a CPU. - `turbostat`: SMI = 0, no thermal or power limit. - No swap-in or disk read at stall times. - During a stall CPU load and frequency drop: the game is waiting, not starved. - Other Vulkan games and `vkcube` run fine on the same machine. ### 4. Tried without effect | Test | Result | |---|---| | V-Sync On/Off, 30 fps cap, minimum graphics, windowed, Openbox (no compositor), compositing on/off, FG on/off, DLSS on/off | no change | | Kernels 7.0.0-43 / 6.8.0-149, NVIDIA 595.99.02 / 615.71.09 | no change | | NVIDIA shader disk cache off, Wi-Fi/Bluetooth off, SDL audio drivers | no change | | Controller: wired, Bluetooth, and no controller at all | no change | | Monitoring tools stopped (`nvidia-smi`, `pidstat`, MangoHud) | no change | | `RX_VK_PRESENT_THREAD=1`, threaded rendering off | no change | | `RX_VK_COMPACT=0`, `HOS_TEX_COMPACT_MS=off` / `10000` | no measurable change | | `RX_VK_UPLOAD_BUDGET_MS=10`, `RX_VK_ASTC_RGBA8=1` | no change | | `HOS_CORES` 8 / 10 / 12 | no change | | `HOS_RESERVE_CORE=0` | worse | The sparse-mirror destruction fix described in your docs is active here (`0 destroyed`). ### 5. Thread-level observations - **Virtual-core waits** ("vcores: X waited N ms for a core ... never asked") are frequent: 94 waits >= 50 ms in the 88 s profile, and in another run 128 in 86 s (65 >= 100 ms), with chains `NavMeshCAStepDepWorker1` → `tera::FileResourceMgr` → `PreActorModelCompose` → `Porter` on core 1. Only 7 of the 94 overlap a stall, and none of the 11 largest stalls has one within ±4 log lines (14 % expected by chance). The waiting threads are loaders (`tera::FileResourceMgr`, `PreActorModelCompose`, `GeneralFormatAsync_`), blocked while `sead::MainThread` holds the core. I read this as a consequence of the main thread being blocked rather than the cause. - **Replay thread (`mkw_vk replay`)**: scheduled ~3000 times per second (~100 per frame, a `sched_yield` spin around `Recorder::wait_idle`). In a 45 s `bpftrace` capture with stalls it showed no off-CPU wait >= 40 ms. A 40 s `perf record` shows 60 % on CPU, no continuous on-CPU stretch >= 80 ms, longest gap 85 ms. It appears to wait for work rather than cause the stall; the in-game overlay shows "Waiting on: Game's main thread" rising during the holes. This is hard to reconcile with `tex_lookup_ms` of 165 ms. - **`bpftrace` on all threads, 58 s, no controller connected**: of all `ioctl` / `mmap` / `mprotect` / `munmap` / `mremap` / `madvise` calls, only one lasted >= 40 ms (a 42 ms `ioctl` in `mkw_vk replay`). So there is no long NVIDIA ioctl or memory syscall during the stalls. - **`futex` waits >= 60 ms in `Presentation Th`** (60-101 ms) occur at seconds 3, 13, 18, 22, 27, 32 and 46 of the capture (period fit 4.76 s, R = 0.94, 1 s clock resolution). At 3, 22 and 46 s, `sead::MainThread` (52-63 ms) and `mkw_vk recorder` (41 ms) also wait at the same moment, so the whole pipeline stalls at once. The larger events fall on the same ~4.76 s grid every 4-5 cycles. The main replay thread never waits >= 40 ms in `futex`; the presentation thread is starved while the replay does not sleep. - **Guest thread `OptimizeChangeBanc`**: in a separate 37 s `perf` capture it was the only thread with CPU bursts >= 80 ms: 94 ms at t = 22.05 s and 203 ms at t = 26.51 s, 4.46 s apart (hot function `f_805aeea0`, 18 % of its samples). In the profile above it is active in 2 of the 8 largest stalls. Two events are too few to conclude. - **Render-target churn is high** in this scene: 7591 targets and 5271 MB "let go" in the first run, while the pool holds 7-12 MB of 768 MB. ### 6. Side note (not the same problem) A performance power profile (`powerprofilesctl set performance`, `min_perf_pct=100`) cuts replay CPU per frame from 15.8 to 10.5 ms and 50-65 ms frames from ~85/min to ~35/min (30 fps cap). It does not remove the periodic 100-200 ms stalls. ## Questions 1. Is a ~4.6 s cadence (SD ≈ 0.5 s), tied to camera movement, known? Is there a periodic job in the game or the host whose cost grows with the scene (map placement optimisation, navmesh update, actor streaming, render-target / texture release, descriptor or sampler rebuild, a timer)? 2. What does the main thread wait for during these stalls? The sampler only shows `libc.so.6`. Is there a switch, a log or a Tracy build that names the object it waits on? Could the log break down `slow present` (acquire vs queue present) and `texture lookups` (per-call time)? 3. How can `tex_lookup_ms` reach 165 ms with no upload? Could `present` and `texture lookups` be waiting on the same lock shared with a game thread, or on the same driver operation? 4. In the `perf` stacks the replay thread is often in `Recorder::spin<…wait_idle…>` / `sched_yield` before long gaps. What can keep it waiting 100+ ms there, and what is the recorder doing at that time? 5. I tested Nightly 649 (846e5399) and the stalls are unchanged. Are #994 (render-target reuse), #838 (render thread leaving its core) and the occlusion virtual-core fix included in that nightly? If not, is there a branch or build with them that I can test? I can also run a debug build or any switch you suggest, and report. ### Where in the game Kokiri Forest, Town Market, while moving the camera ### How often Every time ### Graphics card and driver version RTX 3060 12 GB, NVIDIA 595.99.02 and 615.71.09 (same result) ### CPU, RAM and operating system i5-12400F (6 cores, 12 threads), 32 GB DDR4, Linux Mint 22.3, X11 (Cinnamon; Openbox without compositor also tried) ### Changes from the defaults Frame generation: Auto (Built-in FG) (default: Off) Frame generation by: Built-in FG (default: FSR 3.1 FG) Fullscreen: On (default: Off) GPU device: 0: NVIDIA GeForce RTX 3060 (default: Auto) Volume: 50% (default: 100%) FPS counter: Dev (default: Off) Menu size: Small (default: Medium) Depends on this PC (the values now, not counted as changes): Quality preset: Auto (this PC) (default depends on this PC) Output resolution: 1920x1200 (native) (default depends on this PC) Profile: Docked (default depends on this PC) Upscaler: NVIDIA DLSS (default depends on this PC) Render resolution: Quality 67% (1280x800) (default depends on this PC) Sharpness: 40% (default depends on this PC) Window scaling filter: Not needed (window = output) (default depends on this PC) Game speed: Real time (default depends on this PC) Frame pacing: On (default depends on this PC) Switches set: LONGSHOT_NO_UPDATE_CHECK=1 SDL_VIDEO_WAYLAND_WMCLASS=(value left out) SDL_VIDEO_X11_WMCLASS=(value left out) ### longshot_log.txt _No response_ ### Screenshots or video _No response_ ### Before you send it - [x] The problem is still there on the newest release or nightly. - [x] I searched the open issues for it. - [x] I'm not attaching game files, keys, the game's executable, its .bin caches or saves.
Member

Possibly related, for context:

  • issue #1268 Game: 200-470 ms stutters when many new textures arrive: the main thread waits ~0.3 s on a GPU fence after a frame of uploads (RTX 3090, Linux) — Linux NVIDIA stalls with main-thread blocking in renderer; this report has no uploads and a different cadence/call path
  • issue #1639 Game: Inconsistent performance and severe stuttering in Hyrule Castle Town / Temple of Time — Broader stutter in Castle Town and Kokiri, but different platform and no established matching cause
  • issue #1521 Over half of the field's hitches have the cause "other", and "game" is only the time nothing else claimed — Performance issue discussing attribution of main-thread waits and frame stalls
  • PR #1553 perf(runtime): reduce Market frame stalls and Vulkan capture overhead — Merged Market stall and Vulkan capture improvements; not shown to fix these periodic stalls
  • issue #1831 Game: stutter at the Castle Town fountain on quick direction changes: Hyrule Field and Hyrule Castle are loaded ahead and dropped again at every turn (related: #996, #1639) — Fountain-area movement-triggered stutter, but linked to repeated adjacent-area loads rather than this issue's camera-dependent periodic stalls
Possibly related, for context: - issue #1268 Game: 200-470 ms stutters when many new textures arrive: the main thread waits ~0.3 s on a GPU fence after a frame of uploads (RTX 3090, Linux) — Linux NVIDIA stalls with main-thread blocking in renderer; this report has no uploads and a different cadence/call path - issue #1639 Game: Inconsistent performance and severe stuttering in Hyrule Castle Town / Temple of Time — Broader stutter in Castle Town and Kokiri, but different platform and no established matching cause - issue #1521 Over half of the field's hitches have the cause "other", and "game" is only the time nothing else claimed — Performance issue discussing attribution of main-thread waits and frame stalls - PR #1553 perf(runtime): reduce Market frame stalls and Vulkan capture overhead — Merged Market stall and Vulkan capture improvements; not shown to fix these periodic stalls - issue #1831 Game: stutter at the Castle Town fountain on quick direction changes: Hyrule Field and Hyrule Castle are loaded ahead and dropped again at every turn (related: #996, #1639) — Fountain-area movement-triggered stutter, but linked to repeated adjacent-area loads rather than this issue's camera-dependent periodic stalls
Sign in to join this conversation.
No milestone
No project
No assignees
2 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
MagicBeanDev/Longshot#1988
No description provided.