Game: rumble continues 5-10 s after releasing a charged sword spin (Y) (Linux, nightly-627 + PRs) #1924

Closed
opened 2026-11-20 21:58:49 +00:00 by nidrix_glandda · 2 comments

What happens

Holding Y to charge the sword spin and then releasing it leaves the controller rumbling for about 5–10 seconds after the spin and its sound have ended.

Build

v3.0.0-n627-g93e8237d, built from nightly-627 afc4d1ad + #1642, #1870, #1893, #1914, #1909/#1911 (off), Linux, RX 9070 XT. bin/ holds libSDL2-2.0.so.0, longshot, longshot_vk.so — no renhdrum.so, so this is the SDL motor path (HOS_RUMBLE_BACKEND auto resolves to motors for the pads here).

Settings and pads

longshot_options.ini: vibration=0 (Auto), rumble_strength=100. Controllers connected during the session: Flydigi Vader 4 Pro (SDL, gyro) and two Xbox userspace pads; which one rumbles needs confirming.

What I would look at

host/pad_sdl.c sends SDL a 1 s rumble effect and re-sends it every 0.5 s while the amplitude is nonzero; a change to zero is sent at once. A 5–10 s tail therefore means either the game keeps sending a nonzero value after the release (the vibration handle/pattern is not being stopped), or the zero does not reach this pad. The hid: vibration #… lines (first six) and the every-4096 hid: … vibration values, strongest … line should say which.

Worth testing: HOS_HD_RUMBLE=0 and HOS_RUMBLE_BACKEND=motors vs hd — I can do that here when the PC is free and add the log.

Expected

The rumble stops when the charged spin ends.

I have no log from the session yet; I can capture one with the switches above on request.

### What happens Holding **Y** to charge the sword spin and then releasing it leaves the controller **rumbling for about 5–10 seconds** after the spin and its sound have ended. ### Build `v3.0.0-n627-g93e8237d`, built from **nightly-627 `afc4d1ad`** + #1642, #1870, #1893, #1914, #1909/#1911 (off), Linux, RX 9070 XT. `bin/` holds `libSDL2-2.0.so.0`, `longshot`, `longshot_vk.so` — **no `renhdrum.so`**, so this is the SDL motor path (`HOS_RUMBLE_BACKEND` auto resolves to motors for the pads here). ### Settings and pads `longshot_options.ini`: `vibration=0` (Auto), `rumble_strength=100`. Controllers connected during the session: Flydigi Vader 4 Pro (SDL, gyro) and two Xbox userspace pads; which one rumbles needs confirming. ### What I would look at `host/pad_sdl.c` sends SDL a **1 s** rumble effect and re-sends it every 0.5 s while the amplitude is nonzero; a change to zero is sent at once. A 5–10 s tail therefore means either the game keeps sending a nonzero value after the release (the vibration handle/pattern is not being stopped), or the zero does not reach this pad. The `hid: vibration #…` lines (first six) and the every-4096 `hid: … vibration values, strongest …` line should say which. Worth testing: `HOS_HD_RUMBLE=0` and `HOS_RUMBLE_BACKEND=motors` vs `hd` — I can do that here when the PC is free and add the log. ### Expected The rumble stops when the charged spin ends. I have no log from the session yet; I can capture one with the switches above on request.
Member

Possibly related, for context:

  • PR #1888 build(renhdrum): RenHDrum 0.5.0, its DualSense HD Haptics, and RenHDrum in Auto for every pad it plays — Merged rumble backend selection change; symptoms may depend on whether Auto selects RenHDrum or SDL motors
Possibly related, for context: - PR #1888 build(renhdrum): RenHDrum 0.5.0, its DualSense HD Haptics, and RenHDrum in Auto for every pad it plays — Merged rumble backend selection change; symptoms may depend on whether Auto selects RenHDrum or SDL motors
Author

Update, and closing as not a confirmed bug: with the game's Vibration row Off the stuck rumble does not happen. Controller: Flydigi Vader 4 Pro (SDL, gyro). I filed this without a reproduction or log, which I should not have done. If it is reported again with Vibration on, it should come with the hid: vibration lines and the HOS_HD_RUMBLE=0 / HOS_RUMBLE_BACKEND=motors comparison. The workaround for now is Vibration Off.

Update, and closing as not a confirmed bug: with the game's **Vibration row Off** the stuck rumble does not happen. Controller: **Flydigi Vader 4 Pro** (SDL, gyro). I filed this without a reproduction or log, which I should not have done. If it is reported again with Vibration on, it should come with the `hid: vibration` lines and the `HOS_HD_RUMBLE=0` / `HOS_RUMBLE_BACKEND=motors` comparison. The workaround for now is Vibration Off.
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#1924
No description provided.