Listing 9. An Example of the lperfex Counter Tool at Work
make
pgf77 -fast -Mcache_align
<@cont_arrow><\#229><@$p>-Mvect=assoc,nosse,cachesize:524288
<@cont_arrow><\#229><@$p>-c cdnz3d.f -o cdnz3d.o
pgf77 -fast -Mcache_align
<@cont_arrow><\#229><@$p>-Mvect=assoc,nosse,cachesize:524288
<@cont_arrow><\#229><@$p> -c sdbdax.f -o sdbdax.o
pgf77 -fast -Mcache_align
<@cont_arrow><\#229><@$p>-Mvect=assoc,nosse,cachesize:524288
<@cont_arrow><\#229><@$p> -o cdnz3d cdnz3d.o sdbdax.o
lperfex -e 41 -e 13 -y ./cdnz3d
./cdnz3d: 188.28 seconds of CPU time and
<@cont_arrow><\#229><@$p>193.68 seconds of real time elapsed
(97.21% CPU utilization) using 0.000000 MB of memory
<@cont_arrow><\#229><@$p>on oscbw01.osc.edu
Event # Event Events Counted
------- ----- --------------
41 Floating point operations
retired (counter 0 only) 23576884423
13 L2 cache lines loaded 151390479
Statistics (aggregated across all threads):
---------------------------------------------
MFLOPS 121.731126
Main memory -> L2 cache bandwidth 25.012884 MB/s
Copyright © 1994 - 2018 Linux Journal. All rights reserved.