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:   Mon, 31 May 2021 15:21:21 +0200
From:   Michal Hocko <mhocko@suse.com>
To:     Vlastimil Babka <vbabka@suse.cz>
Cc:     Aaron Tomlin <atomlin@redhat.com>, linux-mm@kvack.org,
        akpm@linux-foundation.org, willy@infradead.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH v4] mm/page_alloc: bail out on fatal signal during
 reclaim/compaction retry attempt
Message-ID: <YLTi0XXHbntlCNuS@dhcp22.suse.cz>
References: <YKZObDpduqwWi/Zm@casper.infradead.org>
 <20210520142901.3371299-1-atomlin@redhat.com>
 <YLTJjJqemt5Uv9vP@dhcp22.suse.cz>
 <c7b69523-c141-d06c-bc02-953c7a939d91@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <c7b69523-c141-d06c-bc02-953c7a939d91@suse.cz>
Precedence: bulk

On Mon 31-05-21 13:35:31, Vlastimil Babka wrote:
> On 5/31/21 1:33 PM, Michal Hocko wrote:
> > On Thu 20-05-21 15:29:01, Aaron Tomlin wrote:
> >> A customer experienced a low-memory situation and decided to issue a
> >> SIGKILL (i.e. a fatal signal). Instead of promptly terminating as one
> >> would expect, the aforementioned task remained unresponsive.
> >> 
> >> Further investigation indicated that the task was "stuck" in the
> >> reclaim/compaction retry loop. Now, it does not make sense to retry
> >> compaction when a fatal signal is pending.
> > 
> > Is this really true in general? The memory reclaim is retried even when
> > fatal signals are pending. Why should be compaction different? I do
> > agree that retrying way too much is bad but is there any reason why this
> > special case doesn't follow the max retry logic?
> 
> Compaction doesn't do anything if fatal signal is pending, it bails out
> immediately and the checks are rather frequent. So why retry?

OK, I was not aware of that and it would be helpful to have that
mentioned in the changelog.

-- 
Michal Hocko
SUSE Labs