Re: [PATCH] cleanup ACPI numa warnings

[Dave Hansen]

On Sun, 2004-08-08 at 22:10, Alex Williamson wrote:
> On Sun, 2004-08-08 at 21:52 -0700, Dave Hansen wrote:
> > On Sun, 2004-08-08 at 21:19, Alex Williamson wrote:
> > >    Ok, I was all set to switch to static inlines, but it doesn't work.
> > > Compiling w/ debug on, _dbg is undefined, which is part of the
> > > ACPI_DB_INFO macro, but it only gets setup by the ACPI_FUNCTION_NAME
> > > macro.  Guess I got lucky by choosing to do it as a macro.  IMHO, it
> > > doesn't really make sense to make the static inline functions more
> > > complicated or hide where they're getting called to make this all work.
> > > So, I think the choices are to stick with the ugly macros or put #ifdefs
> > > around the code and essentially leave it the way it is.  Sorry I didn't
> > > give it a more thorough look when originally questioned.  Better ideas?
> > > Thanks,
> > 
> > That code is already pretty hideous, so perhaps my original question
> > doesn't have that much impact.  The attached patch at least uses inline
> > functions.  It still has the #ifdefs, but what else do you expect for
> > debugging code?  Is this a feasible approach?
> 
>    If you build with CONFIG_ACPI_DEBUG=y, you'll see the problem I was
> trying to describe above with this approach.
> 
> drivers/acpi/numa.c: In function `acpi_print_srat_processor_affinity':
> drivers/acpi/numa.c:44: error: `_dbg' undeclared (first use in this function)
> drivers/acpi/numa.c:44: error: (Each undeclared identifier is reported only once
> drivers/acpi/numa.c:44: error: for each function it appears in.)
> drivers/acpi/numa.c: At top level:
> drivers/acpi/numa.c:48: error: parse error before '}' token
> drivers/acpi/numa.c: In function `acpi_print_srat_memory_affinity':
> drivers/acpi/numa.c:52: error: `_dbg' undeclared (first use in this function)
> drivers/acpi/numa.c: At top level:
> drivers/acpi/numa.c:58: error: parse error before '}' token
> make[2]: *** [drivers/acpi/numa.o] Error 1
> make[1]: *** [drivers/acpi] Error 2
> make: *** [drivers] Error 2

OK, you win.  ACPI is so insanely disgusting already that you've proved
that you should go ahead and add as many multi-line #defines as you
can.  It can't possibly get any worse. :)

-- Dave

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/