Re: [PATCH 1/3] conf: More thorough error checking when parsing --mtu option

On Wed, 19 Feb 2025 14:14:27 +1100 David Gibson wrote:

We're a bit sloppy with parsing MTU which can lead to some surprising, though fairly harmless, results: * Passing a non-number like '-m xyz' will not give an error and act like -m 0 * Junk after a number (e.g. '-m 1500pqr') will be ignored rather than giving an error * We parse the MTU as a long, then immediately assign to an int, so on some platforms certain ludicrously out of bounds values will be silently truncated, rather than giving an error

Be a bit more thorough with the error checking to avoid that.

Signed-off-by: David Gibson

Applied. -- Stefano