1 ; RUN: llc < %s -mcpu=x86-64 -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
2 ; RUN: llc < %s -mcpu=x86-64 -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
3 ; RUN: llc < %s -mcpu=knl -mattr=+avx512vl | FileCheck %s --check-prefix=AVX512VL
5 target triple = "x86_64-unknown-unknown"
7 define <4 x double> @shuffle_v4f64_0000(<4 x double> %a, <4 x double> %b) {
8 ; AVX1-LABEL: shuffle_v4f64_0000:
10 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
11 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
14 ; AVX2-LABEL: shuffle_v4f64_0000:
16 ; AVX2-NEXT: vbroadcastsd %xmm0, %ymm0
18 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
19 ret <4 x double> %shuffle
22 define <4 x double> @shuffle_v4f64_0001(<4 x double> %a, <4 x double> %b) {
23 ; AVX1-LABEL: shuffle_v4f64_0001:
25 ; AVX1-NEXT: vmovddup {{.*#+}} xmm1 = xmm0[0,0]
26 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
29 ; AVX2-LABEL: shuffle_v4f64_0001:
31 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,1]
33 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 1>
34 ret <4 x double> %shuffle
37 define <4 x double> @shuffle_v4f64_0020(<4 x double> %a, <4 x double> %b) {
38 ; AVX1-LABEL: shuffle_v4f64_0020:
40 ; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
41 ; AVX1-NEXT: vunpcklpd {{.*#+}} xmm1 = xmm1[0],xmm0[0]
42 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
43 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
46 ; AVX2-LABEL: shuffle_v4f64_0020:
48 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,0,2,0]
50 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 0, i32 2, i32 0>
51 ret <4 x double> %shuffle
54 define <4 x double> @shuffle_v4f64_0300(<4 x double> %a, <4 x double> %b) {
55 ; AVX1-LABEL: shuffle_v4f64_0300:
57 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
58 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm1 = ymm1[0,1,2,2]
59 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3]
62 ; AVX2-LABEL: shuffle_v4f64_0300:
64 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,3,0,0]
66 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 3, i32 0, i32 0>
67 ret <4 x double> %shuffle
70 define <4 x double> @shuffle_v4f64_1000(<4 x double> %a, <4 x double> %b) {
71 ; AVX1-LABEL: shuffle_v4f64_1000:
73 ; AVX1-NEXT: vpermilpd {{.*#+}} xmm1 = xmm0[1,0]
74 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
75 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
78 ; AVX2-LABEL: shuffle_v4f64_1000:
80 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[1,0,0,0]
82 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
83 ret <4 x double> %shuffle
86 define <4 x double> @shuffle_v4f64_2200(<4 x double> %a, <4 x double> %b) {
87 ; AVX1-LABEL: shuffle_v4f64_2200:
89 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
90 ; AVX1-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
93 ; AVX2-LABEL: shuffle_v4f64_2200:
95 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[2,2,0,0]
97 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 2, i32 2, i32 0, i32 0>
98 ret <4 x double> %shuffle
101 define <4 x double> @shuffle_v4f64_3330(<4 x double> %a, <4 x double> %b) {
102 ; AVX1-LABEL: shuffle_v4f64_3330:
104 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
105 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0,1,2],ymm0[3]
106 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,1,3,2]
109 ; AVX2-LABEL: shuffle_v4f64_3330:
111 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[3,3,3,0]
113 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 3, i32 3, i32 3, i32 0>
114 ret <4 x double> %shuffle
117 define <4 x double> @shuffle_v4f64_3210(<4 x double> %a, <4 x double> %b) {
118 ; AVX1-LABEL: shuffle_v4f64_3210:
120 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
121 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
124 ; AVX2-LABEL: shuffle_v4f64_3210:
126 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[3,2,1,0]
128 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
129 ret <4 x double> %shuffle
132 define <4 x double> @shuffle_v4f64_0023(<4 x double> %a, <4 x double> %b) {
133 ; ALL-LABEL: shuffle_v4f64_0023:
135 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[0,0,2,3]
138 ; AVX512VL-LABEL: shuffle_v4f64_0023:
140 ; AVX512VL-NEXT: vpermilpd $8, %ymm0, %ymm0
141 ; AVX512VL-NEXT: retq
142 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 0, i32 2, i32 3>
143 ret <4 x double> %shuffle
146 define <4 x double> @shuffle_v4f64_0022(<4 x double> %a, <4 x double> %b) {
147 ; ALL-LABEL: shuffle_v4f64_0022:
149 ; ALL-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
151 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 0, i32 2, i32 2>
152 ret <4 x double> %shuffle
155 define <4 x double> @shuffle_v4f64_1032(<4 x double> %a, <4 x double> %b) {
156 ; ALL-LABEL: shuffle_v4f64_1032:
158 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
160 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 0, i32 3, i32 2>
161 ret <4 x double> %shuffle
164 define <4 x double> @shuffle_v4f64_1133(<4 x double> %a, <4 x double> %b) {
165 ; ALL-LABEL: shuffle_v4f64_1133:
167 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,1,3,3]
169 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 1, i32 3, i32 3>
170 ret <4 x double> %shuffle
173 define <4 x double> @shuffle_v4f64_1023(<4 x double> %a, <4 x double> %b) {
174 ; ALL-LABEL: shuffle_v4f64_1023:
176 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,2,3]
178 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 0, i32 2, i32 3>
179 ret <4 x double> %shuffle
182 define <4 x double> @shuffle_v4f64_1022(<4 x double> %a, <4 x double> %b) {
183 ; ALL-LABEL: shuffle_v4f64_1022:
185 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,2,2]
187 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 0, i32 2, i32 2>
188 ret <4 x double> %shuffle
191 define <4 x double> @shuffle_v4f64_0423(<4 x double> %a, <4 x double> %b) {
192 ; AVX1-LABEL: shuffle_v4f64_0423:
194 ; AVX1-NEXT: vmovddup {{.*#+}} ymm1 = ymm1[0,0,2,2]
195 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2,3]
198 ; AVX2-LABEL: shuffle_v4f64_0423:
200 ; AVX2-NEXT: vbroadcastsd %xmm1, %ymm1
201 ; AVX2-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2,3]
203 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 3>
204 ret <4 x double> %shuffle
207 define <4 x double> @shuffle_v4f64_0462(<4 x double> %a, <4 x double> %b) {
208 ; ALL-LABEL: shuffle_v4f64_0462:
210 ; ALL-NEXT: vmovddup {{.*#+}} ymm1 = ymm1[0,0,2,2]
211 ; ALL-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
212 ; ALL-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2],ymm0[3]
214 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 4, i32 6, i32 2>
215 ret <4 x double> %shuffle
218 define <4 x double> @shuffle_v4f64_0426(<4 x double> %a, <4 x double> %b) {
219 ; ALL-LABEL: shuffle_v4f64_0426:
221 ; ALL-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
223 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
224 ret <4 x double> %shuffle
227 define <4 x double> @shuffle_v4f64_1537(<4 x double> %a, <4 x double> %b) {
228 ; ALL-LABEL: shuffle_v4f64_1537:
230 ; ALL-NEXT: vunpckhpd {{.*#+}} ymm0 = ymm0[1],ymm1[1],ymm0[3],ymm1[3]
232 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
233 ret <4 x double> %shuffle
236 define <4 x double> @shuffle_v4f64_4062(<4 x double> %a, <4 x double> %b) {
237 ; ALL-LABEL: shuffle_v4f64_4062:
239 ; ALL-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
241 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 4, i32 0, i32 6, i32 2>
242 ret <4 x double> %shuffle
245 define <4 x double> @shuffle_v4f64_5173(<4 x double> %a, <4 x double> %b) {
246 ; ALL-LABEL: shuffle_v4f64_5173:
248 ; ALL-NEXT: vunpckhpd {{.*#+}} ymm0 = ymm1[1],ymm0[1],ymm1[3],ymm0[3]
250 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 5, i32 1, i32 7, i32 3>
251 ret <4 x double> %shuffle
254 define <4 x double> @shuffle_v4f64_5163(<4 x double> %a, <4 x double> %b) {
255 ; ALL-LABEL: shuffle_v4f64_5163:
257 ; ALL-NEXT: vshufpd {{.*#+}} ymm0 = ymm1[1],ymm0[1],ymm1[2],ymm0[3]
259 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 5, i32 1, i32 6, i32 3>
260 ret <4 x double> %shuffle
263 define <4 x double> @shuffle_v4f64_0527(<4 x double> %a, <4 x double> %b) {
264 ; ALL-LABEL: shuffle_v4f64_0527:
266 ; ALL-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3]
268 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 5, i32 2, i32 7>
269 ret <4 x double> %shuffle
272 define <4 x double> @shuffle_v4f64_4163(<4 x double> %a, <4 x double> %b) {
273 ; ALL-LABEL: shuffle_v4f64_4163:
275 ; ALL-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0],ymm0[1],ymm1[2],ymm0[3]
277 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 4, i32 1, i32 6, i32 3>
278 ret <4 x double> %shuffle
281 define <4 x double> @shuffle_v4f64_0145(<4 x double> %a, <4 x double> %b) {
282 ; ALL-LABEL: shuffle_v4f64_0145:
284 ; ALL-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
286 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
287 ret <4 x double> %shuffle
290 define <4 x double> @shuffle_v4f64_4501(<4 x double> %a, <4 x double> %b) {
291 ; ALL-LABEL: shuffle_v4f64_4501:
293 ; ALL-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
295 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 4, i32 5, i32 0, i32 1>
296 ret <4 x double> %shuffle
299 define <4 x double> @shuffle_v4f64_0167(<4 x double> %a, <4 x double> %b) {
300 ; ALL-LABEL: shuffle_v4f64_0167:
302 ; ALL-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3]
304 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 1, i32 6, i32 7>
305 ret <4 x double> %shuffle
308 define <4 x double> @shuffle_v4f64_1054(<4 x double> %a, <4 x double> %b) {
309 ; ALL-LABEL: shuffle_v4f64_1054:
311 ; ALL-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
312 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
314 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 0, i32 5, i32 4>
315 ret <4 x double> %shuffle
318 define <4 x double> @shuffle_v4f64_3254(<4 x double> %a, <4 x double> %b) {
319 ; ALL-LABEL: shuffle_v4f64_3254:
321 ; ALL-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3],ymm1[0,1]
322 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
324 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 3, i32 2, i32 5, i32 4>
325 ret <4 x double> %shuffle
328 define <4 x double> @shuffle_v4f64_3276(<4 x double> %a, <4 x double> %b) {
329 ; ALL-LABEL: shuffle_v4f64_3276:
331 ; ALL-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3],ymm1[2,3]
332 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
334 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 3, i32 2, i32 7, i32 6>
335 ret <4 x double> %shuffle
338 define <4 x double> @shuffle_v4f64_1076(<4 x double> %a, <4 x double> %b) {
339 ; ALL-LABEL: shuffle_v4f64_1076:
341 ; ALL-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3]
342 ; ALL-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
344 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 1, i32 0, i32 7, i32 6>
345 ret <4 x double> %shuffle
348 define <4 x double> @shuffle_v4f64_0415(<4 x double> %a, <4 x double> %b) {
349 ; AVX1-LABEL: shuffle_v4f64_0415:
351 ; AVX1-NEXT: vunpckhpd {{.*#+}} xmm2 = xmm0[1],xmm1[1]
352 ; AVX1-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
353 ; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
356 ; AVX2-LABEL: shuffle_v4f64_0415:
358 ; AVX2-NEXT: vpermpd {{.*#+}} ymm1 = ymm1[0,0,2,1]
359 ; AVX2-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,1,1,3]
360 ; AVX2-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3]
362 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
363 ret <4 x double> %shuffle
366 define <4 x double> @shuffle_v4f64_u062(<4 x double> %a, <4 x double> %b) {
367 ; ALL-LABEL: shuffle_v4f64_u062:
369 ; ALL-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
371 %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 undef, i32 0, i32 6, i32 2>
372 ret <4 x double> %shuffle
375 define <4 x i64> @shuffle_v4i64_0000(<4 x i64> %a, <4 x i64> %b) {
376 ; AVX1-LABEL: shuffle_v4i64_0000:
378 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
379 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
382 ; AVX2-LABEL: shuffle_v4i64_0000:
384 ; AVX2-NEXT: vbroadcastsd %xmm0, %ymm0
386 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
387 ret <4 x i64> %shuffle
390 define <4 x i64> @shuffle_v4i64_0001(<4 x i64> %a, <4 x i64> %b) {
391 ; AVX1-LABEL: shuffle_v4i64_0001:
393 ; AVX1-NEXT: vmovddup {{.*#+}} xmm1 = xmm0[0,0]
394 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
397 ; AVX2-LABEL: shuffle_v4i64_0001:
399 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,0,0,1]
401 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 1>
402 ret <4 x i64> %shuffle
405 define <4 x i64> @shuffle_v4i64_0020(<4 x i64> %a, <4 x i64> %b) {
406 ; AVX1-LABEL: shuffle_v4i64_0020:
408 ; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
409 ; AVX1-NEXT: vunpcklpd {{.*#+}} xmm1 = xmm1[0],xmm0[0]
410 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
411 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
414 ; AVX2-LABEL: shuffle_v4i64_0020:
416 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,0,2,0]
418 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 0, i32 2, i32 0>
419 ret <4 x i64> %shuffle
422 define <4 x i64> @shuffle_v4i64_0112(<4 x i64> %a, <4 x i64> %b) {
423 ; AVX1-LABEL: shuffle_v4i64_0112:
425 ; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
426 ; AVX1-NEXT: vshufpd {{.*#+}} xmm1 = xmm0[1],xmm1[0]
427 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
430 ; AVX2-LABEL: shuffle_v4i64_0112:
432 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,1,1,2]
434 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 1, i32 1, i32 2>
435 ret <4 x i64> %shuffle
438 define <4 x i64> @shuffle_v4i64_0300(<4 x i64> %a, <4 x i64> %b) {
439 ; AVX1-LABEL: shuffle_v4i64_0300:
441 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
442 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm1 = ymm1[0,1,2,2]
443 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3]
446 ; AVX2-LABEL: shuffle_v4i64_0300:
448 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,3,0,0]
450 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 3, i32 0, i32 0>
451 ret <4 x i64> %shuffle
454 define <4 x i64> @shuffle_v4i64_1000(<4 x i64> %a, <4 x i64> %b) {
455 ; AVX1-LABEL: shuffle_v4i64_1000:
457 ; AVX1-NEXT: vpermilpd {{.*#+}} xmm1 = xmm0[1,0]
458 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
459 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
462 ; AVX2-LABEL: shuffle_v4i64_1000:
464 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[1,0,0,0]
466 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
467 ret <4 x i64> %shuffle
470 define <4 x i64> @shuffle_v4i64_2200(<4 x i64> %a, <4 x i64> %b) {
471 ; AVX1-LABEL: shuffle_v4i64_2200:
473 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
474 ; AVX1-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
477 ; AVX2-LABEL: shuffle_v4i64_2200:
479 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[2,2,0,0]
481 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 2, i32 2, i32 0, i32 0>
482 ret <4 x i64> %shuffle
485 define <4 x i64> @shuffle_v4i64_3330(<4 x i64> %a, <4 x i64> %b) {
486 ; AVX1-LABEL: shuffle_v4i64_3330:
488 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
489 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0,1,2],ymm0[3]
490 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,1,3,2]
493 ; AVX2-LABEL: shuffle_v4i64_3330:
495 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[3,3,3,0]
497 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 3, i32 3, i32 3, i32 0>
498 ret <4 x i64> %shuffle
501 define <4 x i64> @shuffle_v4i64_3210(<4 x i64> %a, <4 x i64> %b) {
502 ; AVX1-LABEL: shuffle_v4i64_3210:
504 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
505 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
508 ; AVX2-LABEL: shuffle_v4i64_3210:
510 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[3,2,1,0]
512 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
513 ret <4 x i64> %shuffle
516 define <4 x i64> @shuffle_v4i64_0124(<4 x i64> %a, <4 x i64> %b) {
517 ; AVX1-LABEL: shuffle_v4i64_0124:
519 ; AVX1-NEXT: vmovddup {{.*#+}} xmm1 = xmm1[0,0]
520 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm1
521 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1,2],ymm1[3]
524 ; AVX2-LABEL: shuffle_v4i64_0124:
526 ; AVX2-NEXT: vpbroadcastq %xmm1, %ymm1
527 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
529 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 1, i32 2, i32 4>
530 ret <4 x i64> %shuffle
533 define <4 x i64> @shuffle_v4i64_0142(<4 x i64> %a, <4 x i64> %b) {
534 ; AVX1-LABEL: shuffle_v4i64_0142:
536 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1
537 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[0,1,2,2]
538 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2],ymm0[3]
541 ; AVX2-LABEL: shuffle_v4i64_0142:
543 ; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm1, %ymm1
544 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,1,2,2]
545 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm1[4,5],ymm0[6,7]
547 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 2>
548 ret <4 x i64> %shuffle
551 define <4 x i64> @shuffle_v4i64_0412(<4 x i64> %a, <4 x i64> %b) {
552 ; AVX1-LABEL: shuffle_v4i64_0412:
554 ; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm2
555 ; AVX1-NEXT: vshufpd {{.*#+}} xmm2 = xmm0[1],xmm2[0]
556 ; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
557 ; AVX1-NEXT: vmovddup {{.*#+}} ymm1 = ymm1[0,0,2,2]
558 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2,3]
561 ; AVX2-LABEL: shuffle_v4i64_0412:
563 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,1,1,2]
564 ; AVX2-NEXT: vpbroadcastq %xmm1, %ymm1
565 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5,6,7]
567 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 2>
568 ret <4 x i64> %shuffle
571 define <4 x i64> @shuffle_v4i64_4012(<4 x i64> %a, <4 x i64> %b) {
572 ; AVX1-LABEL: shuffle_v4i64_4012:
574 ; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm2
575 ; AVX1-NEXT: vshufpd {{.*#+}} xmm2 = xmm0[1],xmm2[0]
576 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
577 ; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
578 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0],ymm0[1,2,3]
581 ; AVX2-LABEL: shuffle_v4i64_4012:
583 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,0,1,2]
584 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3,4,5,6,7]
586 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 4, i32 0, i32 1, i32 2>
587 ret <4 x i64> %shuffle
590 define <4 x i64> @shuffle_v4i64_0145(<4 x i64> %a, <4 x i64> %b) {
591 ; ALL-LABEL: shuffle_v4i64_0145:
593 ; ALL-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
595 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
596 ret <4 x i64> %shuffle
599 define <4 x i64> @shuffle_v4i64_0451(<4 x i64> %a, <4 x i64> %b) {
600 ; AVX1-LABEL: shuffle_v4i64_0451:
602 ; AVX1-NEXT: vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
603 ; AVX1-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
604 ; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
607 ; AVX2-LABEL: shuffle_v4i64_0451:
609 ; AVX2-NEXT: vinserti128 $1, %xmm0, %ymm0, %ymm0
610 ; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,0,1,3]
611 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3,4,5],ymm0[6,7]
613 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 4, i32 5, i32 1>
614 ret <4 x i64> %shuffle
617 define <4 x i64> @shuffle_v4i64_4501(<4 x i64> %a, <4 x i64> %b) {
618 ; ALL-LABEL: shuffle_v4i64_4501:
620 ; ALL-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
622 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 4, i32 5, i32 0, i32 1>
623 ret <4 x i64> %shuffle
626 define <4 x i64> @shuffle_v4i64_4015(<4 x i64> %a, <4 x i64> %b) {
627 ; AVX1-LABEL: shuffle_v4i64_4015:
629 ; AVX1-NEXT: vunpckhpd {{.*#+}} xmm2 = xmm0[1],xmm1[1]
630 ; AVX1-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm1[0],xmm0[0]
631 ; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
634 ; AVX2-LABEL: shuffle_v4i64_4015:
636 ; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm1, %ymm1
637 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,0,1,3]
638 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3,4,5],ymm1[6,7]
640 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 4, i32 0, i32 1, i32 5>
641 ret <4 x i64> %shuffle
644 define <4 x i64> @shuffle_v4i64_2u35(<4 x i64> %a, <4 x i64> %b) {
645 ; AVX1-LABEL: shuffle_v4i64_2u35:
647 ; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
648 ; AVX1-NEXT: vunpckhpd {{.*#+}} xmm1 = xmm0[1],xmm1[1]
649 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
652 ; AVX2-LABEL: shuffle_v4i64_2u35:
654 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5,6,7]
655 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[2,1,3,1]
657 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 2, i32 undef, i32 3, i32 5>
658 ret <4 x i64> %shuffle
661 define <4 x i64> @shuffle_v4i64_1251(<4 x i64> %a, <4 x i64> %b) {
662 ; AVX1-LABEL: shuffle_v4i64_1251:
664 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3,0,1]
665 ; AVX1-NEXT: vshufpd {{.*#+}} ymm0 = ymm0[1],ymm2[0],ymm0[2],ymm2[3]
666 ; AVX1-NEXT: vpermilpd {{.*#+}} xmm1 = xmm1[1,0]
667 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm1
668 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2],ymm0[3]
671 ; AVX2-LABEL: shuffle_v4i64_1251:
673 ; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,1,1,3]
674 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[1,2,2,1]
675 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm1[4,5],ymm0[6,7]
677 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 1, i32 2, i32 5, i32 1>
678 ret <4 x i64> %shuffle
681 define <4 x i64> @shuffle_v4i64_1054(<4 x i64> %a, <4 x i64> %b) {
682 ; AVX1-LABEL: shuffle_v4i64_1054:
684 ; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
685 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
688 ; AVX2-LABEL: shuffle_v4i64_1054:
690 ; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
691 ; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[2,3,0,1,6,7,4,5]
693 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 1, i32 0, i32 5, i32 4>
694 ret <4 x i64> %shuffle
697 define <4 x i64> @shuffle_v4i64_3254(<4 x i64> %a, <4 x i64> %b) {
698 ; AVX1-LABEL: shuffle_v4i64_3254:
700 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3],ymm1[0,1]
701 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
704 ; AVX2-LABEL: shuffle_v4i64_3254:
706 ; AVX2-NEXT: vperm2i128 {{.*#+}} ymm0 = ymm0[2,3],ymm1[0,1]
707 ; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[2,3,0,1,6,7,4,5]
709 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 3, i32 2, i32 5, i32 4>
710 ret <4 x i64> %shuffle
713 define <4 x i64> @shuffle_v4i64_3276(<4 x i64> %a, <4 x i64> %b) {
714 ; AVX1-LABEL: shuffle_v4i64_3276:
716 ; AVX1-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm0[2,3],ymm1[2,3]
717 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
720 ; AVX2-LABEL: shuffle_v4i64_3276:
722 ; AVX2-NEXT: vperm2i128 {{.*#+}} ymm0 = ymm0[2,3],ymm1[2,3]
723 ; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[2,3,0,1,6,7,4,5]
725 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 3, i32 2, i32 7, i32 6>
726 ret <4 x i64> %shuffle
729 define <4 x i64> @shuffle_v4i64_1076(<4 x i64> %a, <4 x i64> %b) {
730 ; AVX1-LABEL: shuffle_v4i64_1076:
732 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3]
733 ; AVX1-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
736 ; AVX2-LABEL: shuffle_v4i64_1076:
738 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
739 ; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[2,3,0,1,6,7,4,5]
741 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 1, i32 0, i32 7, i32 6>
742 ret <4 x i64> %shuffle
745 define <4 x i64> @shuffle_v4i64_0415(<4 x i64> %a, <4 x i64> %b) {
746 ; AVX1-LABEL: shuffle_v4i64_0415:
748 ; AVX1-NEXT: vunpckhpd {{.*#+}} xmm2 = xmm0[1],xmm1[1]
749 ; AVX1-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
750 ; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
753 ; AVX2-LABEL: shuffle_v4i64_0415:
755 ; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,0,2,1]
756 ; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,1,1,3]
757 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
759 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
760 ret <4 x i64> %shuffle
763 define <4 x i64> @shuffle_v4i64_z4z6(<4 x i64> %a) {
764 ; AVX1-LABEL: shuffle_v4i64_z4z6:
766 ; AVX1-NEXT: vxorpd %ymm1, %ymm1, %ymm1
767 ; AVX1-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
770 ; AVX2-LABEL: shuffle_v4i64_z4z6:
772 ; AVX2-NEXT: vpslldq {{.*#+}} ymm0 = zero,zero,zero,zero,zero,zero,zero,zero,ymm0[0,1,2,3,4,5,6,7],zero,zero,zero,zero,zero,zero,zero,zero,ymm0[16,17,18,19,20,21,22,23]
774 %shuffle = shufflevector <4 x i64> zeroinitializer, <4 x i64> %a, <4 x i32> <i32 0, i32 4, i32 0, i32 6>
775 ret <4 x i64> %shuffle
778 define <4 x i64> @shuffle_v4i64_5zuz(<4 x i64> %a) {
779 ; AVX1-LABEL: shuffle_v4i64_5zuz:
781 ; AVX1-NEXT: vxorpd %ymm1, %ymm1, %ymm1
782 ; AVX1-NEXT: vunpckhpd {{.*#+}} ymm0 = ymm0[1],ymm1[1],ymm0[3],ymm1[3]
785 ; AVX2-LABEL: shuffle_v4i64_5zuz:
787 ; AVX2-NEXT: vpsrldq {{.*#+}} ymm0 = ymm0[8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,ymm0[24,25,26,27,28,29,30,31],zero,zero,zero,zero,zero,zero,zero,zero
789 %shuffle = shufflevector <4 x i64> zeroinitializer, <4 x i64> %a, <4 x i32> <i32 5, i32 0, i32 undef, i32 0>
790 ret <4 x i64> %shuffle
793 define <4 x i64> @shuffle_v4i64_40u2(<4 x i64> %a, <4 x i64> %b) {
794 ; AVX1-LABEL: shuffle_v4i64_40u2:
796 ; AVX1-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
799 ; AVX2-LABEL: shuffle_v4i64_40u2:
801 ; AVX2-NEXT: vpunpcklqdq {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
803 %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 4, i32 0, i32 undef, i32 2>
804 ret <4 x i64> %shuffle
807 define <4 x i64> @stress_test1(<4 x i64> %a, <4 x i64> %b) {
808 ; ALL-LABEL: stress_test1:
810 %c = shufflevector <4 x i64> %b, <4 x i64> undef, <4 x i32> <i32 3, i32 1, i32 1, i32 0>
811 %d = shufflevector <4 x i64> %c, <4 x i64> undef, <4 x i32> <i32 3, i32 undef, i32 2, i32 undef>
812 %e = shufflevector <4 x i64> %b, <4 x i64> undef, <4 x i32> <i32 3, i32 3, i32 1, i32 undef>
813 %f = shufflevector <4 x i64> %d, <4 x i64> %e, <4 x i32> <i32 5, i32 1, i32 1, i32 0>
818 define <4 x i64> @insert_reg_and_zero_v4i64(i64 %a) {
819 ; ALL-LABEL: insert_reg_and_zero_v4i64:
821 ; ALL-NEXT: vmovq %rdi, %xmm0
823 %v = insertelement <4 x i64> undef, i64 %a, i64 0
824 %shuffle = shufflevector <4 x i64> %v, <4 x i64> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
825 ret <4 x i64> %shuffle
828 define <4 x i64> @insert_mem_and_zero_v4i64(i64* %ptr) {
829 ; ALL-LABEL: insert_mem_and_zero_v4i64:
831 ; ALL-NEXT: vmovq {{.*#+}} xmm0 = mem[0],zero
833 %a = load i64, i64* %ptr
834 %v = insertelement <4 x i64> undef, i64 %a, i64 0
835 %shuffle = shufflevector <4 x i64> %v, <4 x i64> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
836 ret <4 x i64> %shuffle
839 define <4 x double> @insert_reg_and_zero_v4f64(double %a) {
840 ; ALL-LABEL: insert_reg_and_zero_v4f64:
842 ; ALL-NEXT: vxorpd %ymm1, %ymm1, %ymm1
843 ; ALL-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3]
845 %v = insertelement <4 x double> undef, double %a, i32 0
846 %shuffle = shufflevector <4 x double> %v, <4 x double> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
847 ret <4 x double> %shuffle
850 define <4 x double> @insert_mem_and_zero_v4f64(double* %ptr) {
851 ; ALL-LABEL: insert_mem_and_zero_v4f64:
853 ; ALL-NEXT: vmovsd {{.*#+}} xmm0 = mem[0],zero
855 %a = load double, double* %ptr
856 %v = insertelement <4 x double> undef, double %a, i32 0
857 %shuffle = shufflevector <4 x double> %v, <4 x double> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
858 ret <4 x double> %shuffle
861 define <4 x double> @splat_mem_v4f64(double* %ptr) {
862 ; ALL-LABEL: splat_mem_v4f64:
864 ; ALL-NEXT: vbroadcastsd (%rdi), %ymm0
866 %a = load double, double* %ptr
867 %v = insertelement <4 x double> undef, double %a, i32 0
868 %shuffle = shufflevector <4 x double> %v, <4 x double> undef, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
869 ret <4 x double> %shuffle
872 define <4 x i64> @splat_mem_v4i64(i64* %ptr) {
873 ; ALL-LABEL: splat_mem_v4i64:
875 ; ALL-NEXT: vbroadcastsd (%rdi), %ymm0
877 %a = load i64, i64* %ptr
878 %v = insertelement <4 x i64> undef, i64 %a, i64 0
879 %shuffle = shufflevector <4 x i64> %v, <4 x i64> undef, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
880 ret <4 x i64> %shuffle
883 define <4 x double> @splat_mem_v4f64_2(double* %p) {
884 ; ALL-LABEL: splat_mem_v4f64_2:
886 ; ALL-NEXT: vbroadcastsd (%rdi), %ymm0
888 %1 = load double, double* %p
889 %2 = insertelement <2 x double> undef, double %1, i32 0
890 %3 = shufflevector <2 x double> %2, <2 x double> undef, <4 x i32> zeroinitializer
894 define <4 x double> @splat_v4f64(<2 x double> %r) {
895 ; AVX1-LABEL: splat_v4f64:
897 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
898 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
901 ; AVX2-LABEL: splat_v4f64:
903 ; AVX2-NEXT: vbroadcastsd %xmm0, %ymm0
905 %1 = shufflevector <2 x double> %r, <2 x double> undef, <4 x i32> zeroinitializer
909 define <4 x i64> @splat_mem_v4i64_from_v2i64(<2 x i64>* %ptr) {
910 ; AVX1-LABEL: splat_mem_v4i64_from_v2i64:
912 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = mem[0,0]
913 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
916 ; AVX2-LABEL: splat_mem_v4i64_from_v2i64:
918 ; AVX2-NEXT: vbroadcastsd (%rdi), %ymm0
920 %v = load <2 x i64>, <2 x i64>* %ptr
921 %shuffle = shufflevector <2 x i64> %v, <2 x i64> undef, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
922 ret <4 x i64> %shuffle
925 define <4 x double> @splat_mem_v4f64_from_v2f64(<2 x double>* %ptr) {
926 ; AVX1-LABEL: splat_mem_v4f64_from_v2f64:
928 ; AVX1-NEXT: vmovddup {{.*#+}} xmm0 = mem[0,0]
929 ; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
932 ; AVX2-LABEL: splat_mem_v4f64_from_v2f64:
934 ; AVX2-NEXT: vbroadcastsd (%rdi), %ymm0
936 %v = load <2 x double>, <2 x double>* %ptr
937 %shuffle = shufflevector <2 x double> %v, <2 x double> undef, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
938 ret <4 x double> %shuffle
941 define <4 x i64> @splat128_mem_v4i64_from_v2i64(<2 x i64>* %ptr) {
942 ; ALL-LABEL: splat128_mem_v4i64_from_v2i64:
944 ; ALL-NEXT: vmovaps (%rdi), %xmm0
945 ; ALL-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
947 %v = load <2 x i64>, <2 x i64>* %ptr
948 %shuffle = shufflevector <2 x i64> %v, <2 x i64> undef, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
949 ret <4 x i64> %shuffle
952 define <4 x double> @splat128_mem_v4f64_from_v2f64(<2 x double>* %ptr) {
953 ; ALL-LABEL: splat128_mem_v4f64_from_v2f64:
955 ; ALL-NEXT: vmovaps (%rdi), %xmm0
956 ; ALL-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
958 %v = load <2 x double>, <2 x double>* %ptr
959 %shuffle = shufflevector <2 x double> %v, <2 x double> undef, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
960 ret <4 x double> %shuffle
963 define <4 x double> @bitcast_v4f64_0426(<4 x double> %a, <4 x double> %b) {
964 ; AVX1-LABEL: bitcast_v4f64_0426:
966 ; AVX1-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
969 ; AVX2-LABEL: bitcast_v4f64_0426:
971 ; AVX2-NEXT: vpunpcklqdq {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
973 %shuffle64 = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 4, i32 0, i32 6, i32 2>
974 %bitcast32 = bitcast <4 x double> %shuffle64 to <8 x float>
975 %shuffle32 = shufflevector <8 x float> %bitcast32, <8 x float> undef, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
976 %bitcast16 = bitcast <8 x float> %shuffle32 to <16 x i16>
977 %shuffle16 = shufflevector <16 x i16> %bitcast16, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 0, i32 1, i32 6, i32 7, i32 4, i32 5, i32 10, i32 11, i32 8, i32 9, i32 14, i32 15, i32 12, i32 13>
978 %bitcast64 = bitcast <16 x i16> %shuffle16 to <4 x double>
979 ret <4 x double> %bitcast64
982 define <4 x i64> @concat_v4i64_0167(<4 x i64> %a0, <4 x i64> %a1) {
983 ; AVX1-LABEL: concat_v4i64_0167:
985 ; AVX1-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3]
988 ; AVX2-LABEL: concat_v4i64_0167:
990 ; AVX2-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
992 %a0lo = shufflevector <4 x i64> %a0, <4 x i64> %a1, <2 x i32> <i32 0, i32 1>
993 %a1hi = shufflevector <4 x i64> %a0, <4 x i64> %a1, <2 x i32> <i32 6, i32 7>
994 %shuffle64 = shufflevector <2 x i64> %a0lo, <2 x i64> %a1hi, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
995 ret <4 x i64> %shuffle64
998 define <4 x i64> @concat_v4i64_0145_bc(<4 x i64> %a0, <4 x i64> %a1) {
999 ; ALL-LABEL: concat_v4i64_0145_bc:
1001 ; ALL-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
1003 %a0lo = shufflevector <4 x i64> %a0, <4 x i64> %a1, <2 x i32> <i32 0, i32 1>
1004 %a1lo = shufflevector <4 x i64> %a0, <4 x i64> %a1, <2 x i32> <i32 4, i32 5>
1005 %bc0lo = bitcast <2 x i64> %a0lo to <4 x i32>
1006 %bc1lo = bitcast <2 x i64> %a1lo to <4 x i32>
1007 %shuffle32 = shufflevector <4 x i32> %bc0lo, <4 x i32> %bc1lo, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
1008 %shuffle64 = bitcast <8 x i32> %shuffle32 to <4 x i64>
1009 ret <4 x i64> %shuffle64
1012 define <4 x i64> @insert_dup_mem_v4i64(i64* %ptr) {
1013 ; ALL-LABEL: insert_dup_mem_v4i64:
1015 ; ALL-NEXT: vbroadcastsd (%rdi), %ymm0
1017 %tmp = load i64, i64* %ptr, align 1
1018 %tmp1 = insertelement <2 x i64> undef, i64 %tmp, i32 0
1019 %tmp2 = shufflevector <2 x i64> %tmp1, <2 x i64> undef, <4 x i32> zeroinitializer