OpenCores
URL https://opencores.org/ocsvn/eco32/eco32/trunk

Subversion Repositories eco32

[/] [eco32/] [trunk/] [doc/] [manual/] [isa/] [comp.tex] - Blame information for rev 22

Details | Compare with Previous | View Log

Line No. Rev Author Line
1 22 hellwig
\section{Computation Instructions}
2
 
3
\newcommand{\rdivzero}{\effect if $R_y=0$ then trigger a \name{Division by Zero Fault}}
4
\newcommand{\idivzero}{\effect if $y=0$ then trigger a \name{Division by Zero Fault}}
5
 
6
The computation instructions compute a function of register values and/or immediate values, and store their result in a general-purpose register.
7
 
8
\subsection{ADD}
9
 
10
The ADD instruction computes the sum of two 32-bit register operands, truncated to 32 bits.\\
11
 
12
\rrrformat{000000}
13
 
14
\regeffects{truncate_{32}(R_x + R_y)}
15
 
16
\subsection{ADDI}
17
 
18
The ADDI instruction computes the sum of a 32-bit register operand and a sign-extended 16-bit immediate operand, truncated to 32 bits.\\
19
 
20
\rriformat{000001}
21
 
22
\regeffects{truncate_{32}(R_x + signext_{32}(y))}
23
 
24
\subsection{SUB}
25
 
26
The SUB instruction computes the difference of two 32-bit register operands, truncated to 32 bits.\\
27
 
28
\rrrformat{000010}
29
 
30
\regeffects{truncate_{32}(R_x - R_y)}
31
 
32
\subsection{SUBI}
33
 
34
The SUBI instruction computes the difference of a 32-bit register operand and a sign-extended 16-bit immediate operand, truncated to 32 bits.\\
35
 
36
\rriformat{000011}
37
 
38
\regeffects{truncate_{32}(R_x - signext_{32}(y))}
39
 
40
\subsection{MUL}
41
 
42
The MUL instruction computes the signed product of two 32-bit register operands, truncated to 32 bits.\\
43
 
44
\rrrformat{000100}
45
 
46
\regeffects{truncate_{32}(R_x *_{signed} R_y)}
47
 
48
\subsection{MULI}
49
 
50
The MULI instruction computes the signed product of a 32-bit register operand and a sign-extended 16-bit immediate operand, truncated to 32 bits.\\
51
 
52
\rriformat{000101}
53
 
54
\regeffects{truncate_{32}(R_x *_{signed} signext_{32}(y))}
55
 
56
\subsection{MULU}
57
 
58
The MULU instruction computes the unsigned product of two 32-bit register operands, truncated to 32 bits.\\
59
 
60
\rrrformat{000110}
61
 
62
\regeffects{truncate_{32}(R_x *_{unsigned} R_y)}
63
 
64
\subsection{MULUI}
65
 
66
The MULUI instruction computes the unsigned product of a 32-bit register operand and a zero-extended 16-bit immediate operand, truncated to 32 bits.\\
67
 
68
\rriformat{000111}
69
 
70
\regeffects{truncate_{32}(R_x *_{unsigned} zeroext_{32}(y))}
71
 
72
\subsection{DIV}
73
 
74
The DIV instruction computes the signed quotient of two 32-bit register operands, truncated to 32 bits.\\
75
 
76
\rrrformat{001000}
77
 
78
\begin{effectize}
79
\rdivzero
80
\regeffect{truncate_{32}(R_x /_{signed} R_y)}
81
\end{effectize}
82
 
83
\subsection{DIVI}
84
 
85
The DIVI instruction computes the signed quotient of a 32-bit register operand and a sign-extended 16-bit immediate operand, truncated to 32 bits.\\
86
 
87
\rrrformat{001001}
88
 
89
\begin{effectize}
90
\idivzero
91
\regeffect{truncate_{32}(R_x /_{signed} signext_{32}(y))}
92
\end{effectize}
93
 
94
\subsection{DIVU}
95
 
96
The DIVU instruction computes the unsigned quotient of two unsigned 32-bit register operands, truncated to 32 bits.\\
97
 
98
\rrrformat{001010}
99
 
100
\begin{effectize}
101
\rdivzero
102
\regeffect{truncate_{32}(R_x /_{unsigned} R_y)}
103
\end{effectize}
104
 
105
\subsection{DIVUI}
106
 
107
The DIVUI instruction computes the unsigned quotient of a 32-bit register operand and a zero-extended 16-bit immediate operand, truncated to 32 bits.\\
108
 
109
\rrrformat{001011}
110
 
111
\begin{effectize}
112
\idivzero
113
\regeffect{truncate_{32}(R_x /_{unsigned} zeroext_{32}(y))}
114
\end{effectize}
115
 
116
\subsection{REM}
117
 
118
The REM instruction computes the signed remainder of two 32-bit register operands, truncated to 32 bits.\\
119
 
120
\rrrformat{001100}
121
 
122
\begin{effectize}
123
\rdivzero
124
\regeffect{truncate_{32}(R_x MOD_{signed} R_y)}
125
\end{effectize}
126
 
127
\subsection{REMI}
128
 
129
The REMI instruction computes the signed remainder of a 32-bit register operand and a sign-extended 16-bit immediate operand, truncated to 32 bits.\\
130
 
131
\rrrformat{001101}
132
 
133
\begin{effectize}
134
\idivzero
135
\regeffect{truncate_{32}(R_x MOD_{signed} signext_{32}(y))}
136
\end{effectize}
137
 
138
\subsection{REMU}
139
 
140
The REMU instruction computes the unsigned remainder of two unsigned 32-bit register operands, truncated to 32 bits.\\
141
 
142
\rrrformat{001110}
143
 
144
\begin{effectize}
145
\rdivzero
146
\regeffect{truncate_{32}(R_x MOD_{unsigned} R_y)}
147
\end{effectize}
148
 
149
\subsection{REMUI}
150
 
151
The REMUI instruction computes the unsigned remainder of a 32-bit register operand and a zero-extended 16-bit immediate operand, truncated to 32 bits.\\
152
 
153
\rrrformat{001111}
154
 
155
\begin{effectize}
156
\idivzero
157
\regeffect{truncate_{32}(R_x MOD_{unsigned} zeroext_{32}(y))}
158
\end{effectize}
159
 
160
\subsection{AND}
161
 
162
The AND instruction computes the bitwise AND of two 32-bit register operands.\\
163
 
164
\rrrformat{010000}
165
 
166
\bitregeffects{R_{x,i} \wedge R_{y,i}}
167
 
168
\subsection{ANDI}
169
 
170
The ANDI instruction computes the bitwise AND of a 32-bit register operand and a zero-extended 16-bit immediate operand.\\
171
 
172
\rriformat{010001}
173
 
174
\bitregeffects{R_{x,i} \wedge zeroext_{32}(y)_i}
175
 
176
\subsection{OR}
177
 
178
The OR instruction computes the bitwise OR of two 32-bit register operands.\\
179
 
180
\rrrformat{010010}
181
 
182
\bitregeffects{R_{x,i} \vee R_{y,i}}
183
 
184
\subsection{ORI}
185
 
186
The ORI instruction computes the bitwise OR of a 32-bit register operand and a zero-extended 16-bit immediate operand.\\
187
 
188
\rriformat{010011}
189
 
190
\bitregeffects{R_{x,i} \vee zeroext_{32}(y)_i}
191
 
192
\subsection{XOR}
193
 
194
The XOR instruction computes the bitwise XOR of two 32-bit register operands.\\
195
 
196
\rrrformat{010100}
197
 
198
\bitregeffects{R_{x,i} \oplus R_{y,i}}
199
 
200
\subsection{XORI}
201
 
202
The XORI instruction computes the bitwise XOR of a 32-bit register operand and a zero-extended 16-bit immediate operand.\\
203
 
204
\rriformat{010101}
205
 
206
\bitregeffects{R_{x,i} \oplus zeroext_{32}(y)_i}
207
 
208
\subsection{XNOR}
209
 
210
The XNOR instruction computes the bitwise XNOR of two 32-bit register operands.\\
211
 
212
\rrrformat{010110}
213
 
214
\bitregeffects{\overline{R_{x,i} \oplus R_{y,i}}}
215
 
216
\subsection{XNORI}
217
 
218
The XNORI instruction computes the bitwise XNOR of a 32-bit register operand and a zero-extended 16-bit immediate operand.\\
219
 
220
\rriformat{010111}
221
 
222
\bitregeffects{\overline{R_{x,i} \oplus zeroext_{32}(y)_i}}
223
 
224
\subsection{SLL}
225
 
226
The SLL instruction computes the result of shifting the first 32-bit register operand to the left by a number of bits specified by the 5 least significant bits of the second 32-bit register operand, and filling up with 0 bits.\\
227
 
228
\rrrformat{011000}
229
 
230
\begin{effectize}
231
\effect $shift \leftarrow unsigned(R_{y,4..0})$
232
\effect $temp_i \leftarrow R_{x,i-shift}$ if $i \geq shift$
233
\effect $temp_i \leftarrow 0$ if $i < shift$
234
\effect $R_r \leftarrow temp$
235
\end{effectize}
236
 
237
\subsection{SLLI}
238
 
239
The SLLI instruction computes the result of shifting the 32-bit register operand to the left by a number of bits specified by the 5 least significant bits of the immediate operand, and filling up with 0 bits.\\
240
 
241
\rriformat{011001}
242
 
243
\begin{effectize}
244
\effect $shift \leftarrow unsigned(y_{4..0})$
245
\effect $temp_i \leftarrow R_{x,i-shift}$ if $i \geq shift$
246
\effect $temp_i \leftarrow 0$ if $i < shift$
247
\effect $R_r \leftarrow temp$
248
\end{effectize}
249
 
250
\subsection{SLR}
251
 
252
The SLR instruction computes the result of shifting the first 32-bit register operand to the right by a number of bits specified by the 5 least significant bits of the second 32-bit register operand, and filling up with 0 bits.\\
253
 
254
\rrrformat{011010}
255
 
256
\begin{effectize}
257
\effect $shift \leftarrow unsigned(R_{y,4..0})$
258
\effect $temp_i \leftarrow R_{x,i+shift}$ if $i + shift < 32$
259
\effect $temp_i \leftarrow 0$ if $i + shift \geq 32$
260
\effect $R_r \leftarrow temp$
261
\end{effectize}
262
 
263
\subsection{SLRI}
264
 
265
The SLRI instruction computes the result of shifting the 32-bit register operand to the right by a number of bits specified by the 5 least significant bits of the immediate operand, and filling up with 0 bits.\\
266
 
267
\rriformat{011011}
268
 
269
\begin{effectize}
270
\effect $shift \leftarrow unsigned(y_{4..0})$
271
\effect $temp_i \leftarrow R_{x,i+shift}$ if $i + shift < 32$
272
\effect $temp_i \leftarrow 0$ if $i + shift \geq 32$
273
\effect $R_r \leftarrow temp$
274
\end{effectize}
275
 
276
\subsection{SAR}
277
 
278
The SAR instruction computes the result of shifting the first 32-bit register operand to the right by a number of bits specified by the 5 least significant bits of the second 32-bit register operand, and replicating the topmost (sign) bit.\\
279
 
280
\rrrformat{011100}
281
 
282
\begin{effectize}
283
\effect $shift \leftarrow unsigned(R_{y,4..0})$
284
\effect $temp_i \leftarrow R_{x,i+shift}$ if $i + shift < 32$
285
\effect $temp_i \leftarrow R_{x,31}$ if $i + shift \geq 32$
286
\effect $R_r \leftarrow temp$
287
\end{effectize}
288
 
289
\subsection{SARI}
290
 
291
The SARI instruction computes the result of shifting the 32-bit register operand to the right by a number of bits specified by the 5 least significant bits of the immediate operand, and replicating the topmost (sign) bit.\\
292
 
293
\rriformat{011101}
294
 
295
\begin{effectize}
296
\effect $shift \leftarrow unsigned(y_{4..0})$
297
\effect $temp_i \leftarrow R_{x,i+shift}$ if $i + shift < 32$
298
\effect $temp_i \leftarrow R_{x,31}$ if $i + shift \geq 32$
299
\effect $R_r \leftarrow temp$
300
\end{effectize}
301
 
302
\subsection{LDHI}
303
 
304
The LDHI instruction is used to generate large constants. The upper 16 bits of the result are taken from the 16-bit immediate operand. The lower 16 bits of the result are 0.\\
305
 
306
\rriformat{011111}
307
 
308
\begin{effectize}
309
\effect $R_{r,31..16} \leftarrow y_{15..0}$
310
\effect $R_{r,15..0} \leftarrow 0$
311
\end{effectize}

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.