From: Eric Hodel <drbrain@...7.net>
Date: 2006-07-20T06:10:00+09:00
Subject: Re: Patch for Unix socket peer credentials

On Jul 19, 2006, at 1:42 PM, Sam Roberts wrote:

> On Thu, Jul 20, 2006 at 03:02:19AM +0900, Eric Hodel wrote:
>> On FreeBSD uid_t and gid_t are unsigned integers.  Searching google
>                                  ^^^^^^^^
> Surely this is a typo? pid_t must be signed:
>
>   http://www.opengroup.org/onlinepubs/009695399/basedefs/sys/ 
> types.h.html
>
>   "blksize_t, pid_t, and ssize_t shall be signed integer types."

Yes, typo, I meant to write "signed integers".  But not pid_t, uid_t.

>> for 'negative uid' reveals that other operating systems also allow
>> negative uids.
>
> Other __UNIX__ operating systems? Are you sure?
>
> Among other things, this would make the kill(2) system call not work:
>
>   http://www.opengroup.org/onlinepubs/009695399/functions/kill.html
>
> Process IDs are positive. It is true that the pid_t type must allow
> negative values, but not when a pid_t describes a process ID.
>
> Btw, presence of -45 in /etc/passwd doesn't mean that a running  
> process
> will ever actually have negative pid. I'll believe that when I see
> the a ps listing showing it.

You are confusing user ids (uid_t) and process ids (pid_t).

-- 
Eric Hodel - drbrain@segment7.net - http://blog.segment7.net
This implementation is HODEL-HASH-9600 compliant

http://trackmap.robotcoop.com