Re: [Syllable-developer] ACPI busmanager

[Zoltan K. <gi...@ny...>]

On Saturday 09 August 2008 15:23:00 Kristian Van Der Vliet wrote:
> On Sat, 2008-08-09 at 15:10 +0200, Zoltan Kovacs wrote:
> > On Saturday 09 August 2008 14:13:10 Kristian Van Der Vliet wrote:
> > > On Sat, 2008-08-09 at 14:03 +0200, Zoltan Kovacs wrote:
> > > > On Saturday 09 August 2008 13:59:40 Kristian Van Der Vliet wrote:
> > > > > On Fri, 2008-08-08 at 23:22 +0200, Zoltan Kovacs wrote:
> > > > > > I just have succesfully compiled the first version of the
> > > > > > synchronized ACPI busmanager. It can be downloaded from
> > > > > > http://giszo.lame.hu/syllable/acpi
> > > > > >
> > > > > > Please test it and post your results back.
> > > > >
> > > > > It stops the boot process on the two machines I tried it with, both
> > > > > stopping as soon as the system starts to use to the ATA controller.
> > > > >
> > > > > Logs from both systems (One Dell GX280, one AMI BIOS Celeron
> > > > > machine) are attached.
> > > >
> > > > Could you please try passing the disable_pci_irq_routing=true kernel
> > > > parameter with the new ACPI busmanager?
> > >
> > > Yes, both systems boot fine if I disable IRQ routing.
> >
> > I just compiled a new PCI busmanager as well because it depends on some
> > headers from the ACPI busmanager. As i also touched the headers this
> > recompile might be needed :)
> >
> > Here is the new PCI busmanager: http://giszo.lame.hu/syllable/pci
> >
> > Vanders, could you try the new ACPI busmanager with this new PCI
> > busmanager together without disabling IRQ routing?
>
> That's working fine.
>
> The only issue I see is the slightly messy kernel output at start: at a
> guess you're using multiple printk()'s. You're better off using
> sprintf() to concatenate the string & then printk() once.
>
> If you can get the source to me today I can get it checked in for
> tomorrows dev. build if you'd like?

Yeah, I also saw the messed kernel log but first i tried to get this to a 
working state and now i'm going to fix that output as well :)

About the sources.. I can send them today, of course. But i don't know if it's 
okey to include in the next development build because it's not really tested. 
Or let's say it's only a development build so it can be unstable? :)