gpd wini mini controller and bluetooth disappears randomly.
not sure when it started, but I frequently to keep things up to date... I have looked at journalctl and sudo dmesg and the results show that xhci_hcd stops working, and says Hc Died cleaning up.. Any help appreciated as I really enjoy using bazzite for both productivity and gaming 🙂
74 Replies
Result is Bluetooth drops out? Asking to see if I'm having this issues too.
yes it seems USB xhci_hcd literally dying... and unable to rmmod/modprobe the driver since it is "built in"
I am pretty certain it has something to do with one of the updates... I have since applied again in the hope the bug is resolved
does it happen randomly?
For me yes.
cant pin on which update did it, but one of them has caused instability in the usb controller... testing in case of hardware by using windows... so far used windows for 8 hours with none of these issues reocuring
not sure if this will help your case, but rebasing to a bazzite version that uses kernel 6.12 has had so far fixed issues with usb drop controller dropouts
hmm notice toasty has posted on #🛟bazzite-help regarding disconnecting usb devices
There was a large USB patch series on 6.13
wonder what advantages the patch entices?
the thing is... if u dont have a secondary os to go to, the symptoms seems to point to a hardware failure
Is only win mini affected or is it also the win 4
How to make this happen fast
not sure, i only have a mini.... my next gpd device is the old win2
How to make it happen fast
the only way is to connect, discponnect mains power, and use sleep mode heaps
i suspect it may have something to do with power profiles for the usb controller when connected to power, vs running on battery
somehow it is "overloading" it and causing it to force stop and powerdown
how long does it take for it to happen
my mini is a 7840u mini.... is ur win4 a 7840u?
8840u
it is random... one time it happened in under an hour... antoher time it happened hours later... initially it could take days
this will be a pain and a half
can you make it happen fast?
if u resume and sleep multiple times, with power cable connected and disconnected, u may be able to simulate.... i also have xr glasses, and i have docks which i connect to.... adding more usb devices!
before i realised it was software, i just thought it was me over using bazzite in a single session, just thought bazzite was not stable
6.12 feels rock solid compared to 6.13...
@antheas https://www.reddit.com/r/Fedora/comments/1jdtj29/usb_system_crashing_on_kernel_613/ like it is definitely not exclusive to bazzite
Fedora Discussion
USB controller crashing
I’ve run into a situation where my USB devices all fail, typically in attempting to return from sleep. I’ve been trying to troubleshoot by eliminating devices and rolling back to an earlier kernel version (6.13.4-200.fc41.x86_64) with limited success. Today, the USB dropped out while the system was awake and in use (streaming a video). When...
seems to be affecting arch too https://bbs.archlinux.org/viewtopic.php?id=303615USB "devices randomly stop working until system is rebooted"
In order to fix this, we need to make it happen fast
might pay to put a hold on other bazzites upgrading to 6.13
just until issue is resolved
as i said it is really hard to replicate... suspect it has to do with power settings powered vs battery
its a bug on the usb controller
also put the full log here so i have it
did u want me to rebase to the latest again?
it would be good
and to make it happen faster
and sad lol
it has been nice having a rock solid system
i will need quite specific instructions on how to tail the log into file...
use the bug reporter from hhd
and do not tail
so like logs from current boot
yes
and hit submit?
and tell u when i do it?
send the link
when does it happen? after sleep?
it is random
you cut the important stuff here
after a sleep it can be working for like an hour, or more, and randomly suddenly the game controller will just not work
yup that is what i cut out immediately after i noticed it was broken
not > you see thisi think one time it happened without a sleep resume in between
you cut the right part
yes
literally the right half is missing
damn
terminal window truncated it
do
lspciis there a way to see that part of the log still?
if you find it yes
but do lspci for me
when it happens?
no
now
i need your controlelr id
do lsusb too
haha u can see i am using my viture pro xr glasses
lspci -nnso 1022:15b9 is being problematic after 6.13
my win 4 has it too
where do i find that 1022;15b9 value?
dont worry about it
is ur win4 a 7840u? if it is, more than likely the hardware will be very similar
8840u
hmm usb hardware should be the same... 7840u and 8840u have same socket or pin layouts
yes but i am also not stressing the hell out of mine
get a full log this time
there is a commit with a quirk
we can apply it here too
by using HHD?
yes or journalctl -b | fpaste
what is fpaste?
did u want me to ujust update now?
man
last version
make it happen
run
journalctl -b | fpastedid u want me to run that now?
no
when it happens
https://lore.kernel.org/all/20241206074817.89189-1-kaihengf@nvidia.com/ seems like this could be the issue
ok
ujust updating now
think i did one yesterday in the hopr there would be an update fix
Make sure it works because if you rebased you'll be on the last version
?
You need to run the rebase command with stable
oh i did it wrong?
i should have just rebased to the previous one
can i cancel out, or is it dangerous to do so?
ok done
rebooted, still on 6,12 wil need to rebase like u said
rebasing to 41.20250325
can i assume that is how i get onto the stable?
Just write stable
Rebase stable
i am in doomsday mode lol
i just rebased to the last one.. but will do ?
yes
hope it does it soon
@antheas something strange. i have done multiple sleeps, gamed on my device more than usual, and used xr glasses, connected to multiple power sources and docks, but so far usb has not fallen over.... will keep u posted
not sure what is happening.. been using the min periodically over a day.. bluetooth and gamepad has not stopped working.. in the past did not fix it even though the revision was updating, but reverting to 6.12, ujust updating, then rpm-ostree reolling back to 6.13, then switching to stable seems to have fixed it?
it is almost like i had a bad rollback someone in the past
Testing kernel has a potential fix
ok