* [BENCHMARK] 2.5.40{-mm1} with contest 0.42
@ 2002-10-01 13:38 Con Kolivas
0 siblings, 0 replies; only message in thread
From: Con Kolivas @ 2002-10-01 13:38 UTC (permalink / raw
To: linux-kernel; +Cc: Andrew Morton
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Here follow the latest benchmarks for 2.5.40 and 2.5.40-mm1:
noload:
Kernel Time CPU% Ratio
2.4.19 67.7 98 1.00
2.5.38 72.4 94 1.07
2.5.38-mm3 73.0 93 1.08
2.5.39 73.2 93 1.08
2.5.39-mm1 73.0 94 1.08
2.5.40 73.2 94 1.08
2.5.40-mm1 73.4 93 1.08
process_load:
Kernel Time CPU% Ratio
2.4.19 110.8 57 1.64
2.5.38 85.7 79 1.27
2.5.38-mm3 96.3 72 1.42
2.5.39 88.9 76 1.31
2.5.39-mm1 99.0 69 1.46
2.5.40 96.4 71 1.43
2.5.40-mm1 94.5 72 1.40
io_load:
Kernel Time CPU% Ratio
2.4.19 216.1 33 3.19
2.5.38 887.8 8 13.11
2.5.38-mm3 105.2 70 1.55
2.5.39 229.4 34 3.39
2.5.39-mm1 239.5 32 3.54
2.5.40 230.3 33 3.40
2.5.40-mm1 117.5 62 1.74
mem_load:
Kernel Time CPU% Ratio
2.4.19 105.4 70 1.56
2.5.38 107.9 73 1.59
2.5.38-mm3 117.1 63 1.73
2.5.39 103.7 72 1.53
2.5.39-mm1 104.6 73 1.54
2.5.40 103.0 73 1.52
2.5.40-mm1 111.3 67 1.64
Summary:
2.5.40 shows no significant difference from 2.5.39 (except in process_load;
now equivalent to 2.5.39-mm1 - the significance of which is not clear of it's
importance)
2.5.40-mm1 shows a substantial improvement under IO load (consistent with the
change in fifo_batch as discussed under the thread "[BENCHMARK] 2.5.39-mm1")
2.5.40-mm1 also shows a slight drop in performance under mem_load. Note of
interest that both 2.5.38-mm3 and 2.5.40-mm1 which have fifo_batch set to 16
instead of 32 have both better IO load performance, and worse mem_load
performance - related?
Con
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
iD8DBQE9maV0F6dfvkL3i1gRAjQZAJ9Ue60e1pe9A/WKigGW3UiMfshngACfT/FW
cpLOyqDqFq7LkaTCMJo51l8=
=yrf+
-----END PGP SIGNATURE-----
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2002-10-01 13:36 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-10-01 13:38 [BENCHMARK] 2.5.40{-mm1} with contest 0.42 Con Kolivas
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.