Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
Date:   Wed, 27 Apr 2022 14:20:22 +0200
From:   Oscar Salvador <osalvador@suse.de>
To:     David Hildenbrand <david@redhat.com>
Cc:     Naoya Horiguchi <naoya.horiguchi@linux.dev>, linux-mm@kvack.org,
        Andrew Morton <akpm@linux-foundation.org>,
        Miaohe Lin <linmiaohe@huawei.com>,
        Mike Kravetz <mike.kravetz@oracle.com>,
        Yang Shi <shy828301@gmail.com>,
        Muchun Song <songmuchun@bytedance.com>,
        Naoya Horiguchi <naoya.horiguchi@nec.com>,
        linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCH v1 0/4] mm, hwpoison: improve handling workload
 related to hugetlb and memory_hotplug
Message-ID: <Ymk1BkSEs5uCgV6e@localhost.localdomain>
References: <20220427042841.678351-1-naoya.horiguchi@linux.dev>
 <54399815-10fe-9d43-7ada-7ddb55e798cb@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <54399815-10fe-9d43-7ada-7ddb55e798cb@redhat.com>
Precedence: bulk

On Wed, Apr 27, 2022 at 12:48:16PM +0200, David Hildenbrand wrote:
> I raised some time ago already that I don't quite see the value of
> allowing memory offlining with poisened pages.
> 
> 1) It overcomplicates the offlining code and seems to be partially
>    broken
> 2) It happens rarely (ever?), so do we even care?
> 3) Once the memory is offline, we can re-online it and lost HWPoison.
>    The memory can be happily used.
> 
> 3) can happen easily if our DIMM consists of multiple memory blocks and
> offlining of some memory block fails -> we'll re-online all already
> offlined ones. We'll happily reuse previously HWPoisoned pages, which
> feels more dangerous to me then just leaving the DIMM around (and
> eventually hwpoisoning all pages on it such that it won't get used
> anymore?).
> 
> So maybe we should just fail offlining once we stumble over a hwpoisoned
> page?
> 
> Yes, we would disallow removing a semi-broken DIMM from the system that
> was onlined MOVABLE. I wonder if we really need that and how often it
> happens in real life. Most systems I am aware of don't allow for
> replacing individual DIMMs, but only complete NUMA nodes. Hm.

I teend to agree with all you said.
And to be honest, the mechanism of making a semi-broken DIMM healthy
again has always been a mistery to me.

One would think that:

1- you hot-remove the memory
2- you fix/remove it
3- you hotplug memory again

but I am not sure how many times this came to be.

And there is also the thing about losing the hwpoison information
between offline<->online transitions, so, the thing is unreliable.

And for that to work, we would have to add a bunch of code
to keep track of "offlined" pages that are hwpoisoned, so we
flag them again once they get onlined, and that means more
room for errors.

So, I would lean towards the fact of not allowing to offline
memory that contain such pages in the first place, unless that
proves to be a no-go.


-- 
Oscar Salvador
SUSE Labs