Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
From:   shankarapailoor <shankarapailoor@gmail.com>
Date:   Tue, 29 May 2018 10:32:35 -0700
Message-ID: <CAB+yDaaXoeQ_hVVVvQ6y+gpa+XSH2YScwp1gR-Rz-2aSQUj74g@mail.gmail.com>
Subject: BUG in mm/page-writeback.c:2652
To:     Chris Mason <clm@fb.com>, jbacik@fb.com, dsterba@suse.com
Cc:     linux-btrfs@vger.kernel.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi,

I am fuzzing linux 4.17-rc4 with BTRFS and I encountered a BUG in
mm/page-writeback.c error described below. This bug has occured 3
times in our fuzzing campaign.

Here is the full stack trace: https://pastebin.com/QcDi9xA2

Truncated Stack Trace: https://pastebin.com/GWRNmhmC

Here is a C reproducer: https://pastebin.com/xTAVB3L2

My linux configs: https://pastebin.com/6J1KMczH

It appears that there is a fault injection inside of create_io_em
during the slab allocation for the path. After that
extent_clear_unlock_delalloc gets called. I don't know if the
extent_map start and end are well defined in this case.

Regards,
Shankara Pailoor