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:   Fri, 9 Jun 2023 15:11:11 -0700
From:   Andrew Morton <akpm@linux-foundation.org>
To:     Khalid Aziz <khalid.aziz@oracle.com>
Cc:     David Hildenbrand <david@redhat.com>,
        Matthew Wilcox <willy@infradead.org>,
        John Hubbard <jhubbard@nvidia.com>,
        Steven Sistare <steven.sistare@oracle.com>,
        ying.huang@intel.com, mgorman@techsingularity.net,
        baolin.wang@linux.alibaba.com, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org, Khalid Aziz <khalid@kernel.org>
Subject: Re: [PATCH v4] mm, compaction: Skip all non-migratable pages during
 scan
Message-Id: <20230609151111.bb7908254853a0879db07746@linux-foundation.org>
In-Reply-To: <3c2ac70b-31a2-4e85-0ade-cb9f701525b3@oracle.com>
References: <ZG/I7tYY4uV/32hP@casper.infradead.org>
        <ZG/To8Z3StoVoenU@casper.infradead.org>
        <60367660-f4a3-06dc-4d17-4dbdc733ef74@oracle.com>
        <ZHDh4Jeb/vKY+nGU@casper.infradead.org>
        <f5a37f8d-d888-9085-2f2b-1e350a267396@redhat.com>
        <ZHD+eOMpZpWXNAig@casper.infradead.org>
        <e31cd404-56ce-4cad-fcc3-3a6695f750fa@redhat.com>
        <846b770c-9f63-90a2-0435-ec82484e3f74@nvidia.com>
        <ZHF2byaRlaX3W6Md@casper.infradead.org>
        <9821bd9c-7c30-8f0c-68e4-6b1d312bc032@nvidia.com>
        <ZHPydXSAfRq8sh0u@casper.infradead.org>
        <4d035744-271d-1ca3-a440-f8b1573eec96@redhat.com>
        <3c2ac70b-31a2-4e85-0ade-cb9f701525b3@oracle.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Tue, 30 May 2023 09:42:33 -0600 Khalid Aziz <khalid.aziz@oracle.com> wrote:

> > I completely agree that we should be careful with such mapcount vs. pagecount checks, and if we can use something 
> > better, let's use something *better*.
> 
> When we have a reliable folio_maybe_dma_longterm_pinned() function, it will be better to call that instead of doing 
> refcount vs mapcount check. Until that better function to check for pinned pages is in place, may I propose that the 
> current patch fixes a customer problem though not optimally and is a good enough working solution. When a better 
> function is in place, page_has_extra_refs() function can be updated to rely on this other function instead of refcount 
> vs mapcount.

We seem rather stuck with this patch.  I think I'll drop it while we
ponder a way forward.