RE: BUG: build on FC4

Well, if Sarge w/ 2.95 gcc was mis-optimizing a line that wouldn't be the
biggest surprise in the universe, now would it?  That won't be the same for
FC4 w/ gcc 3.3 or 3.4, though...

I guess you could copy the memset() from main() in the cvs into 3.1 and give
that a try.  It would at least initialize the fool thing and couldn't hurt.

But w/o some way to make it fail, I'm going to focus on the cvs (3.2).

-----Burton

-----Original Message-----
From: ntop-dev-bounces <at> unipi.it [mailto:ntop-dev-bounces <at> unipi.it] On Behalf
Of Ola Lundqvist
Sent: Wednesday, August 03, 2005 5:09 PM
To: ntop-dev <at> Unipi.IT
Subject: Re: [Ntop-dev] BUG: build on FC4

Hello

I have got the same problem (on the same function InitDev) on Debian.
The solution to by problem was compiling it from cvs and it worked much
better. It is the reason why 3.1 version never got into the stable release.

I debugged it down to an exact line but it must be some compiler issue or
optimization issue as everything looked good in ddd until a very normal line
(an usual assignment) and it broke.

So I hope for the 3.2 version so I can release that for Debian (not sarge
though). :)