Re: [patch 2.5] PCI MWI cacheline size fix

David Brownell (david-b@pacbell.net)
Thu, 20 Mar 2003 07:29:52 -0800


Ivan Kokshaysky wrote:
> This is rather conservative variant of previous patch:

Looks ok, and the previous patch behaved fine on the hardware
I could test it with. The changes being: to set the cacheline
size only on the "prep_mwi" path (vs much earlier), and allow
cacheline size to be a multiple of the "real" one.

> - assume cacheline size of 32 bytes for all x86s except K7/K8 and P4.
> Actually it's good for 386/486s as quite a few PCI devices do not support
> smaller values.

Given the number of people that questioned that K7/K8/P4 logic,
I'd suggest putting that comment into the pcibios_init() code.

> If you all are fine with it, I can make a 2.4 counterpart.

Yes, having that avoid compile-time config is also important.

Thanks for updating this stuff -- it'll be good to know more
drivers can reduce their PCI/cache overheads.

- Dave

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