Re: [RFC 00/15] x86_64: Optimize percpu accesses

[Jeremy Fitzhardinge]

Christoph Lameter wrote:
> That is going to be difficult. The GS register is tied up for the pda area
> as long as you have it. And you cannot get rid of the pda because of the library
> compatibility issues. We would break binary compatibility if we would get rid of the pda.
>
> If one attempts to remove one field after another then the converted accesses will not be able to use GS relative accesses anymore. This can lead to all sorts of complications.
>   

Eh?  Yes they will.  That's the whole point of making the pda a percpu 
variable itself.  You can use %gs:<small> to get to the pda, and 
%gs:<larger> to get to percpu variables.  Converting pda->percpu will 
just have the effect of increasing the %gs offset into the percpu space.

This project isn't interesting to me unless per-cpu variables are 
directly accessible off %gs.

   J
--
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/