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

[Timur Tabi]

Arjan van de Ven wrote:

> this is a myth; linux is free to move the page about in physical memory
> even if it's mlock()ed!!

Then Linux has a very odd definition of the word "locked".

> And even then, the user can munlock the memory from another thread etc
> etc. Not a good idea.

Well, that's okay, because then the app is doing something stupid, so we don't worry about 
that.

> get_user_pages() is used from AIO and other parts of the kernel for
> similar purposes and in fact is designed for it, so it better work. If
> it has bugs those should be fixed, not worked around!

I've been complaining about get_user_pages() not working for a long time now, but I can 
only demonstrate the problem with our hardware.  See my other post in this thread for details.

-- 
Timur Tabi
Staff Software Engineer
timur.tabi at ammasso.com