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
118
119
120
121
122
123
124
125
126
127
128
|
/*
* Copyright 2015 Google Inc.
*
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
#include "TimingStateMachine.h"
#include "SkCanvas.h"
#include "SkCommandLineFlags.h"
DEFINE_int32(gpuFrameLag, 5, "Overestimate of maximum number of frames GPU is allowed to lag.");
DEFINE_int32(frames, 5, "Number of frames of each skp to render per sample.");
DEFINE_double(loopMs, 5, "Each benchmark will be tuned until it takes loopsMs millseconds.");
TimingStateMachine::TimingStateMachine()
: fCurrentFrame(0)
, fLoops(1)
, fLastMeasurement(0.)
, fState(kPreWarmLoopsPerCanvasPreDraw_State) {
}
TimingStateMachine::ParentEvents TimingStateMachine::nextFrame(SkCanvas* canvas,
Benchmark* benchmark) {
switch (fState) {
case kPreWarmLoopsPerCanvasPreDraw_State:
return this->perCanvasPreDraw(canvas, benchmark, kPreWarmLoops_State);
case kPreWarmLoops_State:
return this->preWarm(kTuneLoops_State);
case kTuneLoops_State:
return this->tuneLoops();
case kPreWarmTimingPerCanvasPreDraw_State:
return this->perCanvasPreDraw(canvas, benchmark, kPreWarmTiming_State);
case kPreWarmTiming_State:
return this->preWarm(kTiming_State);
case kTiming_State:
return this->timing(canvas, benchmark);
}
SkFAIL("Incomplete switch\n");
return kTiming_ParentEvents;
}
inline void TimingStateMachine::nextState(State nextState) {
fState = nextState;
}
TimingStateMachine::ParentEvents TimingStateMachine::perCanvasPreDraw(SkCanvas* canvas,
Benchmark* benchmark,
State nextState) {
benchmark->perCanvasPreDraw(canvas);
benchmark->preDraw(canvas);
fCurrentFrame = 0;
this->nextState(nextState);
return kTiming_ParentEvents;
}
TimingStateMachine::ParentEvents TimingStateMachine::preWarm(State nextState) {
if (fCurrentFrame >= FLAGS_gpuFrameLag) {
// we currently time across all frames to make sure we capture all GPU work
this->nextState(nextState);
fCurrentFrame = 0;
fTimer.start();
} else {
fCurrentFrame++;
}
return kTiming_ParentEvents;
}
inline double TimingStateMachine::elapsed() {
fTimer.end();
return fTimer.fWall;
}
void TimingStateMachine::resetTimingState() {
fCurrentFrame = 0;
fTimer = WallTimer();
}
inline TimingStateMachine::ParentEvents TimingStateMachine::tuneLoops() {
if (1 << 30 == fLoops) {
// We're about to wrap. Something's wrong with the bench.
SkDebugf("InnerLoops wrapped\n");
fLoops = 1;
return kTiming_ParentEvents;
} else {
double elapsedMs = this->elapsed();
if (elapsedMs > FLAGS_loopMs) {
this->nextState(kPreWarmTimingPerCanvasPreDraw_State);
} else {
fLoops *= 2;
this->nextState(kPreWarmLoops_State);
}
this->resetTimingState();
return kReset_ParentEvents;
}
}
void TimingStateMachine::recordMeasurement() {
fLastMeasurement = this->elapsed() / (FLAGS_frames * fLoops);
}
inline TimingStateMachine::ParentEvents TimingStateMachine::timing(SkCanvas* canvas,
Benchmark* benchmark) {
if (fCurrentFrame >= FLAGS_frames) {
this->recordMeasurement();
this->resetTimingState();
return kTimingFinished_ParentEvents;
} else {
fCurrentFrame++;
return kTiming_ParentEvents;
}
}
void TimingStateMachine::nextBenchmark(SkCanvas* canvas, Benchmark* benchmark) {
benchmark->postDraw(canvas);
benchmark->perCanvasPostDraw(canvas);
fLoops = 1;
this->nextState(kPreWarmLoopsPerCanvasPreDraw_State);
}
void TimingStateMachine::nextSampleWithPrewarm() {
this->nextState(kPreWarmTimingPerCanvasPreDraw_State);
}
void TimingStateMachine::nextSample() {
fTimer.start();
}
|
