Re: [discuss] Re: 2.6.19-rc4: known unfixed regressions (v3)

[Andi Kleen <ak@suse.de>]

> What a piece of crap. 
> 
> Andi, I'm getting really upset about this kind of thing. You've been very 
> much not careful about MMCFG in general, and are allowing total crap to go 
> into the kernel, without any thought. Just "testing" something isn't good 
> enough, it needs to be thought out.

Sorry, probably should have read the patch more carefully.

I think I agreed with the high level idea but didn't double check
the details.

> 
> I'm going to revert that totally bogus commit that added that broken 
> "pci_mmcfg_insert_resources()" function. It could be done right, but doing 
> it right would require that the function 

Ok fine by me.

> We really should stop using MMCONFIG entirely, until we have a 

Hmm, for .19 at least you mean? 

Entirely stopping it would break the x86 macs minis again I think.
But we can make it "only use if type 1 doesn't work" 

I'm sure some people will be upset again if we don't use it.
Perhaps there are really users who want to use the PCI-E error handling
for example.

> per-southbridge true knowledge of what the real decoding is. The BIOS 
> tables for this are simply too damn unreliable.

My hopes for that are on Vista. Perhaps use a DMI year test again
(>= 2007) and only white lists for older boards.

-Andi