From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Carpenter Date: Thu, 30 Jan 2020 10:00:09 +0000 Subject: Re: Open tasks? Message-Id: <20200130100009.GD1778@kadam> List-Id: References: In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: kernel-janitors@vger.kernel.org On Mon, Jan 20, 2020 at 08:36:32PM -0800, Raymond Deng wrote: > Hi, are there open tasks for reasonably new people to do? Thanks. The new hotest idea is to sign up for te syzbot mailing list and fix bugs. For example: https://syzkaller.appspot.com/bug?extid=C37a14770d51a085a520 It has a c reproducer. It has been bisected but if it's a race condition, the git bisection could be suspect. It affects v5.5-rc7 (It's in the dmesg, you need the version information for the lines to match up). My first thought was that it is an integer overflow but I can't see how that would be possible. Smatch said it was possible, but Smatch assumed that we had set vc->vc_cols to an invalid number before the start. Btw, using the Smatch `smdb.py where vc_cols` command is super helpful for debugging these bugs. Then my next thought was that it is a race condition but we're always holding the console_lock() so I can't see how that works either... Again for these things `smdb.py` really helps for examining the callers. I'm working on recording when the caller is holding a lock but it's a bit complicated and I'm not ready to push that code yet. It's a mystery. regards, dan carpenter