Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Wed, 14 Jul 2021 10:00:19 +0200
From:   Michal Hocko <mhocko@suse.com>
To:     David Hildenbrand <david@redhat.com>
Cc:     Evan Green <evgreen@chromium.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Pavel Machek <pavel@ucw.cz>, Alex Shi <alexs@kernel.org>,
        Alistair Popple <apopple@nvidia.com>,
        Jens Axboe <axboe@kernel.dk>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Joonsoo Kim <iamjoonsoo.kim@lge.com>,
        "Matthew Wilcox (Oracle)" <willy@infradead.org>,
        Miaohe Lin <linmiaohe@huawei.com>,
        Minchan Kim <minchan@kernel.org>,
        Vlastimil Babka <vbabka@suse.cz>, linux-kernel@vger.kernel.org,
        linux-mm@kvack.org, linux-api@vger.kernel.org
Subject: Re: [PATCH v2] mm: Enable suspend-only swap spaces
Message-ID: <YO6Zk5wwpdqwwGUs@dhcp22.suse.cz>
References: <20210709105012.v2.1.I09866d90c6de14f21223a03e9e6a31f8a02ecbaf@changeid>
 <YOvpVRSMJe8NQuS2@dhcp22.suse.cz>
 <30dddfb1-388c-a593-0987-73e821216da9@redhat.com>
 <YO55ZIrgkLXI4BbS@dhcp22.suse.cz>
 <b84dfb7b-9ae7-8cd7-ce65-0b1952e30e8e@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <b84dfb7b-9ae7-8cd7-ce65-0b1952e30e8e@redhat.com>
Precedence: bulk

On Wed 14-07-21 09:51:13, David Hildenbrand wrote:
[...]
> Anyhow, the proposal here does not sound completely crazy to me, although
> it's unfortunate how we decided to mangle hibernation and swapping into the
> same mechanism originally; a different interface to active "hibernation only
> backends" would be cleaner than doing a "swapon ..." without swapping.

Completely agreed! And I suspect that a special swap flag just digs that
hole even deeper. While the flag might look simple enough now I am a bit
worried this will open traps in the future.

I am not saying the idea is crazy either, it is just a hack on top of
the existing hack and as such it requires a very good reasoning. So far
I have heard rather vague justification and I am especially curious
about the "no mixing with the regular swapout" concern. It might be very
well the case that there are more usecases which would benefit from it.
-- 
Michal Hocko
SUSE Labs