Result: Affects on non rt sceduling was not measuable by my (non
scietific method). If anyone wants to they could calculate std.dev and
try to interpolate the results. There might be a slight performance
degradation, but the variance of the results are to big. I am still
amazed that linux did not roll over and die doing 40 paralell compiles.
There are some swapping activity in the last runs, shows up on Page
faults.
Method (Compile the linux kernel over night on an othervise idle P200
128Mb):
#!/bin/csh
foreach f (1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
24 25 26 27 29 29 30 31 32 33 34 35 36 37 38 39 40)
make clean > /dev/null
echo -n "$f\: "
time make -j $f bzImage > /dev/null
end
I did not warm up caches, so ignore first line (make -j 1) and I know I
wrote 29 29 in the script.
Results: (rt kernel last night)
1: 869.500u 53.730s 15:41.98 98.0% 0+0k 0+0io 302989pf+0w
2: 824.650u 52.130s 14:37.14 99.9% 0+0k 0+0io 300081pf+0w
3: 789.210u 54.710s 14:04.49 99.9% 0+0k 0+0io 300081pf+0w
4: 781.130u 53.380s 13:55.62 99.8% 0+0k 0+0io 300081pf+0w
5: 775.440u 53.580s 13:49.57 99.9% 0+0k 0+0io 300081pf+0w
6: 785.250u 52.900s 13:58.71 99.9% 0+0k 0+0io 300081pf+0w
7: 779.140u 53.770s 13:53.39 99.9% 0+0k 0+0io 300081pf+0w
8: 787.350u 54.550s 14:02.84 99.8% 0+0k 0+0io 300081pf+0w
9: 818.360u 52.000s 14:31.36 99.8% 0+0k 0+0io 300081pf+0w
10: 799.980u 52.740s 14:13.16 99.9% 0+0k 0+0io 300081pf+0w
11: 796.510u 53.750s 14:10.92 99.9% 0+0k 0+0io 300081pf+0w
12: 831.900u 52.700s 14:45.60 99.8% 0+0k 0+0io 300081pf+0w
13: 828.110u 53.040s 14:41.80 99.9% 0+0k 0+0io 300081pf+0w
14: 800.120u 52.580s 14:12.97 99.9% 0+0k 0+0io 300081pf+0w
15: 799.160u 53.110s 14:12.84 99.9% 0+0k 0+0io 300081pf+0w
16: 832.740u 54.210s 14:48.02 99.8% 0+0k 0+0io 300081pf+0w
17: 805.140u 54.500s 14:20.08 99.9% 0+0k 0+0io 300081pf+0w
18: 793.820u 53.540s 14:07.87 99.9% 0+0k 0+0io 300081pf+0w
19: 826.120u 53.600s 14:40.50 99.9% 0+0k 0+0io 300081pf+0w
20: 824.880u 54.420s 14:40.29 99.8% 0+0k 0+0io 300081pf+0w
21: 805.680u 53.950s 14:19.98 99.9% 0+0k 0+0io 300081pf+0w
22: 803.490u 54.550s 14:18.55 99.9% 0+0k 0+0io 300081pf+0w
23: 837.440u 52.590s 14:50.88 99.9% 0+0k 0+0io 300081pf+0w
24: 827.710u 51.890s 14:40.04 99.9% 0+0k 0+0io 300081pf+0w
25: 801.690u 53.660s 14:15.75 99.9% 0+0k 0+0io 300081pf+0w
26: 812.340u 55.260s 14:28.08 99.9% 0+0k 0+0io 300081pf+0w
27: 840.220u 53.850s 14:54.88 99.9% 0+0k 0+0io 300081pf+0w
29: 812.350u 53.960s 14:26.61 99.9% 0+0k 0+0io 300081pf+0w
29: 796.200u 54.750s 14:11.27 99.9% 0+0k 0+0io 300081pf+0w
30: 835.900u 53.890s 14:50.22 99.9% 0+0k 0+0io 300081pf+0w
31: 829.820u 53.390s 14:44.11 99.8% 0+0k 0+0io 300081pf+0w
32: 811.030u 52.540s 14:23.98 99.9% 0+0k 0+0io 300081pf+0w
33: 810.590u 51.910s 14:22.83 99.9% 0+0k 0+0io 300081pf+0w
34: 842.850u 53.090s 14:56.27 99.9% 0+0k 0+0io 300081pf+0w
35: 817.400u 54.020s 14:31.76 99.9% 0+0k 0+0io 300081pf+0w
...I stopped here and rebooted on linux-2.1.123 (non rtpach) kernel.
More results:
1: 856.900u 52.540s 15:21.29 98.7% 0+0k 0+0io 302346pf+0w
2: 812.920u 52.720s 14:26.05 99.9% 0+0k 0+0io 300081pf+0w
3: 787.130u 53.920s 14:01.56 99.9% 0+0k 0+0io 300081pf+0w
4: 775.310u 55.180s 13:51.66 99.8% 0+0k 0+0io 300081pf+0w
5: 774.400u 53.170s 13:48.02 99.9% 0+0k 0+0io 300081pf+0w
6: 778.150u 53.610s 13:52.28 99.9% 0+0k 0+0io 300081pf+0w
7: 775.060u 53.410s 13:48.85 99.9% 0+0k 0+0io 300081pf+0w
8: 811.260u 53.320s 14:25.29 99.9% 0+0k 0+0io 300081pf+0w
9: 795.850u 53.560s 14:10.40 99.8% 0+0k 0+0io 300081pf+0w
10: 797.910u 52.700s 14:11.52 99.8% 0+0k 0+0io 300081pf+0w
11: 806.860u 53.170s 14:20.52 99.9% 0+0k 0+0io 300081pf+0w
12: 804.510u 53.290s 14:18.38 99.9% 0+0k 0+0io 300081pf+0w
13: 807.630u 53.410s 14:22.23 99.8% 0+0k 0+0io 300081pf+0w
14: 826.450u 52.740s 14:39.70 99.9% 0+0k 0+0io 300081pf+0w
15: 821.630u 54.020s 14:35.99 99.9% 0+0k 0+0io 300081pf+0w
14: 844.520u 54.400s 15:34.55 96.1% 0+0k 0+0io 303225pf+0w
15: 794.270u 53.080s 14:08.00 99.9% 0+0k 0+0io 300187pf+0w
16: 780.050u 53.490s 13:54.39 99.8% 0+0k 0+0io 300081pf+0w
17: 773.960u 54.440s 13:48.60 99.9% 0+0k 0+0io 300081pf+0w
18: 775.900u 54.040s 13:50.25 99.9% 0+0k 0+0io 300081pf+0w
19: 784.780u 54.140s 13:59.15 99.9% 0+0k 0+0io 300081pf+0w
20: 806.120u 53.020s 14:19.95 99.9% 0+0k 0+0io 300081pf+0w
21: 782.490u 53.570s 13:56.67 99.9% 0+0k 0+0io 300081pf+0w
22: 795.780u 53.460s 14:09.51 99.9% 0+0k 0+0io 300081pf+0w
23: 791.640u 52.820s 14:05.18 99.9% 0+0k 0+0io 300081pf+0w
24: 778.060u 54.320s 13:53.48 99.8% 0+0k 0+0io 300081pf+0w
25: 810.280u 52.710s 14:23.51 99.9% 0+0k 0+0io 300081pf+0w
26: 813.380u 52.580s 14:26.21 99.9% 0+0k 0+0io 300081pf+0w
27: 794.820u 52.190s 14:07.47 99.9% 0+0k 0+0io 300081pf+0w
29: 784.990u 54.300s 13:59.62 99.9% 0+0k 0+0io 300081pf+0w
29: 822.710u 51.760s 14:35.39 99.8% 0+0k 0+0io 300081pf+0w
30: 795.300u 54.040s 14:09.80 99.9% 0+0k 0+0io 300081pf+0w
31: 777.510u 53.230s 13:50.96 99.9% 0+0k 0+0io 300081pf+0w
32: 804.590u 53.820s 14:18.70 99.9% 0+0k 0+0io 300081pf+0w
33: 809.950u 54.640s 14:25.54 99.8% 0+0k 0+0io 300081pf+0w
34: 788.730u 53.590s 14:02.87 99.9% 0+0k 0+0io 300081pf+0w
35: 796.000u 54.290s 14:10.58 99.9% 0+0k 0+0io 300081pf+0w
36: 814.000u 52.710s 14:27.01 99.9% 0+0k 0+0io 300081pf+0w
37: 803.120u 55.750s 14:37.24 97.9% 0+0k 0+0io 300564pf+0w
38: 790.470u 53.910s 14:04.58 99.9% 0+0k 0+0io 300550pf+0w
39: 811.770u 53.280s 14:25.26 99.9% 0+0k 0+0io 300081pf+0w
40: 790.720u 54.610s 14:05.85 99.9% 0+0k 0+0io 300081pf+0w
--Hans Eric
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/