From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrea Arcangeli Subject: Re: [PATCH 3/7] Add a UFFD_SECURE flag to the userfaultfd API. Date: Wed, 23 Oct 2019 19:27:08 -0400 Message-ID: <20191023232708.GA433@redhat.com> References: <20191023224110.GE9902@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org To: Andy Lutomirski Cc: Andy Lutomirski , Jann Horn , Daniel Colascione , Linus Torvalds , Pavel Emelyanov , Lokesh Gidra , Nick Kralevich , Nosh Minwalla , Tim Murray , Mike Rapoport , Linux API , LKML , "Dr. David Alan Gilbert" List-Id: linux-api@vger.kernel.org Archived-At: List-Archive: List-Post: On Wed, Oct 23, 2019 at 04:01:53PM -0700, Andy Lutomirski wrote: > Delivering events through read() is just fine. The problem is when > delivering an event does more than just returning bytes. As far as > I=E2=80=99ve noticed, uffd=E2=80=99s read() just returns bytes as long as= FORK is > disabled. Yes, fork is the only case where read doesn't return bytes. Moving only the fd creation to a separate syscall would then avoid involuntary creation of the fd.