6 //14 VECTOR OPERATIONS
\r
9 //Dreg_hi = Dreg_lo = SIGN ( Dreg_hi ) * Dreg_hi + SIGN ( Dreg_lo ) * Dreg_lo ; /* (b) */
\r
11 r7.h=r7.l=sign(r2.h)*r3.h+sign(r2.l)*r3.l ;
\r
12 r0.h=r0.l=sign(r1.h)*r2.h+sign(r1.l)*r2.l ;
\r
13 r3.h=r3.l=sign(r4.h)*r5.h+sign(r4.l)*r5.l ;
\r
14 r6.h=r6.l=sign(r7.h)*r0.h+sign(r7.l)*r0.l ;
\r
15 r1.h=r1.l=sign(r2.h)*r3.h+sign(r2.l)*r3.l ;
\r
16 r4.h=r4.l=sign(r5.h)*r6.h+sign(r5.l)*r6.l ;
\r
17 r7.h=r7.l=sign(r0.h)*r1.h+sign(r0.l)*r1.l ;
\r
18 r2.h=r2.l=sign(r3.h)*r4.h+sign(r3.l)*r4.l ;
\r
20 //Dual 16-Bit Operation
\r
21 //Dreg = VIT_MAX ( Dreg , Dreg ) (ASL) ; /* shift history bits left (b) */
\r
22 //Dreg = VIT_MAX ( Dreg , Dreg ) (ASR) ; /* shift history bits right (b) */
\r
23 //Single 16-Bit Operation
\r
24 //Dreg_lo = VIT_MAX ( Dreg ) (ASL) ; /* shift history bits left (b) */
\r
25 //Dreg_lo = VIT_MAX ( Dreg ) (ASR) ; /* shift history bits right (b) */
\r
26 r5 = vit_max(r3, r2)(asl) ; /* shift left, dual operation */
\r
27 r7 = vit_max (r1, r0) (asr) ; /* shift right, dual operation */
\r
29 r0 = vit_max(r1, r2)(asl) ; /* shift left, dual operation */
\r
30 r3 = vit_max (r4, r5) (asr) ; /* shift right, dual operation */
\r
31 r6 = vit_max(r7, r0)(asl) ; /* shift left, dual operation */
\r
32 r1 = vit_max (r2, r3) (asr) ; /* shift right, dual operation */
\r
33 r4 = vit_max(r5, r6)(asl) ; /* shift left, dual operation */
\r
34 r7 = vit_max (r0, r1) (asr) ; /* shift right, dual operation */
\r
35 r2 = vit_max(r3, r4)(asl) ; /* shift left, dual operation */
\r
36 r5 = vit_max (r6, r7) (asr) ; /* shift right, dual operation */
\r
39 r3.l = vit_max (r1)(asl) ; /* shift left, single operation */
\r
40 r3.l = vit_max (r1)(asr) ; /* shift right, single operation */
\r
42 r0.l = vit_max (r1)(asl) ; /* shift left, single operation */
\r
43 r2.l = vit_max (r3)(asr) ; /* shift right, single operation */
\r
44 r4.l = vit_max (r5)(asl) ; /* shift left, single operation */
\r
45 r6.l = vit_max (r7)(asr) ; /* shift right, single operation */
\r
46 r1.l = vit_max (r2)(asl) ; /* shift left, single operation */
\r
47 r3.l = vit_max (r4)(asr) ; /* shift right, single operation */
\r
48 r5.l = vit_max (r6)(asl) ; /* shift left, single operation */
\r
49 r7.l = vit_max (r0)(asr) ; /* shift right, single operation */
\r
51 //Dreg = ABS Dreg (V) ; /* (b) */
\r
64 //Dual 16-Bit Operations
\r
65 //Dreg = Dreg +|+ Dreg (opt_mode_0) ; /* add | add (b) */
\r
66 r5=r3 +|+ r4 ; /* dual 16-bit operations, add|add */
\r
96 //Dreg = Dreg
\96|+ Dreg (opt_mode_0) ; /* subtract | add (b) */
\r
97 r6=r0 -|+ r1(s) ; /* same as above, subtract|add with saturation */
\r
115 r6=r7 -|+ r0 (CO) ;
\r
120 r0=r1 -|+ r2 (SCO);
\r
121 r3=r4 -|+ r5 (SCO);
\r
122 r6=r7 -|+ r0 (SCO);
\r
123 r1=r2 -|+ r3 (SCO);
\r
124 r4=r3 -|+ r5 (SCO);
\r
125 r6=r3 -|+ r7 (SCO);
\r
128 //Dreg = Dreg +|
\96 Dreg (opt_mode_0) ; /* add | subtract (b) */
\r
129 r0=r2 +|- r1(co) ; /* add|subtract with half-word results crossed over in the destination register */
\r
147 r6=r7 +|- r0 (CO) ;
\r
152 r0=r1 +|- r2 (SCO);
\r
153 r3=r4 +|- r5 (SCO);
\r
154 r6=r7 +|- r0 (SCO);
\r
155 r1=r2 +|- r3 (SCO);
\r
156 r4=r3 +|- r5 (SCO);
\r
157 r6=r3 +|- r7 (SCO);
\r
159 //Dreg = Dreg
\96|
\96 Dreg (opt_mode_0) ; /* subtract | subtract (b) */
\r
160 r7=r3 -|- r6(sco) ; /* subtract|subtract with saturation and half-word results crossed over in the destination register */
\r
178 r6=r7 -|- r0 (CO) ;
\r
183 r0=r1 -|- r2 (SCO);
\r
184 r3=r4 -|- r5 (SCO);
\r
185 r6=r7 -|- r0 (SCO);
\r
186 r1=r2 -|- r3 (SCO);
\r
187 r4=r3 -|- r5 (SCO);
\r
188 r6=r3 -|- r7 (SCO);
\r
190 //Quad 16-Bit Operations
\r
191 //Dreg = Dreg +|+ Dreg, Dreg = Dreg
\96|
\96 Dreg (opt_mode_0,opt_mode_2) ; /* add | add, subtract | subtract; the set of source registers must be the same for each operation (b) */
\r
192 r5=r3 +|+ r4, r7=r3-|-r4 ; /* quad 16-bit operations, add|add, subtract|subtract */
\r
194 r0=r1 +|+ r2, r7=r1 -|- r2;
\r
195 r3=r4 +|+ r5, r6=r4 -|- r5;
\r
196 r6=r7 +|+ r0, r5=r7 -|- r0;
\r
197 r1=r2 +|+ r3, r4=r2 -|- r3;
\r
198 r4=r3 +|+ r5, r3=r3 -|- r5;
\r
199 r6=r3 +|+ r7, r2=r3 -|- r7;
\r
201 r0=r1 +|+ r2, r7=r1 -|- r2(S);
\r
202 r3=r4 +|+ r5, r6=r4 -|- r5(S);
\r
203 r6=r7 +|+ r0, r5=r7 -|- r0(S);
\r
204 r1=r2 +|+ r3, r4=r2 -|- r3(S);
\r
205 r4=r3 +|+ r5, r3=r3 -|- r5(S);
\r
206 r6=r3 +|+ r7, r2=r3 -|- r7(S);
\r
209 r0=r1 +|+ r2, r7=r1 -|- r2(CO);
\r
210 r3=r4 +|+ r5, r6=r4 -|- r5(CO);
\r
211 r6=r7 +|+ r0, r5=r7 -|- r0(CO);
\r
212 r1=r2 +|+ r3, r4=r2 -|- r3(CO);
\r
213 r4=r3 +|+ r5, r3=r3 -|- r5(CO);
\r
214 r6=r3 +|+ r7, r2=r3 -|- r7(CO);
\r
217 r0=r1 +|+ r2, r7=r1 -|- r2(SCO);
\r
218 r3=r4 +|+ r5, r6=r4 -|- r5(SCO);
\r
219 r6=r7 +|+ r0, r5=r7 -|- r0(SCO);
\r
220 r1=r2 +|+ r3, r4=r2 -|- r3(SCO);
\r
221 r4=r3 +|+ r5, r3=r3 -|- r5(SCO);
\r
222 r6=r3 +|+ r7, r2=r3 -|- r7(SCO);
\r
224 r0=r1 +|+ r2, r7=r1 -|- r2(ASR);
\r
225 r3=r4 +|+ r5, r6=r4 -|- r5(ASR);
\r
226 r6=r7 +|+ r0, r5=r7 -|- r0(ASR);
\r
227 r1=r2 +|+ r3, r4=r2 -|- r3(ASR);
\r
228 r4=r3 +|+ r5, r3=r3 -|- r5(ASR);
\r
229 r6=r3 +|+ r7, r2=r3 -|- r7(ASR);
\r
232 r0=r1 +|+ r2, r7=r1 -|- r2(ASL);
\r
233 r3=r4 +|+ r5, r6=r4 -|- r5(ASL);
\r
234 r6=r7 +|+ r0, r5=r7 -|- r0(ASL);
\r
235 r1=r2 +|+ r3, r4=r2 -|- r3(ASL);
\r
236 r4=r3 +|+ r5, r3=r3 -|- r5(ASL);
\r
237 r6=r3 +|+ r7, r2=r3 -|- r7(ASL);
\r
240 r0=r1 +|+ r2, r7=r1 -|- r2(S,ASR);
\r
241 r3=r4 +|+ r5, r6=r4 -|- r5(S,ASR);
\r
242 r6=r7 +|+ r0, r5=r7 -|- r0(S,ASR);
\r
243 r1=r2 +|+ r3, r4=r2 -|- r3(S,ASR);
\r
244 r4=r3 +|+ r5, r3=r3 -|- r5(S,ASR);
\r
245 r6=r3 +|+ r7, r2=r3 -|- r7(S,ASR);
\r
248 r0=r1 +|+ r2, r7=r1 -|- r2(CO,ASR);
\r
249 r3=r4 +|+ r5, r6=r4 -|- r5(CO,ASR);
\r
250 r6=r7 +|+ r0, r5=r7 -|- r0(CO,ASR);
\r
251 r1=r2 +|+ r3, r4=r2 -|- r3(CO,ASR);
\r
252 r4=r3 +|+ r5, r3=r3 -|- r5(CO,ASR);
\r
253 r6=r3 +|+ r7, r2=r3 -|- r7(CO,ASR);
\r
256 r0=r1 +|+ r2, r7=r1 -|- r2(SCO,ASR);
\r
257 r3=r4 +|+ r5, r6=r4 -|- r5(SCO,ASR);
\r
258 r6=r7 +|+ r0, r5=r7 -|- r0(SCO,ASR);
\r
259 r1=r2 +|+ r3, r4=r2 -|- r3(SCO,ASR);
\r
260 r4=r3 +|+ r5, r3=r3 -|- r5(SCO,ASR);
\r
261 r6=r3 +|+ r7, r2=r3 -|- r7(SCO,ASR);
\r
263 r0=r1 +|+ r2, r7=r1 -|- r2(S,ASL);
\r
264 r3=r4 +|+ r5, r6=r4 -|- r5(S,ASL);
\r
265 r6=r7 +|+ r0, r5=r7 -|- r0(S,ASL);
\r
266 r1=r2 +|+ r3, r4=r2 -|- r3(S,ASL);
\r
267 r4=r3 +|+ r5, r3=r3 -|- r5(S,ASL);
\r
268 r6=r3 +|+ r7, r2=r3 -|- r7(S,ASL);
\r
271 r0=r1 +|+ r2, r7=r1 -|- r2(CO,ASL);
\r
272 r3=r4 +|+ r5, r6=r4 -|- r5(CO,ASL);
\r
273 r6=r7 +|+ r0, r5=r7 -|- r0(CO,ASL);
\r
274 r1=r2 +|+ r3, r4=r2 -|- r3(CO,ASL);
\r
275 r4=r3 +|+ r5, r3=r3 -|- r5(CO,ASL);
\r
276 r6=r3 +|+ r7, r2=r3 -|- r7(CO,ASL);
\r
279 r0=r1 +|+ r2, r7=r1 -|- r2(SCO,ASL);
\r
280 r3=r4 +|+ r5, r6=r4 -|- r5(SCO,ASL);
\r
281 r6=r7 +|+ r0, r5=r7 -|- r0(SCO,ASL);
\r
282 r1=r2 +|+ r3, r4=r2 -|- r3(SCO,ASL);
\r
283 r4=r3 +|+ r5, r3=r3 -|- r5(SCO,ASL);
\r
284 r6=r3 +|+ r7, r2=r3 -|- r7(SCO,ASL);
\r
287 //Dreg = Dreg +|
\96 Dreg, Dreg = Dreg
\96|+ Dreg (opt_mode_0,opt_mode_2) ; /* add | subtract, subtract | add; the set of source registers must be the same for each operation (b) */
\r
288 r5=r3 +|- r4, r7=r3 -|+ r4 ; /* quad 16-bit operations, add|subtract, subtract|add */
\r
290 r0=r1 +|- r2, r7=r1 -|+ r2;
\r
291 r3=r4 +|- r5, r6=r4 -|+ r5;
\r
292 r6=r7 +|- r0, r5=r7 -|+ r0;
\r
293 r1=r2 +|- r3, r4=r2 -|+ r3;
\r
294 r4=r3 +|- r5, r3=r3 -|+ r5;
\r
295 r6=r3 +|- r7, r2=r3 -|+ r7;
\r
297 r0=r1 +|- r2, r7=r1 -|+ r2(S);
\r
298 r3=r4 +|- r5, r6=r4 -|+ r5(S);
\r
299 r6=r7 +|- r0, r5=r7 -|+ r0(S);
\r
300 r1=r2 +|- r3, r4=r2 -|+ r3(S);
\r
301 r4=r3 +|- r5, r3=r3 -|+ r5(S);
\r
302 r6=r3 +|- r7, r2=r3 -|+ r7(S);
\r
305 r0=r1 +|- r2, r7=r1 -|+ r2(CO);
\r
306 r3=r4 +|- r5, r6=r4 -|+ r5(CO);
\r
307 r6=r7 +|- r0, r5=r7 -|+ r0(CO);
\r
308 r1=r2 +|- r3, r4=r2 -|+ r3(CO);
\r
309 r4=r3 +|- r5, r3=r3 -|+ r5(CO);
\r
310 r6=r3 +|- r7, r2=r3 -|+ r7(CO);
\r
313 r0=r1 +|- r2, r7=r1 -|+ r2(SCO);
\r
314 r3=r4 +|- r5, r6=r4 -|+ r5(SCO);
\r
315 r6=r7 +|- r0, r5=r7 -|+ r0(SCO);
\r
316 r1=r2 +|- r3, r4=r2 -|+ r3(SCO);
\r
317 r4=r3 +|- r5, r3=r3 -|+ r5(SCO);
\r
318 r6=r3 +|- r7, r2=r3 -|+ r7(SCO);
\r
320 r0=r1 +|- r2, r7=r1 -|+ r2(ASR);
\r
321 r3=r4 +|- r5, r6=r4 -|+ r5(ASR);
\r
322 r6=r7 +|- r0, r5=r7 -|+ r0(ASR);
\r
323 r1=r2 +|- r3, r4=r2 -|+ r3(ASR);
\r
324 r4=r3 +|- r5, r3=r3 -|+ r5(ASR);
\r
325 r6=r3 +|- r7, r2=r3 -|+ r7(ASR);
\r
328 r0=r1 +|- r2, r7=r1 -|+ r2(ASL);
\r
329 r3=r4 +|- r5, r6=r4 -|+ r5(ASL);
\r
330 r6=r7 +|- r0, r5=r7 -|+ r0(ASL);
\r
331 r1=r2 +|- r3, r4=r2 -|+ r3(ASL);
\r
332 r4=r3 +|- r5, r3=r3 -|+ r5(ASL);
\r
333 r6=r3 +|- r7, r2=r3 -|+ r7(ASL);
\r
336 r0=r1 +|- r2, r7=r1 -|+ r2(S,ASR);
\r
337 r3=r4 +|- r5, r6=r4 -|+ r5(S,ASR);
\r
338 r6=r7 +|- r0, r5=r7 -|+ r0(S,ASR);
\r
339 r1=r2 +|- r3, r4=r2 -|+ r3(S,ASR);
\r
340 r4=r3 +|- r5, r3=r3 -|+ r5(S,ASR);
\r
341 r6=r3 +|- r7, r2=r3 -|+ r7(S,ASR);
\r
344 r0=r1 +|- r2, r7=r1 -|+ r2(CO,ASR);
\r
345 r3=r4 +|- r5, r6=r4 -|+ r5(CO,ASR);
\r
346 r6=r7 +|- r0, r5=r7 -|+ r0(CO,ASR);
\r
347 r1=r2 +|- r3, r4=r2 -|+ r3(CO,ASR);
\r
348 r4=r3 +|- r5, r3=r3 -|+ r5(CO,ASR);
\r
349 r6=r3 +|- r7, r2=r3 -|+ r7(CO,ASR);
\r
352 r0=r1 +|- r2, r7=r1 -|+ r2(SCO,ASR);
\r
353 r3=r4 +|- r5, r6=r4 -|+ r5(SCO,ASR);
\r
354 r6=r7 +|- r0, r5=r7 -|+ r0(SCO,ASR);
\r
355 r1=r2 +|- r3, r4=r2 -|+ r3(SCO,ASR);
\r
356 r4=r3 +|- r5, r3=r3 -|+ r5(SCO,ASR);
\r
357 r6=r3 +|- r7, r2=r3 -|+ r7(SCO,ASR);
\r
359 r0=r1 +|- r2, r7=r1 -|+ r2(S,ASL);
\r
360 r3=r4 +|- r5, r6=r4 -|+ r5(S,ASL);
\r
361 r6=r7 +|- r0, r5=r7 -|+ r0(S,ASL);
\r
362 r1=r2 +|- r3, r4=r2 -|+ r3(S,ASL);
\r
363 r4=r3 +|- r5, r3=r3 -|+ r5(S,ASL);
\r
364 r6=r3 +|- r7, r2=r3 -|+ r7(S,ASL);
\r
367 r0=r1 +|- r2, r7=r1 -|+ r2(CO,ASL);
\r
368 r3=r4 +|- r5, r6=r4 -|+ r5(CO,ASL);
\r
369 r6=r7 +|- r0, r5=r7 -|+ r0(CO,ASL);
\r
370 r1=r2 +|- r3, r4=r2 -|+ r3(CO,ASL);
\r
371 r4=r3 +|- r5, r3=r3 -|+ r5(CO,ASL);
\r
372 r6=r3 +|- r7, r2=r3 -|+ r7(CO,ASL);
\r
375 r0=r1 +|- r2, r7=r1 -|+ r2(SCO,ASL);
\r
376 r3=r4 +|- r5, r6=r4 -|+ r5(SCO,ASL);
\r
377 r6=r7 +|- r0, r5=r7 -|+ r0(SCO,ASL);
\r
378 r1=r2 +|- r3, r4=r2 -|+ r3(SCO,ASL);
\r
379 r4=r3 +|- r5, r3=r3 -|+ r5(SCO,ASL);
\r
380 r6=r3 +|- r7, r2=r3 -|+ r7(SCO,ASL);
\r
384 //Dual 32-Bit Operations
\r
385 //Dreg = Dreg + Dreg, Dreg = Dreg - Dreg (opt_mode_1) ; /* add, subtract; the set of source registers must be the same for each operation (b) */
\r
386 r2=r0+r1, r3=r0-r1 ; /* 32-bit operations */
\r
388 r7=r0+r1, r0=r0-r1 ; /* 32-bit operations */
\r
389 r6=r1+r2, r1=r1-r2 ; /* 32-bit operations */
\r
390 r5=r2+r3, r2=r2-r3 ; /* 32-bit operations */
\r
391 r4=r3+r4, r3=r3-r4 ; /* 32-bit operations */
\r
392 r3=r4+r5, r4=r4-r5 ; /* 32-bit operations */
\r
393 r2=r5+r6, r5=r5-r6 ; /* 32-bit operations */
\r
394 r1=r6+r7, r6=r6-r7 ; /* 32-bit operations */
\r
395 r0=r7+r0, r7=r7-r0 ; /* 32-bit operations */
\r
397 r2=r0+r1, r3=r0-r1(s) ; /* dual 32-bit operations with saturation */
\r
398 r7=r0+r1, r0=r0-r1 (s); /* 32-bit operations */
\r
399 r6=r1+r2, r1=r1-r2 (s); /* 32-bit operations */
\r
400 r5=r2+r3, r2=r2-r3 (s); /* 32-bit operations */
\r
401 r4=r3+r4, r3=r3-r4(s) ; /* 32-bit operations */
\r
402 r3=r4+r5, r4=r4-r5 (s); /* 32-bit operations */
\r
403 r2=r5+r6, r5=r5-r6 (s); /* 32-bit operations */
\r
404 r1=r6+r7, r6=r6-r7 (s); /* 32-bit operations */
\r
405 r0=r7+r0, r7=r7-r0 (s); /* 32-bit operations */
\r
409 //Dual 40-Bit Accumulator Operations
\r
410 //Dreg = A1 + A0, Dreg = A1 - A0 (opt_mode_1) ; /* add, subtract Accumulators; subtract A0 from A1 (b) */
\r
411 r0=a1+a0, r1=a1-a0 ;
\r
412 r2=a1+a0, r3=a1-a0 ;
\r
413 r4=a1+a0, r5=a1-a0 ;
\r
414 r6=a1+a0, r7=a1-a0 ;
\r
415 r1=a1+a0, r0=a1-a0 ;
\r
416 r3=a1+a0, r2=a1-a0 ;
\r
417 r5=a1+a0, r4=a1-a0 ;
\r
419 r0=a1+a0, r1=a1-a0 (s);
\r
420 r2=a1+a0, r3=a1-a0 (s);
\r
421 r4=a1+a0, r5=a1-a0 (s);
\r
422 r6=a1+a0, r7=a1-a0 (s);
\r
423 r1=a1+a0, r0=a1-a0 (s);
\r
424 r3=a1+a0, r2=a1-a0 (s);
\r
425 r5=a1+a0, r4=a1-a0 (s);
\r
427 //Dreg = A0 + A1, Dreg = A0 - A1 (opt_mode_1) ; /* add, subtract Accumulators; subtract A1 from A0 (b) */
\r
428 r4=a0+a1, r6=a0-a1(s);
\r
430 r0=a0+a1, r1=a0-a1 ;
\r
431 r2=a0+a1, r3=a0-a1 ;
\r
432 r4=a0+a1, r5=a0-a1 ;
\r
433 r6=a0+a1, r7=a0-a1 ;
\r
434 r1=a0+a1, r0=a0-a1 ;
\r
435 r3=a0+a1, r2=a0-a1 ;
\r
436 r5=a0+a1, r4=a0-a1 ;
\r
438 r0=a0+a1, r1=a0-a1 (s);
\r
439 r2=a0+a1, r3=a0-a1 (s);
\r
440 r4=a0+a1, r5=a0-a1 (s);
\r
441 r6=a0+a1, r7=a0-a1 (s);
\r
442 r1=a0+a1, r0=a0-a1 (s);
\r
443 r3=a0+a1, r2=a0-a1 (s);
\r
444 r5=a0+a1, r4=a0-a1 (s);
\r
446 //Constant Shift Magnitude
\r
447 //Dreg = Dreg >>> uimm4 (V) ; /* arithmetic shift right, immediate (b) */
\r
460 //Dreg = Dreg << uimm4 (V,S) ; /* arithmetic shift left, immediate with saturation (b) */
\r
471 //Registered Shift Magnitude
\r
472 //Dreg = ASHIFT Dreg BY Dreg_lo (V) ; /* arithmetic shift (b) */
\r
473 r2=ashift r7 by r5.l (v) ;
\r
475 R0 = ASHIFT R1 BY R2.L (V);
\r
476 R3 = ASHIFT R4 BY R5.L (V);
\r
477 R6 = ASHIFT R7 BY R0.L (V);
\r
478 R1 = ASHIFT R2 BY R3.L (V);
\r
479 R4 = ASHIFT R5 BY R6.L (V);
\r
480 R7 = ASHIFT R0 BY R1.L (V);
\r
481 R2 = ASHIFT R3 BY R4.L (V);
\r
482 R5 = ASHIFT R6 BY R7.L (V);
\r
485 //Dreg = ASHIFT Dreg BY Dreg_lo (V, S) ; /* arithmetic shift with saturation (b) */
\r
486 R0 = ASHIFT R1 BY R2.L (V,S);
\r
487 R3 = ASHIFT R4 BY R5.L (V,S);
\r
488 R6 = ASHIFT R7 BY R0.L (V,S);
\r
489 R1 = ASHIFT R2 BY R3.L (V,S);
\r
490 R4 = ASHIFT R5 BY R6.L (V,S);
\r
491 R7 = ASHIFT R0 BY R1.L (V,S);
\r
492 R2 = ASHIFT R3 BY R4.L (V,S);
\r
493 R5 = ASHIFT R6 BY R7.L (V,S);
\r
495 //Constant Shift Magnitude
\r
496 //Dreg = Dreg >> uimm4 (V) ; /* logical shift right, immediate (b) */
\r
506 //Dreg = Dreg << uimm4 (V) ; /* logical shift left, immediate (b) */
\r
517 //Registered Shift Magnitude
\r
518 //Dreg = LSHIFT Dreg BY Dreg_lo (V) ; /* logical shift (b) */
\r
520 R0 = LSHIFT R1 BY R2.L (V);
\r
521 R3 = LSHIFT R4 BY R5.L (V);
\r
522 R6 = LSHIFT R7 BY R0.L (V);
\r
523 R1 = LSHIFT R2 BY R3.L (V);
\r
524 R4 = LSHIFT R5 BY R6.L (V);
\r
525 R7 = LSHIFT R0 BY R1.L (V);
\r
526 R2 = LSHIFT R3 BY R4.L (V);
\r
527 R5 = LSHIFT R6 BY R7.L (V);
\r
529 //Dreg = MAX ( Dreg , Dreg ) (V) ; /* dual 16-bit operations (b) */
\r
530 r7 = max (r1, r0) (v) ;
\r
532 R0 = MAX (R1, R2) (V);
\r
533 R3 = MAX (R4, R5) (V);
\r
534 R6 = MAX (R7, R0) (V);
\r
535 R1 = MAX (R2, R3) (V);
\r
536 R4 = MAX (R5, R6) (V);
\r
537 R7 = MAX (R0, R1) (V);
\r
538 R2 = MAX (R3, R4) (V);
\r
539 R5 = MAX (R6, R7) (V);
\r
541 //Dreg = MIN ( Dreg , Dreg ) (V) ; /* dual 16-bit operation (b) */
\r
542 R0 = MIN (R1, R2) (V);
\r
543 R3 = MIN (R4, R5) (V);
\r
544 R6 = MIN (R7, R0) (V);
\r
545 R1 = MIN (R2, R3) (V);
\r
546 R4 = MIN (R5, R6) (V);
\r
547 R7 = MIN (R0, R1) (V);
\r
548 R2 = MIN (R3, R4) (V);
\r
549 R5 = MIN (R6, R7) (V);
\r
551 r2.h=r7.l*r6.h, r2.l=r7.h*r6.h ;
\r
552 /* simultaneous MAC0 and MAC1 execution, 16-bit results. Both
\r
553 results are signed fractions. */
\r
554 r4.l=r1.l*r0.l, r4.h=r1.h*r0.h ;
\r
555 /* same as above. MAC order is arbitrary. */
\r
556 r0.h=r3.h*r2.l (m), r0.l=r3.l*r2.l ;
\r
558 a1=r2.l*r3.h, a0=r2.h*r3.h ;
\r
559 /* both multiply signed fractions into separate Accumulators */
\r
560 a0=r1.l*r0.l, a1+=r1.h*r0.h ;
\r
561 /* same as above, but sum result into A1. MAC order is arbitrary.
\r
563 a1+=r3.h*r3.l, a0-=r3.h*r3.h ;
\r
564 /* sum product into A1, subtract product from A0 */
\r
565 a1=r3.h*r2.l (m), a0+=r3.l*r2.l ;
\r
566 /* MAC1 multiplies a signed fraction in r3.h by an unsigned fraction
\r
567 in r2.l. MAC0 multiplies two signed fractions. */
\r
568 a1=r7.h*r4.h (m), a0+=r7.l*r4.l (fu) ;
\r
569 /* MAC1 multiplies signed fraction by unsigned fraction. MAC0
\r
570 multiplies and accumulates two unsigned fractions. */
\r
571 a1+=r3.h*r2.h, a0=r3.l*r2.l (is) ;
\r
572 /* both MACs perform signed integer multiplication */
\r
573 a1=r6.h*r7.h, a0+=r6.l*r7.l (w32) ;
\r
574 /* both MACs multiply signed fractions, sign extended, and saturate
\r
575 both Accumulators at bit 31 */
\r
576 r2.h=(a1=r7.l*r6.h), r2.l=(a0=r7.h*r6.h) ; /* simultaneous MAC0
\r
577 and MAC1 execution, both are signed fractions, both products load
\r
578 into the Accumulators,MAC1 into half-word registers. */
\r
579 r4.l=(a0=r1.l*r0.l), r4.h=(a1+=r1.h*r0.h) ; /* same as above,
\r
580 but sum result into A1. ; MAC order is arbitrary. */
\r
581 r7.h=(a1+=r6.h*r5.l), r7.l=(a0=r6.h*r5.h) ; /* sum into A1,
\r
582 subtract into A0 */
\r
583 r0.h=(a1=r7.h*r4.l) (m), r0.l=(a0+=r7.l*r4.l) ; /* MAC1 multiplies
\r
584 a signed fraction by an unsigned fraction. MAC0 multiplies
\r
585 two signed fractions. */
\r
586 r5.h=(a1=r3.h*r2.h) (m), r5.l=(a0+=r3.l*r2.l) (fu) ; /* MAC1
\r
587 multiplies signed fraction by unsigned fraction. MAC0 multiplies
\r
588 two unsigned fractions. */
\r
589 r0.h=(a1+=r3.h*r2.h), r0.l=(a0=r3.l*r2.l) (is) ; /* both MACs
\r
590 perform signed integer multiplication. */
\r
591 r5.h=(a1=r2.h*r1.h), a0+=r2.l*r1.l ; /* both MACs multiply
\r
592 signed fractions. MAC0 does not copy the accum result. */
\r
593 r3.h=(a1=r2.h*r1.h) (m), a0=r2.l*r1.l ; /* MAC1 multiplies
\r
594 signed fraction by unsigned fraction and uses all 40 bits of A1.
\r
595 MAC0 multiplies two signed fractions. */
\r
596 r3.h=a1, r3.l=(a0+=r0.l*r1.l) (s2rnd) ; /* MAC1 copies Accumulator
\r
597 to register half. MAC0 multiplies signed fractions. Both
\r
598 scale the result and round on the way to the destination register.
\r
600 r0.l=(a0+=r7.l*r6.l), r0.h=(a1+=r7.h*r6.h) (iss2) ; /* both
\r
601 MACs process signed integer the way to the destination half-registers.
\r
603 r3=(a1=r6.h*r7.h), r2=(a0=r6.l*r7.l) ; /* simultaneous MAC0 and
\r
604 MAC1 execution, both are signed fractions, both products load
\r
605 into the Accumulators */
\r
606 r4=(a0=r6.l*r7.l), r5=(a1+=r6.h*r7.h) ; /* same as above, but
\r
607 sum result into A1. MAC order is arbitrary. */
\r
608 r7=(a1+=r3.h*r5.h), r6=(a0-=r3.l*r5.l) ; /* sum into A1, subtract
\r
610 r1=(a1=r7.l*r4.l) (m), r0=(a0+=r7.h*r4.h) ; /* MAC1 multiplies
\r
611 a signed fraction by an unsigned fraction. MAC0 multiplies two
\r
612 signed fractions. */
\r
613 r5=(a1=r3.h*r7.h) (m), r4=(a0+=r3.l*r7.l) (fu) ; /* MAC1 multiplies
\r
614 signed fraction by unsigned fraction. MAC0 multiplies two
\r
615 unsigned fractions. */
\r
616 r1=(a1+=r3.h*r2.h), r0=(a0=r3.l*r2.l) (is) ; /* both MACs perform
\r
617 signed integer multiplication */
\r
618 r5=(a1-=r6.h*r7.h), a0+=r6.l*r7.l ; /* both MACs multiply
\r
619 signed fractions. MAC0 does not copy the accum result */
\r
620 r3=(a1=r6.h*r7.h) (m), a0-=r6.l*r7.l ; /* MAC1 multiplies
\r
621 signed fraction by unsigned fraction and uses all 40 bits of A1.
\r
622 MAC0 multiplies two signed fractions. */
\r
623 r3=a1, r2=(a0+=r0.l*r1.l) (s2rnd) ; /* MAC1 moves Accumulator
\r
624 to register. MAC0 multiplies signed fractions. Both scale the
\r
625 result and round on the way to the destination register. */
\r
626 r0=(a0+=r7.l*r6.l), r1=(a1+=r7.h*r6.h) (iss2) ; /* both MACs
\r
627 process signed integer operands and scale the result on the way
\r
628 to the destination registers. */
\r
630 r5 =-r3 (v) ; /* R5.H becomes the negative of R3.H and R5.L
\r
631 becomes the negative of R3.L If r3 = 0x0004 7FFF the result is r5
\r
634 r3=pack(r4.l, r5.l) ; /* pack low / low half-words */
\r
635 r1=pack(r6.l, r4.h) ; /* pack low / high half-words */
\r
636 r0=pack(r2.h, r4.l) ; /* pack high / low half-words */
\r
637 r5=pack(r7.h, r2.h) ; /* pack high / high half-words */
\r
639 (r1,r0) = SEARCH R2 (LE) || R2=[P0++];
\r
640 /* search for the last minimum in all but the
\r
641 last element of the array */
\r
642 (r1,r0) = SEARCH R2 (LE);
\r
644 saa (r1:0, r3:2) || r0=[i0++] || r2=[i1++] ;
\r
645 saa (r1:0, r3:2)(r) || r1=[i0++] || r3=[i1++] ;
\r
646 mnop || r1 = [i0++] || r3 = [i1++] ;
\r
647 r7.h=r7.l=sign(r2.h)*r3.h + sign(r2.l)*r3.l || i0+=m3 || r0=[i0]
\r
650 /* Add/subtract two vector values while incrementing an Ireg and
\r
651 loading a data register. */
\r
652 R2 = R2 +|+ R4, R4 = R2 -|- R4 (ASR) || I0 += M0 (BREV) || R1 = [I0] ;
\r
653 /* Multiply and accumulate to Accumulator while loading a data
\r
654 register and storing a data register using an Ireg pointer. */
\r
655 A1=R2.L*R1.L, A0=R2.H*R1.H || R2.H=W[I2++] || [I3++]=R3 ;
\r
656 /* Multiply and accumulate while loading two data registers. One
\r
657 load uses an Ireg pointer. */
\r
658 A1+=R0.L*R2.H,A0+=R0.L*R2.L || R2.L=W[I2++] || R0=[I1--] ;
\r
659 R3.H=(A1+=R0.L*R1.H), R3.L=(A0+=R0.L*R1.L) || R0=[P0++] || R1=[I0] ;
\r
660 /* Pack two vector values while storing a data register using an
\r
661 Ireg pointer and loading another data register. */
\r
662 R1=PACK(R1.H,R0.H) || [I0++]=R0 || R2.L=W[I2++] ;
\r
664 /* Multiply-Accumulate to a Data register while incrementing an
\r
666 r6=(a0+=r3.h*r2.h)(fu) || i2-=m0 ;
\r
667 /* which the assembler expands into:
\r
668 r6=(a0+=r3.h*r2.h)(fu) || i2-=m0 || nop ; */
\r