Dennis Peterson wrote: > Dennis Peterson wrote: > >> I've rebuilt it with my standard script so it should behave again. Sorry for >> that distraction. I'll update the output of the time tests shortly. >> >> dp > > I see in my previous post I gave myself an extra ghz in cpu speed - it's a > 2gig, > not 3. > > RedHat Linux, AMD Athlon 2Mhz, signatures are uncompressed and include all > Sane > Security signatures and MSRBL signatures. The following is unscientific and > anecdotal, but repeatable: > > time clamscan /dev/null > ERROR: Not supported file type (/dev/null) > > ----------- SCAN SUMMARY ----------- > Known viruses: 563038 > Engine version: 0.95rc1 > Scanned directories: 0 > Scanned files: 0 > Infected files: 0 > Data scanned: 0.00 MB > Time: 2.242 sec (0 m 2 s) > > real 0m2.244s > user 0m2.054s > sys 0m0.190s > > > time /usr/local/sbin/clamd > > real 0m2.094s > user 0m1.905s > sys 0m0.188s > > > Solaris 9, Sparc, single 500Mhz cpu (Dual proc returns similar numbers) > > time clamscan /dev/null > ERROR: Not supported file type (/dev/null) > > ----------- SCAN SUMMARY ----------- > Known viruses: 563038 > Engine version: 0.95rc1 > Scanned directories: 0 > Scanned files: 0 > Infected files: 0 > Data scanned: 0.00 MB > Time: 27.203 sec (0 m 27 s) > > real 0m27.24s > user 0m24.94s > sys 0m1.69s > > > time /usr/local/sbin/clamd > > real 0m26.54s > user 0m24.02s > sys 0m1.59s > > I'll try some different optimizations for the Studio 12 compiler and try > again. >
Here's the result of rebuilding ClamAV as a 64-bit app with -mcpu=ultrasparc and -O2: time clamscan /dev/null ERROR: Not supported file type (/dev/null) ----------- SCAN SUMMARY ----------- Known viruses: 563363 Engine version: 0.95rc1 Scanned directories: 0 Scanned files: 0 Infected files: 0 Data scanned: 0.00 MB Time: 17.289 sec (0 m 17 s) real 0m17.34s user 0m14.62s sys 0m2.18s That's quite a pickup in speed (500Mhz Sparc cpu) - I'm cool with that. This is a single proc system but the same test on a dual proc was pretty much the same within half a second. There's not a lot of threading going on in this test. The signature count changed some between tests as would be expected. I was not able to get a succesful compilation of a 64-bit version in Studio 12 - some configuration that is eluding me perhaps, or the complaint about ClamAV but 670 (??) is a big deal in Studio 12 but not in gcc. And it's refreshing to see I have no infected files in /dev/null :) but maybe when scanned files is 0, infected files should be N/A. dp _______________________________________________ Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net http://www.clamav.net/support/ml
