Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230731171233.1098105-1-surenb@google.com> <20230731171233.1098105-2-surenb@google.com>
 <CAHk-=wjEbJS3OhUu+2sV8Kft8GnGcsNFOhYhXYQuk5nvvqR-NQ@mail.gmail.com>
 <CAJuCfpFWOknMsBmk1RwsX9_0-eZBoF+cy=P-E7xAmOWyeo4rvA@mail.gmail.com>
 <CAHk-=wiFXOJ_6mnuP5h3ZKNM1+SBNZFZz9p8hyS8NaYUGLioEg@mail.gmail.com> <CAJuCfpG4Yk65b=0TLfGRqrO7VpY3ZaYKqbBjEP+45ViC9zySVQ@mail.gmail.com>
In-Reply-To: <CAJuCfpG4Yk65b=0TLfGRqrO7VpY3ZaYKqbBjEP+45ViC9zySVQ@mail.gmail.com>
From:   Suren Baghdasaryan <surenb@google.com>
Date:   Tue, 1 Aug 2023 13:28:56 -0700
Message-ID: <CAJuCfpF6WcJBSix0PD0cOD_MaeLpfGz1ddS6Ug_M+g0QTfkdzw@mail.gmail.com>
Subject: Re: [PATCH 1/6] mm: enable page walking API to lock vmas during the walk
To:     Linus Torvalds <torvalds@linux-foundation.org>
Cc:     akpm@linux-foundation.org, jannh@google.com, willy@infradead.org,
        liam.howlett@oracle.com, david@redhat.com, peterx@redhat.com,
        ldufour@linux.ibm.com, vbabka@suse.cz, michel@lespinasse.org,
        jglisse@google.com, mhocko@suse.com, hannes@cmpxchg.org,
        dave@stgolabs.net, hughd@google.com, linux-kernel@vger.kernel.org,
        linux-mm@kvack.org, stable@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, Jul 31, 2023 at 1:24=E2=80=AFPM Suren Baghdasaryan <surenb@google.c=
om> wrote:
>
> On Mon, Jul 31, 2023 at 12:33=E2=80=AFPM Linus Torvalds
> <torvalds@linux-foundation.org> wrote:
> >
> > On Mon, 31 Jul 2023 at 12:31, Suren Baghdasaryan <surenb@google.com> wr=
ote:
> > >
> > > I got the idea but a couple of modifications, if I may.
> >
> > Ack, sounds sane to me.
>
> Ok, I'll wait for more feedback today and will post an update tomorrow. T=
hanks!

I have the new patchset ready but I see 3 places where we walk the
pages after mmap_write_lock() while *I think* we can tolerate
concurrent page faults (don't need to lock the vmas):

s390_enable_sie()
break_ksm()
clear_refs_write()

In all these walks we lock PTL when modifying the page table entries,
that's why I think we can skip locking the vma but maybe I'm missing
something? Could someone please check these 3 cases and confirm or
deny my claim?
Thanks,
Suren.

>
> >
> >              Linus