I tried a quick hack to turn it on, and it didn't seem to work.
Specifically, in release_uhci, I added the line
pci_write_config_word(uhci->dev, USBLEGSUP,USBLEGSUP_SMIENABLE);
right before
free_uhci(uhci);
where I changed uhci.h to contain
#define USBLEGSUP_SMIENABLE 0x0010 /* enable SMI# generation */
>From the documentation I read, it looked like setting this bit returns
control to the BIOS. I also added some printk()s to make sure this
function was being called when I removed the module. It was, but didn't
free up the keyboard.
Also, one time upon removing the driver I got the following message:
eca0: host controller halted. very bad
repeated twice. Re-inserting and removing the module seemed to work
fine, though, and I couldn't get the message to repeat.
Finally, I managed to borrow a USB keyboard temporarily. That seemed to
work, regardless of whether I had BIOS emulation turned on or not.
Ultimately the solution may be to get a USB keyboard, but I'd like to
get this one to work if I can.
-A.J.
On Tue, 2002-12-17 at 19:52, Johannes Erdfelt wrote:
> On Tue, Dec 17, 2002, A.J. Aranyosi <[EMAIL PROTECTED]> wrote:
> > I've recently picked up a Dell Dimension 8250, and I'm having some strange
> > USB problems in Linux (Gentoo). Specifically, whenever I insert either
> > the uhci or usb-uhci module, my PS/2 keyboard freezes up. If I log in
> > remotely I can see that the module initialized the usb mouse properly, but
> > removing the module doesn't free up the keyboard. The freezing happens
> > regardless of whether I have the mouse plugged in or not, and regardless
> > of whether I have legacy emulation support turned on or off in the bios.
> > I've tried it with vanilla 2.4.18, 2.4.20, and 2.5.51 kernels, and with
> > Gentoo's modified 2.4.19 kernel, all with the same results. The usb
> > chipsets (there are four of them, plus two usb 2.0 ones) are VIA UHCI (rev
> > 80) (the usb 2.0 ones are rev 81), and the motherboard chipset is an Intel
> > 82850. Any idea what might be causing this? I'm planning to poke around
> > in uhci.c to see if I can figure out where the freeze happens. Any
> > pointers on where to focus my search? Thanks,
>
> That's interesting. The UHCI drivers turn off legacy emulation when they
> start because we take over the HC. I don't think either turn it back on
> when they are removed.
>
> Maybe the integration on these chips mix the USB and PS/2 functionality
> and there's some sort of hidden dependency or bug there?
>
> Can you try turning on legacy emulation again when the module is
> removed?
>
> Or if you're not comfortable enough doing that, I can hack up a patch.
>
> JE
-------------------------------------------------------
This SF.NET email is sponsored by: Geek Gift Procrastinating?
Get the perfect geek gift now! Before the Holidays pass you by.
T H I N K G E E K . C O M http://www.thinkgeek.com/sf/
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-users
