You are not logged in.
Pages: 1
My Ryzen build (specs listed in the sig, if that makes a difference to anyone) has been seeing this problem lately, which for a while wasn't that bad but it seems to be slowly getting worse. The problem is every once in a while, a USB device will register a signal twice. For example, if I type the letter D on the keyboard, it will sometimes type "dd" or even "ddd". Sometimes if I play a game and press a button, it'll toggle the action twice (which, depending on the game, can get really annoying). So far it's only really been noticeable with gamepads and keyboards - my mouse (plugged into the keyboard's built-in hub) doesn't appear to have the same problem, but I click the mouse a lot less than I type keys or press gamepad buttons, and I don't think I'd notice this "glitch" by moving the pointer.
Plugging the devices in other USB ports doesn't seem to help, and although I didn't try another keyboard, I did try another gamepad, which had the same problems. I regularly use about 4 different computers and this is the only one with this problem, so I don't think it's a config issue or just shaky fingers.
I would like to point out my CPU is overclocked, but, I have confirmed that the OC is stable. I also didn't tweak the BCLK, specifically because doing that is known to cause USB problems.
Last edited by schmidtbag (2018-09-25 12:58:29)
Offline
What makes you believe that this is a problem w/ the bus and not eg. w/ libinput?
Do you have input devices that are not wired via USB (ps/2??) and thus don't expose this behavior?
Do you have trouble w/ non-input USB devices (IO errors on drives, printers etc.)?
Do you get the double-entry in "sudo libinput debug-events" or "sudo showkey" on the console (isolate the multi-user.target for the latter test)?
Offline
What makes you believe that this is a problem w/ the bus and not eg. w/ libinput?
Do you have input devices that are not wired via USB (ps/2??) and thus don't expose this behavior?
Do you have trouble w/ non-input USB devices (IO errors on drives, printers etc.)?Do you get the double-entry in "sudo libinput debug-events" or "sudo showkey" on the console (isolate the multi-user.target for the latter test)?
Thanks for the quick reply
I actually don't think this is specifically a problem with the bus; I legitimately don't know what the problem is, so it very well could be libinput.
I do have a PS/2 keyboard I can try; I haven't done so yet.
So far, other USB devices appear to be working fine (including USB HDDs and flash drives). However, I'm not sure if that's a good metric, since I'm sure there's some error correction involved in that.
I don't have a printer plugged into this computer, but since this problem is so intermittent (it only happens maybe once every 2 or so minutes) I think the chances are pretty slim that this could affect such a device.
I might also try using a webcam and see if I can notice any corruption issues. The tricky part with that is any corrupt pixels would likely only show up in a random spot for a single frame, which would be very hard to detect (again, considering how intermittent this problem is).
Last edited by schmidtbag (2018-09-25 14:02:25)
Offline
The problem could also be in *hci_hid (usb storage and input use different kernel modules)
If you also have a PS/2 keyboard (and actually still a bus - a ps/2-usb converter won't help ;-), that should be a good indicator (if it happens there as well, it's not USB) and showkey would narrow the thing from the other side (if it doesn't happen there, it's rather libinput)
Offline
Haha yeah, of course I couldn't use a PS/2 to USB adapter, but my motherboard does support 1 built-in port, so I'll give that a shot.
But, suppose the PS/2 keyboard proves to behave normally: what do I do next?
Offline
Pages: 1