[BENCHMARK] scheduler tunables with contest - max_sleep_average

Con Kolivas (conman@kolivas.net)
Thu, 19 Dec 2002 20:14:17 +1100


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Here are a set of contest results for 2.5.52-mm1 with max_sleep_average
scheduler settings (default is 2000) using the osdl hardware
(http://www.osdl.org)

noload:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [8] 39.7 180 0 0 1.10
max_sle1000 [3] 40.0 179 0 0 1.10
max_sle4000 [4] 39.7 180 0 0 1.10
max_sle500 [3] 39.6 180 0 0 1.09

cacherun:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 36.9 194 0 0 1.02
max_sle1000 [3] 36.8 194 0 0 1.02
max_sle4000 [4] 36.6 194 0 0 1.01
max_sle500 [3] 36.7 194 0 0 1.01

process_load:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 49.0 144 10 50 1.35
max_sle1000 [3] 52.8 133 14 61 1.46
max_sle4000 [4] 46.5 154 8 39 1.28
max_sle500 [3] 48.1 150 9 44 1.33

ctar_load:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 55.5 156 1 10 1.53
max_sle1000 [3] 53.2 161 1 10 1.47
max_sle4000 [4] 51.7 159 1 9 1.43
max_sle500 [3] 54.7 161 1 10 1.51

xtar_load:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 77.4 122 1 8 2.14
max_sle1000 [3] 82.6 117 1 9 2.28
max_sle4000 [3] 66.0 127 1 8 1.82
max_sle500 [3] 67.0 131 1 8 1.85

io_load:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 80.5 108 10 19 2.22
max_sle1000 [3] 89.9 94 14 21 2.48
max_sle4000 [3] 65.2 121 8 18 1.80
max_sle500 [3] 77.4 110 11 21 2.14

io_other:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 60.1 131 7 18 1.66
max_sle1000 [3] 66.3 127 9 20 1.83
max_sle4000 [3] 80.0 111 10 20 2.21
max_sle500 [3] 106.6 101 19 25 2.94

read_load:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 49.9 149 5 6 1.38
max_sle1000 [3] 50.2 148 5 6 1.39
max_sle4000 [3] 50.4 149 5 6 1.39
max_sle500 [3] 50.1 148 5 6 1.38

list_load:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 43.8 167 0 9 1.21
max_sle1000 [3] 43.7 167 0 9 1.21
max_sle4000 [3] 43.6 167 0 9 1.20
max_sle500 [3] 44.1 167 0 9 1.22

mem_load:
Kernel [runs] Time CPU% Loads LCPU% Ratio
2.5.52-mm1 [7] 71.1 123 36 2 1.96
max_sle1000 [3] 103.3 81 36 2 2.85
max_sle4000 [3] 104.4 79 36 2 2.88
max_sle500 [3] 101.3 76 35 2 2.80

Pretty weird. No clear relationship with changes here except perhaps a
proportional (small) change in ctar_load.

Con
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.0 (GNU/Linux)

iD8DBQE+AY3qF6dfvkL3i1gRAj1sAKCf80vYSSfEjl9SI3WFSOZE1HV6cgCggFgj
tivVtTQa/1wiytrsNyNAjow=
=rd8R
-----END PGP SIGNATURE-----
-
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/