Re: pnp bug in 2.6.27-rc1
On Tuesday 05 August 2008 01:06:35 pm Uwe Bugla wrote:
> Am Monday 04 August 2008 22:19:21 schrieben Sie:
> > On Monday 04 August 2008 12:51:27 pm Uwe Bugla wrote:
> > > But I noticed another problem:
> > >
> > > Without BIOS changes the following happens:
> > > a. starting kernel 2.6.26 the ISA sound card is on IRQ 10, and the second
> > > parport is on IRQ 5. The MIDI port of AD1816a is there, but
> > > interruptless.
> > >
> > > This is the intended scenery.
> > >
> > > b. starting kernel 2.6.27-rc1 the ISA sound card is on IRQ 5, and the
> > > second parport is in polling mode, as IRQ 5 is preoccupied by the sound
> > > card. The MIDI port is on IRQ 10.
> > >
> > > I haven't tested whether the second parport works, but anyway this
> > > configuration not only differs from the 2.6.26 one but also is not
> > > intended. So something's still wrong in the PnP design.
> > > I expected the same configuration in the IRQ table as in 2.6.26.
> > > Unfortunately this is not the case.
> > > It should be the MIDI port to run interruptless, but not the second
> > > parport (additional ISA card).
> >
...
> after doing a little bit of research I am not sure anymore whether you did
> something wrong:
> If I leave out this
>
> "options parport_pc io=0x378,0x278 irq=7,5"
>
> in etc/modprobe.d/i386 the error with the AD1816a soundcard preoccupying IRQ 5
> happens with the consequence that parport 2 (ISA card) runs in polling mode.
Sorry for the delay in answering; I was on vacation most of last week.
I'm not sure what's going on with the parport IRQs. If you think
I need to look into that, let's start a separate thread for it. If
there's work to do here, please cc: LKML too, since a parallel port
issue could affect lots of people.
I do not want users of 2.6.27 to have to change any parport_pc module
parameters. If you had edit something in /etc/modprobe.d/, I think
that's a bug I should investigate.
Bjorn
Reply to: