home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.sysv386
- Path: sparky!uunet!psinntp!spike.insci.com!augie!ross
- From: ross@augie.insci.com (ross andrus)
- Subject: Re: msgrcv(2) bug in SCO ???
- Organization: Insci Incorporated
- Date: Mon, 23 Nov 1992 13:56:50 GMT
- Message-ID: <1992Nov23.135650.3778@spike.insci.com>
- References: <1992Nov20.192851.25181@spike.insci.com>
- Sender: brad@spike.insci.com (Brad Segal)
- Lines: 36
-
- In article <1992Nov20.192851.25181@spike.insci.com> I wrote:
-
- >Hello world
- >
- > I'm writing because I've come up against what looks like a bug in
- > SCO's msgrcv(2) implementation.
-
- <much deleted>
-
- I hate to admit to being stupid as well as hysterical, but in this
- case I must. The whole thing is my fault. I mis-read the man page
- for msgsnd/msgrcv.
-
- It clearly says that the msgsz is the size of the data in the mtext[]
- buffer and does NOT INCLUDE the mtype field. So I was writing off
- the end of my strucutre and hosing my stack. Not SCO or their
- implementation. Me. The fact that my code works on other architectures
- (e.g. RS/6000, DG Aviion) only implies a certain amount of slack in
- *those* implementations.
-
- Further, I got a helpful and courteous e-mail from someone (not a
- support person) in SCO who saw my flames and took the time to help.
-
- Apologies for crying wolf.
-
- ross andrus
- --
- +-----------------------------------------------------------------------------+
- | ross@augie.insci.com |
- | My opinions and views are not shared or endorsed by my employers. Yet. |
- +-----------------------------------------------------------------------------+
- --
- -------------------------------------------------------------------------------
- Brad Segal, INSCI Inc. brad@insci.com
- 130 Cedar Street, 4th Floor
- NY NY 10006 #include <std/disclaimer.h>
-