-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathperf-testing.log
More file actions
117 lines (117 loc) · 5.52 KB
/
perf-testing.log
File metadata and controls
117 lines (117 loc) · 5.52 KB
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
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
[2024-12-27 05:24:28] do trig once per laser ->
[PF.END] total: 9.9940s mu: 3.9282ms ±0.3450
[2024-12-27 05:35:17] -lto ->
[PF.END] total: 9.9940s mu: 3.9281ms ±0.3462
[2024-12-27 06:37:18] cache norm_pdf coefficient using object ->
[PF.END] total: 9.9939s mu: 3.9073ms ±0.1837
[2024-12-27 06:46:13] assume all particles intersect same obstacle ->
[PF.END] total: 9.9936s mu: 3.5337ms ±0.1920
[2024-12-27 06:49:45] inline norm pdf eval ->
[PF.END] total: 9.9936s mu: 3.5375ms ±0.1940
[2024-12-30 09:30:07] new toolchain version ->
[PF.END] total: 9.9937s mu: 3.5777ms ±0.1660
[2024-12-30 09:37:11] remove -Ofast and try just -fno-math-errno ->
[PF.END] total: 9.9939s mu: 3.7702ms ±0.192Q4
[2024-12-30 09:39:44] remove -fno-math-errno and try just -funsafe-math-optimizations ->
[PF.END] total: 9.9942s mu: 4.0432ms ±0.1842
[2024-12-30 09:41:09] make clean just to be sure ->
[PF.END] total: 9.9942s mu: 4.0436ms ±0.1843
[2024-12-30 09:44:19] try only -fno-math-errno again ->
[PF.END] total: 9.9939s mu: 3.7703ms ±0.1925
[2024-12-30 09:46:10] try only -ffast-math ->
[PF.END] total: 9.9939s mu: 3.7050ms ±0.1809
[2024-12-30 09:48:22] add -O3 ->
[PF.END] total: 9.9937s mu: 3.5774ms ±0.1655
[2024-12-31 09:35:43] try using clang (see commit: 3118283a21044f20aceb587e09db2732961a9299) ->
[PF.END] total: 9.9935s mu: 3.4073ms ±0.1736
poor man's profiling ->
[2025-01-02 19:51:56] LL.update(): 3.6062±0.1857
[2025-01-02 19:51:56] PF.predict: 1.0027±0.0150
[2025-01-02 19:51:56] PF.update: 2.5671±0.1837
control test ->
[2025-01-24 23:07:45] LL.update(): 3.7112±0.2569
[2025-01-24 23:07:45] PF.predict: 0.9978±0.0046
[2025-01-24 23:07:45] PF.update: 2.6764±0.2566
faster expected dist finding ->
[2025-01-24 23:10:46] LL.update(): 3.6907±0.2619
[2025-01-24 23:10:46] PF.predict: 1.0225±0.0055
[2025-01-24 23:10:46] PF.update: 2.6303±0.2610
remove laser telemetry printing ->
[2025-01-24 23:15:43] LL.update(): 3.6888±0.2640
[2025-01-24 23:15:43] PF.predict: 1.0439±0.0152
[2025-01-24 23:15:43] PF.update: 2.6075±0.2633
use eigen for dist math ->
[2025-01-24 23:35:10] LL.update(): 3.6830±0.2766
[2025-01-24 23:35:10] PF.predict: 1.0434±0.0164
[2025-01-24 23:35:10] PF.update: 2.6025±0.2752
profile Laser.update ->
[2025-01-24 23:38:36] LL.update(): 3.6041±0.2558
[2025-01-24 23:38:36] PF.predict: 1.0097±0.0052
[2025-01-24 23:38:36] PF.update: 2.5570±0.2530
[2025-01-24 23:38:36] Laser.update: 1.2681±0.1493
restart brain ->
[2025-01-24 23:39:45] LL.update(): 3.6041±0.2562
[2025-01-24 23:39:45] PF.predict: 1.0096±0.0054
[2025-01-24 23:39:45] PF.update: 2.5572±0.2535
[2025-01-24 23:39:45] Laser.update: 1.2684±0.1498
clean sim ->
[2025-01-24 23:41:32] LL.update(): 3.6030±0.2555
[2025-01-24 23:41:32] PF.predict: 1.0096±0.0054
[2025-01-24 23:41:32] PF.update: 2.5563±0.2531
[2025-01-24 23:41:32] Laser.update: 1.2676±0.1494
remove Laser.update profiling ->
[2025-01-24 23:42:38] LL.update(): 3.6825±0.2763
[2025-01-24 23:42:38] PF.predict: 1.0435±0.0164
[2025-01-24 23:42:38] PF.update: 2.6019±0.2748
profile in PF::update() ->
[2025-01-24 23:49:03] LL.update(): 3.5977±0.2581
[2025-01-24 23:49:03] PF.predict: 0.9977±0.0044
[2025-01-24 23:49:03] PF.update: 2.5621±0.2554
[2025-01-24 23:49:03] PF.updateWeights: 2.5421±0.2455
replace every std::array with Eigen::Array ->
[2025-01-25 00:15:23] LL.update(): 3.6321±0.2469
[2025-01-25 00:15:23] PF.predict: 0.9932±0.0148
[2025-01-25 00:15:23] PF.update: 2.6021±0.2440
[2025-01-25 00:15:23] PF.updateWeights: 2.5827±0.2348
maybe simplify dist finding code ->
[2025-01-25 00:30:39] LL.update(): 3.6357±0.2531
[2025-01-25 00:30:39] PF.predict: 1.0029±0.0157
[2025-01-25 00:30:39] PF.update: 2.5962±0.2509
[2025-01-25 00:30:39] PF.updateWeights: 2.5777±0.2434
Use Eigen methods instead of a loop ->
[2025-01-25 00:46:06] LL.update(): 3.7396±0.2854
[2025-01-25 00:46:06] PF.predict: 0.9653±0.0074
[2025-01-25 00:46:06] PF.update: 2.7375±0.2841
[2025-01-25 00:46:06] PF.updateWeights: 2.7203±0.2793
simplify Eigen dist finding code? ->
[2025-01-25 01:05:08] LL.update(): 3.6743±0.2781
[2025-01-25 01:05:08] PF.predict: 1.0161±0.0045
[2025-01-25 01:05:08] PF.update: 2.6218±0.2781
[2025-01-25 01:05:08] PF.updateWeights: 2.6050±0.2736
use Eigen methods to vectorize pdf ->
[2025-01-25 01:12:52] LL.update(): 2.7219±0.1988
[2025-01-25 01:12:52] PF.predict: 0.9653±0.0045
[2025-01-25 01:12:52] PF.update: 1.7208±0.1993
[2025-01-25 01:12:52] PF.updateWeights: 1.7042±0.1946
extract min coeff from lambda to function ->
[2025-01-25 01:15:11] LL.update(): 2.7350±0.1888
[2025-01-25 01:15:11] PF.predict: 0.9857±0.0149
[2025-01-25 01:15:11] PF.update: 1.7137±0.1888
[2025-01-25 01:15:11] PF.updateWeights: 1.6971±0.1839
Use Eigen's min function ->
[2025-01-25 01:24:12] LL.update(): 1.4130±0.0126
[2025-01-25 01:24:12] PF.predict: 1.0340±0.0046
[2025-01-25 01:24:12] PF.update: 0.3432±0.0117
[2025-01-25 01:24:12] PF.updateWeights: 0.3268±0.0020
no -Ofast ->
[2025-01-25 19:44:56] LL.update(): 1.8063±0.0214
[2025-01-25 19:44:56] PF.predict: 1.2698±0.0168
[2025-01-25 19:44:56] PF.update: 0.4990±0.0127
[2025-01-25 19:44:56] PF.updateWeights: 0.4695±0.0020
[2025-01-25 19:44:56] PF.printParticles: 0.0028±0.0004
with -Ofast ->
[2025-01-25 19:54:33] LL.update(): 1.4151±0.0198
[2025-01-25 19:54:33] PF.predict: 1.0329±0.0159
[2025-01-25 19:54:33] PF.update: 0.3427±0.0117
[2025-01-25 19:54:33] PF.updateWeights: 0.3262±0.0021
[2025-01-25 19:54:33] PF.printParticles: 0.0024±0.0005