Re: [PATCH] mm, gup: introduce concept of "foreign" get_user_pages()
From: Vlastimil Babka
Date: Wed Jan 20 2016 - 12:57:06 EST
On 01/20/2016 06:35 PM, Dave Hansen wrote:
> Here's another revision taking Vlastimil's suggestions about
> keeping __get_user_pages_unlocked() as-is in to account.
> This does, indeed, look nicer. Now, all the "__" variants
> take a full tsk/mm and flags.
> He also noted that the two sites where we called gup with
> tsk=NULL were probably incorrectly changing behavior with respect
> to fault accounting. Long-term, I wonder if we should just add
> a "FOLL_" flag to make that more explicit, but for now, I've
> fixed up those sites.
> From: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
> For protection keys, we need to understand whether protections
> should be enforced in software or not. In general, we enforce
> protections when working on our own task, but not when on others.
> We call these "current" and "foreign" operations.
> This patch introduces a new get_user_pages() variant:
> The plain get_user_pages() can no longer be used on mm/tasks
> other than 'current/current->mm', which is by far the most common
> way it is called. Using it makes a few of the call sites look a
> bit nicer.
> In other words, get_user_pages_foreign() is a replacement for
> when get_user_pages() is called on non-current tsk/mm.
> This also switches get_user_pages_(un)locked() over to be like
> get_user_pages() and not take a tsk/mm. There is no
> get_user_pages_foreign_(un)locked(). If someone wants that
> behavior they just have to use "__" variant and pass in
> FOLL_FOREIGN explicitly.
Hm so this gets a bit ahead of patch "mm: add gup flag to indicate "foreign" mm
access", right? It might be cleaner to postpone passing FOLL_FOREIGN until then,
but not critical.
BTW doesn't that other patch miss passing FOLL_FOREIGN from
get_user_pages_foreign() or something? I see it only uses it from break_ksm(),
am I missing something?