Wireshark-dev: Re: [Wireshark-dev] Lint of packet-tcp.c

From: Jaap Keuter <jaap.keuter@xxxxxxxxx>
Date: Fri, 15 Aug 2008 07:30:28 +0200
Hi Chris,

Thanks for taking a look. I looked at your new log also, and still wonder how lint get these 'Possible use of null pointer' results. Use of msp is guarded by ipfd_head, so this should never cause a problem. I wonder what lint sees as path to get in trouble.

Thanx,
Jaap

Maynard, Chris wrote:
The changes look good to me.  I re-ran lint on the updated file (with
some of the warnings taken out just to make it a little bit easier to
scan).  There are still a couple of "Possible use of null pointer"
warnings that show up that seem to have merit and might also need some
attention.  The first one is:
                        tcpinfo->seq = msp->seq;
epan\dissectors\packet-tcp.c(1615) : Warning 613: Possible use of null
pointer
    'msp' in left argument to operator '->' [Reference: file
    epan\dissectors\packet-tcp.c: lines 1482, 1504]

Thanks,
Chris

-----Original Message-----
From: wireshark-dev-bounces@xxxxxxxxxxxxx [mailto:wireshark-dev-
bounces@xxxxxxxxxxxxx] On Behalf Of Jaap Keuter
Sent: Wednesday, August 13, 2008 5:50 PM
To: Developer support list for Wireshark
Subject: Re: [Wireshark-dev] Lint of packet-tcp.c

Hi,

I've committed revision 26010 with some fixes for the really obvious
cases.
Please review the changes, found here:
http://anonsvn.wireshark.org/viewvc/index.py?view=rev&revision=26010

Most of the lint output is caused by:
- signed / unsigned differences
- TRY / CATCH / RETHROW macro
which could have merit, that is for further study.

Thanx,
Jaap