Re: cvs commit: src/sys/conf options src/sys/i386/acpica Makefile acpi_wakecode.S src/sys/i386/conf NOTES

From: Mitsuru IWASAKI <iwasaki_at_jp.FreeBSD.org>
Date: Wed, 29 Oct 2003 23:38:11 +0900 (JST)
Hi,

I've just committed conf/options and i386/conf/NOTES based on your
suggestion.  Thanks!

> > --- src/sys/i386/acpica/Makefile:1.3	Wed May  1 14:54:07 2002
> > +++ src/sys/i386/acpica/Makefile	Tue Oct 28 19:30:45 2003
> > _at__at_ -1,4 +1,4 _at__at_
> > -# $FreeBSD: /repoman/r/ncvs/src/sys/i386/acpica/Makefile,v 1.3 2002/05/01 21:54:07 peter Exp $
> > +# $FreeBSD: /repoman/r/ncvs/src/sys/i386/acpica/Makefile,v 1.4 2003/10/29 03:30:45 iwasaki Exp $
> >
> >  # Correct path for kernel builds
> >  # Don't rely on the kernel's .depend file
> > _at__at_ -10,6 +10,10 _at__at_
> >  CLEANFILES= acpi_wakecode.h acpi_wakecode.bin acpi_wakecode.o
> >  .endif
> >  CFLAGS+=	-I.
> > +
> > +.ifdef ACPI_NO_RESET_VIDEO
> > +CFLAGS+=	-DACPI_NO_RESET_VIDEO
> > +.endif
> >
> >  all: acpi_wakecode.h
> 
> Is this needed?  options alone should be sufficient and acpi_wakeup.S
> should include opt_acpi.h.

Yes this is needed.  For acpi.ko, options in kernel config file won't
be referred as Warner mentioned.
I usually specify kernel options in make.conf for the kernel modules.

> > Index: src/sys/i386/acpica/acpi_wakecode.S
> > diff -u src/sys/i386/acpica/acpi_wakecode.S:1.4 src/sys/i386/acpica/acpi_wakecode.S:1.5
> > --- src/sys/i386/acpica/acpi_wakecode.S:1.4	Sun Oct 26 22:26:51 2003
> > +++ src/sys/i386/acpica/acpi_wakecode.S	Tue Oct 28 19:30:45 2003
> > _at__at_ -43,6 +43,7 _at__at_
> >  	movw	%ax,%ds
> >  	movw	%ax,%ss
> >
> > +#ifndef ACPI_NO_RESET_VIDEO
> >  	/*
> >  	 * Re-initialize video BIOS.  Restore DS and SS from CS in
> >  	 * case the BIOS modified them.
> > _at__at_ -51,6 +52,7 _at__at_
> >  	movw	%cs, %ax
> >  	movw	%ax, %ds
> >  	movw	%ax, %ss
> > +#endif
> >
> >  	/* Load GDT for real mode */
> >  	lgdt	physical_gdt
> 
> As you can see, I restore DS and SS in two places.  I think we can remove
> the first place (above the lcall), right?

As takawata-san suggested, I also thought we need to setup segment
registers before the BIOS call.

Thanks
Received on Wed Oct 29 2003 - 06:38:23 UTC