Re: PATCH 2.3.41.2: Replace Via support for ACPI

[Jeff Garzik]

Martin Mares wrote:
> 
> Hi Jeff,
> 
> > My approach is due to future expandability -- if you ever need ANY
> > per-chip information besides the init function, then you have to go my
> > way of setting up the PCI table.
> 
>    It's improbable you'll need any -- the rest of the driver should not
> use any chipset specific information since it's driven purely by the
> ACPI table and the whole purpose of the init function is to initialize
> the chip and generate its ACPI table in case the BIOS doesn't provide
> one.

I'm willing to bet that there will be more information in the future,
like chipset name or specific quirks or masks for product variants.

> > Therefore, why not use my method in
> > the first place, and make the maintainer's life easier in the long run?
> > The extra table is __initdata so the extra couple bytes doesn't even
> > matter...
> 
>    So let's at least remove the first column of the table which is
> completely useless as it's equal to the row index.

Agreed.

	Jeff

-- 
Jeff Garzik         | Andre the Giant has a posse.
Building 1024       |
MandrakeSoft, Inc.  |