Re: strscpy
Maxime Villard <[email protected]> Tue, 19 May 2020 07:38:58 +0200
| Newsgroups | gmane.os.netbsd.devel.security |
|---|---|
| Message-ID | <[email protected]> |
Le 19/05/2020 à 07:31, Maxime Villard a écrit : > Le 18/05/2020 à 21:03, Martin Husemann a écrit : >> On Mon, May 18, 2020 at 09:00:01PM +0200, Maxime Villard wrote: >>> I believe this is why copystr(), although safer than strlcpy(), isn't used much: >>> the length returned in a pointer is unusual. >> >> Isn't it used (nowadays) only in ~ a single place where it needs to be >> congruent with copyinstr? And it could be totaly removed quite easily? > > copystr() has a small dozen of users. Yes, it could be removed easily, but > we need a better replacement first. I made this patch, the code is simple. > > https://m00nbsd.net/garbage/libkern/strscpy.diff (Mmh, had put the wrong version of my diff, now it's the correct one; we return either -EINVAL or -E2BIG on error)