Re: [PATCH net 1/3] octeontx2-af: Fix mcs string buffer size

[Subrat Pandey]

________________________________________
From: Jagielski, Jedrzej <jedrzej.jagielski@xxxxxxxxx>
Sent: Thursday, July 30, 2026 12:54 PM
To: Subrat Pandey; linux-kernel@xxxxxxxxxxxxxxx; netdev@xxxxxxxxxxxxxxx
Cc: pabeni@xxxxxxxxxx; kuba@xxxxxxxxxx; edumazet@xxxxxxxxxx; davem@xxxxxxxxxxxxx; andrew+netdev@xxxxxxx; Subbaraya Sundeep Bhatta; Ratheesh Kannoth; Geethasowjanya Akula; Sunil Kovvuri Goutham
Subject: [EXTERNAL] RE: [PATCH net 1/3] octeontx2-af: Fix mcs string buffer size

From: Subrat Pandey subratp@ marvell. com Sent: Thursday, July 30, 2026 8: 30 AM >From: Jagielski, Jedrzej jedrzej. jagielski@ intel. com >Date: Wednesday, 29 July 2026 at 4: 17 PM >To: Subrat Pandey subratp@ marvell. com; linux-kernel@ vger. kernel. org

>From: Subrat Pandey subratp@xxxxxxxxxxx<mailto:subratp@xxxxxxxxxxx>
>Sent: Thursday, July 30, 2026 8:30 AM

>>From: Jagielski, Jedrzej jedrzej.jagielski@xxxxxxxxx<mailto:jedrzej.jagielski@xxxxxxxxx>
>>Date: Wednesday, 29 July 2026 at 4:17 PM
>>To: Subrat Pandey subratp@xxxxxxxxxxx<mailto:subratp@xxxxxxxxxxx>; linux-kernel@xxxxxxxxxxxxxxx<mailto:linux-kernel@xxxxxxxxxxxxxxx> linux-kernel@xxxxxxxxxxxxxxx<mailto:linux-kernel@xxxxxxxxxxxxxxx>; netdev@xxxxxxxxxxxxxxx<mailto:netdev@xxxxxxxxxxxxxxx> netdev@xxxxxxxxxxxxxxx<mailto:netdev@xxxxxxxxxxxxxxx>
>>Cc: pabeni@xxxxxxxxxx<mailto:pabeni@xxxxxxxxxx> pabeni@xxxxxxxxxx<mailto:pabeni@xxxxxxxxxx>; kuba@xxxxxxxxxx<mailto:kuba@xxxxxxxxxx> kuba@xxxxxxxxxx<mailto:kuba@xxxxxxxxxx>; edumazet@xxxxxxxxxx<mailto:edumazet@xxxxxxxxxx> edumazet@xxxxxxxxxx<mailto:edumazet@xxxxxxxxxx>; davem@xxxxxxxxxxxxx<mailto:davem@xxxxxxxxxxxxx> davem@xxxxxxxxxxxxx<mailto:davem@xxxxxxxxxxxxx>; andrew+netdev@xxxxxxx<mailto:andrew+netdev@xxxxxxx> andrew+netdev@xxxxxxx<mailto:andrew+netdev@xxxxxxx>; Subbaraya Sundeep Bhatta sbhatta@xxxxxxxxxxx<mailto:sbhatta@xxxxxxxxxxx>; Ratheesh Kannoth rkannoth@xxxxxxxxxxx<mailto:rkannoth@xxxxxxxxxxx>; Geethasowjanya Akula gakula@xxxxxxxxxxx<mailto:gakula@xxxxxxxxxxx>; Sunil Kovvuri Goutham sgoutham@xxxxxxxxxxx<mailto:sgoutham@xxxxxxxxxxx>
>>Subject: [EXTERNAL] RE: [PATCH net 1/3] octeontx2-af: Fix mcs string buffer size
>>>>From: Subrat Pandey subratp@xxxxxxxxxxx<mailto:subratp@xxxxxxxxxxx>
>>>>Sent: Wednesday, July 29, 2026 9:24 AM
>>
>>>>From: Stefan Wiehler stefan.wiehler@xxxxxxxxx<mailto:stefan.wiehler@xxxxxxxxx>
>>>>
>>>>The "mcs%d" name is built with the unbounded sprintf(), which writes
>>>>without any regard for the size of the destination buffer and can
>>>>overrun it. Switch to snprintf() with sizeof() so the formatted write
>>>>is always clamped to the buffer. The buffer is also enlarged to
>>>>24 bytes to comfortably hold the string for any MCS block index.
>>
>>>Hi Subrat
>>
>>>Is this something that can actually occur during any real scenario?
>>
>>Thanks for the review.
>>
>>No — in practice it can't overflow. The change is defensive hardening rather than a functional fix; we mainly want to avoid the static-analyzer warning.  If you prefer we can also drop this patch.
>>
>>Subrat.
>
>I am just wondering whether we should drop fixes tag for this one if any real
>scenario issue is not fixed in fact. Same for net as a target tree
>

Agreed. We'll drop this patch from the series and re-target it to net-next.

>>
>>>>>>Fixes: d06c2aba5163 ("octeontx2-af: cn10k: mcs: Add debugfs support").
>>>>Signed-off-by: Stefan Wiehler stefan.wiehler@xxxxxxxxx<mailto:stefan.wiehler@xxxxxxxxx>
>>>>Signed-off-by: Subrat Pandey subratp@xxxxxxxxxxx<mailto:subratp@xxxxxxxxxxx>