Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:6 as permitted sender) client-ip=2620:137:e000::3:6;
Date:   Tue, 31 Oct 2023 14:02:43 +0100
From:   Greg KH <gregkh@linuxfoundation.org>
To:     Mukesh Ojha <quic_mojha@quicinc.com>
Cc:     Johannes Berg <johannes@sipsolutions.net>,
        Yu Wang <quic_yyuwang@quicinc.com>, rafael@kernel.org,
        linux-kernel@vger.kernel.org, kernel@quicinc.com
Subject: Re: [PATCH] Devcoredump: fix use-after-free issue when releasing
 devcd device
Message-ID: <2023103119-litigator-yonder-6ee1@gregkh>
References: <20231027055521.2679-1-quic_yyuwang@quicinc.com>
 <2023102730-twins-thieving-d04e@gregkh>
 <22ab53d1ae36d4925732e6e1dc989dc75af126da.camel@sipsolutions.net>
 <ab509035-cf3c-4902-87a8-cbba03ac948c@quicinc.com>
 <d00e8976baa4fb93a1f3f1191998994540bb173b.camel@sipsolutions.net>
 <bb8113aa-1573-5e02-3fcd-bd92b8ac14ba@quicinc.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <bb8113aa-1573-5e02-3fcd-bd92b8ac14ba@quicinc.com>
Precedence: bulk

On Tue, Oct 31, 2023 at 06:16:08PM +0530, Mukesh Ojha wrote:
> 
> 
> On 10/31/2023 2:29 PM, Johannes Berg wrote:
> > On Tue, 2023-10-31 at 16:29 +0800, Yu Wang wrote:
> > > 
> > > In this case, the device is temporarily added for dump only, so we need to
> > > delete it when dump is completed.
> > > The other users doesn't add/delete the device like this.
> > 
> > For good reason, I guess? I think this is probably a bad idea.
> > 
> > The whole point of this was to actually know which device created the
> > coredump? If you make one up on the fly that's ... pointless? Surely you
> > must have _some_ device that already exists?
> 
> Passing device name to be user space looks to be the reason.

Wait, again, why are you creating a fake device just to dump data?
That's not what this api is for at all, why are you abusing it in ways
it was not designed to be used?

And I will strongly argue, that if no in-kernel users are having
problems, perhaps it is your out-of-tree code?

Unless you can show any in-kernel user of this trigging the issue, I
don't think there's anything we need to do here, do you?

thanks,

greg k-h