[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Y0a/1SnV9kTz0hRG@dhcp22.suse.cz>
Date: Wed, 12 Oct 2022 15:23:33 +0200
From: Michal Hocko <mhocko@...e.com>
To: Vinicius Petrucci <vpetrucci@...il.com>
Cc: Frank van der Linden <fvdl@...gle.com>,
Zhongkun He <hezhongkun.hzk@...edance.com>, corbet@....net,
akpm@...ux-foundation.org, linux-mm@...ck.org,
linux-kernel@...r.kernel.org, linux-api@...r.kernel.org,
linux-doc@...r.kernel.org, wuyun.abel@...edance.com
Subject: Re: [RFC] mm: add new syscall pidfd_set_mempolicy()
On Wed 12-10-22 15:07:48, Michal Hocko wrote:
> On Wed 12-10-22 07:34:06, Vinicius Petrucci wrote:
[...]
> > Having a feature like "pidfd/process_mbind" would really simplify our
> > user-level agent implementation moving forward, as right now we are
> > adding a LD_PRELOAD wrapper (for signal handler) to listen and execute
> > "mbind" requests from another process. If there's any other
> > alternative solution to this already (via ptrace?), please let me
> > know.
>
> userfaultfd sounds like the closest match if #PF handling under control
> of an external agent is viable.
And just to clarify. I haven't ever played with using userfaultfd for
numa balancing so I might be completely wrong here.
--
Michal Hocko
SUSE Labs
Powered by blists - more mailing lists