Received-SPF: pass (google.com: domain of linux-kernel+bounces-137293-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Date: Tue, 9 Apr 2024 18:10:18 +0200
From: Oscar Salvador <osalvador@suse.de>
To: Miaohe Lin <linmiaohe@huawei.com>
Cc: akpm@linux-foundation.org, naoya.horiguchi@nec.com, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] mm/memory-failure: fix deadlock when
 hugetlb_optimize_vmemmap is enabled
Message-ID: <ZhVoatdJZ1RWu2r3@localhost.localdomain>
References: <20240407085456.2798193-1-linmiaohe@huawei.com>
 <ZhVMThr9TNeP6SWj@localhost.localdomain>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <ZhVMThr9TNeP6SWj@localhost.localdomain>

On Tue, Apr 09, 2024 at 04:10:22PM +0200, Oscar Salvador wrote:
> On Sun, Apr 07, 2024 at 04:54:56PM +0800, Miaohe Lin wrote:
> > In short, below scene breaks the lock dependency chain:
> > 
> >  memory_failure
> >   __page_handle_poison
> >    zone_pcp_disable -- lock(pcp_batch_high_lock)
> >    dissolve_free_huge_page
> >     __hugetlb_vmemmap_restore_folio
> >      static_key_slow_dec
> >       cpus_read_lock -- rlock(cpu_hotplug_lock)
> > 
> > Fix this by calling drain_all_pages() instead.
> > 
> > Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
> 
> Acked-by: Oscar Salvador <osalvador@suse.de>

On a second though,

disabling pcp via zone_pcp_disable() was a deterministic approach.
Now, with drain_all_pages() we drain PCP queues to buddy, but nothing
guarantees that those pages do not end up in a PCP queue again before we
the call to take_page_off_budy() if we
need refilling, right?

I guess we can live with that because we will let the system know that we
failed to isolate that page.


-- 
Oscar Salvador
SUSE Labs