maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/X86/avx512-intrinsics-upgrade.ll

3801 lines
175 KiB
LLVM
Raw Normal View History

[AVX512] Remove masked store intrinsics. Clang now emits generic masked store intrinsics instead. The intrinsics will be autoupgraded to the same generic masked stores. llvm-svn: 271245
2016-05-31 09:50:02 +08:00
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl | FileCheck %s
[x86] Lowering Mask Set1 intrinsics to LLVM IR This patch, together with a matching clang patch (https://reviews.llvm.org/D37668), implements the lowering of X86 mask set1 intrinsics to IR. Differential Revision: https://reviews.llvm.org/D37669 llvm-svn: 313625
2017-09-19 19:03:06 +08:00
define <16 x i32>@test_int_x86_avx512_mask_pbroadcastd_gpr_512(i32 %x0, <16 x i32> %x1, i16 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcastd_gpr_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpbroadcastd %edi, %zmm1
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpbroadcastd %edi, %zmm0 {%k1}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpbroadcastd %edi, %zmm1 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> %x1, i16 -1)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> %x1, i16 %mask)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> zeroinitializer, i16 %mask)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res2, %res3
ret <16 x i32> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32, <16 x i32>, i16)
define <8 x i64>@test_int_x86_avx512_mask_pbroadcastq_gpr_512(i64 %x0, <8 x i64> %x1, i8 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcastq_gpr_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpbroadcastq %rdi, %zmm1
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpbroadcastq %rdi, %zmm0 {%k1}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpbroadcastq %rdi, %zmm1 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> %x1,i8 -1)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> %x1,i8 %mask)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> zeroinitializer,i8 %mask)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res2, %res3
ret <8 x i64> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64, <8 x i64>, i8)
[X86][AVX512] Autoupgrade the BROADCAST intrinsics llvm-svn: 274550
2016-07-05 21:58:47 +08:00
declare <16 x float> @llvm.x86.avx512.mask.broadcast.ss.ps.512(<4 x float>, <16 x float>, i16) nounwind readonly
define <16 x float> @test_x86_vbroadcast_ss_ps_512(<4 x float> %a0, <16 x float> %a1, i16 %mask ) {
; CHECK-LABEL: test_x86_vbroadcast_ss_ps_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vbroadcastss %xmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vbroadcastss %xmm0, %zmm1 {%k1}
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vaddps %zmm1, %zmm2, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vbroadcastss %xmm0, %zmm0 {%k1} {z}
[X86][AVX512] Autoupgrade the BROADCAST intrinsics llvm-svn: 274550
2016-07-05 21:58:47 +08:00
; CHECK-NEXT: vaddps %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.broadcast.ss.ps.512(<4 x float> %a0, <16 x float> zeroinitializer, i16 -1)
%res1 = call <16 x float> @llvm.x86.avx512.mask.broadcast.ss.ps.512(<4 x float> %a0, <16 x float> %a1, i16 %mask)
%res2 = call <16 x float> @llvm.x86.avx512.mask.broadcast.ss.ps.512(<4 x float> %a0, <16 x float> zeroinitializer, i16 %mask)
%res3 = fadd <16 x float> %res, %res1
%res4 = fadd <16 x float> %res2, %res3
ret <16 x float> %res4
}
declare <8 x double> @llvm.x86.avx512.mask.broadcast.sd.pd.512(<2 x double>, <8 x double>, i8) nounwind readonly
define <8 x double> @test_x86_vbroadcast_sd_pd_512(<2 x double> %a0, <8 x double> %a1, i8 %mask ) {
; CHECK-LABEL: test_x86_vbroadcast_sd_pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vbroadcastsd %xmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vbroadcastsd %xmm0, %zmm1 {%k1}
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vaddpd %zmm1, %zmm2, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vbroadcastsd %xmm0, %zmm0 {%k1} {z}
[X86][AVX512] Autoupgrade the BROADCAST intrinsics llvm-svn: 274550
2016-07-05 21:58:47 +08:00
; CHECK-NEXT: vaddpd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.broadcast.sd.pd.512(<2 x double> %a0, <8 x double> zeroinitializer, i8 -1)
%res1 = call <8 x double> @llvm.x86.avx512.mask.broadcast.sd.pd.512(<2 x double> %a0, <8 x double> %a1, i8 %mask)
%res2 = call <8 x double> @llvm.x86.avx512.mask.broadcast.sd.pd.512(<2 x double> %a0, <8 x double> zeroinitializer, i8 %mask)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res2, %res3
ret <8 x double> %res4
}
declare <16 x i32> @llvm.x86.avx512.pbroadcastd.512(<4 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_pbroadcastd_512(<4 x i32> %x0, <16 x i32> %x1, i16 %mask) {
; CHECK-LABEL: test_int_x86_avx512_pbroadcastd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpbroadcastd %xmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpbroadcastd %xmm0, %zmm1 {%k1}
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vpaddd %zmm1, %zmm2, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpbroadcastd %xmm0, %zmm0 {%k1} {z}
[X86][AVX512] Autoupgrade the BROADCAST intrinsics llvm-svn: 274550
2016-07-05 21:58:47 +08:00
; CHECK-NEXT: vpaddd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.pbroadcastd.512(<4 x i32> %x0, <16 x i32> %x1, i16 -1)
%res1 = call <16 x i32> @llvm.x86.avx512.pbroadcastd.512(<4 x i32> %x0, <16 x i32> %x1, i16 %mask)
%res2 = call <16 x i32> @llvm.x86.avx512.pbroadcastd.512(<4 x i32> %x0, <16 x i32> zeroinitializer, i16 %mask)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res2, %res3
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.pbroadcastq.512(<2 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_pbroadcastq_512(<2 x i64> %x0, <8 x i64> %x1, i8 %mask) {
; CHECK-LABEL: test_int_x86_avx512_pbroadcastq_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpbroadcastq %xmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpbroadcastq %xmm0, %zmm1 {%k1}
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vpaddq %zmm1, %zmm2, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpbroadcastq %xmm0, %zmm0 {%k1} {z}
[X86][AVX512] Autoupgrade the BROADCAST intrinsics llvm-svn: 274550
2016-07-05 21:58:47 +08:00
; CHECK-NEXT: vpaddq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.pbroadcastq.512(<2 x i64> %x0, <8 x i64> %x1,i8 -1)
%res1 = call <8 x i64> @llvm.x86.avx512.pbroadcastq.512(<2 x i64> %x0, <8 x i64> %x1,i8 %mask)
%res2 = call <8 x i64> @llvm.x86.avx512.pbroadcastq.512(<2 x i64> %x0, <8 x i64> zeroinitializer,i8 %mask)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res2, %res3
ret <8 x i64> %res4
}
[X86][AVX512] Autoupgrade the MOVDDUP/MOVSLDUP/MOVSHDUP intrinsics llvm-svn: 274439
2016-07-02 22:42:35 +08:00
declare <16 x float> @llvm.x86.avx512.mask.movsldup.512(<16 x float>, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_movsldup_512(<16 x float> %x0, <16 x float> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_movsldup_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovsldup {{.*#+}} zmm2 = zmm0[0,0,2,2,4,4,6,6,8,8,10,10,12,12,14,14]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for masked shuffle comments This patch adds support for including the avx512 mask register information in the mask/maskz versions of shuffle instruction comments. This initial version just adds support for MOVDDUP/MOVSHDUP/MOVSLDUP to reduce the mass of test regenerations, other shuffle instructions can be added in due course. Differential Revision: http://reviews.llvm.org/D21953 llvm-svn: 274459
2016-07-03 21:08:29 +08:00
; CHECK-NEXT: vmovsldup {{.*#+}} zmm1 {%k1} = zmm0[0,0,2,2,4,4,6,6,8,8,10,10,12,12,14,14]
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vaddps %zmm2, %zmm1, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vmovsldup {{.*#+}} zmm0 {%k1} {z} = zmm0[0,0,2,2,4,4,6,6,8,8,10,10,12,12,14,14]
[X86][AVX512] Autoupgrade the MOVDDUP/MOVSLDUP/MOVSHDUP intrinsics llvm-svn: 274439
2016-07-02 22:42:35 +08:00
; CHECK-NEXT: vaddps %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.movsldup.512(<16 x float> %x0, <16 x float> %x1, i16 %x2)
%res1 = call <16 x float> @llvm.x86.avx512.mask.movsldup.512(<16 x float> %x0, <16 x float> %x1, i16 -1)
%res2 = call <16 x float> @llvm.x86.avx512.mask.movsldup.512(<16 x float> %x0, <16 x float> zeroinitializer, i16 %x2)
%res3 = fadd <16 x float> %res, %res1
%res4 = fadd <16 x float> %res2, %res3
ret <16 x float> %res4
}
declare <16 x float> @llvm.x86.avx512.mask.movshdup.512(<16 x float>, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_movshdup_512(<16 x float> %x0, <16 x float> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_movshdup_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovshdup {{.*#+}} zmm2 = zmm0[1,1,3,3,5,5,7,7,9,9,11,11,13,13,15,15]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for masked shuffle comments This patch adds support for including the avx512 mask register information in the mask/maskz versions of shuffle instruction comments. This initial version just adds support for MOVDDUP/MOVSHDUP/MOVSLDUP to reduce the mass of test regenerations, other shuffle instructions can be added in due course. Differential Revision: http://reviews.llvm.org/D21953 llvm-svn: 274459
2016-07-03 21:08:29 +08:00
; CHECK-NEXT: vmovshdup {{.*#+}} zmm1 {%k1} = zmm0[1,1,3,3,5,5,7,7,9,9,11,11,13,13,15,15]
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vaddps %zmm2, %zmm1, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vmovshdup {{.*#+}} zmm0 {%k1} {z} = zmm0[1,1,3,3,5,5,7,7,9,9,11,11,13,13,15,15]
[X86][AVX512] Autoupgrade the MOVDDUP/MOVSLDUP/MOVSHDUP intrinsics llvm-svn: 274439
2016-07-02 22:42:35 +08:00
; CHECK-NEXT: vaddps %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.movshdup.512(<16 x float> %x0, <16 x float> %x1, i16 %x2)
%res1 = call <16 x float> @llvm.x86.avx512.mask.movshdup.512(<16 x float> %x0, <16 x float> %x1, i16 -1)
%res2 = call <16 x float> @llvm.x86.avx512.mask.movshdup.512(<16 x float> %x0, <16 x float> zeroinitializer, i16 %x2)
%res3 = fadd <16 x float> %res, %res1
%res4 = fadd <16 x float> %res2, %res3
ret <16 x float> %res4
}
declare <8 x double> @llvm.x86.avx512.mask.movddup.512(<8 x double>, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_movddup_512(<8 x double> %x0, <8 x double> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_movddup_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovddup {{.*#+}} zmm2 = zmm0[0,0,2,2,4,4,6,6]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for masked shuffle comments This patch adds support for including the avx512 mask register information in the mask/maskz versions of shuffle instruction comments. This initial version just adds support for MOVDDUP/MOVSHDUP/MOVSLDUP to reduce the mass of test regenerations, other shuffle instructions can be added in due course. Differential Revision: http://reviews.llvm.org/D21953 llvm-svn: 274459
2016-07-03 21:08:29 +08:00
; CHECK-NEXT: vmovddup {{.*#+}} zmm1 {%k1} = zmm0[0,0,2,2,4,4,6,6]
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vaddpd %zmm2, %zmm1, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vmovddup {{.*#+}} zmm0 {%k1} {z} = zmm0[0,0,2,2,4,4,6,6]
[X86][AVX512] Autoupgrade the MOVDDUP/MOVSLDUP/MOVSHDUP intrinsics llvm-svn: 274439
2016-07-02 22:42:35 +08:00
; CHECK-NEXT: vaddpd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.movddup.512(<8 x double> %x0, <8 x double> %x1, i8 %x2)
%res1 = call <8 x double> @llvm.x86.avx512.mask.movddup.512(<8 x double> %x0, <8 x double> %x1, i8 -1)
%res2 = call <8 x double> @llvm.x86.avx512.mask.movddup.512(<8 x double> %x0, <8 x double> zeroinitializer, i8 %x2)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res2, %res3
ret <8 x double> %res4
}
[X86][AVX512] Autoupgrade the VPERMPD/VPERMQ intrinsics llvm-svn: 274506
2016-07-04 22:19:05 +08:00
declare <8 x double> @llvm.x86.avx512.mask.perm.df.512(<8 x double>, i32, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_perm_df_512(<8 x double> %x0, i32 %x1, <8 x double> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_perm_df_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpermpd {{.*#+}} zmm2 = zmm0[3,0,0,0,7,4,4,4]
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpermpd {{.*#+}} zmm1 {%k1} = zmm0[3,0,0,0,7,4,4,4]
; CHECK-NEXT: vpermpd {{.*#+}} zmm0 {%k1} {z} = zmm0[3,0,0,0,7,4,4,4]
; CHECK-NEXT: vaddpd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vaddpd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.perm.df.512(<8 x double> %x0, i32 3, <8 x double> %x2, i8 %x3)
%res1 = call <8 x double> @llvm.x86.avx512.mask.perm.df.512(<8 x double> %x0, i32 3, <8 x double> zeroinitializer, i8 %x3)
%res2 = call <8 x double> @llvm.x86.avx512.mask.perm.df.512(<8 x double> %x0, i32 3, <8 x double> %x2, i8 -1)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res3, %res2
ret <8 x double> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.perm.di.512(<8 x i64>, i32, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_perm_di_512(<8 x i64> %x0, i32 %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_perm_di_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpermq {{.*#+}} zmm2 = zmm0[3,0,0,0,7,4,4,4]
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpermq {{.*#+}} zmm1 {%k1} = zmm0[3,0,0,0,7,4,4,4]
; CHECK-NEXT: vpermq {{.*#+}} zmm0 {%k1} {z} = zmm0[3,0,0,0,7,4,4,4]
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.perm.di.512(<8 x i64> %x0, i32 3, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.perm.di.512(<8 x i64> %x0, i32 3, <8 x i64> zeroinitializer, i8 %x3)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.perm.di.512(<8 x i64> %x0, i32 3, <8 x i64> %x2, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
[AVX512] Remove masked store intrinsics. Clang now emits generic masked store intrinsics instead. The intrinsics will be autoupgraded to the same generic masked stores. llvm-svn: 271245
2016-05-31 09:50:02 +08:00
define void @test_store1(<16 x float> %data, i8* %ptr, i8* %ptr2, i16 %mask) {
; CHECK-LABEL: test_store1:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovups %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovups %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.storeu.ps.512(i8* %ptr, <16 x float> %data, i16 %mask)
call void @llvm.x86.avx512.mask.storeu.ps.512(i8* %ptr2, <16 x float> %data, i16 -1)
ret void
}
declare void @llvm.x86.avx512.mask.storeu.ps.512(i8*, <16 x float>, i16 )
define void @test_store2(<8 x double> %data, i8* %ptr, i8* %ptr2, i8 %mask) {
; CHECK-LABEL: test_store2:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovupd %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovupd %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.storeu.pd.512(i8* %ptr, <8 x double> %data, i8 %mask)
call void @llvm.x86.avx512.mask.storeu.pd.512(i8* %ptr2, <8 x double> %data, i8 -1)
ret void
}
declare void @llvm.x86.avx512.mask.storeu.pd.512(i8*, <8 x double>, i8)
define void @test_mask_store_aligned_ps(<16 x float> %data, i8* %ptr, i8* %ptr2, i16 %mask) {
; CHECK-LABEL: test_mask_store_aligned_ps:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovaps %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovaps %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.store.ps.512(i8* %ptr, <16 x float> %data, i16 %mask)
call void @llvm.x86.avx512.mask.store.ps.512(i8* %ptr2, <16 x float> %data, i16 -1)
ret void
}
declare void @llvm.x86.avx512.mask.store.ps.512(i8*, <16 x float>, i16 )
define void @test_mask_store_aligned_pd(<8 x double> %data, i8* %ptr, i8* %ptr2, i8 %mask) {
; CHECK-LABEL: test_mask_store_aligned_pd:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovapd %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovapd %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.store.pd.512(i8* %ptr, <8 x double> %data, i8 %mask)
call void @llvm.x86.avx512.mask.store.pd.512(i8* %ptr2, <8 x double> %data, i8 -1)
ret void
}
declare void @llvm.x86.avx512.mask.store.pd.512(i8*, <8 x double>, i8)
define void@test_int_x86_avx512_mask_storeu_q_512(i8* %ptr1, i8* %ptr2, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_storeu_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovdqu64 %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovdqu64 %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.storeu.q.512(i8* %ptr1, <8 x i64> %x1, i8 %x2)
call void @llvm.x86.avx512.mask.storeu.q.512(i8* %ptr2, <8 x i64> %x1, i8 -1)
ret void
}
declare void @llvm.x86.avx512.mask.storeu.q.512(i8*, <8 x i64>, i8)
define void@test_int_x86_avx512_mask_storeu_d_512(i8* %ptr1, i8* %ptr2, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_storeu_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovdqu32 %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovdqu32 %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.storeu.d.512(i8* %ptr1, <16 x i32> %x1, i16 %x2)
call void @llvm.x86.avx512.mask.storeu.d.512(i8* %ptr2, <16 x i32> %x1, i16 -1)
ret void
}
declare void @llvm.x86.avx512.mask.storeu.d.512(i8*, <16 x i32>, i16)
define void@test_int_x86_avx512_mask_store_q_512(i8* %ptr1, i8* %ptr2, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_store_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovdqa64 %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovdqa64 %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.store.q.512(i8* %ptr1, <8 x i64> %x1, i8 %x2)
call void @llvm.x86.avx512.mask.store.q.512(i8* %ptr2, <8 x i64> %x1, i8 -1)
ret void
}
declare void @llvm.x86.avx512.mask.store.q.512(i8*, <8 x i64>, i8)
define void@test_int_x86_avx512_mask_store_d_512(i8* %ptr1, i8* %ptr2, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_store_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovdqa32 %zmm0, (%rdi) {%k1}
; CHECK-NEXT: vmovdqa32 %zmm0, (%rsi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.mask.store.d.512(i8* %ptr1, <16 x i32> %x1, i16 %x2)
call void @llvm.x86.avx512.mask.store.d.512(i8* %ptr2, <16 x i32> %x1, i16 -1)
ret void
}
declare void @llvm.x86.avx512.mask.store.d.512(i8*, <16 x i32>, i16)
[AVX512] Remove masked load intrinsics. Clang now emits generic masked load intrinsics instead. The intrinsics will be autoupgraded to the same generic masked loads. llvm-svn: 271478
2016-06-02 12:19:36 +08:00
define <16 x float> @test_mask_load_aligned_ps(<16 x float> %data, i8* %ptr, i16 %mask) {
; CHECK-LABEL: test_mask_load_aligned_ps:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovaps (%rdi), %zmm0
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vmovaps (%rdi), %zmm0 {%k1}
; CHECK-NEXT: vmovaps (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vaddps %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.load.ps.512(i8* %ptr, <16 x float> zeroinitializer, i16 -1)
%res1 = call <16 x float> @llvm.x86.avx512.mask.load.ps.512(i8* %ptr, <16 x float> %res, i16 %mask)
%res2 = call <16 x float> @llvm.x86.avx512.mask.load.ps.512(i8* %ptr, <16 x float> zeroinitializer, i16 %mask)
%res4 = fadd <16 x float> %res2, %res1
ret <16 x float> %res4
}
declare <16 x float> @llvm.x86.avx512.mask.load.ps.512(i8*, <16 x float>, i16)
define <16 x float> @test_mask_load_unaligned_ps(<16 x float> %data, i8* %ptr, i16 %mask) {
; CHECK-LABEL: test_mask_load_unaligned_ps:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovups (%rdi), %zmm0
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vmovups (%rdi), %zmm0 {%k1}
; CHECK-NEXT: vmovups (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vaddps %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.loadu.ps.512(i8* %ptr, <16 x float> zeroinitializer, i16 -1)
%res1 = call <16 x float> @llvm.x86.avx512.mask.loadu.ps.512(i8* %ptr, <16 x float> %res, i16 %mask)
%res2 = call <16 x float> @llvm.x86.avx512.mask.loadu.ps.512(i8* %ptr, <16 x float> zeroinitializer, i16 %mask)
%res4 = fadd <16 x float> %res2, %res1
ret <16 x float> %res4
}
declare <16 x float> @llvm.x86.avx512.mask.loadu.ps.512(i8*, <16 x float>, i16)
define <8 x double> @test_mask_load_aligned_pd(<8 x double> %data, i8* %ptr, i8 %mask) {
; CHECK-LABEL: test_mask_load_aligned_pd:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovapd (%rdi), %zmm0
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vmovapd (%rdi), %zmm0 {%k1}
; CHECK-NEXT: vmovapd (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vaddpd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.load.pd.512(i8* %ptr, <8 x double> zeroinitializer, i8 -1)
%res1 = call <8 x double> @llvm.x86.avx512.mask.load.pd.512(i8* %ptr, <8 x double> %res, i8 %mask)
%res2 = call <8 x double> @llvm.x86.avx512.mask.load.pd.512(i8* %ptr, <8 x double> zeroinitializer, i8 %mask)
%res4 = fadd <8 x double> %res2, %res1
ret <8 x double> %res4
}
declare <8 x double> @llvm.x86.avx512.mask.load.pd.512(i8*, <8 x double>, i8)
define <8 x double> @test_mask_load_unaligned_pd(<8 x double> %data, i8* %ptr, i8 %mask) {
; CHECK-LABEL: test_mask_load_unaligned_pd:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovupd (%rdi), %zmm0
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vmovupd (%rdi), %zmm0 {%k1}
; CHECK-NEXT: vmovupd (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vaddpd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.loadu.pd.512(i8* %ptr, <8 x double> zeroinitializer, i8 -1)
%res1 = call <8 x double> @llvm.x86.avx512.mask.loadu.pd.512(i8* %ptr, <8 x double> %res, i8 %mask)
%res2 = call <8 x double> @llvm.x86.avx512.mask.loadu.pd.512(i8* %ptr, <8 x double> zeroinitializer, i8 %mask)
%res4 = fadd <8 x double> %res2, %res1
ret <8 x double> %res4
}
declare <8 x double> @llvm.x86.avx512.mask.loadu.pd.512(i8*, <8 x double>, i8)
declare <16 x i32> @llvm.x86.avx512.mask.loadu.d.512(i8*, <16 x i32>, i16)
define <16 x i32> @test_mask_load_unaligned_d(i8* %ptr, i8* %ptr2, <16 x i32> %data, i16 %mask) {
; CHECK-LABEL: test_mask_load_unaligned_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovdqu32 (%rdi), %zmm0
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovdqu32 (%rsi), %zmm0 {%k1}
; CHECK-NEXT: vmovdqu32 (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.loadu.d.512(i8* %ptr, <16 x i32> zeroinitializer, i16 -1)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.loadu.d.512(i8* %ptr2, <16 x i32> %res, i16 %mask)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.loadu.d.512(i8* %ptr, <16 x i32> zeroinitializer, i16 %mask)
%res4 = add <16 x i32> %res2, %res1
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.loadu.q.512(i8*, <8 x i64>, i8)
define <8 x i64> @test_mask_load_unaligned_q(i8* %ptr, i8* %ptr2, <8 x i64> %data, i8 %mask) {
; CHECK-LABEL: test_mask_load_unaligned_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovdqu64 (%rdi), %zmm0
; CHECK-NEXT: kmovw %edx, %k1
; CHECK-NEXT: vmovdqu64 (%rsi), %zmm0 {%k1}
; CHECK-NEXT: vmovdqu64 (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.loadu.q.512(i8* %ptr, <8 x i64> zeroinitializer, i8 -1)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.loadu.q.512(i8* %ptr2, <8 x i64> %res, i8 %mask)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.loadu.q.512(i8* %ptr, <8 x i64> zeroinitializer, i8 %mask)
%res4 = add <8 x i64> %res2, %res1
ret <8 x i64> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.load.d.512(i8*, <16 x i32>, i16)
define <16 x i32> @test_mask_load_aligned_d(<16 x i32> %data, i8* %ptr, i16 %mask) {
; CHECK-LABEL: test_mask_load_aligned_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovdqa32 (%rdi), %zmm0
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vmovdqa32 (%rdi), %zmm0 {%k1}
; CHECK-NEXT: vmovdqa32 (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.load.d.512(i8* %ptr, <16 x i32> zeroinitializer, i16 -1)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.load.d.512(i8* %ptr, <16 x i32> %res, i16 %mask)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.load.d.512(i8* %ptr, <16 x i32> zeroinitializer, i16 %mask)
%res4 = add <16 x i32> %res2, %res1
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.load.q.512(i8*, <8 x i64>, i8)
define <8 x i64> @test_mask_load_aligned_q(<8 x i64> %data, i8* %ptr, i8 %mask) {
; CHECK-LABEL: test_mask_load_aligned_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovdqa64 (%rdi), %zmm0
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vmovdqa64 (%rdi), %zmm0 {%k1}
; CHECK-NEXT: vmovdqa64 (%rdi), %zmm1 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.load.q.512(i8* %ptr, <8 x i64> zeroinitializer, i8 -1)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.load.q.512(i8* %ptr, <8 x i64> %res, i8 %mask)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.load.q.512(i8* %ptr, <8 x i64> zeroinitializer, i8 %mask)
%res4 = add <8 x i64> %res2, %res1
ret <8 x i64> %res4
}
[X86][AVX512] Autoupgrade the VPERMILPD/VPERMILPS intrinsics llvm-svn: 274498
2016-07-04 20:40:54 +08:00
declare <8 x double> @llvm.x86.avx512.mask.vpermil.pd.512(<8 x double>, i32, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_vpermil_pd_512(<8 x double> %x0, <8 x double> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_vpermil_pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpermilpd {{.*#+}} zmm2 = zmm0[0,1,3,2,5,4,6,6]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpermilpd {{.*#+}} zmm1 {%k1} = zmm0[0,1,3,2,5,4,6,6]
; CHECK-NEXT: vpermilpd {{.*#+}} zmm0 {%k1} {z} = zmm0[0,1,3,2,5,4,6,6]
; CHECK-NEXT: vaddpd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vaddpd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.vpermil.pd.512(<8 x double> %x0, i32 22, <8 x double> %x2, i8 %x3)
%res1 = call <8 x double> @llvm.x86.avx512.mask.vpermil.pd.512(<8 x double> %x0, i32 22, <8 x double> zeroinitializer, i8 %x3)
%res2 = call <8 x double> @llvm.x86.avx512.mask.vpermil.pd.512(<8 x double> %x0, i32 22, <8 x double> %x2, i8 -1)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res3, %res2
ret <8 x double> %res4
}
declare <16 x float> @llvm.x86.avx512.mask.vpermil.ps.512(<16 x float>, i32, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_vpermil_ps_512(<16 x float> %x0, <16 x float> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_vpermil_ps_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpermilps {{.*#+}} zmm2 = zmm0[2,1,1,0,6,5,5,4,10,9,9,8,14,13,13,12]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpermilps {{.*#+}} zmm1 {%k1} = zmm0[2,1,1,0,6,5,5,4,10,9,9,8,14,13,13,12]
; CHECK-NEXT: vpermilps {{.*#+}} zmm0 {%k1} {z} = zmm0[2,1,1,0,6,5,5,4,10,9,9,8,14,13,13,12]
; CHECK-NEXT: vaddps %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vaddps %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.vpermil.ps.512(<16 x float> %x0, i32 22, <16 x float> %x2, i16 %x3)
%res1 = call <16 x float> @llvm.x86.avx512.mask.vpermil.ps.512(<16 x float> %x0, i32 22, <16 x float> zeroinitializer, i16 %x3)
%res2 = call <16 x float> @llvm.x86.avx512.mask.vpermil.ps.512(<16 x float> %x0, i32 22, <16 x float> %x2, i16 -1)
%res3 = fadd <16 x float> %res, %res1
%res4 = fadd <16 x float> %res3, %res2
ret <16 x float> %res4
}
[AVX512] Remove maksed pshufd, pshuflw, and phufhw intrinsics and autoupgrade them to selects and shufflevector. llvm-svn: 272527
2016-06-13 10:36:48 +08:00
declare <16 x i32> @llvm.x86.avx512.mask.pshuf.d.512(<16 x i32>, i32, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pshuf_d_512(<16 x i32> %x0, i32 %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pshuf_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpshufd {{.*#+}} zmm2 = zmm0[3,0,0,0,7,4,4,4,11,8,8,8,15,12,12,12]
; CHECK-NEXT: kmovw %esi, %k1
[X86][AVX512] Add support for VPALIGNR/PSHUFD/PSHUFHW/PSHUFLW masked shuffle comments llvm-svn: 274466
2016-07-03 23:00:51 +08:00
; CHECK-NEXT: vpshufd {{.*#+}} zmm1 {%k1} = zmm0[3,0,0,0,7,4,4,4,11,8,8,8,15,12,12,12]
; CHECK-NEXT: vpshufd {{.*#+}} zmm0 {%k1} {z} = zmm0[3,0,0,0,7,4,4,4,11,8,8,8,15,12,12,12]
[AVX512] Remove maksed pshufd, pshuflw, and phufhw intrinsics and autoupgrade them to selects and shufflevector. llvm-svn: 272527
2016-06-13 10:36:48 +08:00
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pshuf.d.512(<16 x i32> %x0, i32 3, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pshuf.d.512(<16 x i32> %x0, i32 3, <16 x i32> zeroinitializer, i16 %x3)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.pshuf.d.512(<16 x i32> %x0, i32 3, <16 x i32> %x2, i16 -1)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
define i16 @test_pcmpeq_d(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_pcmpeq_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpeqd %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ax<def> %ax<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i16 @llvm.x86.avx512.mask.pcmpeq.d.512(<16 x i32> %a, <16 x i32> %b, i16 -1)
ret i16 %res
}
define i16 @test_mask_pcmpeq_d(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: test_mask_pcmpeq_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpeqd %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ax<def> %ax<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i16 @llvm.x86.avx512.mask.pcmpeq.d.512(<16 x i32> %a, <16 x i32> %b, i16 %mask)
ret i16 %res
}
declare i16 @llvm.x86.avx512.mask.pcmpeq.d.512(<16 x i32>, <16 x i32>, i16)
define i8 @test_pcmpeq_q(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_pcmpeq_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpeqq %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %al<def> %al<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i8 @llvm.x86.avx512.mask.pcmpeq.q.512(<8 x i64> %a, <8 x i64> %b, i8 -1)
ret i8 %res
}
define i8 @test_mask_pcmpeq_q(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
; CHECK-LABEL: test_mask_pcmpeq_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpeqq %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %al<def> %al<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i8 @llvm.x86.avx512.mask.pcmpeq.q.512(<8 x i64> %a, <8 x i64> %b, i8 %mask)
ret i8 %res
}
declare i8 @llvm.x86.avx512.mask.pcmpeq.q.512(<8 x i64>, <8 x i64>, i8)
define i16 @test_pcmpgt_d(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_pcmpgt_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpgtd %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ax<def> %ax<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i16 @llvm.x86.avx512.mask.pcmpgt.d.512(<16 x i32> %a, <16 x i32> %b, i16 -1)
ret i16 %res
}
define i16 @test_mask_pcmpgt_d(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: test_mask_pcmpgt_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpgtd %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ax<def> %ax<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i16 @llvm.x86.avx512.mask.pcmpgt.d.512(<16 x i32> %a, <16 x i32> %b, i16 %mask)
ret i16 %res
}
declare i16 @llvm.x86.avx512.mask.pcmpgt.d.512(<16 x i32>, <16 x i32>, i16)
define i8 @test_pcmpgt_q(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_pcmpgt_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpgtq %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %al<def> %al<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i8 @llvm.x86.avx512.mask.pcmpgt.q.512(<8 x i64> %a, <8 x i64> %b, i8 -1)
ret i8 %res
}
define i8 @test_mask_pcmpgt_q(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
; CHECK-LABEL: test_mask_pcmpgt_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpgtq %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: kmovw %k0, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %al<def> %al<kill> %eax<kill>
[AVX512] Remove the masked vpcmpeq/vcmpgt intrinsics and autoupgrade them to native icmps. llvm-svn: 273240
2016-06-21 11:53:24 +08:00
; CHECK-NEXT: retq
%res = call i8 @llvm.x86.avx512.mask.pcmpgt.q.512(<8 x i64> %a, <8 x i64> %b, i8 %mask)
ret i8 %res
}
declare i8 @llvm.x86.avx512.mask.pcmpgt.q.512(<8 x i64>, <8 x i64>, i8)
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
declare <8 x double> @llvm.x86.avx512.mask.unpckh.pd.512(<8 x double>, <8 x double>, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_unpckh_pd_512(<8 x double> %x0, <8 x double> %x1, <8 x double> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_unpckh_pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vunpckhpd {{.*#+}} zmm3 = zmm0[1],zmm1[1],zmm0[3],zmm1[3],zmm0[5],zmm1[5],zmm0[7],zmm1[7]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for UNPCK masked shuffle comments llvm-svn: 274464
2016-07-03 22:26:21 +08:00
; CHECK-NEXT: vunpckhpd {{.*#+}} zmm2 {%k1} = zmm0[1],zmm1[1],zmm0[3],zmm1[3],zmm0[5],zmm1[5],zmm0[7],zmm1[7]
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: vaddpd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.unpckh.pd.512(<8 x double> %x0, <8 x double> %x1, <8 x double> %x2, i8 %x3)
%res1 = call <8 x double> @llvm.x86.avx512.mask.unpckh.pd.512(<8 x double> %x0, <8 x double> %x1, <8 x double> %x2, i8 -1)
%res2 = fadd <8 x double> %res, %res1
ret <8 x double> %res2
}
declare <16 x float> @llvm.x86.avx512.mask.unpckh.ps.512(<16 x float>, <16 x float>, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_unpckh_ps_512(<16 x float> %x0, <16 x float> %x1, <16 x float> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_unpckh_ps_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vunpckhps {{.*#+}} zmm3 = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for UNPCK masked shuffle comments llvm-svn: 274464
2016-07-03 22:26:21 +08:00
; CHECK-NEXT: vunpckhps {{.*#+}} zmm2 {%k1} = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: vaddps %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.unpckh.ps.512(<16 x float> %x0, <16 x float> %x1, <16 x float> %x2, i16 %x3)
%res1 = call <16 x float> @llvm.x86.avx512.mask.unpckh.ps.512(<16 x float> %x0, <16 x float> %x1, <16 x float> %x2, i16 -1)
%res2 = fadd <16 x float> %res, %res1
ret <16 x float> %res2
}
declare <8 x double> @llvm.x86.avx512.mask.unpckl.pd.512(<8 x double>, <8 x double>, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_unpckl_pd_512(<8 x double> %x0, <8 x double> %x1, <8 x double> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_unpckl_pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vunpcklpd {{.*#+}} zmm3 = zmm0[0],zmm1[0],zmm0[2],zmm1[2],zmm0[4],zmm1[4],zmm0[6],zmm1[6]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for UNPCK masked shuffle comments llvm-svn: 274464
2016-07-03 22:26:21 +08:00
; CHECK-NEXT: vunpcklpd {{.*#+}} zmm2 {%k1} = zmm0[0],zmm1[0],zmm0[2],zmm1[2],zmm0[4],zmm1[4],zmm0[6],zmm1[6]
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: vaddpd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.unpckl.pd.512(<8 x double> %x0, <8 x double> %x1, <8 x double> %x2, i8 %x3)
%res1 = call <8 x double> @llvm.x86.avx512.mask.unpckl.pd.512(<8 x double> %x0, <8 x double> %x1, <8 x double> %x2, i8 -1)
%res2 = fadd <8 x double> %res, %res1
ret <8 x double> %res2
}
declare <16 x float> @llvm.x86.avx512.mask.unpckl.ps.512(<16 x float>, <16 x float>, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_unpckl_ps_512(<16 x float> %x0, <16 x float> %x1, <16 x float> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_unpckl_ps_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vunpcklps {{.*#+}} zmm3 = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for UNPCK masked shuffle comments llvm-svn: 274464
2016-07-03 22:26:21 +08:00
; CHECK-NEXT: vunpcklps {{.*#+}} zmm2 {%k1} = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: vaddps %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.unpckl.ps.512(<16 x float> %x0, <16 x float> %x1, <16 x float> %x2, i16 %x3)
%res1 = call <16 x float> @llvm.x86.avx512.mask.unpckl.ps.512(<16 x float> %x0, <16 x float> %x1, <16 x float> %x2, i16 -1)
%res2 = fadd <16 x float> %res, %res1
ret <16 x float> %res2
}
declare <8 x i64> @llvm.x86.avx512.mask.punpcklqd.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_punpcklqd_q_512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_punpcklqd_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpunpcklqdq {{.*#+}} zmm3 = zmm0[0],zmm1[0],zmm0[2],zmm1[2],zmm0[4],zmm1[4],zmm0[6],zmm1[6]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpunpcklqdq {{.*#+}} zmm2 = zmm0[0],zmm1[0],zmm0[2],zmm1[2],zmm0[4],zmm1[4],zmm0[6],zmm1[6]
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm2
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: vpunpcklqdq {{.*#+}} zmm0 = zmm0[0],zmm1[0],zmm0[2],zmm1[2],zmm0[4],zmm1[4],zmm0[6],zmm1[6]
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.punpcklqd.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.punpcklqd.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 -1)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.punpcklqd.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> zeroinitializer,i8 %x3)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res2, %res3
ret <8 x i64> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.punpckhqd.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_punpckhqd_q_512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_punpckhqd_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpunpckhqdq {{.*#+}} zmm3 = zmm0[1],zmm1[1],zmm0[3],zmm1[3],zmm0[5],zmm1[5],zmm0[7],zmm1[7]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpunpckhqdq {{.*#+}} zmm2 = zmm0[1],zmm1[1],zmm0[3],zmm1[3],zmm0[5],zmm1[5],zmm0[7],zmm1[7]
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.punpckhqd.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.punpckhqd.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 -1)
%res2 = add <8 x i64> %res, %res1
ret <8 x i64> %res2
}
declare <16 x i32> @llvm.x86.avx512.mask.punpckhd.q.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_punpckhd_q_512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_punpckhd_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpunpckhdq {{.*#+}} zmm3 = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for UNPCK masked shuffle comments llvm-svn: 274464
2016-07-03 22:26:21 +08:00
; CHECK-NEXT: vpunpckhdq {{.*#+}} zmm2 {%k1} = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.punpckhd.q.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.punpckhd.q.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
declare <16 x i32> @llvm.x86.avx512.mask.punpckld.q.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_punpckld_q_512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_punpckld_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpunpckldq {{.*#+}} zmm3 = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
; CHECK-NEXT: kmovw %edi, %k1
[X86][AVX512] Add support for UNPCK masked shuffle comments llvm-svn: 274464
2016-07-03 22:26:21 +08:00
; CHECK-NEXT: vpunpckldq {{.*#+}} zmm2 {%k1} = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
[AVX512] Remove masked unpack intrinsics and autoupgrade to vectorshuffle and selects. llvm-svn: 273543
2016-06-23 15:37:33 +08:00
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.punpckld.q.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.punpckld.q.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
[AVX512] Remove and autoupgrade a duplicate set of 512-bit masked shift intrinsics. I'm not sure if clang ever used these builtin names or not. llvm-svn: 274827
2016-07-08 14:14:47 +08:00
define <16 x i32> @test_x86_avx512_pslli_d(<16 x i32> %a0) {
; CHECK-LABEL: test_x86_avx512_pslli_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpslld $7, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pslli.d(<16 x i32> %a0, i32 7, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_pslli_d(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_pslli_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpslld $7, %zmm0, %zmm1 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
[AVX512] Remove and autoupgrade a duplicate set of 512-bit masked shift intrinsics. I'm not sure if clang ever used these builtin names or not. llvm-svn: 274827
2016-07-08 14:14:47 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pslli.d(<16 x i32> %a0, i32 7, <16 x i32> %a1, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_pslli_d(<16 x i32> %a0, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_pslli_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpslld $7, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pslli.d(<16 x i32> %a0, i32 7, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.pslli.d(<16 x i32>, i32, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_pslli_q(<8 x i64> %a0) {
; CHECK-LABEL: test_x86_avx512_pslli_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsllq $7, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pslli.q(<8 x i64> %a0, i32 7, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_pslli_q(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_pslli_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllq $7, %zmm0, %zmm1 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
[AVX512] Remove and autoupgrade a duplicate set of 512-bit masked shift intrinsics. I'm not sure if clang ever used these builtin names or not. llvm-svn: 274827
2016-07-08 14:14:47 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pslli.q(<8 x i64> %a0, i32 7, <8 x i64> %a1, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_pslli_q(<8 x i64> %a0, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_pslli_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllq $7, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pslli.q(<8 x i64> %a0, i32 7, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.pslli.q(<8 x i64>, i32, <8 x i64>, i8) nounwind readnone
define <16 x i32> @test_x86_avx512_psrli_d(<16 x i32> %a0) {
; CHECK-LABEL: test_x86_avx512_psrli_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrld $7, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrli.d(<16 x i32> %a0, i32 7, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psrli_d(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrli_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrld $7, %zmm0, %zmm1 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
[AVX512] Remove and autoupgrade a duplicate set of 512-bit masked shift intrinsics. I'm not sure if clang ever used these builtin names or not. llvm-svn: 274827
2016-07-08 14:14:47 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrli.d(<16 x i32> %a0, i32 7, <16 x i32> %a1, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psrli_d(<16 x i32> %a0, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrli_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrld $7, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrli.d(<16 x i32> %a0, i32 7, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psrli.d(<16 x i32>, i32, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psrli_q(<8 x i64> %a0) {
; CHECK-LABEL: test_x86_avx512_psrli_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrlq $7, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrli.q(<8 x i64> %a0, i32 7, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psrli_q(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrli_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlq $7, %zmm0, %zmm1 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
[AVX512] Remove and autoupgrade a duplicate set of 512-bit masked shift intrinsics. I'm not sure if clang ever used these builtin names or not. llvm-svn: 274827
2016-07-08 14:14:47 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrli.q(<8 x i64> %a0, i32 7, <8 x i64> %a1, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psrli_q(<8 x i64> %a0, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrli_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlq $7, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrli.q(<8 x i64> %a0, i32 7, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psrli.q(<8 x i64>, i32, <8 x i64>, i8) nounwind readnone
define <16 x i32> @test_x86_avx512_psrai_d(<16 x i32> %a0) {
; CHECK-LABEL: test_x86_avx512_psrai_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrad $7, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrai.d(<16 x i32> %a0, i32 7, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psrai_d(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrai_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrad $7, %zmm0, %zmm1 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
[AVX512] Remove and autoupgrade a duplicate set of 512-bit masked shift intrinsics. I'm not sure if clang ever used these builtin names or not. llvm-svn: 274827
2016-07-08 14:14:47 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrai.d(<16 x i32> %a0, i32 7, <16 x i32> %a1, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psrai_d(<16 x i32> %a0, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrai_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrad $7, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrai.d(<16 x i32> %a0, i32 7, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psrai.d(<16 x i32>, i32, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psrai_q(<8 x i64> %a0) {
; CHECK-LABEL: test_x86_avx512_psrai_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsraq $7, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrai.q(<8 x i64> %a0, i32 7, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psrai_q(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrai_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsraq $7, %zmm0, %zmm1 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
[AVX512] Remove and autoupgrade a duplicate set of 512-bit masked shift intrinsics. I'm not sure if clang ever used these builtin names or not. llvm-svn: 274827
2016-07-08 14:14:47 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrai.q(<8 x i64> %a0, i32 7, <8 x i64> %a1, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psrai_q(<8 x i64> %a0, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrai_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsraq $7, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrai.q(<8 x i64> %a0, i32 7, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psrai.q(<8 x i64>, i32, <8 x i64>, i8) nounwind readnone
[X86] Remove and autoupgrade 512-bit non-temporal store intrinsics. llvm-svn: 274966
2016-07-09 12:38:27 +08:00
declare void @llvm.x86.avx512.storent.q.512(i8*, <8 x i64>)
define void@test_storent_q_512(<8 x i64> %data, i8* %ptr) {
; CHECK-LABEL: test_storent_q_512:
; CHECK: ## BB#0:
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovntps %zmm0, (%rdi)
[X86] Remove and autoupgrade 512-bit non-temporal store intrinsics. llvm-svn: 274966
2016-07-09 12:38:27 +08:00
; CHECK-NEXT: retq
call void @llvm.x86.avx512.storent.q.512(i8* %ptr, <8 x i64> %data)
ret void
}
declare void @llvm.x86.avx512.storent.pd.512(i8*, <8 x double>)
define void @test_storent_pd_512(<8 x double> %data, i8* %ptr) {
; CHECK-LABEL: test_storent_pd_512:
; CHECK: ## BB#0:
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovntps %zmm0, (%rdi)
[X86] Remove and autoupgrade 512-bit non-temporal store intrinsics. llvm-svn: 274966
2016-07-09 12:38:27 +08:00
; CHECK-NEXT: retq
call void @llvm.x86.avx512.storent.pd.512(i8* %ptr, <8 x double> %data)
ret void
}
declare void @llvm.x86.avx512.storent.ps.512(i8*, <16 x float>)
define void @test_storent_ps_512(<16 x float> %data, i8* %ptr) {
; CHECK-LABEL: test_storent_ps_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovntps %zmm0, (%rdi)
; CHECK-NEXT: retq
call void @llvm.x86.avx512.storent.ps.512(i8* %ptr, <16 x float> %data)
ret void
}
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
define <16 x i32> @test_xor_epi32(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_xor_epi32:
; CHECK: ## BB#0:
[AVX512] Don't use 32-bit elements version of AND/OR/XOR/ANDN during isel unless we're matching a masked op or broadcast Selecting 32-bit element logical ops without a select or broadcast requires matching a bitconvert on the inputs to the and. But that's a weird thing to rely on. It's entirely possible that one of the inputs doesn't have a bitcast and one does. Since there's no functional difference, just remove the extra patterns and save some isel table size. Differential Revision: https://reviews.llvm.org/D36854 llvm-svn: 312138
2017-08-31 00:38:33 +08:00
; CHECK-NEXT: vpxorq %zmm1, %zmm0, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pxor.d.512(<16 x i32> %a,<16 x i32> %b, <16 x i32>zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_xor_epi32(<16 x i32> %a,<16 x i32> %b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_xor_epi32:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpxord %zmm1, %zmm0, %zmm2 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pxor.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.pxor.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32> @test_or_epi32(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_or_epi32:
; CHECK: ## BB#0:
[AVX512] Don't use 32-bit elements version of AND/OR/XOR/ANDN during isel unless we're matching a masked op or broadcast Selecting 32-bit element logical ops without a select or broadcast requires matching a bitconvert on the inputs to the and. But that's a weird thing to rely on. It's entirely possible that one of the inputs doesn't have a bitcast and one does. Since there's no functional difference, just remove the extra patterns and save some isel table size. Differential Revision: https://reviews.llvm.org/D36854 llvm-svn: 312138
2017-08-31 00:38:33 +08:00
; CHECK-NEXT: vporq %zmm1, %zmm0, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.por.d.512(<16 x i32> %a,<16 x i32> %b, <16 x i32>zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_or_epi32(<16 x i32> %a,<16 x i32> %b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_or_epi32:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpord %zmm1, %zmm0, %zmm2 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.por.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.por.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32> @test_and_epi32(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_and_epi32:
; CHECK: ## BB#0:
[AVX512] Don't use 32-bit elements version of AND/OR/XOR/ANDN during isel unless we're matching a masked op or broadcast Selecting 32-bit element logical ops without a select or broadcast requires matching a bitconvert on the inputs to the and. But that's a weird thing to rely on. It's entirely possible that one of the inputs doesn't have a bitcast and one does. Since there's no functional difference, just remove the extra patterns and save some isel table size. Differential Revision: https://reviews.llvm.org/D36854 llvm-svn: 312138
2017-08-31 00:38:33 +08:00
; CHECK-NEXT: vpandq %zmm1, %zmm0, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pand.d.512(<16 x i32> %a,<16 x i32> %b, <16 x i32>zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_and_epi32(<16 x i32> %a,<16 x i32> %b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_and_epi32:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpandd %zmm1, %zmm0, %zmm2 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pand.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.pand.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <8 x i64> @test_xor_epi64(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_xor_epi64:
; CHECK: ## BB#0:
[AVX-512] Don't let ExeDependencyFix pass convert VPANDD/Q to VPANDPS/PD unless DQI instructions are supported. Same for ANDN, OR, and XOR. Thanks to Igor Breger for pointing out my mistake. llvm-svn: 277292
2016-08-01 01:15:07 +08:00
; CHECK-NEXT: vpxorq %zmm1, %zmm0, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pxor.q.512(<8 x i64> %a,<8 x i64> %b, <8 x i64>zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_xor_epi64(<8 x i64> %a,<8 x i64> %b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_xor_epi64:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpxorq %zmm1, %zmm0, %zmm2 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pxor.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.pxor.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64> @test_or_epi64(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_or_epi64:
; CHECK: ## BB#0:
[AVX-512] Don't let ExeDependencyFix pass convert VPANDD/Q to VPANDPS/PD unless DQI instructions are supported. Same for ANDN, OR, and XOR. Thanks to Igor Breger for pointing out my mistake. llvm-svn: 277292
2016-08-01 01:15:07 +08:00
; CHECK-NEXT: vporq %zmm1, %zmm0, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.por.q.512(<8 x i64> %a,<8 x i64> %b, <8 x i64>zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_or_epi64(<8 x i64> %a,<8 x i64> %b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_or_epi64:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vporq %zmm1, %zmm0, %zmm2 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.por.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.por.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64> @test_and_epi64(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_and_epi64:
; CHECK: ## BB#0:
[AVX-512] Don't let ExeDependencyFix pass convert VPANDD/Q to VPANDPS/PD unless DQI instructions are supported. Same for ANDN, OR, and XOR. Thanks to Igor Breger for pointing out my mistake. llvm-svn: 277292
2016-08-01 01:15:07 +08:00
; CHECK-NEXT: vpandq %zmm1, %zmm0, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pand.q.512(<8 x i64> %a,<8 x i64> %b, <8 x i64>zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_and_epi64(<8 x i64> %a,<8 x i64> %b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_and_epi64:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpandq %zmm1, %zmm0, %zmm2 {%k1}
[AVX512] Add initial support for the Execution Domain fixing pass to change some EVEX instructions. llvm-svn: 276393
2016-07-22 13:00:52 +08:00
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
[AVX512] Remove masked logic op intrinsics and autoupgrade them to native IR. llvm-svn: 275155
2016-07-12 13:27:53 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pand.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.pand.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
[AVX-512] Remove masked integer add/sub/mull intrinsics and upgrade to native IR. llvm-svn: 280611
2016-09-04 10:09:53 +08:00
define <16 x i32> @test_mask_add_epi32_rr(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_mask_add_epi32_rr:
; CHECK: ## BB#0:
; CHECK-NEXT: vpaddd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rrk(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_add_epi32_rrk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpaddd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rrkz(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: test_mask_add_epi32_rrkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpaddd %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rm(<16 x i32> %a, <16 x i32>* %ptr_b) {
; CHECK-LABEL: test_mask_add_epi32_rm:
; CHECK: ## BB#0:
; CHECK-NEXT: vpaddd (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rmk(<16 x i32> %a, <16 x i32>* %ptr_b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_add_epi32_rmk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddd (%rdi), %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rmkz(<16 x i32> %a, <16 x i32>* %ptr_b, i16 %mask) {
; CHECK-LABEL: test_mask_add_epi32_rmkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddd (%rdi), %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rmb(<16 x i32> %a, i32* %ptr_b) {
; CHECK-LABEL: test_mask_add_epi32_rmb:
; CHECK: ## BB#0:
; CHECK-NEXT: vpaddd (%rdi){1to16}, %zmm0, %zmm0
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rmbk(<16 x i32> %a, i32* %ptr_b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_add_epi32_rmbk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddd (%rdi){1to16}, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_add_epi32_rmbkz(<16 x i32> %a, i32* %ptr_b, i16 %mask) {
; CHECK-LABEL: test_mask_add_epi32_rmbkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddd (%rdi){1to16}, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.padd.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32> @test_mask_sub_epi32_rr(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_mask_sub_epi32_rr:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsubd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rrk(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_sub_epi32_rrk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsubd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rrkz(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: test_mask_sub_epi32_rrkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsubd %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rm(<16 x i32> %a, <16 x i32>* %ptr_b) {
; CHECK-LABEL: test_mask_sub_epi32_rm:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsubd (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rmk(<16 x i32> %a, <16 x i32>* %ptr_b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_sub_epi32_rmk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubd (%rdi), %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rmkz(<16 x i32> %a, <16 x i32>* %ptr_b, i16 %mask) {
; CHECK-LABEL: test_mask_sub_epi32_rmkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubd (%rdi), %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rmb(<16 x i32> %a, i32* %ptr_b) {
; CHECK-LABEL: test_mask_sub_epi32_rmb:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsubd (%rdi){1to16}, %zmm0, %zmm0
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rmbk(<16 x i32> %a, i32* %ptr_b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_sub_epi32_rmbk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubd (%rdi){1to16}, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_sub_epi32_rmbkz(<16 x i32> %a, i32* %ptr_b, i16 %mask) {
; CHECK-LABEL: test_mask_sub_epi32_rmbkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubd (%rdi){1to16}, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psub.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <8 x i64> @test_mask_add_epi64_rr(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_mask_add_epi64_rr:
; CHECK: ## BB#0:
; CHECK-NEXT: vpaddq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rrk(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_add_epi64_rrk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpaddq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rrkz(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
; CHECK-LABEL: test_mask_add_epi64_rrkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpaddq %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rm(<8 x i64> %a, <8 x i64>* %ptr_b) {
; CHECK-LABEL: test_mask_add_epi64_rm:
; CHECK: ## BB#0:
; CHECK-NEXT: vpaddq (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <8 x i64>, <8 x i64>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rmk(<8 x i64> %a, <8 x i64>* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_add_epi64_rmk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddq (%rdi), %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%b = load <8 x i64>, <8 x i64>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rmkz(<8 x i64> %a, <8 x i64>* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_add_epi64_rmkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddq (%rdi), %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%b = load <8 x i64>, <8 x i64>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rmb(<8 x i64> %a, i64* %ptr_b) {
; CHECK-LABEL: test_mask_add_epi64_rmb:
; CHECK: ## BB#0:
; CHECK-NEXT: vpaddq (%rdi){1to8}, %zmm0, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rmbk(<8 x i64> %a, i64* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_add_epi64_rmbk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddq (%rdi){1to8}, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_add_epi64_rmbkz(<8 x i64> %a, i64* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_add_epi64_rmbkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpaddq (%rdi){1to8}, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%res = call <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.padd.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64> @test_mask_sub_epi64_rr(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_mask_sub_epi64_rr:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsubq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rrk(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_sub_epi64_rrk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsubq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rrkz(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
; CHECK-LABEL: test_mask_sub_epi64_rrkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsubq %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rm(<8 x i64> %a, <8 x i64>* %ptr_b) {
; CHECK-LABEL: test_mask_sub_epi64_rm:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsubq (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <8 x i64>, <8 x i64>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rmk(<8 x i64> %a, <8 x i64>* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_sub_epi64_rmk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubq (%rdi), %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%b = load <8 x i64>, <8 x i64>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rmkz(<8 x i64> %a, <8 x i64>* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_sub_epi64_rmkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubq (%rdi), %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%b = load <8 x i64>, <8 x i64>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rmb(<8 x i64> %a, i64* %ptr_b) {
; CHECK-LABEL: test_mask_sub_epi64_rmb:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsubq (%rdi){1to8}, %zmm0, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rmbk(<8 x i64> %a, i64* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_sub_epi64_rmbk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubq (%rdi){1to8}, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_sub_epi64_rmbkz(<8 x i64> %a, i64* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_sub_epi64_rmbkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsubq (%rdi){1to8}, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%res = call <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64> %a, <8 x i64> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psub.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <16 x i32> @test_mask_mullo_epi32_rr_512(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_mask_mullo_epi32_rr_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmulld %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rrk_512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_mullo_epi32_rrk_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmulld %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rrkz_512(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: test_mask_mullo_epi32_rrkz_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmulld %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rm_512(<16 x i32> %a, <16 x i32>* %ptr_b) {
; CHECK-LABEL: test_mask_mullo_epi32_rm_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmulld (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rmk_512(<16 x i32> %a, <16 x i32>* %ptr_b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_mullo_epi32_rmk_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmulld (%rdi), %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rmkz_512(<16 x i32> %a, <16 x i32>* %ptr_b, i16 %mask) {
; CHECK-LABEL: test_mask_mullo_epi32_rmkz_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmulld (%rdi), %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rmb_512(<16 x i32> %a, i32* %ptr_b) {
; CHECK-LABEL: test_mask_mullo_epi32_rmb_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmulld (%rdi){1to16}, %zmm0, %zmm0
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 -1)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rmbk_512(<16 x i32> %a, i32* %ptr_b, <16 x i32> %passThru, i16 %mask) {
; CHECK-LABEL: test_mask_mullo_epi32_rmbk_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmulld (%rdi){1to16}, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passThru, i16 %mask)
ret < 16 x i32> %res
}
define <16 x i32> @test_mask_mullo_epi32_rmbkz_512(<16 x i32> %a, i32* %ptr_b, i16 %mask) {
; CHECK-LABEL: test_mask_mullo_epi32_rmbkz_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmulld (%rdi){1to16}, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%q = load i32, i32* %ptr_b
%vecinit.i = insertelement <16 x i32> undef, i32 %q, i32 0
%b = shufflevector <16 x i32> %vecinit.i, <16 x i32> undef, <16 x i32> zeroinitializer
%res = call <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32> %a, <16 x i32> %b, <16 x i32> zeroinitializer, i16 %mask)
ret < 16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.pmull.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
[x86][AVX512] Lowering shuffle i/f intrinsics to LLVM IR This patch, together with a matching clang patch (https://reviews.llvm.org/D38672), implements the lowering of X86 shuffle i/f intrinsics to IR. Differential Revision: https://reviews.llvm.org/D38671 Change-Id: I1e7d359a74743e995ec356237a85214ce55d3661 llvm-svn: 318026
2017-11-13 17:16:39 +08:00
declare <16 x float> @llvm.x86.avx512.mask.shuf.f32x4(<16 x float>, <16 x float>, i32, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_shuf_f32x4(<16 x float> %x0, <16 x float> %x1, <16 x float> %x3, i16 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_shuf_f32x4:
; CHECK: ## BB#0:
; CHECK-NEXT: vshuff32x4 {{.*#+}} zmm3 = zmm0[8,9,10,11,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vshuff32x4 {{.*#+}} zmm2 {%k1} = zmm0[8,9,10,11,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
; CHECK-NEXT: vaddps %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.shuf.f32x4(<16 x float> %x0, <16 x float> %x1, i32 22, <16 x float> %x3, i16 %x4)
%res1 = call <16 x float> @llvm.x86.avx512.mask.shuf.f32x4(<16 x float> %x0, <16 x float> %x1, i32 22, <16 x float> %x3, i16 -1)
%res2 = fadd <16 x float> %res, %res1
ret <16 x float> %res2
}
declare <8 x double> @llvm.x86.avx512.mask.shuf.f64x2(<8 x double>, <8 x double>, i32, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_shuf_f64x2(<8 x double> %x0, <8 x double> %x1, <8 x double> %x3, i8 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_shuf_f64x2:
; CHECK: ## BB#0:
; CHECK-NEXT: vshuff64x2 {{.*#+}} zmm3 = zmm0[4,5,2,3],zmm1[2,3,0,1]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vshuff64x2 {{.*#+}} zmm2 {%k1} = zmm0[4,5,2,3],zmm1[2,3,0,1]
; CHECK-NEXT: vaddpd %zmm3, %zmm2, %zmm2
; CHECK-NEXT: vshuff64x2 {{.*#+}} zmm0 {%k1} {z} = zmm0[4,5,2,3],zmm1[2,3,0,1]
; CHECK-NEXT: vaddpd %zmm0, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.shuf.f64x2(<8 x double> %x0, <8 x double> %x1, i32 22, <8 x double> %x3, i8 %x4)
%res1 = call <8 x double> @llvm.x86.avx512.mask.shuf.f64x2(<8 x double> %x0, <8 x double> %x1, i32 22, <8 x double> %x3, i8 -1)
%res2 = call <8 x double> @llvm.x86.avx512.mask.shuf.f64x2(<8 x double> %x0, <8 x double> %x1, i32 22, <8 x double> zeroinitializer, i8 %x4)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res3, %res2
ret <8 x double> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.shuf.i32x4(<16 x i32>, <16 x i32>, i32, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_shuf_i32x4(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x3, i16 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_shuf_i32x4:
; CHECK: ## BB#0:
; CHECK-NEXT: vshufi32x4 {{.*#+}} zmm3 = zmm0[8,9,10,11,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vshufi32x4 {{.*#+}} zmm2 {%k1} = zmm0[8,9,10,11,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.shuf.i32x4(<16 x i32> %x0, <16 x i32> %x1, i32 22, <16 x i32> %x3, i16 %x4)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.shuf.i32x4(<16 x i32> %x0, <16 x i32> %x1, i32 22, <16 x i32> %x3, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
declare <8 x i64> @llvm.x86.avx512.mask.shuf.i64x2(<8 x i64>, <8 x i64>, i32, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_shuf_i64x2(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x3, i8 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_shuf_i64x2:
; CHECK: ## BB#0:
; CHECK-NEXT: vshufi64x2 {{.*#+}} zmm3 = zmm0[4,5,2,3],zmm1[2,3,0,1]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vshufi64x2 {{.*#+}} zmm2 {%k1} = zmm0[4,5,2,3],zmm1[2,3,0,1]
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.shuf.i64x2(<8 x i64> %x0, <8 x i64> %x1, i32 22, <8 x i64> %x3, i8 %x4)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.shuf.i64x2(<8 x i64> %x0, <8 x i64> %x1, i32 22, <8 x i64> %x3, i8 -1)
%res2 = add <8 x i64> %res, %res1
ret <8 x i64> %res2
}
[X86] Remove masked shufpd/shufps intrinsics and autoupgrade to native vector shuffles. They were removed from clang previously but accidentally left in the backend. llvm-svn: 281300
2016-09-13 15:40:53 +08:00
declare <8 x double> @llvm.x86.avx512.mask.shuf.pd.512(<8 x double>, <8 x double>, i32, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_shuf_pd_512(<8 x double> %x0, <8 x double> %x1, <8 x double> %x3, i8 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_shuf_pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vshufpd {{.*#+}} zmm3 = zmm0[0],zmm1[1],zmm0[3],zmm1[2],zmm0[5],zmm1[4],zmm0[6],zmm1[6]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vshufpd {{.*#+}} zmm2 {%k1} = zmm0[0],zmm1[1],zmm0[3],zmm1[2],zmm0[5],zmm1[4],zmm0[6],zmm1[6]
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vaddpd %zmm3, %zmm2, %zmm2
[X86] Remove masked shufpd/shufps intrinsics and autoupgrade to native vector shuffles. They were removed from clang previously but accidentally left in the backend. llvm-svn: 281300
2016-09-13 15:40:53 +08:00
; CHECK-NEXT: vshufpd {{.*#+}} zmm0 {%k1} {z} = zmm0[0],zmm1[1],zmm0[3],zmm1[2],zmm0[5],zmm1[4],zmm0[6],zmm1[6]
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vaddpd %zmm0, %zmm2, %zmm0
[X86] Remove masked shufpd/shufps intrinsics and autoupgrade to native vector shuffles. They were removed from clang previously but accidentally left in the backend. llvm-svn: 281300
2016-09-13 15:40:53 +08:00
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.shuf.pd.512(<8 x double> %x0, <8 x double> %x1, i32 22, <8 x double> %x3, i8 %x4)
%res1 = call <8 x double> @llvm.x86.avx512.mask.shuf.pd.512(<8 x double> %x0, <8 x double> %x1, i32 22, <8 x double> %x3, i8 -1)
%res2 = call <8 x double> @llvm.x86.avx512.mask.shuf.pd.512(<8 x double> %x0, <8 x double> %x1, i32 22, <8 x double> zeroinitializer, i8 %x4)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res3, %res2
ret <8 x double> %res4
}
declare <16 x float> @llvm.x86.avx512.mask.shuf.ps.512(<16 x float>, <16 x float>, i32, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_shuf_ps_512(<16 x float> %x0, <16 x float> %x1, <16 x float> %x3, i16 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_shuf_ps_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vshufps {{.*#+}} zmm3 = zmm0[2,1],zmm1[1,0],zmm0[6,5],zmm1[5,4],zmm0[10,9],zmm1[9,8],zmm0[14,13],zmm1[13,12]
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vshufps {{.*#+}} zmm2 {%k1} = zmm0[2,1],zmm1[1,0],zmm0[6,5],zmm1[5,4],zmm0[10,9],zmm1[9,8],zmm0[14,13],zmm1[13,12]
; CHECK-NEXT: vaddps %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.shuf.ps.512(<16 x float> %x0, <16 x float> %x1, i32 22, <16 x float> %x3, i16 %x4)
%res1 = call <16 x float> @llvm.x86.avx512.mask.shuf.ps.512(<16 x float> %x0, <16 x float> %x1, i32 22, <16 x float> %x3, i16 -1)
%res2 = fadd <16 x float> %res, %res1
ret <16 x float> %res2
}
[AVX-512] Remove masked pmin/pmax intrinsics and autoupgrade to native IR. Clang patch to replace 512-bit vector and 64-bit element versions with native IR will follow. llvm-svn: 284955
2016-10-24 12:04:16 +08:00
declare <16 x i32> @llvm.x86.avx512.mask.pmaxs.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pmaxs_d_512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmaxs_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmaxsd %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmaxsd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmaxs.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pmaxs.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
declare <8 x i64> @llvm.x86.avx512.mask.pmaxs.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmaxs_q_512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmaxs_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmaxsq %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmaxsq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmaxs.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmaxs.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 -1)
%res2 = add <8 x i64> %res, %res1
ret <8 x i64> %res2
}
declare <16 x i32> @llvm.x86.avx512.mask.pmaxu.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pmaxu_d_512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmaxu_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmaxud %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmaxud %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmaxu.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pmaxu.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
declare <8 x i64> @llvm.x86.avx512.mask.pmaxu.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmaxu_q_512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmaxu_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmaxuq %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmaxuq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmaxu.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmaxu.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 -1)
%res2 = add <8 x i64> %res, %res1
ret <8 x i64> %res2
}
declare <16 x i32> @llvm.x86.avx512.mask.pmins.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pmins_d_512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmins_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpminsd %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpminsd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmins.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pmins.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
declare <8 x i64> @llvm.x86.avx512.mask.pmins.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmins_q_512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmins_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpminsq %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpminsq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmins.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmins.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 -1)
%res2 = add <8 x i64> %res, %res1
ret <8 x i64> %res2
}
declare <16 x i32> @llvm.x86.avx512.mask.pminu.d.512(<16 x i32>, <16 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pminu_d_512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pminu_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpminud %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpminud %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pminu.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pminu.d.512(<16 x i32> %x0, <16 x i32> %x1, <16 x i32> %x2, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
declare <8 x i64> @llvm.x86.avx512.mask.pminu.q.512(<8 x i64>, <8 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pminu_q_512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_pminu_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpminuq %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpminuq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pminu.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pminu.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 -1)
%res2 = add <8 x i64> %res, %res1
ret <8 x i64> %res2
}
[X86][AVX512] Removing llvm x86 intrinsics for _mm_mask_move_{ss|sd} intrinsics. Differential Revision: https://reviews.llvm.org/D26128 llvm-svn: 287087
2016-11-16 17:00:28 +08:00
define <4 x float> @test_mm_mask_move_ss(<4 x float> %__W, i8 zeroext %__U, <4 x float> %__A, <4 x float> %__B) {
; CHECK-LABEL: test_mm_mask_move_ss:
; CHECK: ## BB#0: ## %entry
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vmovss %xmm2, %xmm1, %xmm0 {%k1}
; CHECK-NEXT: retq
entry:
%res = call <4 x float> @llvm.x86.avx512.mask.move.ss(<4 x float> %__A, <4 x float> %__B, <4 x float> %__W, i8 %__U)
ret <4 x float> %res
}
define <4 x float> @test_mm_maskz_move_ss(i8 zeroext %__U, <4 x float> %__A, <4 x float> %__B) {
; CHECK-LABEL: test_mm_maskz_move_ss:
; CHECK: ## BB#0: ## %entry
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vmovss %xmm1, %xmm0, %xmm0 {%k1} {z}
; CHECK-NEXT: retq
entry:
%res = call <4 x float> @llvm.x86.avx512.mask.move.ss(<4 x float> %__A, <4 x float> %__B, <4 x float> zeroinitializer, i8 %__U)
ret <4 x float> %res
}
define <2 x double> @test_mm_mask_move_sd(<2 x double> %__W, i8 zeroext %__U, <2 x double> %__A, <2 x double> %__B) {
; CHECK-LABEL: test_mm_mask_move_sd:
; CHECK: ## BB#0: ## %entry
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vmovsd %xmm2, %xmm1, %xmm0 {%k1}
; CHECK-NEXT: retq
entry:
%res = call <2 x double> @llvm.x86.avx512.mask.move.sd(<2 x double> %__A, <2 x double> %__B, <2 x double> %__W, i8 %__U)
ret <2 x double> %res
}
define <2 x double> @test_mm_maskz_move_sd(i8 zeroext %__U, <2 x double> %__A, <2 x double> %__B) {
; CHECK-LABEL: test_mm_maskz_move_sd:
; CHECK: ## BB#0: ## %entry
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vmovsd %xmm1, %xmm0, %xmm0 {%k1} {z}
; CHECK-NEXT: retq
entry:
%res = call <2 x double> @llvm.x86.avx512.mask.move.sd(<2 x double> %__A, <2 x double> %__B, <2 x double> zeroinitializer, i8 %__U)
ret <2 x double> %res
}
declare <4 x float> @llvm.x86.avx512.mask.move.ss(<4 x float>, <4 x float>, <4 x float>, i8)
declare <2 x double> @llvm.x86.avx512.mask.move.sd(<2 x double>, <2 x double>, <2 x double>, i8)
[AVX-512] Remove masked pmovzx/pmovsx builtins and autoupgrade them to selects and native zext/sext. This mostly reuses earlier autoupgrade support for the sse and avx equivalents. Just needed to add the code to add the select. llvm-svn: 286092
2016-11-07 10:12:57 +08:00
declare <16 x i32> @llvm.x86.avx512.mask.pmovzxb.d.512(<16 x i8>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pmovzxb_d_512(<16 x i8> %x0, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovzxb_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovzxbd {{.*#+}} zmm2 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovzxbd {{.*#+}} zmm1 {%k1} = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
; CHECK-NEXT: vpmovzxbd {{.*#+}} zmm0 {%k1} {z} = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmovzxb.d.512(<16 x i8> %x0, <16 x i32> %x1, i16 %x2)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pmovzxb.d.512(<16 x i8> %x0, <16 x i32> zeroinitializer, i16 %x2)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.pmovzxb.d.512(<16 x i8> %x0, <16 x i32> %x1, i16 -1)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.pmovzxb.q.512(<16 x i8>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmovzxb_q_512(<16 x i8> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovzxb_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovzxbq {{.*#+}} zmm2 = xmm0[0],zero,zero,zero,zero,zero,zero,zero,xmm0[1],zero,zero,zero,zero,zero,zero,zero,xmm0[2],zero,zero,zero,zero,zero,zero,zero,xmm0[3],zero,zero,zero,zero,zero,zero,zero,xmm0[4],zero,zero,zero,zero,zero,zero,zero,xmm0[5],zero,zero,zero,zero,zero,zero,zero,xmm0[6],zero,zero,zero,zero,zero,zero,zero,xmm0[7],zero,zero,zero,zero,zero,zero,zero
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovzxbq {{.*#+}} zmm1 {%k1} = xmm0[0],zero,zero,zero,zero,zero,zero,zero,xmm0[1],zero,zero,zero,zero,zero,zero,zero,xmm0[2],zero,zero,zero,zero,zero,zero,zero,xmm0[3],zero,zero,zero,zero,zero,zero,zero,xmm0[4],zero,zero,zero,zero,zero,zero,zero,xmm0[5],zero,zero,zero,zero,zero,zero,zero,xmm0[6],zero,zero,zero,zero,zero,zero,zero,xmm0[7],zero,zero,zero,zero,zero,zero,zero
; CHECK-NEXT: vpmovzxbq {{.*#+}} zmm0 {%k1} {z} = xmm0[0],zero,zero,zero,zero,zero,zero,zero,xmm0[1],zero,zero,zero,zero,zero,zero,zero,xmm0[2],zero,zero,zero,zero,zero,zero,zero,xmm0[3],zero,zero,zero,zero,zero,zero,zero,xmm0[4],zero,zero,zero,zero,zero,zero,zero,xmm0[5],zero,zero,zero,zero,zero,zero,zero,xmm0[6],zero,zero,zero,zero,zero,zero,zero,xmm0[7],zero,zero,zero,zero,zero,zero,zero
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmovzxb.q.512(<16 x i8> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmovzxb.q.512(<16 x i8> %x0, <8 x i64> zeroinitializer, i8 %x2)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.pmovzxb.q.512(<16 x i8> %x0, <8 x i64> %x1, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.pmovzxd.q.512(<8 x i32>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmovzxd_q_512(<8 x i32> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovzxd_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovzxdq {{.*#+}} zmm2 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovzxdq {{.*#+}} zmm1 {%k1} = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero
; CHECK-NEXT: vpmovzxdq {{.*#+}} zmm0 {%k1} {z} = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmovzxd.q.512(<8 x i32> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmovzxd.q.512(<8 x i32> %x0, <8 x i64> zeroinitializer, i8 %x2)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.pmovzxd.q.512(<8 x i32> %x0, <8 x i64> %x1, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.pmovzxw.d.512(<16 x i16>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pmovzxw_d_512(<16 x i16> %x0, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovzxw_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovzxwd {{.*#+}} zmm2 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovzxwd {{.*#+}} zmm1 {%k1} = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
; CHECK-NEXT: vpmovzxwd {{.*#+}} zmm0 {%k1} {z} = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmovzxw.d.512(<16 x i16> %x0, <16 x i32> %x1, i16 %x2)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pmovzxw.d.512(<16 x i16> %x0, <16 x i32> zeroinitializer, i16 %x2)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.pmovzxw.d.512(<16 x i16> %x0, <16 x i32> %x1, i16 -1)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.pmovzxw.q.512(<8 x i16>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmovzxw_q_512(<8 x i16> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovzxw_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovzxwq {{.*#+}} zmm2 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovzxwq {{.*#+}} zmm1 {%k1} = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero
; CHECK-NEXT: vpmovzxwq {{.*#+}} zmm0 {%k1} {z} = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmovzxw.q.512(<8 x i16> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmovzxw.q.512(<8 x i16> %x0, <8 x i64> zeroinitializer, i8 %x2)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.pmovzxw.q.512(<8 x i16> %x0, <8 x i64> %x1, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.pmovsxb.d.512(<16 x i8>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pmovsxb_d_512(<16 x i8> %x0, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxb_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovsxbd %xmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovsxbd %xmm0, %zmm1 {%k1}
; CHECK-NEXT: vpmovsxbd %xmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmovsxb.d.512(<16 x i8> %x0, <16 x i32> %x1, i16 %x2)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pmovsxb.d.512(<16 x i8> %x0, <16 x i32> zeroinitializer, i16 %x2)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.pmovsxb.d.512(<16 x i8> %x0, <16 x i32> %x1, i16 -1)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.pmovsxb.q.512(<16 x i8>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmovsxb_q_512(<16 x i8> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxb_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovsxbq %xmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovsxbq %xmm0, %zmm1 {%k1}
; CHECK-NEXT: vpmovsxbq %xmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmovsxb.q.512(<16 x i8> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmovsxb.q.512(<16 x i8> %x0, <8 x i64> zeroinitializer, i8 %x2)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.pmovsxb.q.512(<16 x i8> %x0, <8 x i64> %x1, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.pmovsxd.q.512(<8 x i32>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmovsxd_q_512(<8 x i32> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxd_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovsxdq %ymm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovsxdq %ymm0, %zmm1 {%k1}
; CHECK-NEXT: vpmovsxdq %ymm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmovsxd.q.512(<8 x i32> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmovsxd.q.512(<8 x i32> %x0, <8 x i64> zeroinitializer, i8 %x2)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.pmovsxd.q.512(<8 x i32> %x0, <8 x i64> %x1, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.pmovsxw.d.512(<16 x i16>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pmovsxw_d_512(<16 x i16> %x0, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxw_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovsxwd %ymm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovsxwd %ymm0, %zmm1 {%k1}
; CHECK-NEXT: vpmovsxwd %ymm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pmovsxw.d.512(<16 x i16> %x0, <16 x i32> %x1, i16 %x2)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pmovsxw.d.512(<16 x i16> %x0, <16 x i32> zeroinitializer, i16 %x2)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.pmovsxw.d.512(<16 x i16> %x0, <16 x i32> %x1, i16 -1)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.pmovsxw.q.512(<8 x i16>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pmovsxw_q_512(<8 x i16> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxw_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmovsxwq %xmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmovsxwq %xmm0, %zmm1 {%k1}
; CHECK-NEXT: vpmovsxwq %xmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmovsxw.q.512(<8 x i16> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pmovsxw.q.512(<8 x i16> %x0, <8 x i64> zeroinitializer, i8 %x2)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.pmovsxw.q.512(<8 x i16> %x0, <8 x i64> %x1, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
[AVX-512] Remove the remaining masked shift by immediate or by single value. Autoupgrade them to recently introduced unmasked versions and a select. After this I'll add the unmasked intrinsics to InstCombineCalls to finish making our handling of these types of shuffles consistent between AVX-512 and the legacy intrinsics. llvm-svn: 286725
2016-11-13 02:04:46 +08:00
declare <8 x i64> @llvm.x86.avx512.mask.psrl.qi.512(<8 x i64>, i32, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_psrl_qi_512(<8 x i64> %x0, i32 %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_psrl_qi_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrlq $4, %zmm0, %zmm2
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsrlq $4, %zmm0, %zmm1 {%k1}
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vpaddq %zmm2, %zmm1, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpsrlq $4, %zmm0, %zmm0 {%k1} {z}
[AVX-512] Remove the remaining masked shift by immediate or by single value. Autoupgrade them to recently introduced unmasked versions and a select. After this I'll add the unmasked intrinsics to InstCombineCalls to finish making our handling of these types of shuffles consistent between AVX-512 and the legacy intrinsics. llvm-svn: 286725
2016-11-13 02:04:46 +08:00
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrl.qi.512(<8 x i64> %x0, i32 4, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.psrl.qi.512(<8 x i64> %x0, i32 4, <8 x i64> %x2, i8 -1)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.psrl.qi.512(<8 x i64> %x0, i32 4, <8 x i64> zeroinitializer, i8 %x3)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.psrl.di.512(<16 x i32>, i32, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_psrl_di_512(<16 x i32> %x0, i32 %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_psrl_di_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrld $4, %zmm0, %zmm2
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsrld $4, %zmm0, %zmm1 {%k1}
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vpaddd %zmm2, %zmm1, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpsrld $4, %zmm0, %zmm0 {%k1} {z}
[AVX-512] Remove the remaining masked shift by immediate or by single value. Autoupgrade them to recently introduced unmasked versions and a select. After this I'll add the unmasked intrinsics to InstCombineCalls to finish making our handling of these types of shuffles consistent between AVX-512 and the legacy intrinsics. llvm-svn: 286725
2016-11-13 02:04:46 +08:00
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrl.di.512(<16 x i32> %x0, i32 4, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.psrl.di.512(<16 x i32> %x0, i32 4, <16 x i32> %x2, i16 -1)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.psrl.di.512(<16 x i32> %x0, i32 4, <16 x i32> zeroinitializer, i16 %x3)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.psra.di.512(<16 x i32>, i32, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_psra_di_512(<16 x i32> %x0, i32 %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_psra_di_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrad $3, %zmm0, %zmm2
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsrad $3, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vpsrad $3, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psra.di.512(<16 x i32> %x0, i32 3, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.psra.di.512(<16 x i32> %x0, i32 3, <16 x i32> zeroinitializer, i16 %x3)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.psra.di.512(<16 x i32> %x0, i32 3, <16 x i32> %x2, i16 -1)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.psra.qi.512(<8 x i64>, i32, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_psra_qi_512(<8 x i64> %x0, i32 %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_psra_qi_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsraq $3, %zmm0, %zmm2
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsraq $3, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vpsraq $3, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psra.qi.512(<8 x i64> %x0, i32 3, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.psra.qi.512(<8 x i64> %x0, i32 3, <8 x i64> zeroinitializer, i8 %x3)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.psra.qi.512(<8 x i64> %x0, i32 3, <8 x i64> %x2, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.psll.di.512(<16 x i32>, i32, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_psll_di_512(<16 x i32> %x0, i32 %x1, <16 x i32> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_psll_di_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpslld $3, %zmm0, %zmm2
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpslld $3, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vpslld $3, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psll.di.512(<16 x i32> %x0, i32 3, <16 x i32> %x2, i16 %x3)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.psll.di.512(<16 x i32> %x0, i32 3, <16 x i32> zeroinitializer, i16 %x3)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.psll.di.512(<16 x i32> %x0, i32 3, <16 x i32> %x2, i16 -1)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res3, %res2
ret <16 x i32> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.psll.qi.512(<8 x i64>, i32, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_psll_qi_512(<8 x i64> %x0, i32 %x1, <8 x i64> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_psll_qi_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsllq $3, %zmm0, %zmm2
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpsllq $3, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vpsllq $3, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm0, %zmm1, %zmm0
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psll.qi.512(<8 x i64> %x0, i32 3, <8 x i64> %x2, i8 %x3)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.psll.qi.512(<8 x i64> %x0, i32 3, <8 x i64> zeroinitializer, i8 %x3)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.psll.qi.512(<8 x i64> %x0, i32 3, <8 x i64> %x2, i8 -1)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res3, %res2
ret <8 x i64> %res4
}
[AVX-512] Move masked shift intrinsics tests to the autoupgrade test file. These missed being moved in r286725. llvm-svn: 286746
2016-11-13 11:42:27 +08:00
define <16 x i32> @test_x86_avx512_psll_d(<16 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: test_x86_avx512_psll_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpslld %xmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psll.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psll_d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> %a2, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psll_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpslld %xmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psll.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> %a2, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psll_d(<16 x i32> %a0, <4 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psll_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpslld %xmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psll.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psll.d(<16 x i32>, <4 x i32>, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psll_q(<8 x i64> %a0, <2 x i64> %a1) {
; CHECK-LABEL: test_x86_avx512_psll_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsllq %xmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psll.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psll_q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> %a2, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psll_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllq %xmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psll.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> %a2, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psll_q(<8 x i64> %a0, <2 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psll_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllq %xmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psll.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psll.q(<8 x i64>, <2 x i64>, <8 x i64>, i8) nounwind readnone
define <16 x i32> @test_x86_avx512_psrl_d(<16 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: test_x86_avx512_psrl_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrld %xmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrl.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psrl_d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> %a2, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrl_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrld %xmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrl.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> %a2, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psrl_d(<16 x i32> %a0, <4 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrl_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrld %xmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrl.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psrl.d(<16 x i32>, <4 x i32>, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psrl_q(<8 x i64> %a0, <2 x i64> %a1) {
; CHECK-LABEL: test_x86_avx512_psrl_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrlq %xmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrl.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psrl_q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> %a2, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrl_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlq %xmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrl.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> %a2, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psrl_q(<8 x i64> %a0, <2 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrl_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlq %xmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrl.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psrl.q(<8 x i64>, <2 x i64>, <8 x i64>, i8) nounwind readnone
define <16 x i32> @test_x86_avx512_psra_d(<16 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: test_x86_avx512_psra_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrad %xmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psra.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psra_d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> %a2, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psra_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrad %xmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psra.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> %a2, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psra_d(<16 x i32> %a0, <4 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psra_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrad %xmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psra.d(<16 x i32> %a0, <4 x i32> %a1, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psra.d(<16 x i32>, <4 x i32>, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psra_q(<8 x i64> %a0, <2 x i64> %a1) {
; CHECK-LABEL: test_x86_avx512_psra_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsraq %xmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psra.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psra_q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> %a2, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psra_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsraq %xmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psra.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> %a2, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psra_q(<8 x i64> %a0, <2 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psra_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsraq %xmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psra.q(<8 x i64> %a0, <2 x i64> %a1, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psra.q(<8 x i64>, <2 x i64>, <8 x i64>, i8) nounwind readnone
[AVX-512] Remove and autoupgrade masked dword/qword variable shift intrinsics to the new unmasked versions and selects. llvm-svn: 286786
2016-11-14 09:53:22 +08:00
define <16 x i32> @test_x86_avx512_psllv_d(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK-LABEL: test_x86_avx512_psllv_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsllvd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psllv.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psllv_d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psllv_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllvd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psllv.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psllv_d(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psllv_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllvd %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psllv.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psllv.d(<16 x i32>, <16 x i32>, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psllv_q(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK-LABEL: test_x86_avx512_psllv_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsllvq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psllv.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psllv_q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> %a2, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psllv_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllvq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psllv.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> %a2, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psllv_q(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psllv_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsllvq %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psllv.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psllv.q(<8 x i64>, <8 x i64>, <8 x i64>, i8) nounwind readnone
define <16 x i32> @test_x86_avx512_psrav_d(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK-LABEL: test_x86_avx512_psrav_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsravd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrav.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psrav_d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrav_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsravd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrav.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psrav_d(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrav_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsravd %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrav.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psrav.d(<16 x i32>, <16 x i32>, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psrav_q(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK-LABEL: test_x86_avx512_psrav_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsravq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrav.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psrav_q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> %a2, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrav_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsravq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrav.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> %a2, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psrav_q(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrav_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsravq %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrav.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psrav.q(<8 x i64>, <8 x i64>, <8 x i64>, i8) nounwind readnone
define <16 x i32> @test_x86_avx512_psrlv_d(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK-LABEL: test_x86_avx512_psrlv_d:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrlvd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrlv.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_mask_psrlv_d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrlv_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlvd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrlv.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask)
ret <16 x i32> %res
}
define <16 x i32> @test_x86_avx512_maskz_psrlv_d(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrlv_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlvd %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.psrlv.d(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.psrlv.d(<16 x i32>, <16 x i32>, <16 x i32>, i16) nounwind readnone
define <8 x i64> @test_x86_avx512_psrlv_q(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK-LABEL: test_x86_avx512_psrlv_q:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrlvq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrlv.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_mask_psrlv_q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> %a2, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_mask_psrlv_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlvq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrlv.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> %a2, i8 %mask)
ret <8 x i64> %res
}
define <8 x i64> @test_x86_avx512_maskz_psrlv_q(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_x86_avx512_maskz_psrlv_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpsrlvq %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.psrlv.q(<8 x i64> %a0, <8 x i64> %a1, <8 x i64> zeroinitializer, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.psrlv.q(<8 x i64>, <8 x i64>, <8 x i64>, i8) nounwind readnone
define <8 x i64> @test_x86_avx512_psrlv_q_memop(<8 x i64> %a0, <8 x i64>* %ptr) {
; CHECK-LABEL: test_x86_avx512_psrlv_q_memop:
; CHECK: ## BB#0:
; CHECK-NEXT: vpsrlvq (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <8 x i64>, <8 x i64>* %ptr
%res = call <8 x i64> @llvm.x86.avx512.mask.psrlv.q(<8 x i64> %a0, <8 x i64> %b, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
[X86][AVX512] Autoupgrade lossless i32/u32 to f64 conversion intrinsics with generic IR Both the (V)CVTDQ2PD (i32 to f64) and (V)CVTUDQ2PD (u32 to f64) conversion instructions are lossless and can be safely represented as generic SINT_TO_FP/UINT_TO_FP calls instead of x86 intrinsics without affecting final codegen. LLVM counterpart to D26686 Differential Revision: https://reviews.llvm.org/D26736 llvm-svn: 287108
2016-11-16 22:48:32 +08:00
declare <8 x double> @llvm.x86.avx512.mask.cvtdq2pd.512(<8 x i32>, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_cvt_dq2pd_512(<8 x i32> %x0, <8 x double> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_cvt_dq2pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vcvtdq2pd %ymm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vcvtdq2pd %ymm0, %zmm1 {%k1}
; CHECK-NEXT: vaddpd %zmm2, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.cvtdq2pd.512(<8 x i32> %x0, <8 x double> %x1, i8 %x2)
%res1 = call <8 x double> @llvm.x86.avx512.mask.cvtdq2pd.512(<8 x i32> %x0, <8 x double> %x1, i8 -1)
%res2 = fadd <8 x double> %res, %res1
ret <8 x double> %res2
}
declare <8 x double> @llvm.x86.avx512.mask.cvtudq2pd.512(<8 x i32>, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_cvt_udq2pd_512(<8 x i32> %x0, <8 x double> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_cvt_udq2pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vcvtudq2pd %ymm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vcvtudq2pd %ymm0, %zmm1 {%k1}
; CHECK-NEXT: vaddpd %zmm2, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.cvtudq2pd.512(<8 x i32> %x0, <8 x double> %x1, i8 %x2)
%res1 = call <8 x double> @llvm.x86.avx512.mask.cvtudq2pd.512(<8 x i32> %x0, <8 x double> %x1, i8 -1)
%res2 = fadd <8 x double> %res, %res1
ret <8 x double> %res2
}
[AVX-512] Remove intrinsics for valignd/q and autoupgrade them to native shuffles. llvm-svn: 287744
2016-11-23 14:54:55 +08:00
define <8 x i64> @test_valign_q(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: test_valign_q:
; CHECK: ## BB#0:
; CHECK-NEXT: valignq {{.*#+}} zmm0 = zmm1[2,3,4,5,6,7],zmm0[0,1]
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.valign.q.512(<8 x i64> %a, <8 x i64> %b, i32 2, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
define <8 x i64> @test_mask_valign_q(<8 x i64> %a, <8 x i64> %b, <8 x i64> %src, i8 %mask) {
; CHECK-LABEL: test_mask_valign_q:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: valignq {{.*#+}} zmm2 {%k1} = zmm1[2,3,4,5,6,7],zmm0[0,1]
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.valign.q.512(<8 x i64> %a, <8 x i64> %b, i32 2, <8 x i64> %src, i8 %mask)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.valign.q.512(<8 x i64>, <8 x i64>, i32, <8 x i64>, i8)
define <16 x i32> @test_maskz_valign_d(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: test_maskz_valign_d:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: valignd {{.*#+}} zmm0 {%k1} {z} = zmm1[5,6,7,8,9,10,11,12,13,14,15],zmm0[0,1,2,3,4]
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.valign.d.512(<16 x i32> %a, <16 x i32> %b, i32 5, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.valign.d.512(<16 x i32>, <16 x i32>, i32, <16 x i32>, i16)
[X86] Remove masking from 512-bit VPERMIL intrinsics in preparation for being able to constant fold them in InstCombineCalls like we do for 128/256-bit. llvm-svn: 289350
2016-12-11 09:26:44 +08:00
declare <8 x double> @llvm.x86.avx512.mask.vpermilvar.pd.512(<8 x double>, <8 x i64>, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_vpermilvar_pd_512(<8 x double> %x0, <8 x i64> %x1, <8 x double> %x2, i8 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_vpermilvar_pd_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpermilpd %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpermilpd %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpermilpd %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vaddpd %zmm0, %zmm2, %zmm0
; CHECK-NEXT: vaddpd %zmm0, %zmm3, %zmm0
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.vpermilvar.pd.512(<8 x double> %x0, <8 x i64> %x1, <8 x double> %x2, i8 %x3)
%res1 = call <8 x double> @llvm.x86.avx512.mask.vpermilvar.pd.512(<8 x double> %x0, <8 x i64> %x1, <8 x double> zeroinitializer, i8 %x3)
%res2 = call <8 x double> @llvm.x86.avx512.mask.vpermilvar.pd.512(<8 x double> %x0, <8 x i64> %x1, <8 x double> %x2, i8 -1)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res2, %res3
ret <8 x double> %res4
}
declare <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float>, <16 x i32>, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_vpermilvar_ps_512(<16 x float> %x0, <16 x i32> %x1, <16 x float> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_vpermilvar_ps_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpermilps %zmm1, %zmm0, %zmm3
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpermilps %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vpermilps %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vaddps %zmm0, %zmm2, %zmm0
; CHECK-NEXT: vaddps %zmm0, %zmm3, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float> %x0, <16 x i32> %x1, <16 x float> %x2, i16 %x3)
%res1 = call <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float> %x0, <16 x i32> %x1, <16 x float> zeroinitializer, i16 %x3)
%res2 = call <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float> %x0, <16 x i32> %x1, <16 x float> %x2, i16 -1)
%res3 = fadd <16 x float> %res, %res1
%res4 = fadd <16 x float> %res2, %res3
ret <16 x float> %res4
}
; Test case to make sure we can print shuffle decode comments for constant pool loads.
define <16 x float>@test_int_x86_avx512_mask_vpermilvar_ps_512_constant_pool(<16 x float> %x0, <16 x i32> %x1, <16 x float> %x2, i16 %x3) {
; CHECK-LABEL: test_int_x86_avx512_mask_vpermilvar_ps_512_constant_pool:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpermilps {{.*#+}} zmm2 {%k1} = zmm0[2,3,0,1,7,6,5,4,9,8,11,10,12,13,14,15]
; CHECK-NEXT: vpermilps {{.*#+}} zmm1 {%k1} {z} = zmm0[1,0,3,2,4,5,6,7,9,8,11,10,12,13,14,15]
Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." This also reverts follow-ups r303292 and r303298. It broke some Chromium tests under MSan, and apparently also internal tests at Google. llvm-svn: 303369
2017-05-19 02:50:05 +08:00
; CHECK-NEXT: vaddps %zmm1, %zmm2, %zmm1
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpermilps {{.*#+}} zmm0 = zmm0[1,0,3,2,4,5,6,7,10,11,8,9,14,15,13,12]
[X86] Remove masking from 512-bit VPERMIL intrinsics in preparation for being able to constant fold them in InstCombineCalls like we do for 128/256-bit. llvm-svn: 289350
2016-12-11 09:26:44 +08:00
; CHECK-NEXT: vaddps %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float> %x0, <16 x i32> <i32 2, i32 3, i32 0, i32 1, i32 3, i32 2, i32 1, i32 0, i32 1, i32 0, i32 3, i32 2, i32 0, i32 1, i32 2, i32 3>, <16 x float> %x2, i16 %x3)
%res1 = call <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float> %x0, <16 x i32> <i32 1, i32 0, i32 3, i32 2, i32 0, i32 1, i32 2, i32 3, i32 1, i32 0, i32 3, i32 2, i32 0, i32 1, i32 2, i32 3>, <16 x float> zeroinitializer, i16 %x3)
%res2 = call <16 x float> @llvm.x86.avx512.mask.vpermilvar.ps.512(<16 x float> %x0, <16 x i32> <i32 1, i32 0, i32 3, i32 2, i32 0, i32 1, i32 2, i32 3, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 1, i32 0>, <16 x float> %x2, i16 -1)
%res3 = fadd <16 x float> %res, %res1
%res4 = fadd <16 x float> %res2, %res3
ret <16 x float> %res4
}
[AVX-512] Remove masked pmuldq and pmuludq intrinsics and autoupgrade them to unmasked intrinsics plus a select. llvm-svn: 290583
2016-12-27 13:30:14 +08:00
define <8 x i64> @test_mask_mul_epi32_rr(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_mask_mul_epi32_rr:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmuldq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rrk(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epi32_rrk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmuldq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rrkz(<16 x i32> %a, <16 x i32> %b, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epi32_rrkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmuldq %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rm(<16 x i32> %a, <16 x i32>* %ptr_b) {
; CHECK-LABEL: test_mask_mul_epi32_rm:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmuldq (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rmk(<16 x i32> %a, <16 x i32>* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epi32_rmk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuldq (%rdi), %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rmkz(<16 x i32> %a, <16 x i32>* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epi32_rmkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuldq (%rdi), %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rmb(<16 x i32> %a, i64* %ptr_b) {
; CHECK-LABEL: test_mask_mul_epi32_rmb:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmuldq (%rdi){1to8}, %zmm0, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b64 = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%b = bitcast <8 x i64> %b64 to <16 x i32>
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rmbk(<16 x i32> %a, i64* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epi32_rmbk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuldq (%rdi){1to8}, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b64 = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%b = bitcast <8 x i64> %b64 to <16 x i32>
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epi32_rmbkz(<16 x i32> %a, i64* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epi32_rmbkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuldq (%rdi){1to8}, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b64 = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%b = bitcast <8 x i64> %b64 to <16 x i32>
%res = call <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.pmul.dq.512(<16 x i32>, <16 x i32>, <8 x i64>, i8)
define <8 x i64> @test_mask_mul_epu32_rr(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: test_mask_mul_epu32_rr:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmuludq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rrk(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epu32_rrk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmuludq %zmm1, %zmm0, %zmm2 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm2, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rrkz(<16 x i32> %a, <16 x i32> %b, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epu32_rrkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpmuludq %zmm1, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rm(<16 x i32> %a, <16 x i32>* %ptr_b) {
; CHECK-LABEL: test_mask_mul_epu32_rm:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmuludq (%rdi), %zmm0, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rmk(<16 x i32> %a, <16 x i32>* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epu32_rmk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuludq (%rdi), %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rmkz(<16 x i32> %a, <16 x i32>* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epu32_rmkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuludq (%rdi), %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%b = load <16 x i32>, <16 x i32>* %ptr_b
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rmb(<16 x i32> %a, i64* %ptr_b) {
; CHECK-LABEL: test_mask_mul_epu32_rmb:
; CHECK: ## BB#0:
; CHECK-NEXT: vpmuludq (%rdi){1to8}, %zmm0, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b64 = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%b = bitcast <8 x i64> %b64 to <16 x i32>
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 -1)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rmbk(<16 x i32> %a, i64* %ptr_b, <8 x i64> %passThru, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epu32_rmbk:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuludq (%rdi){1to8}, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b64 = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%b = bitcast <8 x i64> %b64 to <16 x i32>
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> %passThru, i8 %mask)
ret < 8 x i64> %res
}
define <8 x i64> @test_mask_mul_epu32_rmbkz(<16 x i32> %a, i64* %ptr_b, i8 %mask) {
; CHECK-LABEL: test_mask_mul_epu32_rmbkz:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vpmuludq (%rdi){1to8}, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: retq
%q = load i64, i64* %ptr_b
%vecinit.i = insertelement <8 x i64> undef, i64 %q, i32 0
%b64 = shufflevector <8 x i64> %vecinit.i, <8 x i64> undef, <8 x i32> zeroinitializer
%b = bitcast <8 x i64> %b64 to <16 x i32>
%res = call <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32> %a, <16 x i32> %b, <8 x i64> zeroinitializer, i8 %mask)
ret < 8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.pmulu.dq.512(<16 x i32>, <16 x i32>, <8 x i64>, i8)
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
define <4 x float> @test_mask_vextractf32x4(<4 x float> %b, <16 x float> %a, i8 %mask) {
; CHECK-LABEL: test_mask_vextractf32x4:
; CHECK: ## BB#0:
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kmovw %edi, %k0
; CHECK-NEXT: kshiftlw $12, %k0, %k1
; CHECK-NEXT: kshiftrw $15, %k1, %k1
; CHECK-NEXT: kshiftlw $13, %k0, %k2
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kshiftrw $15, %k2, %k2
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kshiftlw $15, %k0, %k3
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kshiftrw $15, %k3, %k3
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kshiftlw $14, %k0, %k0
; CHECK-NEXT: kshiftrw $15, %k0, %k0
; CHECK-NEXT: kmovw %k0, %eax
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kmovw %k3, %ecx
; CHECK-NEXT: vmovd %ecx, %xmm2
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: vpinsrb $4, %eax, %xmm2, %xmm2
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kmovw %k2, %eax
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: vpinsrb $8, %eax, %xmm2, %xmm2
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrb $12, %eax, %xmm2, %xmm2
[X86][Haswell] Updating HSW instruction scheduling information This patch completely replaces the instruction scheduling information for the Haswell architecture target by modifying the file X86SchedHaswell.td located under the X86 Target. We used the scheduling information retrieved from the Haswell architects in order to replace and modify the existing scheduling. The patch continues the scheduling replacement effort started with the SNB target in r307529 and r310792. Information includes latency, number of micro-Ops and used ports by each HSW instruction. Please expect some performance fluctuations due to code alignment effects. Reviewers: RKSimon, zvi, aymanmus, craig.topper, m_zuckerman, igorb, dim, chandlerc, aaboud Differential Revision: https://reviews.llvm.org/D36663 llvm-svn: 311879
2017-08-28 18:04:16 +08:00
; CHECK-NEXT: vextractf32x4 $2, %zmm1, %xmm1
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: vpslld $31, %xmm2, %xmm2
; CHECK-NEXT: vblendvps %xmm2, %xmm1, %xmm0, %xmm0
; CHECK-NEXT: retq
%res = call <4 x float> @llvm.x86.avx512.mask.vextractf32x4.512(<16 x float> %a, i32 2, <4 x float> %b, i8 %mask)
ret <4 x float> %res
}
declare <4 x float> @llvm.x86.avx512.mask.vextractf32x4.512(<16 x float>, i32, <4 x float>, i8)
define <4 x i64> @test_mask_vextracti64x4(<4 x i64> %b, <8 x i64> %a, i8 %mask) {
; CHECK-LABEL: test_mask_vextracti64x4:
; CHECK: ## BB#0:
[AVX-512] Add unmasked subvector inserts and extract to the execution domain tables. llvm-svn: 309632
2017-08-01 06:07:29 +08:00
; CHECK-NEXT: vextractf64x4 $1, %zmm1, %ymm1
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kmovw %edi, %k0
; CHECK-NEXT: kshiftlw $12, %k0, %k1
; CHECK-NEXT: kshiftrw $15, %k1, %k1
; CHECK-NEXT: kshiftlw $13, %k0, %k2
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kshiftrw $15, %k2, %k2
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kshiftlw $15, %k0, %k3
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kshiftrw $15, %k3, %k3
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kshiftlw $14, %k0, %k0
; CHECK-NEXT: kshiftrw $15, %k0, %k0
; CHECK-NEXT: kmovw %k0, %eax
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kmovw %k3, %ecx
; CHECK-NEXT: vmovd %ecx, %xmm2
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: vpinsrb $4, %eax, %xmm2, %xmm2
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kmovw %k2, %eax
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: vpinsrb $8, %eax, %xmm2, %xmm2
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrb $12, %eax, %xmm2, %xmm2
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: vpslld $31, %xmm2, %xmm2
; CHECK-NEXT: vpmovsxdq %xmm2, %ymm2
; CHECK-NEXT: vblendvpd %ymm2, %ymm1, %ymm0, %ymm0
; CHECK-NEXT: retq
[AVX-512] Update vextract64x4 intrinsic upgrade test cases to use a legal immediate so they test the instruction selection correctly. llvm-svn: 291061
2017-01-05 09:34:55 +08:00
%res = call <4 x i64> @llvm.x86.avx512.mask.vextracti64x4.512(<8 x i64> %a, i32 1, <4 x i64> %b, i8 %mask)
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
ret <4 x i64> %res
}
declare <4 x i64> @llvm.x86.avx512.mask.vextracti64x4.512(<8 x i64>, i32, <4 x i64>, i8)
define <4 x i32> @test_maskz_vextracti32x4(<16 x i32> %a, i8 %mask) {
; CHECK-LABEL: test_maskz_vextracti32x4:
; CHECK: ## BB#0:
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kmovw %edi, %k0
; CHECK-NEXT: kshiftlw $12, %k0, %k1
; CHECK-NEXT: kshiftrw $15, %k1, %k1
; CHECK-NEXT: kshiftlw $13, %k0, %k2
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kshiftrw $15, %k2, %k2
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kshiftlw $15, %k0, %k3
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kshiftrw $15, %k3, %k3
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: kshiftlw $14, %k0, %k0
; CHECK-NEXT: kshiftrw $15, %k0, %k0
; CHECK-NEXT: kmovw %k0, %eax
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kmovw %k3, %ecx
; CHECK-NEXT: vmovd %ecx, %xmm1
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: vpinsrb $4, %eax, %xmm1, %xmm1
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: kmovw %k2, %eax
[X86][AVX512] Make i1 illegal in the CodeGen This patch defines the i1 type as illegal in the X86 backend for AVX512. For DAG operations on <N x i1> types (build vector, extract vector element, ...) i8 is used, and should be truncated/extended. This should produce better scalar code for i1 types since GPRs will be used instead of mask registers. Differential Revision: https://reviews.llvm.org/D32273 llvm-svn: 303421
2017-05-19 20:35:15 +08:00
; CHECK-NEXT: vpinsrb $8, %eax, %xmm1, %xmm1
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrb $12, %eax, %xmm1, %xmm1
[X86][Haswell] Updating HSW instruction scheduling information This patch completely replaces the instruction scheduling information for the Haswell architecture target by modifying the file X86SchedHaswell.td located under the X86 Target. We used the scheduling information retrieved from the Haswell architects in order to replace and modify the existing scheduling. The patch continues the scheduling replacement effort started with the SNB target in r307529 and r310792. Information includes latency, number of micro-Ops and used ports by each HSW instruction. Please expect some performance fluctuations due to code alignment effects. Reviewers: RKSimon, zvi, aymanmus, craig.topper, m_zuckerman, igorb, dim, chandlerc, aaboud Differential Revision: https://reviews.llvm.org/D36663 llvm-svn: 311879
2017-08-28 18:04:16 +08:00
; CHECK-NEXT: vextracti32x4 $2, %zmm0, %xmm0
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: vpslld $31, %xmm1, %xmm1
; CHECK-NEXT: vpsrad $31, %xmm1, %xmm1
; CHECK-NEXT: vpand %xmm0, %xmm1, %xmm0
; CHECK-NEXT: retq
%res = call <4 x i32> @llvm.x86.avx512.mask.vextracti32x4.512(<16 x i32> %a, i32 2, <4 x i32> zeroinitializer, i8 %mask)
ret <4 x i32> %res
}
declare <4 x i32> @llvm.x86.avx512.mask.vextracti32x4.512(<16 x i32>, i32, <4 x i32>, i8)
define <4 x double> @test_vextractf64x4(<8 x double> %a) {
; CHECK-LABEL: test_vextractf64x4:
; CHECK: ## BB#0:
[AVX-512] Update vextract64x4 intrinsic upgrade test cases to use a legal immediate so they test the instruction selection correctly. llvm-svn: 291061
2017-01-05 09:34:55 +08:00
; CHECK-NEXT: vextractf64x4 $1, %zmm0, %ymm0
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
; CHECK-NEXT: retq
[AVX-512] Update vextract64x4 intrinsic upgrade test cases to use a legal immediate so they test the instruction selection correctly. llvm-svn: 291061
2017-01-05 09:34:55 +08:00
%res = call <4 x double> @llvm.x86.avx512.mask.vextractf64x4.512(<8 x double> %a, i32 1, <4 x double> zeroinitializer, i8 -1)
[AVX-512] Remove vextract intrinsics and autoupgrade to native shufflevectors. This unfortunately generates some really terrible code without VLX support due to v2i1 and v4i1 not being legal. Hopefully we can improve that in future patches. llvm-svn: 290863
2017-01-03 13:45:46 +08:00
ret <4 x double> %res
}
declare <4 x double> @llvm.x86.avx512.mask.vextractf64x4.512(<8 x double>, i32, <4 x double>, i8)
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
declare <16 x float> @llvm.x86.avx512.mask.insertf32x4.512(<16 x float>, <4 x float>, i32, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_insertf32x4_512(<16 x float> %x0, <4 x float> %x1, <16 x float> %x3, i16 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_insertf32x4_512:
; CHECK: ## BB#0:
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %xmm1<def> %xmm1<kill> %zmm1<def>
[AVX-512] Add support for detecting 512-bit shuffles that contain a 128-bit subvector insertion from the lowest subvector of one of the sources. These are best handled with a vinsert32x4 or vinsert64x2 instruction. llvm-svn: 290946
2017-01-04 15:32:03 +08:00
; CHECK-NEXT: vinsertf32x4 $1, %xmm1, %zmm0, %zmm3
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: kmovw %edi, %k1
[AVX-512] Add support for detecting 512-bit shuffles that contain a 128-bit subvector insertion from the lowest subvector of one of the sources. These are best handled with a vinsert32x4 or vinsert64x2 instruction. llvm-svn: 290946
2017-01-04 15:32:03 +08:00
; CHECK-NEXT: vinsertf32x4 $1, %xmm1, %zmm0, %zmm2 {%k1}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vaddps %zmm3, %zmm2, %zmm2
[AVX-512] Add support for detecting 512-bit shuffles that contain a 128-bit subvector insertion from the lowest subvector of one of the sources. These are best handled with a vinsert32x4 or vinsert64x2 instruction. llvm-svn: 290946
2017-01-04 15:32:03 +08:00
; CHECK-NEXT: vinsertf32x4 $1, %xmm1, %zmm0, %zmm0 {%k1} {z}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vaddps %zmm2, %zmm0, %zmm0
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: retq
%res = call <16 x float> @llvm.x86.avx512.mask.insertf32x4.512(<16 x float> %x0, <4 x float> %x1, i32 1, <16 x float> %x3, i16 %x4)
%res1 = call <16 x float> @llvm.x86.avx512.mask.insertf32x4.512(<16 x float> %x0, <4 x float> %x1, i32 1, <16 x float> %x3, i16 -1)
%res2 = call <16 x float> @llvm.x86.avx512.mask.insertf32x4.512(<16 x float> %x0, <4 x float> %x1, i32 1, <16 x float> zeroinitializer, i16 %x4)
%res3 = fadd <16 x float> %res, %res1
%res4 = fadd <16 x float> %res2, %res3
ret <16 x float> %res4
}
declare <16 x i32> @llvm.x86.avx512.mask.inserti32x4.512(<16 x i32>, <4 x i32>, i32, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_inserti32x4_512(<16 x i32> %x0, <4 x i32> %x1, <16 x i32> %x3, i16 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_inserti32x4_512:
; CHECK: ## BB#0:
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %xmm1<def> %xmm1<kill> %zmm1<def>
[AVX-512] Add support for detecting 512-bit shuffles that contain a 128-bit subvector insertion from the lowest subvector of one of the sources. These are best handled with a vinsert32x4 or vinsert64x2 instruction. llvm-svn: 290946
2017-01-04 15:32:03 +08:00
; CHECK-NEXT: vinserti32x4 $1, %xmm1, %zmm0, %zmm3
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: kmovw %edi, %k1
[AVX-512] Add support for detecting 512-bit shuffles that contain a 128-bit subvector insertion from the lowest subvector of one of the sources. These are best handled with a vinsert32x4 or vinsert64x2 instruction. llvm-svn: 290946
2017-01-04 15:32:03 +08:00
; CHECK-NEXT: vinserti32x4 $1, %xmm1, %zmm0, %zmm2 {%k1}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpaddd %zmm3, %zmm2, %zmm2
[AVX-512] Add support for detecting 512-bit shuffles that contain a 128-bit subvector insertion from the lowest subvector of one of the sources. These are best handled with a vinsert32x4 or vinsert64x2 instruction. llvm-svn: 290946
2017-01-04 15:32:03 +08:00
; CHECK-NEXT: vinserti32x4 $1, %xmm1, %zmm0, %zmm0 {%k1} {z}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpaddd %zmm2, %zmm0, %zmm0
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.inserti32x4.512(<16 x i32> %x0, <4 x i32> %x1, i32 1, <16 x i32> %x3, i16 %x4)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.inserti32x4.512(<16 x i32> %x0, <4 x i32> %x1, i32 1, <16 x i32> %x3, i16 -1)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.inserti32x4.512(<16 x i32> %x0, <4 x i32> %x1, i32 1, <16 x i32> zeroinitializer, i16 %x4)
%res3 = add <16 x i32> %res, %res1
%res4 = add <16 x i32> %res2, %res3
ret <16 x i32> %res4
}
declare <8 x double> @llvm.x86.avx512.mask.insertf64x4.512(<8 x double>, <4 x double>, i32, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_insertf64x4_512(<8 x double> %x0, <4 x double> %x1, <8 x double> %x3, i8 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_insertf64x4_512:
; CHECK: ## BB#0:
[AVX-512] Teach shuffle lowering to use vinsert instructions for shuffles corresponding to 256-bit subvector inserts. llvm-svn: 290870
2017-01-03 15:00:40 +08:00
; CHECK-NEXT: vinsertf64x4 $1, %ymm1, %zmm0, %zmm3
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: kmovw %edi, %k1
[AVX-512] Teach shuffle lowering to use vinsert instructions for shuffles corresponding to 256-bit subvector inserts. llvm-svn: 290870
2017-01-03 15:00:40 +08:00
; CHECK-NEXT: vinsertf64x4 $1, %ymm1, %zmm0, %zmm2 {%k1}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vaddpd %zmm3, %zmm2, %zmm2
[AVX-512] Teach shuffle lowering to use vinsert instructions for shuffles corresponding to 256-bit subvector inserts. llvm-svn: 290870
2017-01-03 15:00:40 +08:00
; CHECK-NEXT: vinsertf64x4 $1, %ymm1, %zmm0, %zmm0 {%k1} {z}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vaddpd %zmm2, %zmm0, %zmm0
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: retq
%res = call <8 x double> @llvm.x86.avx512.mask.insertf64x4.512(<8 x double> %x0, <4 x double> %x1, i32 1, <8 x double> %x3, i8 %x4)
%res1 = call <8 x double> @llvm.x86.avx512.mask.insertf64x4.512(<8 x double> %x0, <4 x double> %x1, i32 1, <8 x double> %x3, i8 -1)
%res2 = call <8 x double> @llvm.x86.avx512.mask.insertf64x4.512(<8 x double> %x0, <4 x double> %x1, i32 1, <8 x double> zeroinitializer, i8 %x4)
%res3 = fadd <8 x double> %res, %res1
%res4 = fadd <8 x double> %res2, %res3
ret <8 x double> %res4
}
declare <8 x i64> @llvm.x86.avx512.mask.inserti64x4.512(<8 x i64>, <4 x i64>, i32, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_inserti64x4_512(<8 x i64> %x0, <4 x i64> %x1, <8 x i64> %x3, i8 %x4) {
; CHECK-LABEL: test_int_x86_avx512_mask_inserti64x4_512:
; CHECK: ## BB#0:
[AVX-512] Teach shuffle lowering to use vinsert instructions for shuffles corresponding to 256-bit subvector inserts. llvm-svn: 290870
2017-01-03 15:00:40 +08:00
; CHECK-NEXT: vinserti64x4 $1, %ymm1, %zmm0, %zmm3
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: kmovw %edi, %k1
[AVX-512] Teach shuffle lowering to use vinsert instructions for shuffles corresponding to 256-bit subvector inserts. llvm-svn: 290870
2017-01-03 15:00:40 +08:00
; CHECK-NEXT: vinserti64x4 $1, %ymm1, %zmm0, %zmm2 {%k1}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpaddq %zmm3, %zmm2, %zmm2
[AVX-512] Teach shuffle lowering to use vinsert instructions for shuffles corresponding to 256-bit subvector inserts. llvm-svn: 290870
2017-01-03 15:00:40 +08:00
; CHECK-NEXT: vinserti64x4 $1, %ymm1, %zmm0, %zmm0 {%k1} {z}
Add LiveRangeShrink pass to shrink live range within BB. Summary: LiveRangeShrink pass moves instruction right after the definition with the same BB if the instruction and its operands all have more than one use. This pass is inexpensive and guarantees optimal live-range within BB. Reviewers: davidxl, wmi, hfinkel, MatzeB, andreadb Reviewed By: MatzeB, andreadb Subscribers: hiraditya, jyknight, sanjoy, skatkov, gberry, jholewinski, qcolombet, javed.absar, krytarowski, atrick, spatel, RKSimon, andreadb, MatzeB, mehdi_amini, mgorny, efriedma, davide, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D32563 llvm-svn: 304371
2017-06-01 07:25:25 +08:00
; CHECK-NEXT: vpaddq %zmm2, %zmm0, %zmm0
[AVX-512] Remove vinsert intrinsics and autoupgrade to native shufflevectors. There are some codegen problems here that I'll try to fix in future commits. llvm-svn: 290864
2017-01-03 13:45:57 +08:00
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.inserti64x4.512(<8 x i64> %x0, <4 x i64> %x1, i32 1, <8 x i64> %x3, i8 %x4)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.inserti64x4.512(<8 x i64> %x0, <4 x i64> %x1, i32 1, <8 x i64> %x3, i8 -1)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.inserti64x4.512(<8 x i64> %x0, <4 x i64> %x1, i32 1, <8 x i64> zeroinitializer, i8 %x4)
%res3 = add <8 x i64> %res, %res1
%res4 = add <8 x i64> %res2, %res3
ret <8 x i64> %res4
}
[X86][SSE] Update MOVNTDQA non-temporal loads to generic implementation (LLVM) MOVNTDQA non-temporal aligned vector loads can be correctly represented using generic builtin loads, allowing us to remove the existing x86 intrinsics. Clang companion patch: D31766. Differential Revision: https://reviews.llvm.org/D31767 llvm-svn: 300325
2017-04-14 23:05:35 +08:00
define <8 x i64> @test_x86_avx512_movntdqa(i8* %a0) {
; CHECK-LABEL: test_x86_avx512_movntdqa:
; CHECK: ## BB#0:
; CHECK-NEXT: vmovntdqa (%rdi), %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.movntdqa(i8* %a0)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.movntdqa(i8*) nounwind readonly
[AVX-512] Remove and autoupgrade the masked integer compare intrinsics Summary: These intrinsics aren't used by clang and haven't been for a while. There's some really terrible codegen in the 32-bit target for avx512bw due to i64 not being legal. But as I said these intrinsics aren't used by clang even before this patch so this codegen reflects our clang behavior today. Reviewers: spatel, RKSimon, zvi, igorb Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D34389 llvm-svn: 306047
2017-06-23 04:11:01 +08:00
define <8 x i16> @test_cmp_d_512(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK-LABEL: test_cmp_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpeqd %zmm1, %zmm0, %k0
; CHECK-NEXT: vpcmpgtd %zmm0, %zmm1, %k1
; CHECK-NEXT: vpcmpled %zmm1, %zmm0, %k2
; CHECK-NEXT: vpcmpneqd %zmm1, %zmm0, %k3
; CHECK-NEXT: vpcmpled %zmm0, %zmm1, %k4
; CHECK-NEXT: vpcmpgtd %zmm1, %zmm0, %k5
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpxor %xmm0, %xmm0, %xmm0
; CHECK-NEXT: vpinsrw $0, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: kxnorw %k0, %k0, %k0
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 0, i16 -1)
%vec0 = insertelement <8 x i16> undef, i16 %res0, i32 0
%res1 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 1, i16 -1)
%vec1 = insertelement <8 x i16> %vec0, i16 %res1, i32 1
%res2 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 2, i16 -1)
%vec2 = insertelement <8 x i16> %vec1, i16 %res2, i32 2
%res3 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 3, i16 -1)
%vec3 = insertelement <8 x i16> %vec2, i16 %res3, i32 3
%res4 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 4, i16 -1)
%vec4 = insertelement <8 x i16> %vec3, i16 %res4, i32 4
%res5 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 5, i16 -1)
%vec5 = insertelement <8 x i16> %vec4, i16 %res5, i32 5
%res6 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 6, i16 -1)
%vec6 = insertelement <8 x i16> %vec5, i16 %res6, i32 6
%res7 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 7, i16 -1)
%vec7 = insertelement <8 x i16> %vec6, i16 %res7, i32 7
ret <8 x i16> %vec7
}
define <8 x i16> @test_mask_cmp_d_512(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_mask_cmp_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpeqd %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: vpcmpgtd %zmm0, %zmm1, %k2 {%k1}
; CHECK-NEXT: vpcmpled %zmm1, %zmm0, %k3 {%k1}
; CHECK-NEXT: kxorw %k0, %k0, %k4
; CHECK-NEXT: vpcmpneqd %zmm1, %zmm0, %k5 {%k1}
; CHECK-NEXT: vpcmpled %zmm0, %zmm1, %k6 {%k1}
; CHECK-NEXT: vpcmpgtd %zmm1, %zmm0, %k1 {%k1}
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: kmovw %k0, %ecx
; CHECK-NEXT: vmovd %ecx, %xmm0
; CHECK-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrw $3, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k6, %eax
; CHECK-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: vpinsrw $7, %edi, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 0, i16 %mask)
%vec0 = insertelement <8 x i16> undef, i16 %res0, i32 0
%res1 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 1, i16 %mask)
%vec1 = insertelement <8 x i16> %vec0, i16 %res1, i32 1
%res2 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 2, i16 %mask)
%vec2 = insertelement <8 x i16> %vec1, i16 %res2, i32 2
%res3 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 3, i16 %mask)
%vec3 = insertelement <8 x i16> %vec2, i16 %res3, i32 3
%res4 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 4, i16 %mask)
%vec4 = insertelement <8 x i16> %vec3, i16 %res4, i32 4
%res5 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 5, i16 %mask)
%vec5 = insertelement <8 x i16> %vec4, i16 %res5, i32 5
%res6 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 6, i16 %mask)
%vec6 = insertelement <8 x i16> %vec5, i16 %res6, i32 6
%res7 = call i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 7, i16 %mask)
%vec7 = insertelement <8 x i16> %vec6, i16 %res7, i32 7
ret <8 x i16> %vec7
}
declare i16 @llvm.x86.avx512.mask.cmp.d.512(<16 x i32>, <16 x i32>, i32, i16) nounwind readnone
define <8 x i16> @test_ucmp_d_512(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK-LABEL: test_ucmp_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpeqd %zmm1, %zmm0, %k0
; CHECK-NEXT: vpcmpltud %zmm1, %zmm0, %k1
; CHECK-NEXT: vpcmpleud %zmm1, %zmm0, %k2
; CHECK-NEXT: vpcmpneqd %zmm1, %zmm0, %k3
; CHECK-NEXT: vpcmpnltud %zmm1, %zmm0, %k4
; CHECK-NEXT: vpcmpnleud %zmm1, %zmm0, %k5
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpxor %xmm0, %xmm0, %xmm0
; CHECK-NEXT: vpinsrw $0, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: kxnorw %k0, %k0, %k0
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 0, i16 -1)
%vec0 = insertelement <8 x i16> undef, i16 %res0, i32 0
%res1 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 1, i16 -1)
%vec1 = insertelement <8 x i16> %vec0, i16 %res1, i32 1
%res2 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 2, i16 -1)
%vec2 = insertelement <8 x i16> %vec1, i16 %res2, i32 2
%res3 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 3, i16 -1)
%vec3 = insertelement <8 x i16> %vec2, i16 %res3, i32 3
%res4 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 4, i16 -1)
%vec4 = insertelement <8 x i16> %vec3, i16 %res4, i32 4
%res5 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 5, i16 -1)
%vec5 = insertelement <8 x i16> %vec4, i16 %res5, i32 5
%res6 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 6, i16 -1)
%vec6 = insertelement <8 x i16> %vec5, i16 %res6, i32 6
%res7 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 7, i16 -1)
%vec7 = insertelement <8 x i16> %vec6, i16 %res7, i32 7
ret <8 x i16> %vec7
}
define <8 x i16> @test_mask_ucmp_d_512(<16 x i32> %a0, <16 x i32> %a1, i16 %mask) {
; CHECK-LABEL: test_mask_ucmp_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpeqd %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: vpcmpltud %zmm1, %zmm0, %k2 {%k1}
; CHECK-NEXT: vpcmpleud %zmm1, %zmm0, %k3 {%k1}
; CHECK-NEXT: kxorw %k0, %k0, %k4
; CHECK-NEXT: vpcmpneqd %zmm1, %zmm0, %k5 {%k1}
; CHECK-NEXT: vpcmpnltud %zmm1, %zmm0, %k6 {%k1}
; CHECK-NEXT: vpcmpnleud %zmm1, %zmm0, %k1 {%k1}
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: kmovw %k0, %ecx
; CHECK-NEXT: vmovd %ecx, %xmm0
; CHECK-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrw $3, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k6, %eax
; CHECK-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: vpinsrw $7, %edi, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 0, i16 %mask)
%vec0 = insertelement <8 x i16> undef, i16 %res0, i32 0
%res1 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 1, i16 %mask)
%vec1 = insertelement <8 x i16> %vec0, i16 %res1, i32 1
%res2 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 2, i16 %mask)
%vec2 = insertelement <8 x i16> %vec1, i16 %res2, i32 2
%res3 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 3, i16 %mask)
%vec3 = insertelement <8 x i16> %vec2, i16 %res3, i32 3
%res4 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 4, i16 %mask)
%vec4 = insertelement <8 x i16> %vec3, i16 %res4, i32 4
%res5 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 5, i16 %mask)
%vec5 = insertelement <8 x i16> %vec4, i16 %res5, i32 5
%res6 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 6, i16 %mask)
%vec6 = insertelement <8 x i16> %vec5, i16 %res6, i32 6
%res7 = call i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32> %a0, <16 x i32> %a1, i32 7, i16 %mask)
%vec7 = insertelement <8 x i16> %vec6, i16 %res7, i32 7
ret <8 x i16> %vec7
}
declare i16 @llvm.x86.avx512.mask.ucmp.d.512(<16 x i32>, <16 x i32>, i32, i16) nounwind readnone
define <8 x i8> @test_cmp_q_512(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK-LABEL: test_cmp_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpeqq %zmm1, %zmm0, %k0
; CHECK-NEXT: vpcmpgtq %zmm0, %zmm1, %k1
; CHECK-NEXT: vpcmpleq %zmm1, %zmm0, %k2
; CHECK-NEXT: vpcmpneqq %zmm1, %zmm0, %k3
; CHECK-NEXT: vpcmpleq %zmm0, %zmm1, %k4
; CHECK-NEXT: vpcmpgtq %zmm1, %zmm0, %k5
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpxor %xmm0, %xmm0, %xmm0
; CHECK-NEXT: vpinsrw $0, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: kxnorw %k0, %k0, %k0
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 0, i8 -1)
%vec0 = insertelement <8 x i8> undef, i8 %res0, i32 0
%res1 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 1, i8 -1)
%vec1 = insertelement <8 x i8> %vec0, i8 %res1, i32 1
%res2 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 2, i8 -1)
%vec2 = insertelement <8 x i8> %vec1, i8 %res2, i32 2
%res3 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 3, i8 -1)
%vec3 = insertelement <8 x i8> %vec2, i8 %res3, i32 3
%res4 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 4, i8 -1)
%vec4 = insertelement <8 x i8> %vec3, i8 %res4, i32 4
%res5 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 5, i8 -1)
%vec5 = insertelement <8 x i8> %vec4, i8 %res5, i32 5
%res6 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 6, i8 -1)
%vec6 = insertelement <8 x i8> %vec5, i8 %res6, i32 6
%res7 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 7, i8 -1)
%vec7 = insertelement <8 x i8> %vec6, i8 %res7, i32 7
ret <8 x i8> %vec7
}
define <8 x i8> @test_mask_cmp_q_512(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_mask_cmp_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpeqq %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: vpcmpgtq %zmm0, %zmm1, %k2 {%k1}
; CHECK-NEXT: vpcmpleq %zmm1, %zmm0, %k3 {%k1}
; CHECK-NEXT: kxorw %k0, %k0, %k4
; CHECK-NEXT: vpcmpneqq %zmm1, %zmm0, %k5 {%k1}
; CHECK-NEXT: vpcmpleq %zmm0, %zmm1, %k6 {%k1}
; CHECK-NEXT: vpcmpgtq %zmm1, %zmm0, %k1 {%k1}
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: kmovw %k0, %ecx
; CHECK-NEXT: vmovd %ecx, %xmm0
; CHECK-NEXT: vpinsrb $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrb $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrb $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrb $8, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k6, %eax
; CHECK-NEXT: vpinsrb $10, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrb $12, %eax, %xmm0, %xmm0
; CHECK-NEXT: vpinsrb $14, %edi, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 0, i8 %mask)
%vec0 = insertelement <8 x i8> undef, i8 %res0, i32 0
%res1 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 1, i8 %mask)
%vec1 = insertelement <8 x i8> %vec0, i8 %res1, i32 1
%res2 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 2, i8 %mask)
%vec2 = insertelement <8 x i8> %vec1, i8 %res2, i32 2
%res3 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 3, i8 %mask)
%vec3 = insertelement <8 x i8> %vec2, i8 %res3, i32 3
%res4 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 4, i8 %mask)
%vec4 = insertelement <8 x i8> %vec3, i8 %res4, i32 4
%res5 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 5, i8 %mask)
%vec5 = insertelement <8 x i8> %vec4, i8 %res5, i32 5
%res6 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 6, i8 %mask)
%vec6 = insertelement <8 x i8> %vec5, i8 %res6, i32 6
%res7 = call i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 7, i8 %mask)
%vec7 = insertelement <8 x i8> %vec6, i8 %res7, i32 7
ret <8 x i8> %vec7
}
declare i8 @llvm.x86.avx512.mask.cmp.q.512(<8 x i64>, <8 x i64>, i32, i8) nounwind readnone
define <8 x i8> @test_ucmp_q_512(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK-LABEL: test_ucmp_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpcmpeqq %zmm1, %zmm0, %k0
; CHECK-NEXT: vpcmpltuq %zmm1, %zmm0, %k1
; CHECK-NEXT: vpcmpleuq %zmm1, %zmm0, %k2
; CHECK-NEXT: vpcmpneqq %zmm1, %zmm0, %k3
; CHECK-NEXT: vpcmpnltuq %zmm1, %zmm0, %k4
; CHECK-NEXT: vpcmpnleuq %zmm1, %zmm0, %k5
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpxor %xmm0, %xmm0, %xmm0
; CHECK-NEXT: vpinsrw $0, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: kxnorw %k0, %k0, %k0
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 0, i8 -1)
%vec0 = insertelement <8 x i8> undef, i8 %res0, i32 0
%res1 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 1, i8 -1)
%vec1 = insertelement <8 x i8> %vec0, i8 %res1, i32 1
%res2 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 2, i8 -1)
%vec2 = insertelement <8 x i8> %vec1, i8 %res2, i32 2
%res3 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 3, i8 -1)
%vec3 = insertelement <8 x i8> %vec2, i8 %res3, i32 3
%res4 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 4, i8 -1)
%vec4 = insertelement <8 x i8> %vec3, i8 %res4, i32 4
%res5 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 5, i8 -1)
%vec5 = insertelement <8 x i8> %vec4, i8 %res5, i32 5
%res6 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 6, i8 -1)
%vec6 = insertelement <8 x i8> %vec5, i8 %res6, i32 6
%res7 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 7, i8 -1)
%vec7 = insertelement <8 x i8> %vec6, i8 %res7, i32 7
ret <8 x i8> %vec7
}
define <8 x i8> @test_mask_ucmp_q_512(<8 x i64> %a0, <8 x i64> %a1, i8 %mask) {
; CHECK-LABEL: test_mask_ucmp_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpcmpeqq %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: vpcmpltuq %zmm1, %zmm0, %k2 {%k1}
; CHECK-NEXT: vpcmpleuq %zmm1, %zmm0, %k3 {%k1}
; CHECK-NEXT: kxorw %k0, %k0, %k4
; CHECK-NEXT: vpcmpneqq %zmm1, %zmm0, %k5 {%k1}
; CHECK-NEXT: vpcmpnltuq %zmm1, %zmm0, %k6 {%k1}
; CHECK-NEXT: vpcmpnleuq %zmm1, %zmm0, %k1 {%k1}
; CHECK-NEXT: kmovw %k2, %eax
; CHECK-NEXT: kmovw %k0, %ecx
; CHECK-NEXT: vmovd %ecx, %xmm0
; CHECK-NEXT: vpinsrb $2, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k3, %eax
; CHECK-NEXT: vpinsrb $4, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k4, %eax
; CHECK-NEXT: vpinsrb $6, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k5, %eax
; CHECK-NEXT: vpinsrb $8, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k6, %eax
; CHECK-NEXT: vpinsrb $10, %eax, %xmm0, %xmm0
; CHECK-NEXT: kmovw %k1, %eax
; CHECK-NEXT: vpinsrb $12, %eax, %xmm0, %xmm0
; CHECK-NEXT: vpinsrb $14, %edi, %xmm0, %xmm0
; CHECK-NEXT: retq
%res0 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 0, i8 %mask)
%vec0 = insertelement <8 x i8> undef, i8 %res0, i32 0
%res1 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 1, i8 %mask)
%vec1 = insertelement <8 x i8> %vec0, i8 %res1, i32 1
%res2 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 2, i8 %mask)
%vec2 = insertelement <8 x i8> %vec1, i8 %res2, i32 2
%res3 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 3, i8 %mask)
%vec3 = insertelement <8 x i8> %vec2, i8 %res3, i32 3
%res4 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 4, i8 %mask)
%vec4 = insertelement <8 x i8> %vec3, i8 %res4, i32 4
%res5 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 5, i8 %mask)
%vec5 = insertelement <8 x i8> %vec4, i8 %res5, i32 5
%res6 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 6, i8 %mask)
%vec6 = insertelement <8 x i8> %vec5, i8 %res6, i32 6
%res7 = call i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64> %a0, <8 x i64> %a1, i32 7, i8 %mask)
%vec7 = insertelement <8 x i8> %vec6, i8 %res7, i32 7
ret <8 x i8> %vec7
}
declare i8 @llvm.x86.avx512.mask.ucmp.q.512(<8 x i64>, <8 x i64>, i32, i8) nounwind readnone
[AVX512] Remove and autoupgrade many of the broadcast intrinsics Summary: This autoupgrades most of the broadcast intrinsics. They've been unused in clang for some time. This leaves the 32x2 intrinsics because they are still used in clang. Reviewers: RKSimon, zvi, igorb Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D36606 llvm-svn: 310725
2017-08-12 00:22:45 +08:00
declare <16 x float> @llvm.x86.avx512.mask.broadcastf32x4.512(<4 x float>, <16 x float>, i16)
define <16 x float>@test_int_x86_avx512_mask_broadcastf32x4_512(<4 x float> %x0, <16 x float> %x2, i16 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcastf32x4_512:
; CHECK: ## BB#0:
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %xmm0<def> %xmm0<kill> %ymm0<def>
[AVX512] Remove and autoupgrade many of the broadcast intrinsics Summary: This autoupgrades most of the broadcast intrinsics. They've been unused in clang for some time. This leaves the 32x2 intrinsics because they are still used in clang. Reviewers: RKSimon, zvi, igorb Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D36606 llvm-svn: 310725
2017-08-12 00:22:45 +08:00
; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
; CHECK-NEXT: vinsertf64x4 $1, %ymm0, %zmm0, %zmm0
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vmovaps %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovaps %zmm0, %zmm2 {%k1} {z}
; CHECK-NEXT: vaddps %zmm1, %zmm0, %zmm0
; CHECK-NEXT: vaddps %zmm0, %zmm2, %zmm0
; CHECK-NEXT: retq
%res1 = call <16 x float> @llvm.x86.avx512.mask.broadcastf32x4.512(<4 x float> %x0, <16 x float> %x2, i16 -1)
%res2 = call <16 x float> @llvm.x86.avx512.mask.broadcastf32x4.512(<4 x float> %x0, <16 x float> %x2, i16 %mask)
%res3 = call <16 x float> @llvm.x86.avx512.mask.broadcastf32x4.512(<4 x float> %x0, <16 x float> zeroinitializer, i16 %mask)
%res4 = fadd <16 x float> %res1, %res2
%res5 = fadd <16 x float> %res3, %res4
ret <16 x float> %res5
}
define <16 x float>@test_int_x86_avx512_mask_broadcastf32x4_512_load(<4 x float>* %x0ptr, <16 x float> %x2, i16 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcastf32x4_512_load:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vbroadcastf32x4 {{.*#+}} zmm0 {%k1} = mem[0,1,2,3,0,1,2,3,0,1,2,3,0,1,2,3]
; CHECK-NEXT: retq
%x0 = load <4 x float>, <4 x float>* %x0ptr
%res = call <16 x float> @llvm.x86.avx512.mask.broadcastf32x4.512(<4 x float> %x0, <16 x float> %x2, i16 %mask)
ret <16 x float> %res
}
declare <8 x double> @llvm.x86.avx512.mask.broadcastf64x4.512(<4 x double>, <8 x double>, i8)
define <8 x double>@test_int_x86_avx512_mask_broadcastf64x4_512(<4 x double> %x0, <8 x double> %x2, i8 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcastf64x4_512:
; CHECK: ## BB#0:
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ymm0<def> %ymm0<kill> %zmm0<def>
[AVX512] Remove and autoupgrade many of the broadcast intrinsics Summary: This autoupgrades most of the broadcast intrinsics. They've been unused in clang for some time. This leaves the 32x2 intrinsics because they are still used in clang. Reviewers: RKSimon, zvi, igorb Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D36606 llvm-svn: 310725
2017-08-12 00:22:45 +08:00
; CHECK-NEXT: vinsertf64x4 $1, %ymm0, %zmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vinsertf64x4 $1, %ymm0, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vaddpd %zmm1, %zmm2, %zmm1
; CHECK-NEXT: vinsertf64x4 $1, %ymm0, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vaddpd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res1 = call <8 x double> @llvm.x86.avx512.mask.broadcastf64x4.512(<4 x double> %x0, <8 x double> %x2, i8 -1)
%res2 = call <8 x double> @llvm.x86.avx512.mask.broadcastf64x4.512(<4 x double> %x0, <8 x double> %x2, i8 %mask)
%res3 = call <8 x double> @llvm.x86.avx512.mask.broadcastf64x4.512(<4 x double> %x0, <8 x double> zeroinitializer, i8 %mask)
%res4 = fadd <8 x double> %res1, %res2
%res5 = fadd <8 x double> %res3, %res4
ret <8 x double> %res5
}
define <8 x double>@test_int_x86_avx512_mask_broadcastf64x4_512_load(<4 x double>* %x0ptr, <8 x double> %x2, i8 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcastf64x4_512_load:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vbroadcastf64x4 {{.*#+}} zmm0 {%k1} = mem[0,1,2,3,0,1,2,3]
; CHECK-NEXT: retq
%x0 = load <4 x double>, <4 x double>* %x0ptr
%res = call <8 x double> @llvm.x86.avx512.mask.broadcastf64x4.512(<4 x double> %x0, <8 x double> %x2, i8 %mask)
ret <8 x double> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.broadcasti32x4.512(<4 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_broadcasti32x4_512(<4 x i32> %x0, <16 x i32> %x2, i16 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcasti32x4_512:
; CHECK: ## BB#0:
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %xmm0<def> %xmm0<kill> %ymm0<def>
[AVX512] Remove and autoupgrade many of the broadcast intrinsics Summary: This autoupgrades most of the broadcast intrinsics. They've been unused in clang for some time. This leaves the 32x2 intrinsics because they are still used in clang. Reviewers: RKSimon, zvi, igorb Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D36606 llvm-svn: 310725
2017-08-12 00:22:45 +08:00
; CHECK-NEXT: vinserti128 $1, %xmm0, %ymm0, %ymm0
; CHECK-NEXT: vinserti64x4 $1, %ymm0, %zmm0, %zmm0
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vmovdqa32 %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vmovdqa32 %zmm0, %zmm2 {%k1} {z}
; CHECK-NEXT: vpaddd %zmm1, %zmm0, %zmm0
; CHECK-NEXT: vpaddd %zmm0, %zmm2, %zmm0
; CHECK-NEXT: retq
%res1 = call <16 x i32> @llvm.x86.avx512.mask.broadcasti32x4.512(<4 x i32> %x0, <16 x i32> %x2, i16 -1)
%res2 = call <16 x i32> @llvm.x86.avx512.mask.broadcasti32x4.512(<4 x i32> %x0, <16 x i32> %x2, i16 %mask)
%res3 = call <16 x i32> @llvm.x86.avx512.mask.broadcasti32x4.512(<4 x i32> %x0, <16 x i32> zeroinitializer, i16 %mask)
%res4 = add <16 x i32> %res1, %res2
%res5 = add <16 x i32> %res3, %res4
ret <16 x i32> %res5
}
define <16 x i32>@test_int_x86_avx512_mask_broadcasti32x4_512_load(<4 x i32>* %x0ptr, <16 x i32> %x2, i16 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcasti32x4_512_load:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vbroadcasti32x4 {{.*#+}} zmm0 {%k1} = mem[0,1,2,3,0,1,2,3,0,1,2,3,0,1,2,3]
; CHECK-NEXT: retq
%x0 = load <4 x i32>, <4 x i32>* %x0ptr
%res = call <16 x i32> @llvm.x86.avx512.mask.broadcasti32x4.512(<4 x i32> %x0, <16 x i32> %x2, i16 %mask)
ret <16 x i32> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.broadcasti64x4.512(<4 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_broadcasti64x4_512(<4 x i64> %x0, <8 x i64> %x2, i8 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcasti64x4_512:
; CHECK: ## BB#0:
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ymm0<def> %ymm0<kill> %zmm0<def>
[AVX512] Remove and autoupgrade many of the broadcast intrinsics Summary: This autoupgrades most of the broadcast intrinsics. They've been unused in clang for some time. This leaves the 32x2 intrinsics because they are still used in clang. Reviewers: RKSimon, zvi, igorb Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D36606 llvm-svn: 310725
2017-08-12 00:22:45 +08:00
; CHECK-NEXT: vinserti64x4 $1, %ymm0, %zmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vinserti64x4 $1, %ymm0, %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vpaddq %zmm1, %zmm2, %zmm1
; CHECK-NEXT: vinserti64x4 $1, %ymm0, %zmm0, %zmm0 {%k1} {z}
; CHECK-NEXT: vpaddq %zmm1, %zmm0, %zmm0
; CHECK-NEXT: retq
%res1 = call <8 x i64> @llvm.x86.avx512.mask.broadcasti64x4.512(<4 x i64> %x0, <8 x i64> %x2, i8 -1)
%res2 = call <8 x i64> @llvm.x86.avx512.mask.broadcasti64x4.512(<4 x i64> %x0, <8 x i64> %x2, i8 %mask)
%res3 = call <8 x i64> @llvm.x86.avx512.mask.broadcasti64x4.512(<4 x i64> %x0, <8 x i64> zeroinitializer, i8 %mask)
%res4 = add <8 x i64> %res1, %res2
%res5 = add <8 x i64> %res3, %res4
ret <8 x i64> %res5
}
define <8 x i64>@test_int_x86_avx512_mask_broadcasti64x4_512_load(<4 x i64>* %x0ptr, <8 x i64> %x2, i8 %mask) {
; CHECK-LABEL: test_int_x86_avx512_mask_broadcasti64x4_512_load:
; CHECK: ## BB#0:
; CHECK-NEXT: kmovw %esi, %k1
; CHECK-NEXT: vbroadcasti64x4 {{.*#+}} zmm0 {%k1} = mem[0,1,2,3,0,1,2,3]
; CHECK-NEXT: retq
%x0 = load <4 x i64>, <4 x i64>* %x0ptr
%res = call <8 x i64> @llvm.x86.avx512.mask.broadcasti64x4.512(<4 x i64> %x0, <8 x i64> %x2, i8 %mask)
ret <8 x i64> %res
}
[X86] [PATCH] [intrinsics] Lowering X86 ABS intrinsics to IR. (llvm) This patch, together with a matching clang patch (https://reviews.llvm.org/D37694), implements the lowering of X86 ABS intrinsics to IR. differential revision: https://reviews.llvm.org/D37693. llvm-svn: 313134
2017-09-13 17:02:36 +08:00
declare <16 x i32> @llvm.x86.avx512.mask.pabs.d.512(<16 x i32>, <16 x i32>, i16)
define <16 x i32>@test_int_x86_avx512_mask_pabs_d_512(<16 x i32> %x0, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pabs_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpabsd %zmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpabsd %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vpaddd %zmm2, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <16 x i32> @llvm.x86.avx512.mask.pabs.d.512(<16 x i32> %x0, <16 x i32> %x1, i16 %x2)
%res1 = call <16 x i32> @llvm.x86.avx512.mask.pabs.d.512(<16 x i32> %x0, <16 x i32> %x1, i16 -1)
%res2 = add <16 x i32> %res, %res1
ret <16 x i32> %res2
}
declare <8 x i64> @llvm.x86.avx512.mask.pabs.q.512(<8 x i64>, <8 x i64>, i8)
define <8 x i64>@test_int_x86_avx512_mask_pabs_q_512(<8 x i64> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_mask_pabs_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vpabsq %zmm0, %zmm2
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vpabsq %zmm0, %zmm1 {%k1}
; CHECK-NEXT: vpaddq %zmm2, %zmm1, %zmm0
; CHECK-NEXT: retq
%res = call <8 x i64> @llvm.x86.avx512.mask.pabs.q.512(<8 x i64> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call <8 x i64> @llvm.x86.avx512.mask.pabs.q.512(<8 x i64> %x0, <8 x i64> %x1, i8 -1)
%res2 = add <8 x i64> %res, %res1
ret <8 x i64> %res2
}
[X86] test/testn intrinsics lowering to IR. llvm part. Remove builtins from llvm and add AutoUpgrade support. Also add fast-isel tests for the TEST and TESTN instructions. Differential Revision: https://reviews.llvm.org/D38736 llvm-svn: 318036
2017-11-13 20:51:18 +08:00
define i8 @test_vptestmq(<8 x i64> %a0, <8 x i64> %a1, i8 %m) {
; CHECK-LABEL: test_vptestmq:
; CHECK: ## BB#0:
; CHECK-NEXT: vptestmq %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %k0, %ecx
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vptestmq %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: addb %cl, %al
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %al<def> %al<kill> %eax<kill>
[X86] test/testn intrinsics lowering to IR. llvm part. Remove builtins from llvm and add AutoUpgrade support. Also add fast-isel tests for the TEST and TESTN instructions. Differential Revision: https://reviews.llvm.org/D38736 llvm-svn: 318036
2017-11-13 20:51:18 +08:00
; CHECK-NEXT: retq
%res = call i8 @llvm.x86.avx512.ptestm.q.512(<8 x i64> %a0, <8 x i64> %a1, i8 -1)
%res1 = call i8 @llvm.x86.avx512.ptestm.q.512(<8 x i64> %a0, <8 x i64> %a1, i8 %m)
%res2 = add i8 %res1, %res
ret i8 %res2
}
declare i8 @llvm.x86.avx512.ptestm.q.512(<8 x i64>, <8 x i64>, i8)
define i16 @test_vptestmd(<16 x i32> %a0, <16 x i32> %a1, i16 %m) {
; CHECK-LABEL: test_vptestmd:
; CHECK: ## BB#0:
; CHECK-NEXT: vptestmd %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %k0, %ecx
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vptestmd %zmm1, %zmm0, %k0 {%k1}
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: addl %ecx, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ax<def> %ax<kill> %eax<kill>
[X86] test/testn intrinsics lowering to IR. llvm part. Remove builtins from llvm and add AutoUpgrade support. Also add fast-isel tests for the TEST and TESTN instructions. Differential Revision: https://reviews.llvm.org/D38736 llvm-svn: 318036
2017-11-13 20:51:18 +08:00
; CHECK-NEXT: retq
%res = call i16 @llvm.x86.avx512.ptestm.d.512(<16 x i32> %a0, <16 x i32> %a1, i16 -1)
%res1 = call i16 @llvm.x86.avx512.ptestm.d.512(<16 x i32> %a0, <16 x i32> %a1, i16 %m)
%res2 = add i16 %res1, %res
ret i16 %res2
}
declare i16 @llvm.x86.avx512.ptestm.d.512(<16 x i32>, <16 x i32>, i16)
declare i16 @llvm.x86.avx512.ptestnm.d.512(<16 x i32>, <16 x i32>, i16 %x2)
define i16@test_int_x86_avx512_ptestnm_d_512(<16 x i32> %x0, <16 x i32> %x1, i16 %x2) {
; CHECK-LABEL: test_int_x86_avx512_ptestnm_d_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vptestnmd %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vptestnmd %zmm1, %zmm0, %k1 {%k1}
; CHECK-NEXT: kmovw %k1, %ecx
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: addl %ecx, %eax
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %ax<def> %ax<kill> %eax<kill>
[X86] test/testn intrinsics lowering to IR. llvm part. Remove builtins from llvm and add AutoUpgrade support. Also add fast-isel tests for the TEST and TESTN instructions. Differential Revision: https://reviews.llvm.org/D38736 llvm-svn: 318036
2017-11-13 20:51:18 +08:00
; CHECK-NEXT: retq
%res = call i16 @llvm.x86.avx512.ptestnm.d.512(<16 x i32> %x0, <16 x i32> %x1, i16 %x2)
%res1 = call i16 @llvm.x86.avx512.ptestnm.d.512(<16 x i32> %x0, <16 x i32> %x1, i16-1)
%res2 = add i16 %res, %res1
ret i16 %res2
}
declare i8 @llvm.x86.avx512.ptestnm.q.512(<8 x i64>, <8 x i64>, i8 %x2)
define i8@test_int_x86_avx512_ptestnm_q_512(<8 x i64> %x0, <8 x i64> %x1, i8 %x2) {
; CHECK-LABEL: test_int_x86_avx512_ptestnm_q_512:
; CHECK: ## BB#0:
; CHECK-NEXT: vptestnmq %zmm1, %zmm0, %k0
; CHECK-NEXT: kmovw %edi, %k1
; CHECK-NEXT: vptestnmq %zmm1, %zmm0, %k1 {%k1}
; CHECK-NEXT: kmovw %k1, %ecx
; CHECK-NEXT: kmovw %k0, %eax
; CHECK-NEXT: addb %cl, %al
[CodeGen] Print register names in lowercase in both MIR and debug output As part of the unification of the debug format and the MIR format, always print registers as lowercase. * Only debug printing is affected. It now follows MIR. Differential Revision: https://reviews.llvm.org/D40417 llvm-svn: 319187
2017-11-29 01:15:09 +08:00
; CHECK-NEXT: ## kill: %al<def> %al<kill> %eax<kill>
[X86] test/testn intrinsics lowering to IR. llvm part. Remove builtins from llvm and add AutoUpgrade support. Also add fast-isel tests for the TEST and TESTN instructions. Differential Revision: https://reviews.llvm.org/D38736 llvm-svn: 318036
2017-11-13 20:51:18 +08:00
; CHECK-NEXT: retq
%res = call i8 @llvm.x86.avx512.ptestnm.q.512(<8 x i64> %x0, <8 x i64> %x1, i8 %x2)
%res1 = call i8 @llvm.x86.avx512.ptestnm.q.512(<8 x i64> %x0, <8 x i64> %x1, i8-1)
%res2 = add i8 %res, %res1
ret i8 %res2
}