Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <3b7f0d1d-57ec-7870-fc6e-0449e3112461@kernel.org>
Date:   Wed, 6 Sep 2023 10:11:29 +0900
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.13.0
Subject: Re: [PATCH] ata: sata_mv: Fix incorrect string length computation in
 mv_dump_mem()
Content-Language: en-US
To:     Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Cc:     linux-kernel@vger.kernel.org, kernel-janitors@vger.kernel.org,
        linux-ide@vger.kernel.org
References: <1a35e114a3dcc33053ca7cca41cb06b8426d8c40.1693857262.git.christophe.jaillet@wanadoo.fr>
 <b5dcc85d-f709-a3a3-e7ad-9c71f278842a@kernel.org>
 <9150d691-49b8-c102-a5de-d1fa207d61ed@wanadoo.fr>
From:   Damien Le Moal <dlemoal@kernel.org>
Organization: Western Digital Research
In-Reply-To: <9150d691-49b8-c102-a5de-d1fa207d61ed@wanadoo.fr>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 9/6/23 00:28, Christophe JAILLET wrote:
> 
> 
> Le 05/09/2023 à 07:04, Damien Le Moal a écrit :
>> On 9/5/23 04:54, Christophe JAILLET wrote:
>>> snprintf() returns the "number of characters which *would* be generated for
>>> the given input", not the size *really* generated.
>>>
>>> In order to avoid too large values for 'o' (and potential negative values
>>> for "sizeof(linebuf) o") use scnprintf() instead of snprintf().
>>>
>>> Note that given the "w < 4" in the for loop, the buffer can NOT
>>> overflow, but using the *right* function is always better.
>>>
>>> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
>>
>> Doesn't this need Fixes and CC stable tags ?
> 
> I don't think so.
> As said in the commit message :
>     Note that given the "w < 4" in the for loop, the buffer can NOT
>     overflow, but using the *right* function is always better.
> 
> linebuf is 38 chars.
> In each iteration, we write 9 bytes + NULL.
> We write only 4 elements per line (because of w < 4), so 9 * 4 + 1 = 37 
> bytes are needed.
> 9 is for %08x<space>
> 
> It can't overflow.
> Moreover, it is really unlikely that the size of linebuf or the number 
> of elements on each line change in a stable kernel.
> 
> So, from my POV, this patch is more a clean-up than anything else.
> 
> I would even agree that it is maybe not even needed. But should someone 
> cut'n'paste it one day, then using the correct function could maybe help 
> him.

OK. Fine. But then maybe the patch title should be something like "Improve
string length computation in mv_dump_mem()" as the "Fix" word you used seems to
be somewhat misleading. With the patch title as is, the stable bot will likely
pick up that patch for stable. Fine with me, unless you see an issue with that.

> 
> CJ
> 
>>
>>> ---
>>>   drivers/ata/sata_mv.c | 4 ++--
>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
>>> index d105db5c7d81..45e48d653c60 100644
>>> --- a/drivers/ata/sata_mv.c
>>> +++ b/drivers/ata/sata_mv.c
>>> @@ -1255,8 +1255,8 @@ static void mv_dump_mem(struct device *dev, void __iomem *start, unsigned bytes)
>>>   
>>>   	for (b = 0; b < bytes; ) {
>>>   		for (w = 0, o = 0; b < bytes && w < 4; w++) {
>>> -			o += snprintf(linebuf + o, sizeof(linebuf) - o,
>>> -				      "%08x ", readl(start + b));
>>> +			o += scnprintf(linebuf + o, sizeof(linebuf) - o,
>>> +				       "%08x ", readl(start + b));
>>>   			b += sizeof(u32);
>>>   		}
>>>   		dev_dbg(dev, "%s: %p: %s\n",
>>

-- 
Damien Le Moal
Western Digital Research