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, 9 Sep 2020 22:49:12 +0100
From:   Matthew Wilcox <willy@infradead.org>
To:     Vlastimil Babka <vbabka@suse.cz>
Cc:     zangchunxin@bytedance.com, akpm@linux-foundation.org,
        linux-mm@kvack.org, linux-kernel@vger.kernel.org,
        Muchun Song <songmuchun@bytedance.com>
Subject: Re: [PATCH v2] mm/vmscan: fix infinite loop in drop_slab_node
Message-ID: <20200909214912.GO6583@casper.infradead.org>
References: <20200909152047.27905-1-zangchunxin@bytedance.com>
 <16906d44-9e3c-76a1-f1a9-ced61e865467@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <16906d44-9e3c-76a1-f1a9-ced61e865467@suse.cz>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Sep 09, 2020 at 07:59:44PM +0200, Vlastimil Babka wrote:
> - Exit also on other signals such as SIGABRT, SIGTERM? If I write to drop_caches
> and think it's too long, I would prefer to kill it by ctrl-c and not just kill
> -9. Dunno if the canonical way of testing for this is if
> (signal_pending(current)) or differently.

fatal_signal_pending() is the canonical way to do it.  If your task has
installed a signal handler for ABRT or TERM, that's its prerogative,
but it's chosen not to get killed, and it's not allowed to see short
reads/writes, so we can't break out early.