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;
Date:   Tue, 19 Apr 2022 15:44:50 -0400
From:   Peter Xu <peterx@redhat.com>
To:     Alistair Popple <apopple@nvidia.com>
Cc:     linux-kernel@vger.kernel.org, linux-mm@kvack.org,
        Mike Kravetz <mike.kravetz@oracle.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        David Hildenbrand <david@redhat.com>,
        Matthew Wilcox <willy@infradead.org>,
        Nadav Amit <nadav.amit@gmail.com>,
        Axel Rasmussen <axelrasmussen@google.com>,
        Andrea Arcangeli <aarcange@redhat.com>,
        "Kirill A . Shutemov" <kirill@shutemov.name>,
        Hugh Dickins <hughd@google.com>,
        Jerome Glisse <jglisse@redhat.com>,
        Mike Rapoport <rppt@linux.vnet.ibm.com>
Subject: Re: [PATCH v8 01/23] mm: Introduce PTE_MARKER swap entry
Message-ID: <Yl8RMjgIoW0r2TFM@xz-m1.local>
References: <20220405014646.13522-1-peterx@redhat.com>
 <20220405014646.13522-2-peterx@redhat.com>
 <87o80xjwvd.fsf@nvdebian.thelocal>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <87o80xjwvd.fsf@nvdebian.thelocal>
Precedence: bulk

On Tue, Apr 19, 2022 at 06:25:31PM +1000, Alistair Popple wrote:
> Hi Peter,

Hi, Alistair,

> 
> Is there something I have missed that means PTE markers can only be used with
> file-backed memory? Obviously that's all you care about for this patch series,
> but if we needed to mark some anonymous PTE for special processing is there
> anything that would prevent us using a PTE marker? Specifically I was thinking
> about it in relation to this series:
> <https://lore.kernel.org/linux-mm/87pmldjxiq.fsf@nvdebian.thelocal/>

It's not necessarily to be restricted to file-backed.  All the file-backed
check here in this series was just for safety purpose and nothing else.

I think it's a very good example of that swap-read-error case that pte
marker does sound like a great fit, but let's see whether people would
still like to stick with hwpoison which makes some sense too.  Then let's
keep the discussion in that thread.

Thanks for the pointer!

-- 
Peter Xu