On Fri, Aug 19, 2016 at 04:52:51PM +0300, Pavel Emelyanov wrote:

On 08/19/2016 04:43 PM, Andrea Arcangeli wrote:

quoted

On Fri, Aug 19, 2016 at 04:20:22PM +0300, Pavel Emelyanov wrote:

quoted

And (!) after non-cooperative patches are functional too.

I merged your non-cooperative patches in my tree although there's no
testcase to exercise them yet.

Hm... Are you talking about some in-kernel test, or just any? We have
tests in CRIU tree for UFFD (not sure we've wired up the non-cooperative
part though).

Well, CRIU is by definition non-cooperative :)
Still, we don't have fork() and other events in CRIU lazy restore yet.
I have some brute force additions to the selftests/vm/userfaultfd.c that
verify that the events work, and I'm trying now to get a clean version.

BTW, with addition of hugetlbfs and tmpfs support to userfaultfd, we'd need
MADV_REMOVE and fallocate(PUNCH_HOLE) events in addition to
MADV_DONTNEED...
 

quoted

Thanks,
Andrea
.

-- Pavel

--
Sincerely yours,
Mike.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>