Lines Matching +full:0 +full:x92
7 define <2 x i64> @test_mm_cvtne2ps2bf16_128(<4 x float> %A, <4 x float> %B) local_unnamed_addr #0 {
9 ; CHECK: # %bb.0: # %entry
10 ; CHECK-NEXT: vcvtne2ps2bf16 %xmm1, %xmm0, %xmm0 # encoding: [0x62,0xf2,0x7f,0x08,0x72,0xc1]
11 ; CHECK-NEXT: ret{{[l|q]}} # encoding: [0xc3]
13 %0 = tail call <8 x i16> @llvm.x86.avx512bf16.cvtne2ps2bf16.128(<4 x float> %A, <4 x float> %B) #2
14 %1 = bitcast <8 x i16> %0 to <2 x i64>
18 …t_mm_maskz_cvtne2ps2bf16_128(<4 x float> %A, <4 x float> %B, i8 zeroext %U) local_unnamed_addr #0 {
20 ; X86: # %bb.0: # %entry
21 ; X86-NEXT: vcvtne2ps2bf16 %xmm1, %xmm0, %xmm0 # encoding: [0x62,0xf2,0x7f,0x08,0x72,0xc1]
22 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
23 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
24 ; X86-NEXT: vmovdqu16 %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf1,0xff,0x89,0x6f,0xc0]
25 ; X86-NEXT: retl # encoding: [0xc3]
28 ; X64: # %bb.0: # %entry
29 ; X64-NEXT: vcvtne2ps2bf16 %xmm1, %xmm0, %xmm0 # encoding: [0x62,0xf2,0x7f,0x08,0x72,0xc1]
30 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
31 ; X64-NEXT: vmovdqu16 %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf1,0xff,0x89,0x6f,0xc0]
32 ; X64-NEXT: retq # encoding: [0xc3]
34 %0 = tail call <8 x i16> @llvm.x86.avx512bf16.cvtne2ps2bf16.128(<4 x float> %A, <4 x float> %B) #2
36 %2 = select <8 x i1> %1, <8 x i16> %0, <8 x i16> zeroinitializer
41 …ne2ps2bf16_128(<2 x i64> %C, i8 zeroext %U, <4 x float> %A, <4 x float> %B) local_unnamed_addr #0 {
43 ; X86: # %bb.0: # %entry
44 ; X86-NEXT: vcvtne2ps2bf16 %xmm2, %xmm1, %xmm1 # encoding: [0x62,0xf2,0x77,0x08,0x72,0xca]
45 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
46 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
47 ; X86-NEXT: vmovdqu16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf1,0xff,0x09,0x6f,0xc1]
48 ; X86-NEXT: retl # encoding: [0xc3]
51 ; X64: # %bb.0: # %entry
52 ; X64-NEXT: vcvtne2ps2bf16 %xmm2, %xmm1, %xmm1 # encoding: [0x62,0xf2,0x77,0x08,0x72,0xca]
53 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
54 ; X64-NEXT: vmovdqu16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf1,0xff,0x09,0x6f,0xc1]
55 ; X64-NEXT: retq # encoding: [0xc3]
57 %0 = tail call <8 x i16> @llvm.x86.avx512bf16.cvtne2ps2bf16.128(<4 x float> %A, <4 x float> %B) #2
60 %3 = select <8 x i1> %2, <8 x i16> %0, <8 x i16> %1
69 ; CHECK: # %bb.0: # %entry
70 ; CHECK-NEXT: vcvtne2ps2bf16 %ymm1, %ymm0, %ymm0 # encoding: [0x62,0xf2,0x7f,0x28,0x72,0xc1]
71 ; CHECK-NEXT: ret{{[l|q]}} # encoding: [0xc3]
73 …%0 = tail call <16 x i16> @llvm.x86.avx512bf16.cvtne2ps2bf16.256(<8 x float> %A, <8 x float> %B) #4
74 %1 = bitcast <16 x i16> %0 to <4 x i64>
80 ; X86: # %bb.0: # %entry
81 ; X86-NEXT: vcvtne2ps2bf16 %ymm1, %ymm0, %ymm0 # encoding: [0x62,0xf2,0x7f,0x28,0x72,0xc1]
82 ; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1 # encoding: [0xc5,0xf8,0x90,0x4c,0x24,0x04]
83 ; X86-NEXT: vmovdqu16 %ymm0, %ymm0 {%k1} {z} # encoding: [0x62,0xf1,0xff,0xa9,0x6f,0xc0]
84 ; X86-NEXT: retl # encoding: [0xc3]
87 ; X64: # %bb.0: # %entry
88 ; X64-NEXT: vcvtne2ps2bf16 %ymm1, %ymm0, %ymm0 # encoding: [0x62,0xf2,0x7f,0x28,0x72,0xc1]
89 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
90 ; X64-NEXT: vmovdqu16 %ymm0, %ymm0 {%k1} {z} # encoding: [0x62,0xf1,0xff,0xa9,0x6f,0xc0]
91 ; X64-NEXT: retq # encoding: [0xc3]
93 …%0 = tail call <16 x i16> @llvm.x86.avx512bf16.cvtne2ps2bf16.256(<8 x float> %A, <8 x float> %B) #4
95 %2 = select <16 x i1> %1, <16 x i16> %0, <16 x i16> zeroinitializer
102 ; X86: # %bb.0: # %entry
103 ; X86-NEXT: vcvtne2ps2bf16 %ymm2, %ymm1, %ymm1 # encoding: [0x62,0xf2,0x77,0x28,0x72,0xca]
104 ; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1 # encoding: [0xc5,0xf8,0x90,0x4c,0x24,0x04]
105 ; X86-NEXT: vmovdqu16 %ymm1, %ymm0 {%k1} # encoding: [0x62,0xf1,0xff,0x29,0x6f,0xc1]
106 ; X86-NEXT: retl # encoding: [0xc3]
109 ; X64: # %bb.0: # %entry
110 ; X64-NEXT: vcvtne2ps2bf16 %ymm2, %ymm1, %ymm1 # encoding: [0x62,0xf2,0x77,0x28,0x72,0xca]
111 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
112 ; X64-NEXT: vmovdqu16 %ymm1, %ymm0 {%k1} # encoding: [0x62,0xf1,0xff,0x29,0x6f,0xc1]
113 ; X64-NEXT: retq # encoding: [0xc3]
115 …%0 = tail call <16 x i16> @llvm.x86.avx512bf16.cvtne2ps2bf16.256(<8 x float> %A, <8 x float> %B) #4
118 %3 = select <16 x i1> %2, <16 x i16> %0, <16 x i16> %1
127 ; CHECK: # %bb.0: # %entry
128 ; CHECK-NEXT: vcvtneps2bf16 %ymm0, %xmm0 # encoding: [0x62,0xf2,0x7e,0x28,0x72,0xc0]
129 ; CHECK-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77]
130 ; CHECK-NEXT: ret{{[l|q]}} # encoding: [0xc3]
132 %0 = tail call <8 x i16> @llvm.x86.avx512bf16.cvtneps2bf16.256(<8 x float> %A) #4
133 %1 = bitcast <8 x i16> %0 to <2 x i64>
139 ; X86: # %bb.0: # %entry
140 ; X86-NEXT: vcvtneps2bf16 %ymm0, %xmm0 # encoding: [0x62,0xf2,0x7e,0x28,0x72,0xc0]
141 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
142 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
143 ; X86-NEXT: vmovdqu16 %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf1,0xff,0x89,0x6f,0xc0]
144 ; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77]
145 ; X86-NEXT: retl # encoding: [0xc3]
148 ; X64: # %bb.0: # %entry
149 ; X64-NEXT: vcvtneps2bf16 %ymm0, %xmm0 # encoding: [0x62,0xf2,0x7e,0x28,0x72,0xc0]
150 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
151 ; X64-NEXT: vmovdqu16 %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf1,0xff,0x89,0x6f,0xc0]
152 ; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77]
153 ; X64-NEXT: retq # encoding: [0xc3]
155 %0 = tail call <8 x i16> @llvm.x86.avx512bf16.cvtneps2bf16.256(<8 x float> %A) #4
157 %2 = select <8 x i1> %1, <8 x i16> %0, <8 x i16> zeroinitializer
164 ; X86: # %bb.0: # %entry
165 ; X86-NEXT: vcvtneps2bf16 %ymm1, %xmm1 # encoding: [0x62,0xf2,0x7e,0x28,0x72,0xc9]
166 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
167 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
168 ; X86-NEXT: vmovdqu16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf1,0xff,0x09,0x6f,0xc1]
169 ; X86-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77]
170 ; X86-NEXT: retl # encoding: [0xc3]
173 ; X64: # %bb.0: # %entry
174 ; X64-NEXT: vcvtneps2bf16 %ymm1, %xmm1 # encoding: [0x62,0xf2,0x7e,0x28,0x72,0xc9]
175 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
176 ; X64-NEXT: vmovdqu16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf1,0xff,0x09,0x6f,0xc1]
177 ; X64-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77]
178 ; X64-NEXT: retq # encoding: [0xc3]
180 %0 = tail call <8 x i16> @llvm.x86.avx512bf16.cvtneps2bf16.256(<8 x float> %A) #4
183 %3 = select <8 x i1> %2, <8 x i16> %0, <8 x i16> %1
192 ; CHECK: # %bb.0: # %entry
193 ; CHECK-NEXT: vcvtneps2bf16 %xmm0, %xmm0 # encoding: [0x62,0xf2,0x7e,0x08,0x72,0xc0]
194 ; CHECK-NEXT: ret{{[l|q]}} # encoding: [0xc3]
196 …%0 = tail call <8 x i16> @llvm.x86.avx512bf16.mask.cvtneps2bf16.128(<4 x float> %A, <8 x i16> unde…
197 %1 = bitcast <8 x i16> %0 to <2 x i64>
203 ; X86: # %bb.0: # %entry
204 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
205 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
206 ; X86-NEXT: vcvtneps2bf16 %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0x7e,0x89,0x72,0xc0]
207 ; X86-NEXT: retl # encoding: [0xc3]
210 ; X64: # %bb.0: # %entry
211 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
212 ; X64-NEXT: vcvtneps2bf16 %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0x7e,0x89,0x72,0xc0]
213 ; X64-NEXT: retq # encoding: [0xc3]
215 %0 = bitcast i8 %U to <8 x i1>
216 %1 = shufflevector <8 x i1> %0, <8 x i1> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
224 ; X86: # %bb.0: # %entry
225 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
226 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
227 ; X86-NEXT: vcvtneps2bf16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf2,0x7e,0x09,0x72,0xc1]
228 ; X86-NEXT: retl # encoding: [0xc3]
231 ; X64: # %bb.0: # %entry
232 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
233 ; X64-NEXT: vcvtneps2bf16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf2,0x7e,0x09,0x72,0xc1]
234 ; X64-NEXT: retq # encoding: [0xc3]
236 %0 = bitcast i8 %U to <8 x i1>
237 %1 = shufflevector <8 x i1> %0, <8 x i1> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
248 ; X86: # %bb.0: # %entry
249 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
250 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
251 ; X86-NEXT: vcvtneps2bf16 %xmm1, %xmm1 # encoding: [0x62,0xf2,0x7e,0x08,0x72,0xc9]
252 ; X86-NEXT: vmovdqu16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf1,0xff,0x09,0x6f,0xc1]
253 ; X86-NEXT: retl # encoding: [0xc3]
256 ; X64: # %bb.0: # %entry
257 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
258 ; X64-NEXT: vcvtneps2bf16 %xmm1, %xmm1 # encoding: [0x62,0xf2,0x7e,0x08,0x72,0xc9]
259 ; X64-NEXT: vmovdqu16 %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf1,0xff,0x09,0x6f,0xc1]
260 ; X64-NEXT: retq # encoding: [0xc3]
262 %0 = bitcast i8 %U to <8 x i1>
265 %3 = select <8 x i1> %0, <8 x i16> %2, <8 x i16> %1
274 ; CHECK: # %bb.0: # %entry
275 ; CHECK-NEXT: vdpbf16ps %ymm2, %ymm1, %ymm0 # encoding: [0x62,0xf2,0x76,0x28,0x52,0xc2]
276 ; CHECK-NEXT: ret{{[l|q]}} # encoding: [0xc3]
278 …%0 = tail call <8 x float> @llvm.x86.avx512bf16.dpbf16ps.256(<8 x float> %E, <8 x i32> %A, <8 x i3…
279 ret <8 x float> %0
284 ; X86: # %bb.0: # %entry
285 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
286 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
287 ; X86-NEXT: vdpbf16ps %ymm2, %ymm1, %ymm0 {%k1} {z} # encoding: [0x62,0xf2,0x76,0xa9,0x52,0xc2]
288 ; X86-NEXT: retl # encoding: [0xc3]
291 ; X64: # %bb.0: # %entry
292 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
293 ; X64-NEXT: vdpbf16ps %ymm2, %ymm1, %ymm0 {%k1} {z} # encoding: [0x62,0xf2,0x76,0xa9,0x52,0xc2]
294 ; X64-NEXT: retq # encoding: [0xc3]
296 …%0 = tail call <8 x float> @llvm.x86.avx512bf16.dpbf16ps.256(<8 x float> %E, <8 x i32> %A, <8 x i3…
298 %2 = select <8 x i1> %1, <8 x float> %0, <8 x float> zeroinitializer
303 ; X86: # %bb.0: # %entry
304 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
305 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
306 ; X86-NEXT: vdpbf16ps %ymm2, %ymm1, %ymm0 {%k1} # encoding: [0x62,0xf2,0x76,0x29,0x52,0xc2]
307 ; X86-NEXT: retl # encoding: [0xc3]
310 ; X64: # %bb.0: # %entry
311 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
312 ; X64-NEXT: vdpbf16ps %ymm2, %ymm1, %ymm0 {%k1} # encoding: [0x62,0xf2,0x76,0x29,0x52,0xc2]
313 ; X64-NEXT: retq # encoding: [0xc3]
315 …%0 = tail call <8 x float> @llvm.x86.avx512bf16.dpbf16ps.256(<8 x float> %E, <8 x i32> %A, <8 x i3…
317 %2 = select <8 x i1> %1, <8 x float> %0, <8 x float> %E
325 ; CHECK: # %bb.0: # %entry
326 ; CHECK-NEXT: vdpbf16ps %xmm2, %xmm1, %xmm0 # encoding: [0x62,0xf2,0x76,0x08,0x52,0xc2]
327 ; CHECK-NEXT: ret{{[l|q]}} # encoding: [0xc3]
329 …%0 = tail call <4 x float> @llvm.x86.avx512bf16.dpbf16ps.128(<4 x float> %E, <4 x i32> %A, <4x i32…
330 ret <4 x float> %0
335 ; X86: # %bb.0: # %entry
336 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
337 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
338 ; X86-NEXT: vdpbf16ps %xmm2, %xmm1, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0x76,0x89,0x52,0xc2]
339 ; X86-NEXT: retl # encoding: [0xc3]
342 ; X64: # %bb.0: # %entry
343 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
344 ; X64-NEXT: vdpbf16ps %xmm2, %xmm1, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0x76,0x89,0x52,0xc2]
345 ; X64-NEXT: retq # encoding: [0xc3]
347 …%0 = tail call <4 x float> @llvm.x86.avx512bf16.dpbf16ps.128(<4 x float> %E, <4 x i32> %A, <4 x i3…
349 %2 = select <4 x i1> %1, <4 x float> %0, <4 x float> zeroinitializer
354 ; X86: # %bb.0: # %entry
355 ; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax # encoding: [0x0f,0xb6,0x44,0x24,0x04]
356 ; X86-NEXT: kmovd %eax, %k1 # encoding: [0xc5,0xfb,0x92,0xc8]
357 ; X86-NEXT: vdpbf16ps %xmm2, %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf2,0x76,0x09,0x52,0xc2]
358 ; X86-NEXT: retl # encoding: [0xc3]
361 ; X64: # %bb.0: # %entry
362 ; X64-NEXT: kmovd %edi, %k1 # encoding: [0xc5,0xfb,0x92,0xcf]
363 ; X64-NEXT: vdpbf16ps %xmm2, %xmm1, %xmm0 {%k1} # encoding: [0x62,0xf2,0x76,0x09,0x52,0xc2]
364 ; X64-NEXT: retq # encoding: [0xc3]
366 …%0 = tail call <4 x float> @llvm.x86.avx512bf16.dpbf16ps.128(<4 x float> %E, <4 x i32> %A, <4 x i3…
368 %2 = select <4 x i1> %1, <4 x float> %0, <4 x float> %E