Re: PR/50746 CVS commit: src/usr.bin/mail

[Steffen Nurpmeso <steffen%sdaoden.eu@localhost>]

The following reply was made to PR bin/50746; it has been noted by GNATS.

From: Steffen Nurpmeso <steffen%sdaoden.eu@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: dcb314%hotmail.com@localhost, gnats-admin%netbsd.org@localhost, netbsd-bugs%netbsd.org@localhost,
        Steffen (Daode) Nurpmeso <steffen%sdaoden.eu@localhost>
Subject: Re: PR/50746 CVS commit: src/usr.bin/mail
Date: Wed, 03 Feb 2016 12:18:45 +0100

 "Christos Zoulas" <christos%netbsd.org@localhost> wrote:
  |The following reply was made to PR bin/50746; it has been noted by GNATS.
  |
  |From: "Christos Zoulas" <christos%netbsd.org@localhost>
  |To: gnats-bugs%gnats.NetBSD.org@localhost
  |Cc: 
  |Subject: PR/50746 CVS commit: src/usr.bin/mail
  |Date: Wed, 3 Feb 2016 00:18:58 -0500
  |
  | Module Name: src
  | Committed By:        christos
  | Date:                Wed Feb  3 05:18:58 UTC 2016
  | 
  | Modified Files:
  |  src/usr.bin/mail: edit.c
  | 
  | Log Message:
  | PR/50746: David Binderman: Check bounds before dereferencing.
 
 That is unnecessary since you will not get there if the argument
 vector is empty, and if you get there it'll be 0 terminated.
 So in a sense the original code is more true.  In fact the check
 against msgCount is what is really fishy here.
 That change seems to originate in DragonFly, the original code
 was already in 2BSD, and i wildly guess that there was a time when
 it wasn't clear how things would proceed, but already then
 getmsglist() ensured the 0 termination.
 
 --steffen