[BENCHMARK ctxbench 2.5.59 preempt, no preempt, ingo sched

From: Bill Davidsen (davidsen@tmr.com)
Date: Tue Jan 21 2003 - 17:48:24 EST


The interesting thing about this is that with the Ingo scheduler the
semiphone IPC test hung solid.

All test uni kernel, 59n = nopreempt, 59p = preempt, 59ni = no preempt
Ingo scheduler patch. Response testing will follow, it take a l-o-n-g time
to run.

================================================================
    Run information
================================================================

Run: ALL
  CPU_MHz 348.395
  CPUtype Pentium II (Deschutes)
  HostName oddball.prodigy.com
  KernelName 2.5.59-1n
  Ncpu 1

================================================================
    Results by IPC type
================================================================

                                   loops/sec
SIGUSR1 low high average avg/MHz
  2.5.59n.out 41310 41865 41542 119.239
  2.5.59ni.out 40032 40096 40054 114.946
  2.5.59p.out 39992 40127 40053 114.959

                                   loops/sec
message queue low high average avg/MHz
  2.5.59n.out 76381 76537 76472 219.500
  2.5.59ni.out 71558 73266 72299 207.483
  2.5.59p.out 72676 73391 73059 209.689

                                   loops/sec
pipes low high average avg/MHz
  2.5.59n.out 67564 68632 67952 195.044
  2.5.59ni.out 68234 68307 68258 195.888
  2.5.59p.out 65904 73331 70731 203.006

                                   loops/sec
semiphore low high average avg/MHz
  2.5.59n.out 83412 83720 83536 239.774
  2.5.59ni.out 81495 81572 81535 233.989
  2.5.59p.out 78079 78199 78140 224.271

                                   loops/sec
spin+yield low high average avg/MHz
  2.5.59n.out 181186 181848 181446 520.806
  2.5.59ni.out 158614 176824 170709 489.899
  2.5.59p.out 174010 174036 174023 499.466

                                   loops/sec
spinlock low high average avg/MHz
  2.5.59n.out 3 3 3 0.009
  2.5.59ni.out 4 5 4 0.014
  2.5.59p.out 3 3 3 0.009

-- 
bill davidsen

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Thu Jan 23 2003 - 22:00:27 EST