Re: Performance thoughts

[Fyodor <fyodor () insecure org>]

On Wed, Jul 05, 2006 at 12:57:10AM +0200, majek04 wrote:
> Of course you're right, here are some tests:(Debian sid on x86)
> Compilation: "./configure --without-openssl"
> Test: "callgrind ./nmap -sS -n -T5 127.0.0.1/22 -p22"
>
> Default nmap compilation:
> C(XX)FLAGS=""         32.51s user 0.39s system 99% cpu 33.115 total
> The same with debugging information:
> C(XX)FLAGS="-g"               30.85s user 0.37s system 99% cpu 31.317 total
>
> And with forced optimizations:
> C(XX)FLAGS="-O0"      30.44s user 0.37s system 99% cpu 31.019 total
> C(XX)FLAGS="-O1"      8.10s user 0.28s system 99% cpu 8.432 total
> C(XX)FLAGS="-O2"      7.07s user 0.29s system 99% cpu 7.403 total

Does your patch make a difference with -O2?

> We see that removing "-g" changes nothing.
> It seems that default nmap compilation is done without any "-O".

Something is wrong if you aren't getting -O2.  I just tried the nmap
4.20ALPHA3 tarball (and also my SVN version) and get this for the .cc files:

g++ -c  -Ilibdnet-stripped/include -g -O2 -Wall  -fno-strict-aliasing -Ilibpcap -Inbase -Insock/include  
-DHAVE_CONFIG_H -DNMAP_VERSION=\"4.20ALPHA3\" -DNMAP_NAME=\"Nmap\" -DNMAP_URL=\"http://www.insecure.org/nmap/\"; 
-DNMAP_PLATFORM=\"x86_64-unknown-linux-gnu\" -DNMAPDATADIR=\"/usr/local/share/nmap\" nmap.cc -o nmap.o

Now the final linking command probably doesn't use -O2, but I don't
think it would do anything there.

> > I like these graphs.  What are you using to create them?
> Like Richard Moore suggested, callgrind and kcachegrind.

Neat.

Cheers,
-F


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