Re: LOCKDEBUG panic in ucom/uvisor

[Andrew Doran <ad%netbsd.org@localhost>]

On Mon, Jun 02, 2008 at 05:37:51PM -0500, David Young wrote:

> On Tue, Jun 03, 2008 at 02:32:16AM +0400, Aleksej Saushev wrote:
> >   Hello!
> > 
> > For several recent days (2 or 3) I can't write anything via uvisor
> > (reading from uvisor is OK), the kernel panics:
> > 
> > #23 0xc02b4202 in panic ()
> > #24 0xc02adf13 in lockdebug_abort1 ()
> > #25 0xc027b76d in cv_broadcast ()
> > #26 0xc02cb9a4 in ttypull ()
> > #27 0xc0399176 in ucomstart ()
> > #28 0xc02ca2d8 in ttstart ()
> > #29 0xc03990d2 in ucomwritecb ()
> > #30 0xc0393bf3 in usb_transfer_complete ()
> > #31 0xc01b610e in uhci_idone ()
> > #32 0xc01b6345 in uhci_softintr ()
> > #33 0xc0299c69 in softint_dispatch ()
> > #34 0xc0100eed in Xsoftintr ()
> > #35 0x00000000 in ?? ()
> > 
> > The panic is 100% reproducible: I can read from Palm device
> > ("pilot-xfer -u"), but cannot write to it ("pilot-xfer -i").
> > Does anyone know how to fix this?

Unfortunatley I haven't been seeing this one. I've a got a bunch of changes
in my local tree to finish off the tty locking properly, 5 years after we
started it. :-). I'll sort out a workaround tomorrow because it will take a
few days to fix it properly. For now the system does work OK as-is, despite
the complaint from LOCKDEBUG.

Thanks,
Andrew