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
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
|
/*
* This file is part of tela the Tensor Language.
* Copyright (c) 1994-1996 Pekka Janhunen
*/
#include "prg.H"
#define disp(x) cout << #x " = " << (x) << '\n'
#define disp2(w,x) cout << w " = " << (x) << '\n';
/*
First speed measurements (LINUX) 19.12.1993:
11 ADD sum,i
14 ADD i,#I(1)
17 LT tmp,i,#I(100000)
21 JIF 11,tmp
yielded 1223 clock cycles per iteration, i.e. about 300 cc per instruction,
or about 0.1MIPS. -O6 was used. If no optimization, about 3 times slower.
*/
#if 0
int main()
{
Tprg prg;
Tsymbol sym1("AA");
Tobject obj1(5.6,-2.1);
sym1.value() = &obj1;
Toperand op1(Klit,234);
Toperand op2(sym1);
Toperand op3(Kpar,0);
Toperand op4(Kpar,1);
Toperand op5(Tobject(3.14,2));
Toperand op6(1111);
Tcomplex ztab[6] = {1,2,3,4,5,6};
Toperand op7(Tobject(ztab,TDimPack(2,3)));
disp(op1);
disp(op2);
disp(op3);
disp(op4);
disp(op5);
disp(op6);
disp(op7);
prg.add(Fneg,op1);
prg.add(Fpri,op2);
prg.add(Fdiv,op3,op4,op5);
prg.add(Fmul,op3,op7,op6);
disp(prg);
prg.execute();
return 0;
}
#endif
void gaussian()
{
Tprg prg;
Treal xtab[100];
for (int i=0; i<100; i++) xtab[i] = sin(0.01*i);
Tobject objPi(3.141592654);
Tobject objUnit(1.0);
Tobject objTwo(2);
Tobject objX(xtab,100);
Tsymbol reg0("reg0");
Tsymbol reg1("reg1");
Tsymbol reg2("reg2");
Tsymbol reg3("reg3");
prg.add(Fmove,Toperand(reg1),Toperand(objX));
prg.add(Fmove,Toperand(reg2),Toperand(objPi));
prg.add(Fsqrt,Toperand(reg2));
prg.add(Fdiv,Toperand(reg2),Toperand(objUnit),Toperand(reg2));
prg.add(Fpow,Toperand(reg3),Toperand(reg1),Toperand(objTwo));
prg.add(Fneg,Toperand(reg3));
//prg.add(Fbra,Toperand(Klit,3));
prg.add(Fexp,Toperand(reg3));
prg.add(Fmul,Toperand(reg0),Toperand(reg2),Toperand(reg3));
prg.add(Fpri,Toperand(reg0));
//prg.add(Fbra,Toperand(Klit,-29));
disp(prg);
prg.execute();
}
void looping()
{
Tprg prg;
Tobject objZero(0);
Tobject objUnit(1);
Tobject objTen(10);
Tsymbol i("i");
Tsymbol tmp("tmp");
prg.add(Fmove,Toperand(i),Toperand(objZero));
int pc = prg.length();
prg.add(Fpri,Toperand(i));
prg.add(Fadd,Toperand(i),Toperand(objUnit));
prg.add(Flt,Toperand(tmp),Toperand(i),Toperand(objTen));
//prg.add(Fbif,Toperand(Klit,-11),Toperand(tmp));
prg.add(Fjif,Toperand(Klit,pc),Toperand(tmp));
disp(prg);
prg.execute();
}
void summing()
{
const int N=100000;
Tprg prg;
Tobject objN(N);
Tobject objZero(0);
Tobject objRzero(0.0);
Tobject objUnit(1);
Tsymbol i("i");
Tsymbol V("V");
Tsymbol sum("sum");
Tsymbol tmp("tmp");
prg.add(Fpri,Toperand(objZero));
prg.add(Fmove,Toperand(sum),Toperand(objRzero));
prg.add(Frzeros,Toperand(V),Toperand(objN));
prg.add(Fmove,Toperand(i),Toperand(objZero));
int pc = prg.length();
prg.add(Fadd,Toperand(sum),Toperand(i));
prg.add(Fadd,Toperand(i),Toperand(objUnit));
prg.add(Flt,Toperand(tmp),Toperand(i),Toperand(objN));
prg.add(Fjif,Toperand(Klit,pc),Toperand(tmp));
prg.add(Fpri,Toperand(sum));
disp(prg);
float t1=CPUSeconds();
prg.execute();
float t2=CPUSeconds();
float cc=(t2-t1)*Hz/N;
disp(cc);
}
void flopping()
{
const int N=4000;
const int M=1000;
Tprg prg;
Tobject objN(N);
Tobject objM(M);
Tobject objZero(0);
Tobject objUnit(1);
Tsymbol i("i");
Tsymbol V("V");
Tsymbol tmp("tmp");
prg.add(Fizeros,Toperand(V),Toperand(objN));
prg.add(Fmove,Toperand(i),Toperand(objZero));
int pc = prg.length();
prg.add(Fadd,Toperand(V),Toperand(V));
prg.add(Fadd,Toperand(i),Toperand(objUnit));
prg.add(Flt,Toperand(tmp),Toperand(i),Toperand(objM));
prg.add(Fjif,Toperand(Klit,pc),Toperand(tmp));
disp(prg);
float t1=CPUSeconds();
prg.execute();
float t2=CPUSeconds();
float cc=(t2-t1)*Hz/(N*M);
disp(cc);
}
void indexing()
{
Tprg prg;
const int N=10;
Treal xtab[N];
for (int i=0; i<N; i++) xtab[i] = sin(i/Treal(N));
Tobject objX(xtab,TDimPack(2,5));
Tobject ii(1);
Tobject jj(3);
Tobject newval(-2.0);
Tsymbol reg0("reg0");
Tsymbol reg1("reg1");
prg.add(Fmove,Toperand(reg1),Toperand(objX));
prg.add(Fgath,Toperand(reg0),Toperand(reg1),Toperand(ii),Toperand(jj));
prg.add(Fpri,Toperand(reg0));
prg.add(Fscat,Toperand(newval),Toperand(reg1),Toperand(ii),Toperand(jj));
prg.add(Fpri,Toperand(reg1));
Tint itab[2],jtab[2];
itab[0]=1; itab[1]=0; jtab[0]=2; jtab[1]=3;
Tobject iitab(itab,2), jjtab(jtab,2);
prg.add(Fmgath,Toperand(reg0),Toperand(reg1),Toperand(iitab),Toperand(jjtab));
prg.add(Fpri,Toperand(reg0));
prg.add(Fscat,Toperand(newval),Toperand(reg0),Toperand(ii));
prg.add(Fpri,Toperand(reg0));
prg.add(Fmscat,Toperand(reg0),Toperand(reg1),Toperand(iitab),Toperand(jjtab));
prg.add(Fpri,Toperand(reg1));
disp(prg);
prg.execute();
}
int main()
{
//gaussian();
//looping();
//summing();
//flopping();
indexing();
PerformanceReport();
return 0;
}
|
