ACPI problem - broken debug output

Doug White dwhite at gumbysoft.com
Thu Oct 23 09:18:35 PDT 2003


On Thu, 23 Oct 2003, Michail Vidiassov wrote:

> On Wed, 22 Oct 2003, Doug White wrote:
>
> > > But after I have defined ACPI_DEBUG_OUTPUT in
> > > /sys/contrib/dev/acpica/acconfig.h
> > > I have got broken debug output like
> > >
> > > Table [DSDT] - %hd Objects with %hd Devices %hd Methods %hd Regions
> > > %hd Devices found containing: %hd _STA, %hd _INI methods
> > > Initialized %hd/%hd Regions %hd/%hd Fields %hd/%hd Buffers %hd/%hd
> > > Packages (%hd nodes)
> >
> > Typically, acpi debugging is enabled using a couple of loader tunables, as
> > described on acpi(4).  You just need to compile with ACPI_DEBUG to get the
> > output functionality. I think ACPI_DEBUG_OUTPUT is some internal option
> > that you shouldn't be messing with directly.
> >
>
> No, /sys/contrib/dev/acpica/CHANGES.txt told me:
> Replace ACPI_DEBUG define with ACPI_DEBUG_OUTPUT.

Wonder why this change wasn't propagated to -current.  This might not be
relevant for the FreeBSD import.

> The broken output is caused by the lack of support for
> "h" (short int) optional character in the kvprintf
> function, "Scaled down version of printf", residing in
> /sys/kern/subr_prf.c
>
> Thus the solution is to replace %hd with %d in
> /sys/contirb/dev/acpica/* (since short is converted to
> int anyway) or enhancing kvprintf to handle "h",
> to make porting from Linux easier.
>
> Is it worth sending a pr?

Oh, this is -STABLE. Hm.  That might be a problem.

Go ahead and send the pr.

>                     Sincerely, Michail
> PS. BTW. acpi(4) does not mention the need to use
> acpi_dsdt_type="acpi_dsdt", since it is a loader.conf
> default on FreeBSD-CURRENT. But not on 4.9-??.
> May be it is also worth reporting?

Sure.

-- 
Doug White                    |  FreeBSD: The Power to Serve
dwhite at gumbysoft.com          |  www.FreeBSD.org


More information about the freebsd-stable mailing list