Re: patch(1): infinite loop on malformed ed script input

On 06/04/2026 18:01, Theo de Raadt wrote:
> Don't you dare break patch this close to an OpenBSD release.
> If this program fails to apply a patch, it will be deeply painful.
> It must work EXACTLY like it did before.  If it cannot work like it
> worked before, there must be a very strong justification as to
> why it is changing.

Only hunk 1 has been committed as far as I know, there are already some 
builds using it, so an obvious issue should have been catched. But if 
you look at it, the second hunk should only solve an issue when original 
source would break anyway and it hasn't been committed yet, so we are 
fine with this one.

Although I agree, both chunks would better be off source until release 
to be 100% sure.

To all devs, please refrain committing the 2nd hunk before release.