You are not logged in.
When I rebooted after updating to the 3.10 kernel my mouse stopped working. The cursor is displayed but it doesn't move. I downgraded back to 3.9.9-1 and it works fine again. As far as I could see it was detecting it in the dmesg output, so I'm not sure of the cause (maybe just a regression in the driver). I've never had any problems with this mouse before.
The mouse is a Logitech M515 connected wirelessly with a Logitech Unifying receiver.
This is the dmesg line (on 3.9.9):
logitech-djreceiver 0003:046D:C52B.0003: hiddev0,hidraw0: USB HID v1.11 Device [Logitech USB Receiver] on usb-0000:00:14.0-2/input2
Last edited by shaurz (2013-07-25 19:20:53)
Offline
Ahh sorry, looks like this is already a known issue:
Offline
Hmm, mine is working.
$ lsusb | grep -i unify
Bus 001 Device 002: ID 046d:c52b Logitech, Inc. Unifying Receiver
$ uname -a
Linux t1000 3.10.2-1-ck #1 SMP PREEMPT Mon Jul 22 16:14:50 EDT 2013 x86_64 GNU/Linux
maybe I shouldn't update?
It's a very deadly weapon to know what you're doing
--- William Murderface
Offline
It's a problem in all of the 3.10 series so far. There is a patch available but I don't think it's applied upstream yet.
Offline
This is still happening for me with 3.10.3
Where can I get the patchset? I just compiled 3.9.11 to have as a backup till this is fixed.
Offline
Mine works fine with Logitech Anywhere MX, never noticed any problems. Is this still an issue?
According to the bug report, Logitech Anywhere MX should be broken I guess. If you need any info dumped to figure out why it's working for some, tell me what you need.
Kernel:
Linux xarch 3.10.3-1-ARCH #1 SMP PREEMPT Fri Jul 26 11:26:59 CEST 2013 x86_64 GNU/Linux
lsusb:
Bus 007 Device 002: ID 046d:c52b Logitech, Inc. Unifying Receiver
dmesg:
[ 3.332661] logitech-djreceiver 0003:046D:C52B.0003: hiddev0,hidraw0: USB HID v1.11 Device [Logitech USB Receiver] on usb-0000:00:13.0-1/input2
( I have a keyboard attached to the same unifying receiver. Paired them up using a windows laptop. Maybe that's the difference? I guess pairing hardware to the receiver using logitech software could theoretically mess with... uhm... magical internal mapping stuff that I have no idea how it works? Ok, I'll stop guessing making things up now , good luck with the bug :3 )
Last edited by whoops (2013-07-29 10:54:39)
Offline
I've got the same issue with an M705 mouse and K750 keyboard running through a unifying receiver, though I've found an (albeit annoying) workaround (assuming you're on a laptop, I suppose). If you install ltunify from the aur, you can pair and unpair objects. You can also list the connected devices. Oddly enough, upon boot, ltunify will list the two devices as connected...
$ ltunify list
Devices count: 2
Connected devices:
idx=1 Mouse M705
idx=2 Keyboard K750
...however they don't respond. I use my laptop keyboard to unpair both and re-pair them using ltunify unpair/pair commands. After that, devices work fine. Was trying to think if there is a script to run upon boot but ltunify requires that you turn the devices off and on in order to pair, so it's a bit janky. Any suggestions would be great.
And for reference:
$ uname -a
Linux spooky 3.10.3-1-ARCH #1 SMP PREEMPT Fri Jul 26 11:26:59 CEST 2013 x86_64 GNU/Linux
$ lsusb | grep Logitech
Bus 001 Device 002: ID 046d:c52b Logitech, Inc. Unifying Receiver
arch + gnome on dell xps13
Offline
Oooo, pairing tool for linux does finally work? Didn't know that, thx! Last time I had to use a windows laptop.
Hmm... I don't really know if it even makes sense to do something like that... but I'm curious: did you try pairing them the other way around? Or to the other unused unifying controller you probably have around from the second device?
Strangely mine says:
Devices count: 3
Connected devices:
idx=1 Mouse Anywhere MX
idx=2 Keyboard K230
idx=3 Mouse Anywhere MX
This could be, because I once had another Anywhere Mouse that... had a very tragic accident T_T
Anyway: Maybe that has something to do with why my setup works fine / didn't problems at all.
Last edited by whoops (2013-07-30 17:12:11)
Offline
Yep - there are a few tools available, actually. https://wiki.archlinux.org/index.php/Lo … g_Receiver
I suppose I should explore those to see if they are any more conducive to being scriptable. Solaar looks promising.
arch + gnome on dell xps13
Offline
I haven't noticed an issue with 3.10 on my laptop (plugged into 2.0), but I did notice sporadic issues on the desktop (plugged into USB 3.0). Removing the receiver and reinserting or using the USB2 port works for me.
Toggling off/on before pairing is a hardware limitation. What you can try is unbinding/binding the USB receiver itself, for example:
cd /sys/bus/hid/drivers/logitech-djreceiver
dev=$(echo *:*) # assume a single receiver, e.g. 0003:046D:C52B.0003
echo $dev > unbind
echo $dev > bind
I vaguely remember a protocol option to trigger re-enumeration, you might find something in my notes on https://git.lekensteyn.nl/ltunify/ or https://lekensteyn.nl/logitech-unifying.html. (or just reverse apply https://git.kernel.org/cgit/linux/kerne … 56e068887)
Offline
I also have a keyboard and mouse paired:
Devices count: 2
Connected devices:
idx=1 Mouse M515
idx=2 Keyboard K270
Offline
I compiled kernel 3.11-RC3 and the problem was still present with a Logitech 510 mouse. With this kernel unplugging and re-plugging the receiver fixes the problem.
I compiled kernel 3.10.3 and the mouse and receiver were dead dead, no matter what, no replugging
I compiled kernel 3.9.11 which I am currently running, because the mouse still works flawlessly
tomorrow I will examine source code, and see what I can find
Last edited by Super-Nathan (2013-07-31 15:09:59)
Offline
Here is the diff between the two kernels (3.9.11 and 3.10.4) http://pastebin.super-nathan.com/14
I am going to try something real dumb and patch the 3.10.4 kernel with the drivers from 3.9.11. I'll report back
Offline
Unpairing and pairing worked for me (3.10.5-1-ARCH) :
sudo ltunify -d /dev/hidraw0 unpair 1
sudo ltunify -d /dev/hidraw0 pair 10
I've found this trick on this website : http://www.code-corner.de/?p=179
Offline
Quick update: Just upgraded to 3.10.5 and devices are working after one or two unplug/plugs of the unifying receiver. So not perfect, but progress from the manual disconnect/connect of the devices themselves. Worth noting, I had uninstalled ltunify in favor of solaar, given that solaar struck me as more scriptable, if need be (haven't yet). Curious to hear from ltunify users here if they have similar improvement with the kernel update.
And again, for reference:
$ uname -a
Linux spooky 3.10.5-1-ARCH #1 SMP PREEMPT Mon Aug 5 08:04:22 CEST 2013 x86_64 GNU/Linux
arch + gnome on dell xps13
Offline
Quick update: Just upgraded to 3.10.5 and devices are working after one or two unplug/plugs of the unifying receiver. So not perfect, but progress from the manual disconnect/connect of the devices themselves. Worth noting, I had uninstalled ltunify in favor of solaar, given that solaar struck me as more scriptable, if need be (haven't yet). Curious to hear from ltunify users here if they have similar improvement with the kernel update.
For me if I plug and unplug with different ports it will eventually work. I'm not using any unifying software, just the receiver (which like others worked perfectly in all kernels before 3.10).
Offline
Are you sure you don't just need to set the pairing of the mouse to a different "idx"?
As written above, I had 2 mice:
Devices count: 3
Connected devices:
idx=1 Mouse Anywhere MX
idx=2 Keyboard K230
idx=3 Mouse Anywhere MX
The one with idx 3 works. Unless I missed something, all mice that had problems so far were connected to idx1.
Last edited by whoops (2013-08-08 14:54:22)
Offline
DolphinTsar, kernel 3.10.5 does not change the HID code nor does it seem to change relevant things in the USB subsystem.
Besides unbind/bind, what about (as root/sudo):
modprobe -v -r hid-logitech-dj
modprobe -v hid-logitech-dj
Or to force a re-enumeration without reloading the driver:
# should output /dev/hidrawN where N is usually 0
hidraw=/dev/$(cd /sys/bus/hid/drivers/logitech-djreceiver/*/hidraw && echo hidraw*)
printf '\x20\xff\x81\0\0\0\0\0\0\0\0\0\0\0\0' | sudo tee "$hidraw" >/dev/null
Offline
Fedora has a patch which looks relevant: HID-Revert-Revert-HID-Fix-logitech-dj-missing-Unifying-device-issue.patch
Offline
Got two machines with same latest kernel update, works on one, and not the other. Ofcourse, broke on the machine I actually use it. None of the workarounds help ..
Offline
Lekensteyn, your latest suggestion to re-enumerate without reloading the driver worked very well.
I'm using openbox so I simply added:
printf '\x20\xff\x81\0\0\0\0\0\0\0\0\0\0\0\0' | tee /dev/hidraw0 >/dev/null &
to my ~/.config/openbox/autostart file. Now it works on every boot.
Offline
Can you guys please mention what port the USB receiver was plugged into? USB 2.0, USB 3.0 or something else?
Offline
On my part is a usb 2.0 port
Offline
@cellisten, are you sure? If yes, please generate some debugging details:
1. Kernel version, uname -r
2. Output of: dmesg > dmesg.txt
3. Output of: lsusb -v > lsusb.txt
4. Output of: lsmod > lsmod.txt
5. usbmon capture from the plug attempt.
To get (5), you have to:
1. Determine which bus the USB receiver is connected to. See https://lekensteyn.nl/logitech-unifying … epare-qemu for an example.
2. Unplug the USB receiver.
3. As root, run:
mount -t debugfs none /sys/kernel/debug
cat /sys/kernel/debug/usb/usbmon/5u > usbmon.txt
4. Confirm that the mouse/keyboard is not working (press a key, move/click).
5. Ctrl+C to stop the usbmon capture.
If the mouse is working, cancel the usbmon capture and restart from step 3 (or 2 if the receiver was already plugged) until you can get a successful trace. If the device keeps working, remove the receiver, reboot and try again. You can mail the attachments to me at lekensteyn@gmail.com or create a new bug on https://bugzilla.kernel.org/
Offline
I seem to have the same problem: I'm trying to install Arch from the latest install disk (which runs 3.10.3). I get through the boot menu, but then when Arch boots up, I have no keyboard (Logitech K800, has a unifying receiver). I've tried both USB 3 and USB 2 ports, no difference. The keyboard works when I boot off the latest (01/08/2013) Gentoo minimal install CD.
Without a keyboard, it's hard to generate debugging output (I could get some by booting from the Gentoo disk, but would that help?) However, when I went into Hardware Info (from the Arch boot disk) I noticed that the USB controllers had no "module" entry.
Is this a problem with the pre-compiled kernel or the kernel source? I'm happy to compile a new kernel once I have a running system. I could just borrow a keyboard until I get it done.
Offline