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
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=arm-eabi -mattr=+v6t2 | FileCheck %s --check-prefixes=CHECK,SCALAR
; RUN: llc < %s -mtriple=arm-eabi -mattr=+v6t2 -mattr=+neon | FileCheck %s --check-prefixes=CHECK,NEON
declare i8 @llvm.fshl.i8(i8, i8, i8)
declare i16 @llvm.fshl.i16(i16, i16, i16)
declare i32 @llvm.fshl.i32(i32, i32, i32)
declare i64 @llvm.fshl.i64(i64, i64, i64)
declare <4 x i32> @llvm.fshl.v4i32(<4 x i32>, <4 x i32>, <4 x i32>)
declare i8 @llvm.fshr.i8(i8, i8, i8)
declare i16 @llvm.fshr.i16(i16, i16, i16)
declare i32 @llvm.fshr.i32(i32, i32, i32)
declare i64 @llvm.fshr.i64(i64, i64, i64)
declare <4 x i32> @llvm.fshr.v4i32(<4 x i32>, <4 x i32>, <4 x i32>)
; General case - all operands can be variables.
define i16 @fshl_i16(i16 %x, i16 %y, i16 %z) {
; CHECK-LABEL: fshl_i16:
; CHECK: @ %bb.0:
; CHECK-NEXT: pkhbt r0, r1, r0, lsl #16
; CHECK-NEXT: and r1, r2, #15
; CHECK-NEXT: lsl r0, r0, r1
; CHECK-NEXT: lsr r0, r0, #16
; CHECK-NEXT: bx lr
%f = call i16 @llvm.fshl.i16(i16 %x, i16 %y, i16 %z)
ret i16 %f
}
define i32 @fshl_i32(i32 %x, i32 %y, i32 %z) {
; CHECK-LABEL: fshl_i32:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r3, #31
; CHECK-NEXT: lsr r1, r1, #1
; CHECK-NEXT: bic r3, r3, r2
; CHECK-NEXT: and r2, r2, #31
; CHECK-NEXT: lsl r0, r0, r2
; CHECK-NEXT: orr r0, r0, r1, lsr r3
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshl.i32(i32 %x, i32 %y, i32 %z)
ret i32 %f
}
; Verify that weird types are minimally supported.
declare i37 @llvm.fshl.i37(i37, i37, i37)
define i37 @fshl_i37(i37 %x, i37 %y, i37 %z) {
; SCALAR-LABEL: fshl_i37:
; SCALAR: @ %bb.0:
; SCALAR-NEXT: .save {r4, r5, r6, r7, r8, lr}
; SCALAR-NEXT: push {r4, r5, r6, r7, r8, lr}
; SCALAR-NEXT: mov r8, r0
; SCALAR-NEXT: ldr r0, [sp, #28]
; SCALAR-NEXT: mov r4, r1
; SCALAR-NEXT: mov r5, r3
; SCALAR-NEXT: and r1, r0, #31
; SCALAR-NEXT: ldr r0, [sp, #24]
; SCALAR-NEXT: mov r6, r2
; SCALAR-NEXT: mov r2, #37
; SCALAR-NEXT: mov r3, #0
; SCALAR-NEXT: bl __aeabi_uldivmod
; SCALAR-NEXT: lsl r0, r5, #27
; SCALAR-NEXT: tst r2, #32
; SCALAR-NEXT: orr r0, r0, r6, lsr #5
; SCALAR-NEXT: mov r1, r8
; SCALAR-NEXT: and r3, r2, #31
; SCALAR-NEXT: mov r7, #31
; SCALAR-NEXT: movne r1, r0
; SCALAR-NEXT: lslne r0, r6, #27
; SCALAR-NEXT: bic r2, r7, r2
; SCALAR-NEXT: lsl r5, r1, r3
; SCALAR-NEXT: lsr r0, r0, #1
; SCALAR-NEXT: movne r4, r8
; SCALAR-NEXT: lsr r1, r1, #1
; SCALAR-NEXT: lsl r3, r4, r3
; SCALAR-NEXT: orr r0, r5, r0, lsr r2
; SCALAR-NEXT: orr r1, r3, r1, lsr r2
; SCALAR-NEXT: pop {r4, r5, r6, r7, r8, pc}
;
; NEON-LABEL: fshl_i37:
; NEON: @ %bb.0:
; NEON-NEXT: .save {r4, r5, r6, r7, r8, lr}
; NEON-NEXT: push {r4, r5, r6, r7, r8, lr}
; NEON-NEXT: mov r4, r1
; NEON-NEXT: ldr r1, [sp, #28]
; NEON-NEXT: mov r8, r0
; NEON-NEXT: ldr r0, [sp, #24]
; NEON-NEXT: and r1, r1, #31
; NEON-NEXT: mov r5, r3
; NEON-NEXT: mov r6, r2
; NEON-NEXT: mov r2, #37
; NEON-NEXT: mov r3, #0
; NEON-NEXT: bl __aeabi_uldivmod
; NEON-NEXT: lsl r0, r5, #27
; NEON-NEXT: tst r2, #32
; NEON-NEXT: orr r0, r0, r6, lsr #5
; NEON-NEXT: mov r1, r8
; NEON-NEXT: and r3, r2, #31
; NEON-NEXT: mov r7, #31
; NEON-NEXT: movne r1, r0
; NEON-NEXT: lslne r0, r6, #27
; NEON-NEXT: bic r2, r7, r2
; NEON-NEXT: lsl r5, r1, r3
; NEON-NEXT: lsr r0, r0, #1
; NEON-NEXT: movne r4, r8
; NEON-NEXT: lsr r1, r1, #1
; NEON-NEXT: lsl r3, r4, r3
; NEON-NEXT: orr r0, r5, r0, lsr r2
; NEON-NEXT: orr r1, r3, r1, lsr r2
; NEON-NEXT: pop {r4, r5, r6, r7, r8, pc}
%f = call i37 @llvm.fshl.i37(i37 %x, i37 %y, i37 %z)
ret i37 %f
}
; extract(concat(0b1110000, 0b1111111) << 2) = 0b1000011
declare i7 @llvm.fshl.i7(i7, i7, i7)
define i7 @fshl_i7_const_fold() {
; CHECK-LABEL: fshl_i7_const_fold:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #67
; CHECK-NEXT: bx lr
%f = call i7 @llvm.fshl.i7(i7 112, i7 127, i7 2)
ret i7 %f
}
define i8 @fshl_i8_const_fold_overshift_1() {
; CHECK-LABEL: fshl_i8_const_fold_overshift_1:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #128
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshl.i8(i8 255, i8 0, i8 15)
ret i8 %f
}
define i8 @fshl_i8_const_fold_overshift_2() {
; CHECK-LABEL: fshl_i8_const_fold_overshift_2:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #120
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshl.i8(i8 15, i8 15, i8 11)
ret i8 %f
}
define i8 @fshl_i8_const_fold_overshift_3() {
; CHECK-LABEL: fshl_i8_const_fold_overshift_3:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #0
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshl.i8(i8 0, i8 225, i8 8)
ret i8 %f
}
; With constant shift amount, this is 'extr'.
define i32 @fshl_i32_const_shift(i32 %x, i32 %y) {
; CHECK-LABEL: fshl_i32_const_shift:
; CHECK: @ %bb.0:
; CHECK-NEXT: lsl r0, r0, #9
; CHECK-NEXT: orr r0, r0, r1, lsr #23
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshl.i32(i32 %x, i32 %y, i32 9)
ret i32 %f
}
; Check modulo math on shift amount.
define i32 @fshl_i32_const_overshift(i32 %x, i32 %y) {
; CHECK-LABEL: fshl_i32_const_overshift:
; CHECK: @ %bb.0:
; CHECK-NEXT: lsl r0, r0, #9
; CHECK-NEXT: orr r0, r0, r1, lsr #23
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshl.i32(i32 %x, i32 %y, i32 41)
ret i32 %f
}
; 64-bit should also work.
define i64 @fshl_i64_const_overshift(i64 %x, i64 %y) {
; CHECK-LABEL: fshl_i64_const_overshift:
; CHECK: @ %bb.0:
; CHECK-NEXT: lsl r1, r3, #9
; CHECK-NEXT: orr r2, r1, r2, lsr #23
; CHECK-NEXT: lsl r0, r0, #9
; CHECK-NEXT: orr r1, r0, r3, lsr #23
; CHECK-NEXT: mov r0, r2
; CHECK-NEXT: bx lr
%f = call i64 @llvm.fshl.i64(i64 %x, i64 %y, i64 105)
ret i64 %f
}
; This should work without any node-specific logic.
define i8 @fshl_i8_const_fold() {
; CHECK-LABEL: fshl_i8_const_fold:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #128
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshl.i8(i8 255, i8 0, i8 7)
ret i8 %f
}
; Repeat everything for funnel shift right.
; General case - all operands can be variables.
define i16 @fshr_i16(i16 %x, i16 %y, i16 %z) {
; CHECK-LABEL: fshr_i16:
; CHECK: @ %bb.0:
; CHECK-NEXT: pkhbt r0, r1, r0, lsl #16
; CHECK-NEXT: and r1, r2, #15
; CHECK-NEXT: lsr r0, r0, r1
; CHECK-NEXT: bx lr
%f = call i16 @llvm.fshr.i16(i16 %x, i16 %y, i16 %z)
ret i16 %f
}
define i32 @fshr_i32(i32 %x, i32 %y, i32 %z) {
; CHECK-LABEL: fshr_i32:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r3, #31
; CHECK-NEXT: lsl r0, r0, #1
; CHECK-NEXT: bic r3, r3, r2
; CHECK-NEXT: and r2, r2, #31
; CHECK-NEXT: lsl r0, r0, r3
; CHECK-NEXT: orr r0, r0, r1, lsr r2
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshr.i32(i32 %x, i32 %y, i32 %z)
ret i32 %f
}
; Verify that weird types are minimally supported.
declare i37 @llvm.fshr.i37(i37, i37, i37)
define i37 @fshr_i37(i37 %x, i37 %y, i37 %z) {
; SCALAR-LABEL: fshr_i37:
; SCALAR: @ %bb.0:
; SCALAR-NEXT: .save {r4, r5, r6, r7, r11, lr}
; SCALAR-NEXT: push {r4, r5, r6, r7, r11, lr}
; SCALAR-NEXT: mov r5, r0
; SCALAR-NEXT: ldr r0, [sp, #28]
; SCALAR-NEXT: mov r4, r1
; SCALAR-NEXT: mov r6, r3
; SCALAR-NEXT: and r1, r0, #31
; SCALAR-NEXT: ldr r0, [sp, #24]
; SCALAR-NEXT: mov r7, r2
; SCALAR-NEXT: mov r2, #37
; SCALAR-NEXT: mov r3, #0
; SCALAR-NEXT: bl __aeabi_uldivmod
; SCALAR-NEXT: add r0, r2, #27
; SCALAR-NEXT: lsl r2, r6, #27
; SCALAR-NEXT: orr r2, r2, r7, lsr #5
; SCALAR-NEXT: mov r1, #31
; SCALAR-NEXT: tst r0, #32
; SCALAR-NEXT: mov r3, r5
; SCALAR-NEXT: moveq r3, r2
; SCALAR-NEXT: lsleq r2, r7, #27
; SCALAR-NEXT: bic r1, r1, r0
; SCALAR-NEXT: and r7, r0, #31
; SCALAR-NEXT: lsl r6, r3, #1
; SCALAR-NEXT: moveq r4, r5
; SCALAR-NEXT: lsl r6, r6, r1
; SCALAR-NEXT: orr r0, r6, r2, lsr r7
; SCALAR-NEXT: lsl r2, r4, #1
; SCALAR-NEXT: lsl r1, r2, r1
; SCALAR-NEXT: orr r1, r1, r3, lsr r7
; SCALAR-NEXT: pop {r4, r5, r6, r7, r11, pc}
;
; NEON-LABEL: fshr_i37:
; NEON: @ %bb.0:
; NEON-NEXT: .save {r4, r5, r6, r7, r11, lr}
; NEON-NEXT: push {r4, r5, r6, r7, r11, lr}
; NEON-NEXT: mov r4, r1
; NEON-NEXT: ldr r1, [sp, #28]
; NEON-NEXT: mov r5, r0
; NEON-NEXT: ldr r0, [sp, #24]
; NEON-NEXT: and r1, r1, #31
; NEON-NEXT: mov r6, r3
; NEON-NEXT: mov r7, r2
; NEON-NEXT: mov r2, #37
; NEON-NEXT: mov r3, #0
; NEON-NEXT: bl __aeabi_uldivmod
; NEON-NEXT: add r0, r2, #27
; NEON-NEXT: lsl r2, r6, #27
; NEON-NEXT: orr r2, r2, r7, lsr #5
; NEON-NEXT: mov r1, #31
; NEON-NEXT: tst r0, #32
; NEON-NEXT: mov r3, r5
; NEON-NEXT: moveq r3, r2
; NEON-NEXT: lsleq r2, r7, #27
; NEON-NEXT: bic r1, r1, r0
; NEON-NEXT: and r7, r0, #31
; NEON-NEXT: lsl r6, r3, #1
; NEON-NEXT: moveq r4, r5
; NEON-NEXT: lsl r6, r6, r1
; NEON-NEXT: orr r0, r6, r2, lsr r7
; NEON-NEXT: lsl r2, r4, #1
; NEON-NEXT: lsl r1, r2, r1
; NEON-NEXT: orr r1, r1, r3, lsr r7
; NEON-NEXT: pop {r4, r5, r6, r7, r11, pc}
%f = call i37 @llvm.fshr.i37(i37 %x, i37 %y, i37 %z)
ret i37 %f
}
; extract(concat(0b1110000, 0b1111111) >> 2) = 0b0011111
declare i7 @llvm.fshr.i7(i7, i7, i7)
define i7 @fshr_i7_const_fold() {
; CHECK-LABEL: fshr_i7_const_fold:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #31
; CHECK-NEXT: bx lr
%f = call i7 @llvm.fshr.i7(i7 112, i7 127, i7 2)
ret i7 %f
}
define i8 @fshr_i8_const_fold_overshift_1() {
; CHECK-LABEL: fshr_i8_const_fold_overshift_1:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #254
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshr.i8(i8 255, i8 0, i8 15)
ret i8 %f
}
define i8 @fshr_i8_const_fold_overshift_2() {
; CHECK-LABEL: fshr_i8_const_fold_overshift_2:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #225
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshr.i8(i8 15, i8 15, i8 11)
ret i8 %f
}
define i8 @fshr_i8_const_fold_overshift_3() {
; CHECK-LABEL: fshr_i8_const_fold_overshift_3:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #255
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshr.i8(i8 0, i8 255, i8 8)
ret i8 %f
}
; With constant shift amount, this is 'extr'.
define i32 @fshr_i32_const_shift(i32 %x, i32 %y) {
; CHECK-LABEL: fshr_i32_const_shift:
; CHECK: @ %bb.0:
; CHECK-NEXT: lsl r0, r0, #23
; CHECK-NEXT: orr r0, r0, r1, lsr #9
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshr.i32(i32 %x, i32 %y, i32 9)
ret i32 %f
}
; Check modulo math on shift amount. 41-32=9.
define i32 @fshr_i32_const_overshift(i32 %x, i32 %y) {
; CHECK-LABEL: fshr_i32_const_overshift:
; CHECK: @ %bb.0:
; CHECK-NEXT: lsl r0, r0, #23
; CHECK-NEXT: orr r0, r0, r1, lsr #9
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshr.i32(i32 %x, i32 %y, i32 41)
ret i32 %f
}
; 64-bit should also work. 105-64 = 41.
define i64 @fshr_i64_const_overshift(i64 %x, i64 %y) {
; CHECK-LABEL: fshr_i64_const_overshift:
; CHECK: @ %bb.0:
; CHECK-NEXT: lsl r2, r0, #23
; CHECK-NEXT: lsl r1, r1, #23
; CHECK-NEXT: orr r2, r2, r3, lsr #9
; CHECK-NEXT: orr r1, r1, r0, lsr #9
; CHECK-NEXT: mov r0, r2
; CHECK-NEXT: bx lr
%f = call i64 @llvm.fshr.i64(i64 %x, i64 %y, i64 105)
ret i64 %f
}
; This should work without any node-specific logic.
define i8 @fshr_i8_const_fold() {
; CHECK-LABEL: fshr_i8_const_fold:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, #254
; CHECK-NEXT: bx lr
%f = call i8 @llvm.fshr.i8(i8 255, i8 0, i8 7)
ret i8 %f
}
define i32 @fshl_i32_shift_by_bitwidth(i32 %x, i32 %y) {
; CHECK-LABEL: fshl_i32_shift_by_bitwidth:
; CHECK: @ %bb.0:
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshl.i32(i32 %x, i32 %y, i32 32)
ret i32 %f
}
define i32 @fshr_i32_shift_by_bitwidth(i32 %x, i32 %y) {
; CHECK-LABEL: fshr_i32_shift_by_bitwidth:
; CHECK: @ %bb.0:
; CHECK-NEXT: mov r0, r1
; CHECK-NEXT: bx lr
%f = call i32 @llvm.fshr.i32(i32 %x, i32 %y, i32 32)
ret i32 %f
}
define <4 x i32> @fshl_v4i32_shift_by_bitwidth(<4 x i32> %x, <4 x i32> %y) {
; CHECK-LABEL: fshl_v4i32_shift_by_bitwidth:
; CHECK: @ %bb.0:
; CHECK-NEXT: bx lr
%f = call <4 x i32> @llvm.fshl.v4i32(<4 x i32> %x, <4 x i32> %y, <4 x i32> <i32 32, i32 32, i32 32, i32 32>)
ret <4 x i32> %f
}
define <4 x i32> @fshr_v4i32_shift_by_bitwidth(<4 x i32> %x, <4 x i32> %y) {
; SCALAR-LABEL: fshr_v4i32_shift_by_bitwidth:
; SCALAR: @ %bb.0:
; SCALAR-NEXT: ldm sp, {r0, r1, r2, r3}
; SCALAR-NEXT: bx lr
;
; NEON-LABEL: fshr_v4i32_shift_by_bitwidth:
; NEON: @ %bb.0:
; NEON-NEXT: mov r0, sp
; NEON-NEXT: vld1.64 {d16, d17}, [r0]
; NEON-NEXT: vmov r0, r1, d16
; NEON-NEXT: vmov r2, r3, d17
; NEON-NEXT: bx lr
%f = call <4 x i32> @llvm.fshr.v4i32(<4 x i32> %x, <4 x i32> %y, <4 x i32> <i32 32, i32 32, i32 32, i32 32>)
ret <4 x i32> %f
}
|
