Summary: | Linux detects invalid USB Interfaces as Joypads | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Andre Ryan <andreryan908> |
Component: | [OLD] Core system | Assignee: | Gentoo Kernel Bug Wranglers and Kernel Maintainers <kernel> |
Status: | RESOLVED UPSTREAM | ||
Severity: | normal | ||
Priority: | High | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | Patch to fix joydev module in kernel |
Description
Andre Ryan
2009-05-16 09:09:49 UTC
Created attachment 191472 [details, diff]
Patch to fix joydev module in kernel
After fiddling around with the Kernel sources for a while, I tried to modify the USBHID subsystem (hid-core.c) to ignore devices that have no sub class and protocol which did seem to fix the joystick issue at first but I then discovered that the interface does have an actual purpose - to provide the Multimedia keys on the keyboard (play, volume up/down/mute) so that didn't help as much I'd have liked.
The real problem ultimately comes down to the joydev module being a little overenthusiastic about what constitutes a joystick, I added a new blacklist entry that seems to have done the trick (My xpad360 still works at least), patch is attached.
(In reply to comment #1) > Created an attachment (id=191472) [edit] > Patch to fix joydev module in kernel > > After fiddling around with the Kernel sources for a while, I tried to modify > the USBHID subsystem (hid-core.c) to ignore devices that have no sub class and > protocol which did seem to fix the joystick issue at first but I then > discovered that the interface does have an actual purpose - to provide the > Multimedia keys on the keyboard (play, volume up/down/mute) so that didn't help > as much I'd have liked. > > The real problem ultimately comes down to the joydev module being a little > overenthusiastic about what constitutes a joystick, I added a new blacklist > entry that seems to have done the trick (My xpad360 still works at least), > patch is attached. > We don't usually accept user-made patches that were not accepted upstream, so you will have to take this to the linux-input mailing list: http://vger.kernel.org/vger-lists.html#linux-input Thank you! Yep, please get this accepted upstream... Sounds like a purely cosmetic issue if no problems are actually being caused? |