Re: [m68k] Final part of architecture cleanup

[Ben Elliston <bje+dated+1143608734 dot 2faef9 at air dot net dot au>]

Hi Nathan

> This is the final part of the architecture switch cleanup.

This patch is okay, with the exception of:

> I reworked find_cf_chip for the new ordering.  Previously that was
> using strncat to try and avoid overrunning the provided array and
> then giving an as_fatal if it did.  However, that had a logic bug in
> that strncat does not write the trailing NUL char if the array gets
> full.  Thus the following strlen's and subsequent strncat's will go
> wandering off anyway.  This patch simply uses strcat and then checks
> if we wandered off or not.  As we never return if that happens, I
> think this is ok.

The problem is that you detect the buffer overflow after it has
occurred; it will likely be too late.  I'd like to see this code made
more robust.

How about adding some assertions, at the very least (as the code only
processes compile-time strings and not external untrusted data)?

Cheers, Ben

Attachment: pgp00000.pgp
Description: PGP signature