Re: usbconfig / hal-device no longer lists usb devices

On Sunday 22 March 2009 05:24:01 Andrew Thompson wrote:
> On Sun, Mar 22, 2009 at 01:52:54AM +0100, Pieter de Goeje wrote:
> > On Sunday 22 March 2009 01:03:31 Andrew Thompson wrote:
> > > On Sun, Mar 22, 2009 at 02:52:33AM +0300, Anonymous wrote:
> > > > > I added a bunch of printf()s to libusb, specifically
> > > > > ugen20_enumerate(). Both ugen0.2 and ugen1.2 failed at ioctl(f,
> > > > > USB_GET_PLUGTIME, &plugtime) because it returned EINVAL. The
> > > > > ugenX.2 files were opened successfully.
> > > > >
> > > > > At this point it looks like the problem lies somewhere in the
> > > > > kernel, which makes it a lot harder for me to debug.
> > > >
> > > > Can you try to back out r189906? Doing so makes my keyboard to appear
> > > > in usbconfig output again. Here is a ktrace diff for `usbconfig -u 0
> > > > -a 3'.
> >
> > I'll give it a shot. It's rather late so the results will probably have
> > to wait until tomorrow.

That didn't make a difference.

> >
> > > What does sysctl hw.usb2.dev.debug=2 show with usbconfig on the latest
> > > HEAD code?
> >
> > Output is quite long so I put it on the web:
> > http://unforgiven.student.utwente.nl/~pyotr/dump/usbconfig_debug_2.txt
>
> I cant see anything obviously wrong with r189906 so I have committed
> some more debugging traces. Can you update your sources/kernel and run
> usbconfig again with hw.usb2.dev.debug=5

http://unforgiven.student.utwente.nl/~pyotr/dump/usbconfig_debug_5.txt
At line 494 & 511 the ioctls fail.

I also did what HPS asked, I manually copied the usb headers, but that didnt 
help either.  (I deleted /usr/obj and ran a full buildworld / buildkernel / 
installkernel / installworld before that).

-- 
Pieter de Goeje