Messages in this thread | | | Date | Sat, 22 Dec 2007 14:10:26 -0700 | From | Matthew Wilcox <> | Subject | Re: [patch] Make MMCONFIG space (extended PCI config space) a driver opt-in issue |
| |
On Sat, Dec 22, 2007 at 08:02:49AM -0800, Arjan van de Ven wrote: > even then.. it's technically not correct; you're not supposed to mix access types for the same device.. > Just doing opt-in also allows us to do quirks (forbid access) as well.
I think what this specification means is that you can't mix both _at the same time_. ie, doing a type 1 cycle, then before it returns, do an mmconfig cycle. We never try to do that because there's a spinlock to prevent more than one config space access at a time. It *has* to work to do them sequentially -- for example, the BIOS may do type1 config accesses, then the OS may do mmconfig.
I've occasionally wondered if that spinlock needs to get split up, but for the amount of pain that could ensue, I can't imagine it ever being worthwhile.
-- Intel are signing my paycheques ... these opinions are still mine "Bill, look, we understand that you're interested in selling us this operating system, but compare it to ours. We can't possibly take such a retrograde step."
| |