Re: [GIT PULL] execve updates for v6.13-rc1 (take 2)

[Al Viro]

On Fri, Nov 29, 2024 at 10:42:47PM +0100, Vegard Nossum wrote:

> Probably a silly question, but why not do the same thing in all cases?
> file->f_path.dentry->d_name.name _is_ the basename of whatever the
> normal path resolution machinery ended up with; why not use just it
> unconditionally? It handled empty paths already. This also seems to
> match what you wrote in https://lore.kernel.org/all/CAHk-=wgKgi5eqo6oW0bBS2-Cr+d4jraoKfVq6wbmdiWWsZbMLw@xxxxxxxxxxxxxx/
> (no argv[0], no magic strings, no symlinks):

User-visible change of behaviour, quite possibly confusing existing
scripts...

Let's not go there, it's a silly place.