Re: [PATCH] Remove hardcoded static string length

From: Jeffrey E Altman
Date: Sat May 27 2023 - 10:45:45 EST

On 5/24/2023 1:43 PM, Andrew Lunn wrote:
On Wed, May 24, 2023 at 01:02:36PM -0400, Kenny Ho wrote:
On Wed, May 24, 2023 at 12:02 PM Andrew Lunn <andrew@xxxxxxx> wrote:
So the compiler warning/error needs to be fixed a different want.
Understood. Would caping the length at iov_len with a ternary be sufficient?
The quoted text said 'string'. It is not clear if that means c-string,
with a trailing \0. If you just cap iov_len you could end up with a
string which is not terminated.
The expected buffer is a NUL terminated c-string.
The other end of the socket should not blow up, because that would be
an obvious DOS or buffer overwrite attack vector.

This is a valid concern because all versions of IBM AFS 3.6 Rx and OpenAFS Rx prior to 1.6.23 are susceptible to read beyond the end of buffer if either the received data is longer than 65 octets or the received data is 65 octets but not NUL terminated.

Jeffrey Altman

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature