Re: [PATCH] Don't compile non-OPENSSL code when using OPENSSL andvice versa

["Kris Katterjohn" <kjak () ispwest com>]

From: Andreas Ericsson
Sent: 2/28/2006 1:37:13 PM
> I'm not sure why your replies don't get threaded, but it's fairly 
> annoying. Do you have some weird settings in your MUA?

I think the Subject was too long and some the words "and vice versa" sometimes
get crammed together resulting in a "new" Subject. I went and looked at the list
on insecure.org and the past few have been together. I'll try not to make the
Subjects so long anymore (I didn't really want to make it that long, but thought
the other Subjects I tried of weren't very descriptive at all). Sorry about it.

> Kris Katterjohn wrote:
> > I'm not sure what you mean by two "return within ifdefs".
>
> A return within ifdef is one that makes the compiler remove all code 
> following the #endif, i.e. unconditional return inside an #ifdef block.
>
>
> > The one in
> > service_scan.cc is, but there are two sections of code with more than just a return toward the end of 
> > get_random_bytes() in nbase_rnd.c: one for OpenSSL and
> > not. The #ifdef separates them and also allows for tmp to be separated using #if,
> > too. Or maybe I just didn't understand what you were saying.
>
> Perhaps not. I was most likely unclear. I didn't mean it as criticism, 
> though. I just made an observation.

Okay, now I get what you were saying. It seems like it's just me who was confused
more than you were unclear.


I don't know. I guess if most compilers will optimize it (or whatever you would
say it does in this situation), there's not much need for the patch. Other than
maybe the moving of tmp (and the bottom part which makes it possible) in
nbase_rnd.c.

Thanks,
Kris Katterjohn



_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev