Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) client-ip=23.128.96.37;
Date:   Tue, 12 Dec 2023 15:20:20 -0800
From:   Jakub Kicinski <kuba@kernel.org>
To:     Jesse Brandeburg <jesse.brandeburg@intel.com>
Cc:     Kunwu Chan <chentao@kylinos.cn>, <anthony.l.nguyen@intel.com>,
        <davem@davemloft.net>, <edumazet@google.com>, <pabeni@redhat.com>,
        <jacob.e.keller@intel.com>, <przemyslaw.kitszel@intel.com>,
        <intel-wired-lan@lists.osuosl.org>, <netdev@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>, Kunwu Chan <kunwu.chan@hotmail.com>
Subject: Re: [PATCH] iavf: Fix null pointer dereference in
 iavf_print_link_message
Message-ID: <20231212152020.4a24b2f7@kernel.org>
In-Reply-To: <6ba1e424-9903-43db-b567-32a864b896c9@intel.com>
References: <20231211025927.233449-1-chentao@kylinos.cn>
        <20231212132851.59054654@kernel.org>
        <6ba1e424-9903-43db-b567-32a864b896c9@intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Tue, 12 Dec 2023 15:05:19 -0800 Jesse Brandeburg wrote:
> On 12/12/2023 1:28 PM, Jakub Kicinski wrote:
> > On Mon, 11 Dec 2023 10:59:27 +0800 Kunwu Chan wrote:  
> >> kasprintf() returns a pointer to dynamically allocated memory
> >> which can be NULL upon failure.
> >>
> >> Fixes: 1978d3ead82c ("intel: fix string truncation warnings")  
> > 
> > No need for the allocation here, print to a buffer on the stack.  
> 
> Sure, but I think that just takes us full circle back to where we
> started. reverting this to the previous code will add back W=1 warnings.
> 
> The whole point of the commit mentioned above was to get a reasonable
> implementation that won't cause string truncation warnings. Is there
> some trick I don't know about to get an allocation which will not
> trigger snprintf and friends to print warnings from -Wformat-truncation

Hm, it'd be nice if there was a flavor of snprintf which explicitly
doesn't trigger this warning. Or perhaps a marking for the output
buffer that says "truncation OK".

Absent that, can we print to a buffer on the stack and copy?
The link message is probably meh, but automation may get quite
confused if a NIC suddenly stops reporting FW version..