Possible bug in USB or HID on asus mobo with via kt266 a chipset

Possible bug in USB or HID on asus mobo with via kt266 a chipset

[Michael West]

    I recently changed motherboards on my linux box and one of my hid 
controllers (a psx-usb converter) stopped functioning correctly.  I was 
running a 2.4.18 kernel on both boards, and with the new asus board, 
apps reading from the /dev/input/js0 file seem to halt after the first 
19 joystick messages are read.  I tried reproducing the problem on other 
kernel versions, and experienced the same problem with a smattering of 
previous kernels.  I'm using a hid mouse, as well as another hid 
controller, and both work correctly.  Not sure if its related or not, 
but I also seem to have some apparent irq problems, as newly plugged in 
usb devices (any) and by that I mean after the usb-uhci or uhci driver 
is loaded, throw "USB device not accepting new address - * (error = 
-110)" errors.  The situation in 2.4.19-pre2 changed a bit by completely 
breaking the psx-converter (joydev driver assigns no device) only on 
usb-uhci.  pre3 has the same origional problem, as well as 4 and 5. 
 Pre6 and Pre7 seem to completely break all usb hid devices.  The irq 
(or whatever) problems go away and devices are hotplugged fine, but no 
hid devices are ever registered.  Sorry for my infamiliarity with the 
linux kernel source and terminology.  Thanks in advance.
 Michael

Re: Possible bug in USB or HID on asus mobo with via kt266 a chipset

[Johannes Erdfelt]

On Thu, Apr 18, 2002, Michael West <neovorbis@comcast.net> wrote:
>     I recently changed motherboards on my linux box and one of my hid 
> controllers (a psx-usb converter) stopped functioning correctly.  I was 
> running a 2.4.18 kernel on both boards, and with the new asus board, 
> apps reading from the /dev/input/js0 file seem to halt after the first 
> 19 joystick messages are read.  I tried reproducing the problem on other 
> kernel versions, and experienced the same problem with a smattering of 
> previous kernels.  I'm using a hid mouse, as well as another hid 
> controller, and both work correctly.  Not sure if its related or not, 
> but I also seem to have some apparent irq problems, as newly plugged in 
> usb devices (any) and by that I mean after the usb-uhci or uhci driver 
> is loaded, throw "USB device not accepting new address - * (error = 
> -110)" errors.  The situation in 2.4.19-pre2 changed a bit by completely 
> breaking the psx-converter (joydev driver assigns no device) only on 
> usb-uhci.  pre3 has the same origional problem, as well as 4 and 5. 
>  Pre6 and Pre7 seem to completely break all usb hid devices.  The irq 
> (or whatever) problems go away and devices are hotplugged fine, but no 
> hid devices are ever registered.  Sorry for my infamiliarity with the 
> linux kernel source and terminology.  Thanks in advance.

Ahh, VIA. Their USB chipsets have some problems. Do you happen to get a
BABBLE sometime before you start getting the error = -110 messages?

JE

Re: Possible bug in USB or HID on asus mobo with via kt266 a chipset

[Greg KH]

On Thu, Apr 18, 2002 at 07:10:42PM -0400, Michael West wrote:
>    I recently changed motherboards on my linux box and one of my hid 
> controllers (a psx-usb converter) stopped functioning correctly.  I was 
> running a 2.4.18 kernel on both boards, and with the new asus board, 
> apps reading from the /dev/input/js0 file seem to halt after the first 
> 19 joystick messages are read.  I tried reproducing the problem on other 
> kernel versions, and experienced the same problem with a smattering of 
> previous kernels.  I'm using a hid mouse, as well as another hid 
> controller, and both work correctly.  Not sure if its related or not, 
> but I also seem to have some apparent irq problems, as newly plugged in 
> usb devices (any) and by that I mean after the usb-uhci or uhci driver 
> is loaded, throw "USB device not accepting new address - * (error = 
> -110)" errors.  The situation in 2.4.19-pre2 changed a bit by completely 
> breaking the psx-converter (joydev driver assigns no device) only on 
> usb-uhci.  pre3 has the same origional problem, as well as 4 and 5. 
> Pre6 and Pre7 seem to completely break all usb hid devices.  The irq 
> (or whatever) problems go away and devices are hotplugged fine, but no 
> hid devices are ever registered.  Sorry for my infamiliarity with the 
> linux kernel source and terminology.  Thanks in advance.

What is your .config for -pre7?  And what does the kernel log say when
your devices are recognized.  And what does /proc/bus/usb/devices say
when your devices are plugged in?

thanks,

greg k-h

Re: Possible bug in USB or HID on asus mobo with via kt266 a chipset

[Michael West]

Greg KH wrote

>What is your .config for -pre7?  And what does the kernel log say when
>your devices are recognized.  And what does /proc/bus/usb/devices say
>when your devices are plugged in?
>
>thanks,
>
>greg k-h
>

To all whom are interested: I posted my .config file and 
/proc/bus/usb/devices files to ftp://ranma.wox.org/devices.txt and 
ftp://ranma.wox.org/config.txt.  I forgot to check the kernel log, and 
I'm tired of recompiling kernels.  I had origionally planned to switch 
motherboards and cpus with my brother, as I was putting a new comp 
together for him but, out of frustration with my current dilemma, have 
decided to switch back (the amd seems unstable anyway).  Regardless, I 
will not be able to follow up on this bug report.  And to Johannes, no, 
I don't think I've ever even seen a BABBLE error, and I'm not sure what 
that is.