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:   Sun, 1 Aug 2021 22:35:13 +0300
From:   Pavel Skripkin <paskripkin@gmail.com>
To:     Petko Manolov <petkan@nucleusys.com>
Cc:     davem@davemloft.net, kuba@kernel.org, linux-usb@vger.kernel.org,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
        syzbot+02c9f70f3afae308464a@syzkaller.appspotmail.com
Subject: Re: [PATCH] net: pegasus: fix uninit-value in
 get_interrupt_interval
Message-ID: <20210801223513.06bede26@gmail.com>
In-Reply-To: <YQaVS5UwG6RFsL4t@carbon>
References: <20210730214411.1973-1-paskripkin@gmail.com>
        <YQaVS5UwG6RFsL4t@carbon>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Sun, 1 Aug 2021 15:36:27 +0300
Petko Manolov <petkan@nucleusys.com> wrote:

> On 21-07-31 00:44:11, Pavel Skripkin wrote:
> > Syzbot reported uninit value pegasus_probe(). The problem was in
> > missing error handling.
> > 
> > get_interrupt_interval() internally calls read_eprom_word() which
> > can fail in some cases. For example: failed to receive usb control
> > message. These cases should be handled to prevent uninit value bug,
> > since read_eprom_word() will not initialize passed stack variable
> > in case of internal failure.
> 
> Well, this is most definitelly a bug.
> 
> ACK!
> 
> 
> 		Petko
> 
> 

Thank you, Petko!


BTW: I found a lot uses of {get,set}_registers without error
checking. I think, some of them could be fixed easily (like in
enable_eprom_write), but, I guess, disable_eprom_write is not so easy.
For example, if we cannot disable eprom should we retry? If not, will
device get in some unexpected state?

Im not familiar with this device, but I can prepare a patch to wrap all
these calls with proper error checking


With regards,
Pavel Skripkin