Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Wed, 12 Feb 2020 09:09:45 +0100
From:   Michal Hocko <mhocko@kernel.org>
To:     Linus Torvalds <torvalds@linux-foundation.org>
Cc:     Andrew Morton <akpm@linux-foundation.org>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Rik van Riel <riel@surriel.com>,
        linux-fsdevel <linux-fsdevel@vger.kernel.org>,
        Linux-MM <linux-mm@kvack.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Dave Chinner <david@fromorbit.com>,
        Yafang Shao <laoar.shao@gmail.com>,
        Roman Gushchin <guro@fb.com>,
        Al Viro <viro@zeniv.linux.org.uk>, kernel-team@fb.com
Subject: Re: [PATCH] vfs: keep inodes with page cache off the inode shrinker
 LRU
Message-ID: <20200212080945.GA11353@dhcp22.suse.cz>
References: <20200211175507.178100-1-hannes@cmpxchg.org>
 <29b6e848ff4ad69b55201751c9880921266ec7f4.camel@surriel.com>
 <20200211193101.GA178975@cmpxchg.org>
 <20200211154438.14ef129db412574c5576facf@linux-foundation.org>
 <CAHk-=wiGbz3oRvAVFtN-whW-d2F-STKsP1MZT4m_VeycAr1_VQ@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAHk-=wiGbz3oRvAVFtN-whW-d2F-STKsP1MZT4m_VeycAr1_VQ@mail.gmail.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue 11-02-20 16:28:39, Linus Torvalds wrote:
> On Tue, Feb 11, 2020 at 3:44 PM Andrew Morton <akpm@linux-foundation.org> wrote:
> >
> > Testing this will be a challenge, but the issue was real - a 7GB
> > highmem machine isn't crazy and I expect the inode has become larger
> > since those days.
> 
> Hmm. I would say that in the intening years a 7GB highmem machine has
> indeed become crazy.

Absolutely agreed.

> It used to be something we kind of supported.

And it's been few years since we have been actively discouraging people
from using 32b kernels with a lot of memory. There are bug reports
popping out from time to time but I do not remember any case where using
64b kernel would be a no-go. So my strong suspicion is that people
simply keep their kernels on 32b without a good reason because it tends
to work most of the time until they hit one of the lowmem problems and
they move over to 64b.

> But we really should consider HIGHMEM to be something that is on the
> deprecation list. In this day and age, there is no excuse for running
> a 32-bit kernel with lots of physical memory.
> 
> And if you really want to do that, and have some legacy hardware with
> a legacy use case, maybe you should be using a legacy kernel.
> 
> I'd personally be perfectly happy to start removing HIGHMEM support again.

I wouldn't be opposed at all.
-- 
Michal Hocko
SUSE Labs