[openib-general] Re: [PATCH][RFC][0/4] InfiniBand userspace verbs implementation

[Greg KH]

On Sun, Apr 24, 2005 at 04:52:31PM -0500, Timur Tabi wrote:
> Greg KH wrote:
> 
> >You don't "support" i386 or ia64 or x86-64 or ppc64 systems?  What
> >hardware do you support? 
> 
> I've never seen or heard of any x86-32 or x86-64 system that supports 
> hot-swap RAM.

I know of at least 1 x86-32 box from a three-letter-named company with
this feature that has been shipping for a few _years_ now.  That box is
pretty much everywhere now, and I know that other versions of it are
also quite popular (despite the high cost...)

> Our hardware does not support PPC, and our software doesn't support
> ia-64.

Your hardware is just a pci card, right?  Why wouldn't it work on ppc64
and ia64 then?

> > And what about the fact that you are aiming to
> >get this code into mainline, right?  If not, why are you asking here?
> >:)
> 
> Well, our primary concern is getting our stuff to work.  Since 
> get_user_pages() doesn't work, but mlock() does, that's what we use.  I 
> don't know how to fix get_user_pages(), and I don't have the time right now 
> to figure it out.  I know that technically mlock() is not the right way to 
> do it, and so we're not going to be submitting our code for the mainline 
> until get_user_pages() works and our code uses it instead of mlock().

Wait, what _is_ "your stuff"?  The open-ib code?  Or some other, private
fork?  Any pointers to this stuff?

thanks,

greg k-h