[llvm] 0662791 - [X86] vector-interleaved tests - add AVX512-SLOW/AVX512-FAST common prefixes to reduce duplication

Wed Sep 20 04:31:09 PDT 2023

Author: Simon Pilgrim
Date: 2023-09-20T12:17:51+01:00
New Revision: 0662791a1371c8ebc31817e9018d6afedc0195c5

URL: https://github.com/llvm/llvm-project/commit/0662791a1371c8ebc31817e9018d6afedc0195c5
DIFF: https://github.com/llvm/llvm-project/commit/0662791a1371c8ebc31817e9018d6afedc0195c5.diff

LOG: [X86] vector-interleaved tests - add AVX512-SLOW/AVX512-FAST common prefixes to reduce duplication

These aren't always used but its lot more manageable to keep the vector-interleaved files using the same RUN lines wherever possible

Added: 
    

Modified: 
    llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-8.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-8.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-8.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-8.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-8.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-8.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-8.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-2.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-3.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-4.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-5.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-6.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-7.ll
    llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-8.ll

Removed: 
    


################################################################################
diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-2.ll
index fcc68f09c1d20ea..900847c8a191fe1 100644

--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -965,6 +965,8 @@ define void @load_i16_stride2_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1) no
 ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-3.ll
index 2fdc04bd73455cf..1b47c0f3193d3a9 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -77,53 +77,29 @@ define void @load_i16_stride3_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST-PERLANE-NEXT:    vmovd %xmm0, (%rcx)
 ; AVX2-FAST-PERLANE-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i16_stride3_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm0[0,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,2,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[1,2,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,1,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[2,1,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vmovd %xmm1, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovd %xmm2, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm0, (%rcx)
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i16_stride3_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm0[0,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,2,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[1,2,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,1,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[2,1,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vmovd %xmm1, (%rsi)
+; AVX512-SLOW-NEXT:    vmovd %xmm2, (%rdx)
+; AVX512-SLOW-NEXT:    vmovd %xmm0, (%rcx)
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i16_stride3_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm0[0,3,2,3,4,5,6,7]
-; AVX512F-FAST-NEXT:    vpshufb {{.*#+}} xmm2 = xmm0[2,3,8,9,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512F-FAST-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[4,5,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512F-FAST-NEXT:    vmovd %xmm1, (%rsi)
-; AVX512F-FAST-NEXT:    vmovd %xmm2, (%rdx)
-; AVX512F-FAST-NEXT:    vmovd %xmm0, (%rcx)
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i16_stride3_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm0[0,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,2,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[1,2,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,1,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[2,1,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vmovd %xmm1, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm2, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm0, (%rcx)
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i16_stride3_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm0[0,3,2,3,4,5,6,7]
-; AVX512BW-FAST-NEXT:    vpshufb {{.*#+}} xmm2 = xmm0[2,3,8,9,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512BW-FAST-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[4,5,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512BW-FAST-NEXT:    vmovd %xmm1, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovd %xmm2, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovd %xmm0, (%rcx)
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i16_stride3_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm0[0,3,2,3,4,5,6,7]
+; AVX512-FAST-NEXT:    vpshufb {{.*#+}} xmm2 = xmm0[2,3,8,9,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX512-FAST-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[4,5,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX512-FAST-NEXT:    vmovd %xmm1, (%rsi)
+; AVX512-FAST-NEXT:    vmovd %xmm2, (%rdx)
+; AVX512-FAST-NEXT:    vmovd %xmm0, (%rcx)
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <6 x i16>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <6 x i16> %wide.vec, <6 x i16> poison, <2 x i32> <i32 0, i32 3>
   %strided.vec1 = shufflevector <6 x i16> %wide.vec, <6 x i16> poison, <2 x i32> <i32 1, i32 4>

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-4.ll
index 6670de17c6cde00..5a7abf8cdf9fbfe 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -46,59 +46,32 @@ define void @load_i16_stride4_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX1-NEXT:    vmovd %xmm0, (%r8)
 ; AVX1-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i16_stride4_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[0,2,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[1,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[3,1,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[2,0,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[3,1,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpmovqw %xmm0, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovd %xmm1, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm3, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm2, (%r8)
-; AVX512F-SLOW-NEXT:    retq
-;
-; AVX512F-FAST-LABEL: load_i16_stride4_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[2,3,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512F-FAST-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[3,1,2,3]
-; AVX512F-FAST-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[2,0,2,3,4,5,6,7]
-; AVX512F-FAST-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[3,1,2,3,4,5,6,7]
-; AVX512F-FAST-NEXT:    vpmovqw %xmm0, (%rsi)
-; AVX512F-FAST-NEXT:    vmovd %xmm1, (%rdx)
-; AVX512F-FAST-NEXT:    vmovd %xmm3, (%rcx)
-; AVX512F-FAST-NEXT:    vmovd %xmm2, (%r8)
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i16_stride4_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[0,2,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[1,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[3,1,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[2,0,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[3,1,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpmovqw %xmm0, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm1, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm3, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm2, (%r8)
-; AVX512BW-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i16_stride4_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[0,2,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[1,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[3,1,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[2,0,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[3,1,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpmovqw %xmm0, (%rsi)
+; AVX512-SLOW-NEXT:    vmovd %xmm1, (%rdx)
+; AVX512-SLOW-NEXT:    vmovd %xmm3, (%rcx)
+; AVX512-SLOW-NEXT:    vmovd %xmm2, (%r8)
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512BW-FAST-LABEL: load_i16_stride4_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[2,3,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512BW-FAST-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[3,1,2,3]
-; AVX512BW-FAST-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[2,0,2,3,4,5,6,7]
-; AVX512BW-FAST-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[3,1,2,3,4,5,6,7]
-; AVX512BW-FAST-NEXT:    vpmovqw %xmm0, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovd %xmm1, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovd %xmm3, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovd %xmm2, (%r8)
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i16_stride4_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[2,3,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX512-FAST-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[3,1,2,3]
+; AVX512-FAST-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[2,0,2,3,4,5,6,7]
+; AVX512-FAST-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[3,1,2,3,4,5,6,7]
+; AVX512-FAST-NEXT:    vpmovqw %xmm0, (%rsi)
+; AVX512-FAST-NEXT:    vmovd %xmm1, (%rdx)
+; AVX512-FAST-NEXT:    vmovd %xmm3, (%rcx)
+; AVX512-FAST-NEXT:    vmovd %xmm2, (%r8)
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <8 x i16>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <8 x i16> %wide.vec, <8 x i16> poison, <2 x i32> <i32 0, i32 4>
   %strided.vec1 = shufflevector <8 x i16> %wide.vec, <8 x i16> poison, <2 x i32> <i32 1, i32 5>
@@ -5451,8 +5424,10 @@ define void @load_i16_stride4_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX: {{.*}}
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-5.ll
index 30ee74b33575224..8cfd55b3044f569 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -116,83 +116,44 @@ define void @load_i16_stride5_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST-PERLANE-NEXT:    vmovd %xmm1, (%r9)
 ; AVX2-FAST-PERLANE-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i16_stride5_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,2,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[0,3,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm3[1,2,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm4 = xmm0[3,1,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm4 = xmm4[2,1,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpsrlq $48, %xmm0, %xmm0
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm5
-; AVX512F-SLOW-NEXT:    vpblendw {{.*#+}} xmm1 = xmm5[0],xmm1[1],xmm5[2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vmovd %xmm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovd %xmm3, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm4, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm0, (%r8)
-; AVX512F-SLOW-NEXT:    vmovd %xmm1, (%r9)
-; AVX512F-SLOW-NEXT:    retq
-;
-; AVX512F-FAST-LABEL: load_i16_stride5_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vpshufb {{.*#+}} xmm2 = xmm0[0,1,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512F-FAST-NEXT:    vpshufb {{.*#+}} xmm3 = xmm0[2,3,12,13,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512F-FAST-NEXT:    vpshufb {{.*#+}} xmm4 = xmm0[4,5,14,15,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512F-FAST-NEXT:    vpsrlq $48, %xmm0, %xmm0
-; AVX512F-FAST-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-FAST-NEXT:    vpbroadcastw 8(%rdi), %xmm5
-; AVX512F-FAST-NEXT:    vpblendw {{.*#+}} xmm1 = xmm5[0],xmm1[1],xmm5[2,3,4,5,6,7]
-; AVX512F-FAST-NEXT:    vmovd %xmm2, (%rsi)
-; AVX512F-FAST-NEXT:    vmovd %xmm3, (%rdx)
-; AVX512F-FAST-NEXT:    vmovd %xmm4, (%rcx)
-; AVX512F-FAST-NEXT:    vmovd %xmm0, (%r8)
-; AVX512F-FAST-NEXT:    vmovd %xmm1, (%r9)
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i16_stride5_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,2,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[0,3,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm3[1,2,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm4 = xmm0[3,1,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm4 = xmm4[2,1,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpsrlq $48, %xmm0, %xmm0
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm5
-; AVX512BW-SLOW-NEXT:    vpblendw {{.*#+}} xmm1 = xmm5[0],xmm1[1],xmm5[2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vmovd %xmm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm3, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm4, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm0, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm1, (%r9)
-; AVX512BW-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i16_stride5_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,2,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[0,3,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm3[1,2,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm4 = xmm0[3,1,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm4 = xmm4[2,1,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpsrlq $48, %xmm0, %xmm0
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm5
+; AVX512-SLOW-NEXT:    vpblendw {{.*#+}} xmm1 = xmm5[0],xmm1[1],xmm5[2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vmovd %xmm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovd %xmm3, (%rdx)
+; AVX512-SLOW-NEXT:    vmovd %xmm4, (%rcx)
+; AVX512-SLOW-NEXT:    vmovd %xmm0, (%r8)
+; AVX512-SLOW-NEXT:    vmovd %xmm1, (%r9)
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512BW-FAST-LABEL: load_i16_stride5_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-FAST-NEXT:    vpshufb {{.*#+}} xmm2 = xmm0[0,1,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512BW-FAST-NEXT:    vpshufb {{.*#+}} xmm3 = xmm0[2,3,12,13,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512BW-FAST-NEXT:    vpshufb {{.*#+}} xmm4 = xmm0[4,5,14,15,u,u,u,u,u,u,u,u,u,u,u,u]
-; AVX512BW-FAST-NEXT:    vpsrlq $48, %xmm0, %xmm0
-; AVX512BW-FAST-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-FAST-NEXT:    vpbroadcastw 8(%rdi), %xmm5
-; AVX512BW-FAST-NEXT:    vpblendw {{.*#+}} xmm1 = xmm5[0],xmm1[1],xmm5[2,3,4,5,6,7]
-; AVX512BW-FAST-NEXT:    vmovd %xmm2, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovd %xmm3, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovd %xmm4, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovd %xmm0, (%r8)
-; AVX512BW-FAST-NEXT:    vmovd %xmm1, (%r9)
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i16_stride5_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-FAST-NEXT:    vpshufb {{.*#+}} xmm2 = xmm0[0,1,10,11,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX512-FAST-NEXT:    vpshufb {{.*#+}} xmm3 = xmm0[2,3,12,13,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX512-FAST-NEXT:    vpshufb {{.*#+}} xmm4 = xmm0[4,5,14,15,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX512-FAST-NEXT:    vpsrlq $48, %xmm0, %xmm0
+; AVX512-FAST-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-FAST-NEXT:    vpbroadcastw 8(%rdi), %xmm5
+; AVX512-FAST-NEXT:    vpblendw {{.*#+}} xmm1 = xmm5[0],xmm1[1],xmm5[2,3,4,5,6,7]
+; AVX512-FAST-NEXT:    vmovd %xmm2, (%rsi)
+; AVX512-FAST-NEXT:    vmovd %xmm3, (%rdx)
+; AVX512-FAST-NEXT:    vmovd %xmm4, (%rcx)
+; AVX512-FAST-NEXT:    vmovd %xmm0, (%r8)
+; AVX512-FAST-NEXT:    vmovd %xmm1, (%r9)
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <10 x i16>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <10 x i16> %wide.vec, <10 x i16> poison, <2 x i32> <i32 0, i32 5>
   %strided.vec1 = shufflevector <10 x i16> %wide.vec, <10 x i16> poison, <2 x i32> <i32 1, i32 6>
@@ -7766,8 +7727,10 @@ define void @load_i16_stride5_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
 ; AVX512: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-6.ll
index 57fab33b13fa35c..3decaa366af785f 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -150,32 +150,32 @@ define void @load_i16_stride6_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST-PERLANE-NEXT:    vmovd %xmm0, (%rax)
 ; AVX2-FAST-PERLANE-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i16_stride6_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[0,2,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[1,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpbroadcastw 4(%rdi), %xmm4
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm4 = xmm4[0],xmm1[0],xmm4[1],xmm1[1],xmm4[2],xmm1[2],xmm4[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm5 = xmm5[3,1,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm5 = xmm5[0,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpbroadcastw 20(%rdi), %xmm6
-; AVX512F-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm7
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm6 = xmm7[0],xmm6[0],xmm7[1],xmm6[1],xmm7[2],xmm6[2],xmm7[3],xmm6[3]
-; AVX512F-SLOW-NEXT:    vpsrlq $48, %xmm1, %xmm1
-; AVX512F-SLOW-NEXT:    vpsrldq {{.*#+}} xmm0 = xmm0[10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vmovd %xmm3, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovd %xmm2, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm4, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm5, (%r8)
-; AVX512F-SLOW-NEXT:    vmovd %xmm6, (%r9)
-; AVX512F-SLOW-NEXT:    vmovd %xmm0, (%rax)
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i16_stride6_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[0,2,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[1,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpbroadcastw 4(%rdi), %xmm4
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm4 = xmm4[0],xmm1[0],xmm4[1],xmm1[1],xmm4[2],xmm1[2],xmm4[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm5 = xmm5[3,1,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm5 = xmm5[0,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpbroadcastw 20(%rdi), %xmm6
+; AVX512-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm7
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm6 = xmm7[0],xmm6[0],xmm7[1],xmm6[1],xmm7[2],xmm6[2],xmm7[3],xmm6[3]
+; AVX512-SLOW-NEXT:    vpsrlq $48, %xmm1, %xmm1
+; AVX512-SLOW-NEXT:    vpsrldq {{.*#+}} xmm0 = xmm0[10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vmovd %xmm3, (%rsi)
+; AVX512-SLOW-NEXT:    vmovd %xmm2, (%rdx)
+; AVX512-SLOW-NEXT:    vmovd %xmm4, (%rcx)
+; AVX512-SLOW-NEXT:    vmovd %xmm5, (%r8)
+; AVX512-SLOW-NEXT:    vmovd %xmm6, (%r9)
+; AVX512-SLOW-NEXT:    vmovd %xmm0, (%rax)
+; AVX512-SLOW-NEXT:    retq
 ;
 ; AVX512F-FAST-LABEL: load_i16_stride6_vf2:
 ; AVX512F-FAST:       # %bb.0:
@@ -203,33 +203,6 @@ define void @load_i16_stride6_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX512F-FAST-NEXT:    vmovd %xmm0, (%rax)
 ; AVX512F-FAST-NEXT:    retq
 ;
-; AVX512BW-SLOW-LABEL: load_i16_stride6_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm3 = xmm2[0,2,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[1,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpbroadcastw 4(%rdi), %xmm4
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm4 = xmm4[0],xmm1[0],xmm4[1],xmm1[1],xmm4[2],xmm1[2],xmm4[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm5 = xmm5[3,1,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm5 = xmm5[0,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpbroadcastw 20(%rdi), %xmm6
-; AVX512BW-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm7
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm6 = xmm7[0],xmm6[0],xmm7[1],xmm6[1],xmm7[2],xmm6[2],xmm7[3],xmm6[3]
-; AVX512BW-SLOW-NEXT:    vpsrlq $48, %xmm1, %xmm1
-; AVX512BW-SLOW-NEXT:    vpsrldq {{.*#+}} xmm0 = xmm0[10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vmovd %xmm3, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm2, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm4, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm5, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm6, (%r9)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm0, (%rax)
-; AVX512BW-SLOW-NEXT:    retq
-;
 ; AVX512BW-FAST-LABEL: load_i16_stride6_vf2:
 ; AVX512BW-FAST:       # %bb.0:
 ; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
@@ -11906,8 +11879,10 @@ define void @load_i16_stride6_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-7.ll
index 19530e02a99af5e..0771fcea0714cda 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -166,35 +166,35 @@ define void @load_i16_stride7_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST-PERLANE-NEXT:    vmovd %xmm0, (%rax)
 ; AVX2-FAST-PERLANE-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i16_stride7_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpsrld $16, %xmm0, %xmm3
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm4 = xmm3[0],xmm1[0],xmm3[1],xmm1[1],xmm3[2],xmm1[2],xmm3[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm6 = xmm5[2,1,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm6 = xmm6[0,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpshufhw {{.*#+}} xmm5 = xmm5[0,1,2,3,6,5,6,7]
-; AVX512F-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm7
-; AVX512F-SLOW-NEXT:    vpsrlq $48, %xmm1, %xmm8
-; AVX512F-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm7 = xmm7[0],xmm8[0],xmm7[1],xmm8[1],xmm7[2],xmm8[2],xmm7[3],xmm8[3]
-; AVX512F-SLOW-NEXT:    vpunpckhwd {{.*#+}} xmm3 = xmm3[4],xmm1[4],xmm3[5],xmm1[5],xmm3[6],xmm1[6],xmm3[7],xmm1[7]
-; AVX512F-SLOW-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,1,2,3]
-; AVX512F-SLOW-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,3,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vmovd %xmm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovd %xmm4, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovd %xmm6, (%rcx)
-; AVX512F-SLOW-NEXT:    vpextrd $2, %xmm5, (%r8)
-; AVX512F-SLOW-NEXT:    vmovd %xmm7, (%r9)
-; AVX512F-SLOW-NEXT:    vmovd %xmm3, (%r10)
-; AVX512F-SLOW-NEXT:    vmovd %xmm0, (%rax)
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i16_stride7_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpsrld $16, %xmm0, %xmm3
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm4 = xmm3[0],xmm1[0],xmm3[1],xmm1[1],xmm3[2],xmm1[2],xmm3[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm6 = xmm5[2,1,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm6 = xmm6[0,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpshufhw {{.*#+}} xmm5 = xmm5[0,1,2,3,6,5,6,7]
+; AVX512-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm7
+; AVX512-SLOW-NEXT:    vpsrlq $48, %xmm1, %xmm8
+; AVX512-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm7 = xmm7[0],xmm8[0],xmm7[1],xmm8[1],xmm7[2],xmm8[2],xmm7[3],xmm8[3]
+; AVX512-SLOW-NEXT:    vpunpckhwd {{.*#+}} xmm3 = xmm3[4],xmm1[4],xmm3[5],xmm1[5],xmm3[6],xmm1[6],xmm3[7],xmm1[7]
+; AVX512-SLOW-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,1,2,3]
+; AVX512-SLOW-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,3,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vmovd %xmm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovd %xmm4, (%rdx)
+; AVX512-SLOW-NEXT:    vmovd %xmm6, (%rcx)
+; AVX512-SLOW-NEXT:    vpextrd $2, %xmm5, (%r8)
+; AVX512-SLOW-NEXT:    vmovd %xmm7, (%r9)
+; AVX512-SLOW-NEXT:    vmovd %xmm3, (%r10)
+; AVX512-SLOW-NEXT:    vmovd %xmm0, (%rax)
+; AVX512-SLOW-NEXT:    retq
 ;
 ; AVX512F-FAST-LABEL: load_i16_stride7_vf2:
 ; AVX512F-FAST:       # %bb.0:
@@ -224,36 +224,6 @@ define void @load_i16_stride7_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX512F-FAST-NEXT:    vmovd %xmm0, (%rax)
 ; AVX512F-FAST-NEXT:    retq
 ;
-; AVX512BW-SLOW-LABEL: load_i16_stride7_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpsrld $16, %xmm0, %xmm3
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm4 = xmm3[0],xmm1[0],xmm3[1],xmm1[1],xmm3[2],xmm1[2],xmm3[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm6 = xmm5[2,1,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm6 = xmm6[0,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpshufhw {{.*#+}} xmm5 = xmm5[0,1,2,3,6,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpbroadcastw 8(%rdi), %xmm7
-; AVX512BW-SLOW-NEXT:    vpsrlq $48, %xmm1, %xmm8
-; AVX512BW-SLOW-NEXT:    vpunpcklwd {{.*#+}} xmm7 = xmm7[0],xmm8[0],xmm7[1],xmm8[1],xmm7[2],xmm8[2],xmm7[3],xmm8[3]
-; AVX512BW-SLOW-NEXT:    vpunpckhwd {{.*#+}} xmm3 = xmm3[4],xmm1[4],xmm3[5],xmm1[5],xmm3[6],xmm1[6],xmm3[7],xmm1[7]
-; AVX512BW-SLOW-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,1,2,3]
-; AVX512BW-SLOW-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,3,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vmovd %xmm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm4, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm6, (%rcx)
-; AVX512BW-SLOW-NEXT:    vpextrd $2, %xmm5, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm7, (%r9)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm3, (%r10)
-; AVX512BW-SLOW-NEXT:    vmovd %xmm0, (%rax)
-; AVX512BW-SLOW-NEXT:    retq
-;
 ; AVX512BW-FAST-LABEL: load_i16_stride7_vf2:
 ; AVX512BW-FAST:       # %bb.0:
 ; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
@@ -16204,8 +16174,10 @@ define void @load_i16_stride7_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-8.ll
index 91eba6c88037647..47f5ac1c5fc4065 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i16-stride-8.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -9893,6 +9893,8 @@ define void @load_i16_stride8_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-2.ll
index b85824d860c847f..85b318556a5b907 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -124,49 +124,27 @@ define void @load_i32_stride2_vf8(ptr %in.vec, ptr %out.vec0, ptr %out.vec1) nou
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i32_stride2_vf8:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512F-SLOW-NEXT:    vmovaps (%rdi), %ymm1
-; AVX512F-SLOW-NEXT:    vshufps {{.*#+}} ymm1 = ymm1[1,3],mem[1,3],ymm1[5,7],mem[5,7]
-; AVX512F-SLOW-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,1,3]
-; AVX512F-SLOW-NEXT:    vpmovqd %zmm0, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovaps %ymm1, (%rdx)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i32_stride2_vf8:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm0
+; AVX512-SLOW-NEXT:    vmovaps (%rdi), %ymm1
+; AVX512-SLOW-NEXT:    vshufps {{.*#+}} ymm1 = ymm1[1,3],mem[1,3],ymm1[5,7],mem[5,7]
+; AVX512-SLOW-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,1,3]
+; AVX512-SLOW-NEXT:    vpmovqd %zmm0, (%rsi)
+; AVX512-SLOW-NEXT:    vmovaps %ymm1, (%rdx)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i32_stride2_vf8:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %ymm1
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm2 = [1,3,5,7,9,11,13,15]
-; AVX512F-FAST-NEXT:    vpermi2d 32(%rdi), %ymm1, %ymm2
-; AVX512F-FAST-NEXT:    vpmovqd %zmm0, (%rsi)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm2, (%rdx)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i32_stride2_vf8:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512BW-SLOW-NEXT:    vmovaps (%rdi), %ymm1
-; AVX512BW-SLOW-NEXT:    vshufps {{.*#+}} ymm1 = ymm1[1,3],mem[1,3],ymm1[5,7],mem[5,7]
-; AVX512BW-SLOW-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,1,3]
-; AVX512BW-SLOW-NEXT:    vpmovqd %zmm0, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovaps %ymm1, (%rdx)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i32_stride2_vf8:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %ymm1
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm2 = [1,3,5,7,9,11,13,15]
-; AVX512BW-FAST-NEXT:    vpermi2d 32(%rdi), %ymm1, %ymm2
-; AVX512BW-FAST-NEXT:    vpmovqd %zmm0, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm2, (%rdx)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i32_stride2_vf8:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa64 (%rdi), %zmm0
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %ymm1
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm2 = [1,3,5,7,9,11,13,15]
+; AVX512-FAST-NEXT:    vpermi2d 32(%rdi), %ymm1, %ymm2
+; AVX512-FAST-NEXT:    vpmovqd %zmm0, (%rsi)
+; AVX512-FAST-NEXT:    vmovdqa %ymm2, (%rdx)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <16 x i32>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <16 x i32> %wide.vec, <16 x i32> poison, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
   %strided.vec1 = shufflevector <16 x i32> %wide.vec, <16 x i32> poison, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
@@ -761,15 +739,19 @@ define void @load_i32_stride2_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1) no
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-3.ll
index f6a81c61b87a760..e812fb903a808e1 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -59,61 +59,33 @@ define void @load_i32_stride3_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vmovlps %xmm1, (%rcx)
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i32_stride3_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovaps (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovaps 16(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vshufps {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512F-SLOW-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1],xmm1[2,3]
-; AVX512F-SLOW-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[1,0,2,3]
-; AVX512F-SLOW-NEXT:    vbroadcastss 8(%rdi), %xmm3
-; AVX512F-SLOW-NEXT:    vblendps {{.*#+}} xmm1 = xmm3[0],xmm1[1],xmm3[2,3]
-; AVX512F-SLOW-NEXT:    vmovlps %xmm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm0, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm1, (%rcx)
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i32_stride3_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovaps (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovaps 16(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vshufps {{.*#+}} xmm2 = xmm0[0,3,2,3]
+; AVX512-SLOW-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1],xmm1[2,3]
+; AVX512-SLOW-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[1,0,2,3]
+; AVX512-SLOW-NEXT:    vbroadcastss 8(%rdi), %xmm3
+; AVX512-SLOW-NEXT:    vblendps {{.*#+}} xmm1 = xmm3[0],xmm1[1],xmm3[2,3]
+; AVX512-SLOW-NEXT:    vmovlps %xmm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovlps %xmm0, (%rdx)
+; AVX512-SLOW-NEXT:    vmovlps %xmm1, (%rcx)
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i32_stride3_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [5,0,5,0]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm0, %xmm1, %xmm3
-; AVX512F-FAST-NEXT:    vpbroadcastd 8(%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512F-FAST-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512F-FAST-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512F-FAST-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i32_stride3_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovaps (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovaps 16(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vshufps {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512BW-SLOW-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1],xmm1[2,3]
-; AVX512BW-SLOW-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[1,0,2,3]
-; AVX512BW-SLOW-NEXT:    vbroadcastss 8(%rdi), %xmm3
-; AVX512BW-SLOW-NEXT:    vblendps {{.*#+}} xmm1 = xmm3[0],xmm1[1],xmm3[2,3]
-; AVX512BW-SLOW-NEXT:    vmovlps %xmm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovlps %xmm0, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovlps %xmm1, (%rcx)
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i32_stride3_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-FAST-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [5,0,5,0]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm0, %xmm1, %xmm3
-; AVX512BW-FAST-NEXT:    vpbroadcastd 8(%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BW-FAST-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i32_stride3_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-FAST-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,3,2,3]
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [5,0,5,0]
+; AVX512-FAST-NEXT:    vpermi2d %xmm0, %xmm1, %xmm3
+; AVX512-FAST-NEXT:    vpbroadcastd 8(%rdi), %xmm0
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
+; AVX512-FAST-NEXT:    vmovq %xmm2, (%rsi)
+; AVX512-FAST-NEXT:    vmovq %xmm3, (%rdx)
+; AVX512-FAST-NEXT:    vmovq %xmm0, (%rcx)
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <6 x i32>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <6 x i32> %wide.vec, <6 x i32> poison, <2 x i32> <i32 0, i32 3>
   %strided.vec1 = shufflevector <6 x i32> %wide.vec, <6 x i32> poison, <2 x i32> <i32 1, i32 4>
@@ -3049,15 +3021,19 @@ define void @load_i32_stride3_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512BW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-4.ll
index eba201399b0b9bf..7b4df6a350e3374 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -59,61 +59,33 @@ define void @load_i32_stride4_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vpextrq $1, %xmm0, (%r8)
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i32_stride4_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm3[0],xmm1[1],xmm3[2,3]
-; AVX512F-SLOW-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512F-SLOW-NEXT:    vpextrq $1, %xmm0, (%r8)
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i32_stride4_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm3[0],xmm1[1],xmm3[2,3]
+; AVX512-SLOW-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vmovq %xmm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovq %xmm3, (%rdx)
+; AVX512-SLOW-NEXT:    vmovq %xmm0, (%rcx)
+; AVX512-SLOW-NEXT:    vpextrq $1, %xmm0, (%r8)
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i32_stride4_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,5,1,5]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
-; AVX512F-FAST-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-FAST-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512F-FAST-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512F-FAST-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512F-FAST-NEXT:    vpextrq $1, %xmm0, (%r8)
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i32_stride4_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm3[0],xmm1[1],xmm3[2,3]
-; AVX512BW-SLOW-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512BW-SLOW-NEXT:    vpextrq $1, %xmm0, (%r8)
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i32_stride4_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-FAST-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,5,1,5]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
-; AVX512BW-FAST-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-FAST-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512BW-FAST-NEXT:    vpextrq $1, %xmm0, (%r8)
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i32_stride4_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-FAST-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,5,1,5]
+; AVX512-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
+; AVX512-FAST-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-FAST-NEXT:    vmovq %xmm2, (%rsi)
+; AVX512-FAST-NEXT:    vmovq %xmm3, (%rdx)
+; AVX512-FAST-NEXT:    vmovq %xmm0, (%rcx)
+; AVX512-FAST-NEXT:    vpextrq $1, %xmm0, (%r8)
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <8 x i32>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <8 x i32> %wide.vec, <8 x i32> poison, <2 x i32> <i32 0, i32 4>
   %strided.vec1 = shufflevector <8 x i32> %wide.vec, <8 x i32> poison, <2 x i32> <i32 1, i32 5>
@@ -3466,15 +3438,19 @@ define void @load_i32_stride4_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-5.ll
index f658848e511c93b..70c84aa6d2c998b 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -78,93 +78,49 @@ define void @load_i32_stride5_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i32_stride5_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm2
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512F-SLOW-NEXT:    vpextrd $2, %xmm1, %eax
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm4 = xmm0[1,1,1,1]
-; AVX512F-SLOW-NEXT:    vpinsrd $1, %eax, %xmm4, %xmm4
-; AVX512F-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm5
-; AVX512F-SLOW-NEXT:    vpextrd $3, %xmm1, %eax
-; AVX512F-SLOW-NEXT:    vpinsrd $1, %eax, %xmm5, %xmm1
-; AVX512F-SLOW-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm2[0,1,2,3,4,5,6,7,8,9,10,11]
-; AVX512F-SLOW-NEXT:    vpbroadcastd 16(%rdi), %ymm5
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm2 = xmm5[0],xmm2[1],xmm5[2,3]
-; AVX512F-SLOW-NEXT:    vmovq %xmm3, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovq %xmm4, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm1, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm0, (%r8)
-; AVX512F-SLOW-NEXT:    vmovq %xmm2, (%r9)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i32_stride5_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm2
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm0[0],xmm1[1],xmm0[2,3]
+; AVX512-SLOW-NEXT:    vpextrd $2, %xmm1, %eax
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm4 = xmm0[1,1,1,1]
+; AVX512-SLOW-NEXT:    vpinsrd $1, %eax, %xmm4, %xmm4
+; AVX512-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm5
+; AVX512-SLOW-NEXT:    vpextrd $3, %xmm1, %eax
+; AVX512-SLOW-NEXT:    vpinsrd $1, %eax, %xmm5, %xmm1
+; AVX512-SLOW-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm2[0,1,2,3,4,5,6,7,8,9,10,11]
+; AVX512-SLOW-NEXT:    vpbroadcastd 16(%rdi), %ymm5
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm2 = xmm5[0],xmm2[1],xmm5[2,3]
+; AVX512-SLOW-NEXT:    vmovq %xmm3, (%rsi)
+; AVX512-SLOW-NEXT:    vmovq %xmm4, (%rdx)
+; AVX512-SLOW-NEXT:    vmovq %xmm1, (%rcx)
+; AVX512-SLOW-NEXT:    vmovq %xmm0, (%r8)
+; AVX512-SLOW-NEXT:    vmovq %xmm2, (%r9)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i32_stride5_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vmovdqa 32(%rdi), %xmm2
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} xmm3 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm4 = [1,6,1,6]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm4
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [2,7,2,7]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm5
-; AVX512F-FAST-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm2[0,1,2,3,4,5,6,7,8,9,10,11]
-; AVX512F-FAST-NEXT:    vpbroadcastd 16(%rdi), %ymm1
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} xmm1 = xmm1[0],xmm2[1],xmm1[2,3]
-; AVX512F-FAST-NEXT:    vmovq %xmm3, (%rsi)
-; AVX512F-FAST-NEXT:    vmovq %xmm4, (%rdx)
-; AVX512F-FAST-NEXT:    vmovq %xmm5, (%rcx)
-; AVX512F-FAST-NEXT:    vmovq %xmm0, (%r8)
-; AVX512F-FAST-NEXT:    vmovq %xmm1, (%r9)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i32_stride5_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm2
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BW-SLOW-NEXT:    vpextrd $2, %xmm1, %eax
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm4 = xmm0[1,1,1,1]
-; AVX512BW-SLOW-NEXT:    vpinsrd $1, %eax, %xmm4, %xmm4
-; AVX512BW-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm5
-; AVX512BW-SLOW-NEXT:    vpextrd $3, %xmm1, %eax
-; AVX512BW-SLOW-NEXT:    vpinsrd $1, %eax, %xmm5, %xmm1
-; AVX512BW-SLOW-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm2[0,1,2,3,4,5,6,7,8,9,10,11]
-; AVX512BW-SLOW-NEXT:    vpbroadcastd 16(%rdi), %ymm5
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm2 = xmm5[0],xmm2[1],xmm5[2,3]
-; AVX512BW-SLOW-NEXT:    vmovq %xmm3, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm4, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm1, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm0, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm2, (%r9)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i32_stride5_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
-; AVX512BW-FAST-NEXT:    vmovdqa 32(%rdi), %xmm2
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} xmm3 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm4 = [1,6,1,6]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm4
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [2,7,2,7]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm5
-; AVX512BW-FAST-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm2[0,1,2,3,4,5,6,7,8,9,10,11]
-; AVX512BW-FAST-NEXT:    vpbroadcastd 16(%rdi), %ymm1
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} xmm1 = xmm1[0],xmm2[1],xmm1[2,3]
-; AVX512BW-FAST-NEXT:    vmovq %xmm3, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovq %xmm4, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm5, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm0, (%r8)
-; AVX512BW-FAST-NEXT:    vmovq %xmm1, (%r9)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i32_stride5_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vmovdqa 16(%rdi), %xmm1
+; AVX512-FAST-NEXT:    vmovdqa 32(%rdi), %xmm2
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} xmm3 = xmm0[0],xmm1[1],xmm0[2,3]
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm4 = [1,6,1,6]
+; AVX512-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm4
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [2,7,2,7]
+; AVX512-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm5
+; AVX512-FAST-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm2[0,1,2,3,4,5,6,7,8,9,10,11]
+; AVX512-FAST-NEXT:    vpbroadcastd 16(%rdi), %ymm1
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} xmm1 = xmm1[0],xmm2[1],xmm1[2,3]
+; AVX512-FAST-NEXT:    vmovq %xmm3, (%rsi)
+; AVX512-FAST-NEXT:    vmovq %xmm4, (%rdx)
+; AVX512-FAST-NEXT:    vmovq %xmm5, (%rcx)
+; AVX512-FAST-NEXT:    vmovq %xmm0, (%r8)
+; AVX512-FAST-NEXT:    vmovq %xmm1, (%r9)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <10 x i32>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <10 x i32> %wide.vec, <10 x i32> poison, <2 x i32> <i32 0, i32 5>
   %strided.vec1 = shufflevector <10 x i32> %wide.vec, <10 x i32> poison, <2 x i32> <i32 1, i32 6>
@@ -5295,14 +5251,18 @@ define void @load_i32_stride5_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-6.ll
index 2d8ada07be8251d..8c709508720a44b 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -102,69 +102,69 @@ define void @load_i32_stride6_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i32_stride6_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovaps 16(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm2
-; AVX512F-SLOW-NEXT:    vextractps $2, %xmm1, %r10d
-; AVX512F-SLOW-NEXT:    vpinsrd $1, %r10d, %xmm0, %xmm3
-; AVX512F-SLOW-NEXT:    vextractps $3, %xmm1, %r10d
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
-; AVX512F-SLOW-NEXT:    vpinsrd $1, %r10d, %xmm1, %xmm1
-; AVX512F-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm4
-; AVX512F-SLOW-NEXT:    vmovd %xmm2, %r10d
-; AVX512F-SLOW-NEXT:    vpinsrd $1, %r10d, %xmm4, %xmm4
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[3,3,3,3]
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm2[1],xmm0[2,3]
-; AVX512F-SLOW-NEXT:    vmovddup {{.*#+}} xmm2 = [4,2,4,2]
-; AVX512F-SLOW-NEXT:    # xmm2 = mem[0,0]
-; AVX512F-SLOW-NEXT:    vmovaps 32(%rdi), %ymm5
-; AVX512F-SLOW-NEXT:    vblendps {{.*#+}} ymm5 = ymm5[0,1,2,3],mem[4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpermps %ymm5, %ymm2, %ymm2
-; AVX512F-SLOW-NEXT:    vmovddup {{.*#+}} xmm6 = [5,3,5,3]
-; AVX512F-SLOW-NEXT:    # xmm6 = mem[0,0]
-; AVX512F-SLOW-NEXT:    vpermps %ymm5, %ymm6, %ymm5
-; AVX512F-SLOW-NEXT:    vmovq %xmm3, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovq %xmm1, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm4, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm0, (%r8)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm2, (%r9)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm5, (%rax)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i32_stride6_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovaps 16(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm2
+; AVX512-SLOW-NEXT:    vextractps $2, %xmm1, %r10d
+; AVX512-SLOW-NEXT:    vpinsrd $1, %r10d, %xmm0, %xmm3
+; AVX512-SLOW-NEXT:    vextractps $3, %xmm1, %r10d
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
+; AVX512-SLOW-NEXT:    vpinsrd $1, %r10d, %xmm1, %xmm1
+; AVX512-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm4
+; AVX512-SLOW-NEXT:    vmovd %xmm2, %r10d
+; AVX512-SLOW-NEXT:    vpinsrd $1, %r10d, %xmm4, %xmm4
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm2[1],xmm0[2,3]
+; AVX512-SLOW-NEXT:    vmovddup {{.*#+}} xmm2 = [4,2,4,2]
+; AVX512-SLOW-NEXT:    # xmm2 = mem[0,0]
+; AVX512-SLOW-NEXT:    vmovaps 32(%rdi), %ymm5
+; AVX512-SLOW-NEXT:    vblendps {{.*#+}} ymm5 = ymm5[0,1,2,3],mem[4,5,6,7]
+; AVX512-SLOW-NEXT:    vpermps %ymm5, %ymm2, %ymm2
+; AVX512-SLOW-NEXT:    vmovddup {{.*#+}} xmm6 = [5,3,5,3]
+; AVX512-SLOW-NEXT:    # xmm6 = mem[0,0]
+; AVX512-SLOW-NEXT:    vpermps %ymm5, %ymm6, %ymm5
+; AVX512-SLOW-NEXT:    vmovq %xmm3, (%rsi)
+; AVX512-SLOW-NEXT:    vmovq %xmm1, (%rdx)
+; AVX512-SLOW-NEXT:    vmovq %xmm4, (%rcx)
+; AVX512-SLOW-NEXT:    vmovq %xmm0, (%r8)
+; AVX512-SLOW-NEXT:    vmovlps %xmm2, (%r9)
+; AVX512-SLOW-NEXT:    vmovlps %xmm5, (%rax)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i32_stride6_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm0 = [0,6,0,6]
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vmovdqa 16(%rdi), %xmm2
-; AVX512F-FAST-NEXT:    vmovdqa 32(%rdi), %xmm3
-; AVX512F-FAST-NEXT:    vpermi2d %xmm2, %xmm1, %xmm0
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm4 = [1,7,1,7]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm2, %xmm1, %xmm4
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm2 = [2,4,2,4]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm3, %xmm1, %xmm2
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [3,5,3,5]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm3, %xmm1, %xmm5
-; AVX512F-FAST-NEXT:    vmovddup {{.*#+}} xmm1 = [4,2,4,2]
-; AVX512F-FAST-NEXT:    # xmm1 = mem[0,0]
-; AVX512F-FAST-NEXT:    vmovaps 32(%rdi), %ymm3
-; AVX512F-FAST-NEXT:    vblendps {{.*#+}} ymm3 = ymm3[0,1,2,3],mem[4,5,6,7]
-; AVX512F-FAST-NEXT:    vpermps %ymm3, %ymm1, %ymm1
-; AVX512F-FAST-NEXT:    vmovddup {{.*#+}} xmm6 = [5,3,5,3]
-; AVX512F-FAST-NEXT:    # xmm6 = mem[0,0]
-; AVX512F-FAST-NEXT:    vpermps %ymm3, %ymm6, %ymm3
-; AVX512F-FAST-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512F-FAST-NEXT:    vmovq %xmm4, (%rdx)
-; AVX512F-FAST-NEXT:    vmovq %xmm2, (%rcx)
-; AVX512F-FAST-NEXT:    vmovq %xmm5, (%r8)
-; AVX512F-FAST-NEXT:    vmovlps %xmm1, (%r9)
-; AVX512F-FAST-NEXT:    vmovlps %xmm3, (%rax)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i32_stride6_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm0 = [0,6,0,6]
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm1
+; AVX512-FAST-NEXT:    vmovdqa 16(%rdi), %xmm2
+; AVX512-FAST-NEXT:    vmovdqa 32(%rdi), %xmm3
+; AVX512-FAST-NEXT:    vpermi2d %xmm2, %xmm1, %xmm0
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm4 = [1,7,1,7]
+; AVX512-FAST-NEXT:    vpermi2d %xmm2, %xmm1, %xmm4
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm2 = [2,4,2,4]
+; AVX512-FAST-NEXT:    vpermi2d %xmm3, %xmm1, %xmm2
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [3,5,3,5]
+; AVX512-FAST-NEXT:    vpermi2d %xmm3, %xmm1, %xmm5
+; AVX512-FAST-NEXT:    vmovddup {{.*#+}} xmm1 = [4,2,4,2]
+; AVX512-FAST-NEXT:    # xmm1 = mem[0,0]
+; AVX512-FAST-NEXT:    vmovaps 32(%rdi), %ymm3
+; AVX512-FAST-NEXT:    vblendps {{.*#+}} ymm3 = ymm3[0,1,2,3],mem[4,5,6,7]
+; AVX512-FAST-NEXT:    vpermps %ymm3, %ymm1, %ymm1
+; AVX512-FAST-NEXT:    vmovddup {{.*#+}} xmm6 = [5,3,5,3]
+; AVX512-FAST-NEXT:    # xmm6 = mem[0,0]
+; AVX512-FAST-NEXT:    vpermps %ymm3, %ymm6, %ymm3
+; AVX512-FAST-NEXT:    vmovq %xmm0, (%rsi)
+; AVX512-FAST-NEXT:    vmovq %xmm4, (%rdx)
+; AVX512-FAST-NEXT:    vmovq %xmm2, (%rcx)
+; AVX512-FAST-NEXT:    vmovq %xmm5, (%r8)
+; AVX512-FAST-NEXT:    vmovlps %xmm1, (%r9)
+; AVX512-FAST-NEXT:    vmovlps %xmm3, (%rax)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
 ;
 ; AVX512BW-SLOW-LABEL: load_i32_stride6_vf2:
 ; AVX512BW-SLOW:       # %bb.0:
@@ -198,37 +198,6 @@ define void @load_i32_stride6_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX512BW-SLOW-NEXT:    vmovlps %xmm5, (%rax)
 ; AVX512BW-SLOW-NEXT:    vzeroupper
 ; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i32_stride6_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm0 = [0,6,0,6]
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm1
-; AVX512BW-FAST-NEXT:    vmovdqa 16(%rdi), %xmm2
-; AVX512BW-FAST-NEXT:    vmovdqa 32(%rdi), %xmm3
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm2, %xmm1, %xmm0
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm4 = [1,7,1,7]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm2, %xmm1, %xmm4
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm2 = [2,4,2,4]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm3, %xmm1, %xmm2
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [3,5,3,5]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm3, %xmm1, %xmm5
-; AVX512BW-FAST-NEXT:    vmovddup {{.*#+}} xmm1 = [4,2,4,2]
-; AVX512BW-FAST-NEXT:    # xmm1 = mem[0,0]
-; AVX512BW-FAST-NEXT:    vmovaps 32(%rdi), %ymm3
-; AVX512BW-FAST-NEXT:    vblendps {{.*#+}} ymm3 = ymm3[0,1,2,3],mem[4,5,6,7]
-; AVX512BW-FAST-NEXT:    vpermps %ymm3, %ymm1, %ymm1
-; AVX512BW-FAST-NEXT:    vmovddup {{.*#+}} xmm6 = [5,3,5,3]
-; AVX512BW-FAST-NEXT:    # xmm6 = mem[0,0]
-; AVX512BW-FAST-NEXT:    vpermps %ymm3, %ymm6, %ymm3
-; AVX512BW-FAST-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovq %xmm4, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm2, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm5, (%r8)
-; AVX512BW-FAST-NEXT:    vmovlps %xmm1, (%r9)
-; AVX512BW-FAST-NEXT:    vmovlps %xmm3, (%rax)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
   %wide.vec = load <12 x i32>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <12 x i32> %wide.vec, <12 x i32> poison, <2 x i32> <i32 0, i32 6>
   %strided.vec1 = shufflevector <12 x i32> %wide.vec, <12 x i32> poison, <2 x i32> <i32 1, i32 7>
@@ -10329,14 +10298,17 @@ define void @load_i32_stride6_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX: {{.*}}
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-7.ll
index ab44439f98ff3f8..7c98c94de19be04 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -120,143 +120,74 @@ define void @load_i32_stride7_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i32_stride7_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vpinsrd $1, 28(%rdi), %xmm0, %xmm2
-; AVX512F-SLOW-NEXT:    vmovd %xmm1, %r11d
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
-; AVX512F-SLOW-NEXT:    vpinsrd $1, %r11d, %xmm3, %xmm3
-; AVX512F-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm4
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm4 = xmm4[0],xmm1[1],xmm4[2,3]
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[3,2,2,3]
-; AVX512F-SLOW-NEXT:    vpbroadcastq {{.*#+}} xmm1 = [4,11,4,11]
-; AVX512F-SLOW-NEXT:    vmovdqa 32(%rdi), %ymm5
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %ymm6
-; AVX512F-SLOW-NEXT:    vpermi2d %ymm5, %ymm6, %ymm1
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm5[0],ymm6[1],ymm5[2,3,4],ymm6[5],ymm5[6,7]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} ymm7 = ymm7[1,0,2,3,5,4,6,7]
-; AVX512F-SLOW-NEXT:    vextracti128 $1, %ymm7, %xmm7
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} ymm6 = ymm6[2,3,2,3,6,7,6,7]
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm6[0],ymm5[1],ymm6[2,3,4],ymm5[5],ymm6[6,7]
-; AVX512F-SLOW-NEXT:    vextracti128 $1, %ymm5, %xmm5
-; AVX512F-SLOW-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm4, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm0, (%r8)
-; AVX512F-SLOW-NEXT:    vmovq %xmm1, (%r9)
-; AVX512F-SLOW-NEXT:    vmovq %xmm7, (%r10)
-; AVX512F-SLOW-NEXT:    vmovq %xmm5, (%rax)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i32_stride7_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vpinsrd $1, 28(%rdi), %xmm0, %xmm2
+; AVX512-SLOW-NEXT:    vmovd %xmm1, %r11d
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
+; AVX512-SLOW-NEXT:    vpinsrd $1, %r11d, %xmm3, %xmm3
+; AVX512-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm4
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm4 = xmm4[0],xmm1[1],xmm4[2,3]
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[3,2,2,3]
+; AVX512-SLOW-NEXT:    vpbroadcastq {{.*#+}} xmm1 = [4,11,4,11]
+; AVX512-SLOW-NEXT:    vmovdqa 32(%rdi), %ymm5
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %ymm6
+; AVX512-SLOW-NEXT:    vpermi2d %ymm5, %ymm6, %ymm1
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm5[0],ymm6[1],ymm5[2,3,4],ymm6[5],ymm5[6,7]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} ymm7 = ymm7[1,0,2,3,5,4,6,7]
+; AVX512-SLOW-NEXT:    vextracti128 $1, %ymm7, %xmm7
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} ymm6 = ymm6[2,3,2,3,6,7,6,7]
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm6[0],ymm5[1],ymm6[2,3,4],ymm5[5],ymm6[6,7]
+; AVX512-SLOW-NEXT:    vextracti128 $1, %ymm5, %xmm5
+; AVX512-SLOW-NEXT:    vmovq %xmm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovq %xmm3, (%rdx)
+; AVX512-SLOW-NEXT:    vmovq %xmm4, (%rcx)
+; AVX512-SLOW-NEXT:    vmovq %xmm0, (%r8)
+; AVX512-SLOW-NEXT:    vmovq %xmm1, (%r9)
+; AVX512-SLOW-NEXT:    vmovq %xmm7, (%r10)
+; AVX512-SLOW-NEXT:    vmovq %xmm5, (%rax)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i32_stride7_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vpinsrd $1, 28(%rdi), %xmm0, %xmm2
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,4,1,4]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
-; AVX512F-FAST-NEXT:    vpbroadcastd 8(%rdi), %xmm4
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} xmm4 = xmm4[0],xmm1[1],xmm4[2,3]
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [7,2,7,2]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm0, %xmm1, %xmm5
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm0 = [4,11,4,11]
-; AVX512F-FAST-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %ymm6
-; AVX512F-FAST-NEXT:    vpermi2d %ymm1, %ymm6, %ymm0
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm7 = [13,4,6,7,13,4,6,7]
-; AVX512F-FAST-NEXT:    # ymm7 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vpermi2d %ymm6, %ymm1, %ymm7
-; AVX512F-FAST-NEXT:    vextracti128 $1, %ymm7, %xmm7
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm8 = [6,13,6,7,6,13,6,7]
-; AVX512F-FAST-NEXT:    # ymm8 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vpermi2d %ymm1, %ymm6, %ymm8
-; AVX512F-FAST-NEXT:    vextracti128 $1, %ymm8, %xmm1
-; AVX512F-FAST-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512F-FAST-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512F-FAST-NEXT:    vmovq %xmm4, (%rcx)
-; AVX512F-FAST-NEXT:    vmovq %xmm5, (%r8)
-; AVX512F-FAST-NEXT:    vmovq %xmm0, (%r9)
-; AVX512F-FAST-NEXT:    vmovq %xmm7, (%r10)
-; AVX512F-FAST-NEXT:    vmovq %xmm1, (%rax)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i32_stride7_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vpinsrd $1, 28(%rdi), %xmm0, %xmm2
-; AVX512BW-SLOW-NEXT:    vmovd %xmm1, %r11d
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
-; AVX512BW-SLOW-NEXT:    vpinsrd $1, %r11d, %xmm3, %xmm3
-; AVX512BW-SLOW-NEXT:    vpbroadcastd 8(%rdi), %xmm4
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm4 = xmm4[0],xmm1[1],xmm4[2,3]
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[3,2,2,3]
-; AVX512BW-SLOW-NEXT:    vpbroadcastq {{.*#+}} xmm1 = [4,11,4,11]
-; AVX512BW-SLOW-NEXT:    vmovdqa 32(%rdi), %ymm5
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %ymm6
-; AVX512BW-SLOW-NEXT:    vpermi2d %ymm5, %ymm6, %ymm1
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm5[0],ymm6[1],ymm5[2,3,4],ymm6[5],ymm5[6,7]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} ymm7 = ymm7[1,0,2,3,5,4,6,7]
-; AVX512BW-SLOW-NEXT:    vextracti128 $1, %ymm7, %xmm7
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} ymm6 = ymm6[2,3,2,3,6,7,6,7]
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm6[0],ymm5[1],ymm6[2,3,4],ymm5[5],ymm6[6,7]
-; AVX512BW-SLOW-NEXT:    vextracti128 $1, %ymm5, %xmm5
-; AVX512BW-SLOW-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm4, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm0, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm1, (%r9)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm7, (%r10)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm5, (%rax)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i32_stride7_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-FAST-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512BW-FAST-NEXT:    vpinsrd $1, 28(%rdi), %xmm0, %xmm2
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,4,1,4]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
-; AVX512BW-FAST-NEXT:    vpbroadcastd 8(%rdi), %xmm4
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} xmm4 = xmm4[0],xmm1[1],xmm4[2,3]
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [7,2,7,2]
-; AVX512BW-FAST-NEXT:    vpermi2d %xmm0, %xmm1, %xmm5
-; AVX512BW-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm0 = [4,11,4,11]
-; AVX512BW-FAST-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %ymm6
-; AVX512BW-FAST-NEXT:    vpermi2d %ymm1, %ymm6, %ymm0
-; AVX512BW-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm7 = [13,4,6,7,13,4,6,7]
-; AVX512BW-FAST-NEXT:    # ymm7 = mem[0,1,0,1]
-; AVX512BW-FAST-NEXT:    vpermi2d %ymm6, %ymm1, %ymm7
-; AVX512BW-FAST-NEXT:    vextracti128 $1, %ymm7, %xmm7
-; AVX512BW-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm8 = [6,13,6,7,6,13,6,7]
-; AVX512BW-FAST-NEXT:    # ymm8 = mem[0,1,0,1]
-; AVX512BW-FAST-NEXT:    vpermi2d %ymm1, %ymm6, %ymm8
-; AVX512BW-FAST-NEXT:    vextracti128 $1, %ymm8, %xmm1
-; AVX512BW-FAST-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm4, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovq %xmm5, (%r8)
-; AVX512BW-FAST-NEXT:    vmovq %xmm0, (%r9)
-; AVX512BW-FAST-NEXT:    vmovq %xmm7, (%r10)
-; AVX512BW-FAST-NEXT:    vmovq %xmm1, (%rax)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i32_stride7_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vmovdqa 32(%rdi), %xmm1
+; AVX512-FAST-NEXT:    vpinsrd $1, 28(%rdi), %xmm0, %xmm2
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,4,1,4]
+; AVX512-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
+; AVX512-FAST-NEXT:    vpbroadcastd 8(%rdi), %xmm4
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} xmm4 = xmm4[0],xmm1[1],xmm4[2,3]
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm5 = [7,2,7,2]
+; AVX512-FAST-NEXT:    vpermi2d %xmm0, %xmm1, %xmm5
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm0 = [4,11,4,11]
+; AVX512-FAST-NEXT:    vmovdqa 32(%rdi), %ymm1
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %ymm6
+; AVX512-FAST-NEXT:    vpermi2d %ymm1, %ymm6, %ymm0
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm7 = [13,4,6,7,13,4,6,7]
+; AVX512-FAST-NEXT:    # ymm7 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vpermi2d %ymm6, %ymm1, %ymm7
+; AVX512-FAST-NEXT:    vextracti128 $1, %ymm7, %xmm7
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm8 = [6,13,6,7,6,13,6,7]
+; AVX512-FAST-NEXT:    # ymm8 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vpermi2d %ymm1, %ymm6, %ymm8
+; AVX512-FAST-NEXT:    vextracti128 $1, %ymm8, %xmm1
+; AVX512-FAST-NEXT:    vmovq %xmm2, (%rsi)
+; AVX512-FAST-NEXT:    vmovq %xmm3, (%rdx)
+; AVX512-FAST-NEXT:    vmovq %xmm4, (%rcx)
+; AVX512-FAST-NEXT:    vmovq %xmm5, (%r8)
+; AVX512-FAST-NEXT:    vmovq %xmm0, (%r9)
+; AVX512-FAST-NEXT:    vmovq %xmm7, (%r10)
+; AVX512-FAST-NEXT:    vmovq %xmm1, (%rax)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <14 x i32>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <14 x i32> %wide.vec, <14 x i32> poison, <2 x i32> <i32 0, i32 7>
   %strided.vec1 = shufflevector <14 x i32> %wide.vec, <14 x i32> poison, <2 x i32> <i32 1, i32 8>
@@ -12700,14 +12631,18 @@ define void @load_i32_stride7_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX: {{.*}}
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-8.ll
index 25db5a66461e8a4..a47102b6f7c0ae5 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i32-stride-8.ll
@@ -4,13 +4,13 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
@@ -112,105 +112,72 @@ define void @load_i32_stride8_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i32_stride8_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; AVX512F-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512F-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm3[0],xmm1[1],xmm3[2,3]
-; AVX512F-SLOW-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
-; AVX512F-SLOW-NEXT:    vmovaps (%rdi), %ymm4
-; AVX512F-SLOW-NEXT:    vunpcklps {{.*#+}} ymm5 = ymm4[0],ymm1[0],ymm4[1],ymm1[1],ymm4[4],ymm1[4],ymm4[5],ymm1[5]
-; AVX512F-SLOW-NEXT:    vextractf128 $1, %ymm5, %xmm5
-; AVX512F-SLOW-NEXT:    vshufps {{.*#+}} ymm6 = ymm4[1,1,1,1,5,5,5,5]
-; AVX512F-SLOW-NEXT:    vblendps {{.*#+}} ymm6 = ymm6[0],ymm1[1],ymm6[2,3,4],ymm1[5],ymm6[6,7]
-; AVX512F-SLOW-NEXT:    vextractf128 $1, %ymm6, %xmm6
-; AVX512F-SLOW-NEXT:    vunpckhps {{.*#+}} ymm1 = ymm4[2],ymm1[2],ymm4[3],ymm1[3],ymm4[6],ymm1[6],ymm4[7],ymm1[7]
-; AVX512F-SLOW-NEXT:    vextractf128 $1, %ymm1, %xmm4
-; AVX512F-SLOW-NEXT:    vshufps {{.*#+}} ymm1 = ymm1[2,3,2,3,6,7,6,7]
-; AVX512F-SLOW-NEXT:    vextractf128 $1, %ymm1, %xmm1
-; AVX512F-SLOW-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512F-SLOW-NEXT:    vpextrq $1, %xmm0, (%r8)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm5, (%r9)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm6, (%r11)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm4, (%r10)
-; AVX512F-SLOW-NEXT:    vmovlps %xmm1, (%rax)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i32_stride8_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r11
+; AVX512-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm3[0],xmm1[1],xmm3[2,3]
+; AVX512-SLOW-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
+; AVX512-SLOW-NEXT:    vmovaps (%rdi), %ymm4
+; AVX512-SLOW-NEXT:    vunpcklps {{.*#+}} ymm5 = ymm4[0],ymm1[0],ymm4[1],ymm1[1],ymm4[4],ymm1[4],ymm4[5],ymm1[5]
+; AVX512-SLOW-NEXT:    vextractf128 $1, %ymm5, %xmm5
+; AVX512-SLOW-NEXT:    vshufps {{.*#+}} ymm6 = ymm4[1,1,1,1,5,5,5,5]
+; AVX512-SLOW-NEXT:    vblendps {{.*#+}} ymm6 = ymm6[0],ymm1[1],ymm6[2,3,4],ymm1[5],ymm6[6,7]
+; AVX512-SLOW-NEXT:    vextractf128 $1, %ymm6, %xmm6
+; AVX512-SLOW-NEXT:    vunpckhps {{.*#+}} ymm1 = ymm4[2],ymm1[2],ymm4[3],ymm1[3],ymm4[6],ymm1[6],ymm4[7],ymm1[7]
+; AVX512-SLOW-NEXT:    vextractf128 $1, %ymm1, %xmm4
+; AVX512-SLOW-NEXT:    vshufps {{.*#+}} ymm1 = ymm1[2,3,2,3,6,7,6,7]
+; AVX512-SLOW-NEXT:    vextractf128 $1, %ymm1, %xmm1
+; AVX512-SLOW-NEXT:    vmovq %xmm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovq %xmm3, (%rdx)
+; AVX512-SLOW-NEXT:    vmovq %xmm0, (%rcx)
+; AVX512-SLOW-NEXT:    vpextrq $1, %xmm0, (%r8)
+; AVX512-SLOW-NEXT:    vmovlps %xmm5, (%r9)
+; AVX512-SLOW-NEXT:    vmovlps %xmm6, (%r11)
+; AVX512-SLOW-NEXT:    vmovlps %xmm4, (%r10)
+; AVX512-SLOW-NEXT:    vmovlps %xmm1, (%rax)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i32_stride8_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-FAST-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512F-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,5,1,5]
-; AVX512F-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
-; AVX512F-FAST-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512F-FAST-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %ymm4
-; AVX512F-FAST-NEXT:    vpunpckldq {{.*#+}} ymm5 = ymm4[0],ymm1[0],ymm4[1],ymm1[1],ymm4[4],ymm1[4],ymm4[5],ymm1[5]
-; AVX512F-FAST-NEXT:    vextracti128 $1, %ymm5, %xmm5
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm6 = [5,13,5,5,5,13,5,5]
-; AVX512F-FAST-NEXT:    # ymm6 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vpermi2d %ymm1, %ymm4, %ymm6
-; AVX512F-FAST-NEXT:    vextracti128 $1, %ymm6, %xmm6
-; AVX512F-FAST-NEXT:    vpunpckhdq {{.*#+}} ymm1 = ymm4[2],ymm1[2],ymm4[3],ymm1[3],ymm4[6],ymm1[6],ymm4[7],ymm1[7]
-; AVX512F-FAST-NEXT:    vextracti128 $1, %ymm1, %xmm4
-; AVX512F-FAST-NEXT:    vpshufd {{.*#+}} ymm1 = ymm1[2,3,2,3,6,7,6,7]
-; AVX512F-FAST-NEXT:    vextracti128 $1, %ymm1, %xmm1
-; AVX512F-FAST-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512F-FAST-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512F-FAST-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512F-FAST-NEXT:    vpextrq $1, %xmm0, (%r8)
-; AVX512F-FAST-NEXT:    vmovq %xmm5, (%r9)
-; AVX512F-FAST-NEXT:    vmovq %xmm6, (%r11)
-; AVX512F-FAST-NEXT:    vmovq %xmm4, (%r10)
-; AVX512F-FAST-NEXT:    vmovq %xmm1, (%rax)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i32_stride8_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; AVX512BW-SLOW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512BW-SLOW-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[1,1,1,1]
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm3 = xmm3[0],xmm1[1],xmm3[2,3]
-; AVX512BW-SLOW-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512BW-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
-; AVX512BW-SLOW-NEXT:    vmovaps (%rdi), %ymm4
-; AVX512BW-SLOW-NEXT:    vunpcklps {{.*#+}} ymm5 = ymm4[0],ymm1[0],ymm4[1],ymm1[1],ymm4[4],ymm1[4],ymm4[5],ymm1[5]
-; AVX512BW-SLOW-NEXT:    vextractf128 $1, %ymm5, %xmm5
-; AVX512BW-SLOW-NEXT:    vshufps {{.*#+}} ymm6 = ymm4[1,1,1,1,5,5,5,5]
-; AVX512BW-SLOW-NEXT:    vblendps {{.*#+}} ymm6 = ymm6[0],ymm1[1],ymm6[2,3,4],ymm1[5],ymm6[6,7]
-; AVX512BW-SLOW-NEXT:    vextractf128 $1, %ymm6, %xmm6
-; AVX512BW-SLOW-NEXT:    vunpckhps {{.*#+}} ymm1 = ymm4[2],ymm1[2],ymm4[3],ymm1[3],ymm4[6],ymm1[6],ymm4[7],ymm1[7]
-; AVX512BW-SLOW-NEXT:    vextractf128 $1, %ymm1, %xmm4
-; AVX512BW-SLOW-NEXT:    vshufps {{.*#+}} ymm1 = ymm1[2,3,2,3,6,7,6,7]
-; AVX512BW-SLOW-NEXT:    vextractf128 $1, %ymm1, %xmm1
-; AVX512BW-SLOW-NEXT:    vmovq %xmm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm3, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovq %xmm0, (%rcx)
-; AVX512BW-SLOW-NEXT:    vpextrq $1, %xmm0, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovlps %xmm5, (%r9)
-; AVX512BW-SLOW-NEXT:    vmovlps %xmm6, (%r11)
-; AVX512BW-SLOW-NEXT:    vmovlps %xmm4, (%r10)
-; AVX512BW-SLOW-NEXT:    vmovlps %xmm1, (%rax)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
+; AVX512-FAST-LABEL: load_i32_stride8_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r11
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %xmm0
+; AVX512-FAST-NEXT:    vmovdqa 32(%rdi), %xmm1
+; AVX512-FAST-NEXT:    vpunpckldq {{.*#+}} xmm2 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512-FAST-NEXT:    vpbroadcastq {{.*#+}} xmm3 = [1,5,1,5]
+; AVX512-FAST-NEXT:    vpermi2d %xmm1, %xmm0, %xmm3
+; AVX512-FAST-NEXT:    vpunpckhdq {{.*#+}} xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512-FAST-NEXT:    vmovdqa 32(%rdi), %ymm1
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %ymm4
+; AVX512-FAST-NEXT:    vpunpckldq {{.*#+}} ymm5 = ymm4[0],ymm1[0],ymm4[1],ymm1[1],ymm4[4],ymm1[4],ymm4[5],ymm1[5]
+; AVX512-FAST-NEXT:    vextracti128 $1, %ymm5, %xmm5
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm6 = [5,13,5,5,5,13,5,5]
+; AVX512-FAST-NEXT:    # ymm6 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vpermi2d %ymm1, %ymm4, %ymm6
+; AVX512-FAST-NEXT:    vextracti128 $1, %ymm6, %xmm6
+; AVX512-FAST-NEXT:    vpunpckhdq {{.*#+}} ymm1 = ymm4[2],ymm1[2],ymm4[3],ymm1[3],ymm4[6],ymm1[6],ymm4[7],ymm1[7]
+; AVX512-FAST-NEXT:    vextracti128 $1, %ymm1, %xmm4
+; AVX512-FAST-NEXT:    vpshufd {{.*#+}} ymm1 = ymm1[2,3,2,3,6,7,6,7]
+; AVX512-FAST-NEXT:    vextracti128 $1, %ymm1, %xmm1
+; AVX512-FAST-NEXT:    vmovq %xmm2, (%rsi)
+; AVX512-FAST-NEXT:    vmovq %xmm3, (%rdx)
+; AVX512-FAST-NEXT:    vmovq %xmm0, (%rcx)
+; AVX512-FAST-NEXT:    vpextrq $1, %xmm0, (%r8)
+; AVX512-FAST-NEXT:    vmovq %xmm5, (%r9)
+; AVX512-FAST-NEXT:    vmovq %xmm6, (%r11)
+; AVX512-FAST-NEXT:    vmovq %xmm4, (%r10)
+; AVX512-FAST-NEXT:    vmovq %xmm1, (%rax)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
 ;
 ; AVX512BW-FAST-LABEL: load_i32_stride8_vf2:
 ; AVX512BW-FAST:       # %bb.0:
@@ -11129,6 +11096,9 @@ define void @load_i32_stride8_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-2.ll
index 8fa1dd6a30b523e..eb2a1cc227f43ab 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -88,57 +88,31 @@ define void @load_i64_stride2_vf4(ptr %in.vec, ptr %out.vec0, ptr %out.vec1) nou
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i64_stride2_vf4:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovaps (%rdi), %ymm0
-; AVX512F-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
-; AVX512F-SLOW-NEXT:    vunpcklpd {{.*#+}} ymm2 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
-; AVX512F-SLOW-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,2,1,3]
-; AVX512F-SLOW-NEXT:    vunpckhpd {{.*#+}} ymm0 = ymm0[1],ymm1[1],ymm0[3],ymm1[3]
-; AVX512F-SLOW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512F-SLOW-NEXT:    vmovaps %ymm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovaps %ymm0, (%rdx)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i64_stride2_vf4:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovaps (%rdi), %ymm0
+; AVX512-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
+; AVX512-SLOW-NEXT:    vunpcklpd {{.*#+}} ymm2 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
+; AVX512-SLOW-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,2,1,3]
+; AVX512-SLOW-NEXT:    vunpckhpd {{.*#+}} ymm0 = ymm0[1],ymm1[1],ymm0[3],ymm1[3]
+; AVX512-SLOW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
+; AVX512-SLOW-NEXT:    vmovaps %ymm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovaps %ymm0, (%rdx)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i64_stride2_vf4:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm0 = [0,2,4,6]
-; AVX512F-FAST-NEXT:    vmovdqa (%rdi), %ymm1
-; AVX512F-FAST-NEXT:    vmovdqa 32(%rdi), %ymm2
-; AVX512F-FAST-NEXT:    vpermi2q %ymm2, %ymm1, %ymm0
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm3 = [1,3,5,7]
-; AVX512F-FAST-NEXT:    vpermi2q %ymm2, %ymm1, %ymm3
-; AVX512F-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm3, (%rdx)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i64_stride2_vf4:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovaps (%rdi), %ymm0
-; AVX512BW-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
-; AVX512BW-SLOW-NEXT:    vunpcklpd {{.*#+}} ymm2 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
-; AVX512BW-SLOW-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,2,1,3]
-; AVX512BW-SLOW-NEXT:    vunpckhpd {{.*#+}} ymm0 = ymm0[1],ymm1[1],ymm0[3],ymm1[3]
-; AVX512BW-SLOW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512BW-SLOW-NEXT:    vmovaps %ymm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovaps %ymm0, (%rdx)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i64_stride2_vf4:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm0 = [0,2,4,6]
-; AVX512BW-FAST-NEXT:    vmovdqa (%rdi), %ymm1
-; AVX512BW-FAST-NEXT:    vmovdqa 32(%rdi), %ymm2
-; AVX512BW-FAST-NEXT:    vpermi2q %ymm2, %ymm1, %ymm0
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm3 = [1,3,5,7]
-; AVX512BW-FAST-NEXT:    vpermi2q %ymm2, %ymm1, %ymm3
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm3, (%rdx)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i64_stride2_vf4:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm0 = [0,2,4,6]
+; AVX512-FAST-NEXT:    vmovdqa (%rdi), %ymm1
+; AVX512-FAST-NEXT:    vmovdqa 32(%rdi), %ymm2
+; AVX512-FAST-NEXT:    vpermi2q %ymm2, %ymm1, %ymm0
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm3 = [1,3,5,7]
+; AVX512-FAST-NEXT:    vpermi2q %ymm2, %ymm1, %ymm3
+; AVX512-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
+; AVX512-FAST-NEXT:    vmovdqa %ymm3, (%rdx)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <8 x i64>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <8 x i64> %wide.vec, <8 x i64> poison, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
   %strided.vec1 = shufflevector <8 x i64> %wide.vec, <8 x i64> poison, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
@@ -1454,15 +1428,19 @@ define void @load_i64_stride2_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1) no
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-3.ll
index ac1be65682a7b7c..b83571698ea2e65 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -56,57 +56,31 @@ define void @load_i64_stride3_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vmovdqa %xmm1, (%rcx)
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i64_stride3_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vpermpd {{.*#+}} zmm0 = mem[0,3,2,3,4,7,6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vpalignr {{.*#+}} xmm2 = mem[8,9,10,11,12,13,14,15],xmm1[0,1,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm1 = mem[0,1],xmm1[2,3]
-; AVX512F-SLOW-NEXT:    vmovaps %xmm0, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovdqa %xmm2, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovdqa %xmm1, (%rcx)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i64_stride3_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vpermpd {{.*#+}} zmm0 = mem[0,3,2,3,4,7,6,7]
+; AVX512-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vpalignr {{.*#+}} xmm2 = mem[8,9,10,11,12,13,14,15],xmm1[0,1,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm1 = mem[0,1],xmm1[2,3]
+; AVX512-SLOW-NEXT:    vmovaps %xmm0, (%rsi)
+; AVX512-SLOW-NEXT:    vmovdqa %xmm2, (%rdx)
+; AVX512-SLOW-NEXT:    vmovdqa %xmm1, (%rcx)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i64_stride3_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovaps {{.*#+}} xmm0 = [1,4]
-; AVX512F-FAST-NEXT:    vmovaps (%rdi), %zmm1
-; AVX512F-FAST-NEXT:    vpermpd %zmm1, %zmm0, %zmm0
-; AVX512F-FAST-NEXT:    vpermpd {{.*#+}} zmm1 = zmm1[0,3,2,3,4,7,6,7]
-; AVX512F-FAST-NEXT:    vmovaps 16(%rdi), %xmm2
-; AVX512F-FAST-NEXT:    vblendps {{.*#+}} xmm2 = xmm2[0,1],mem[2,3]
-; AVX512F-FAST-NEXT:    vmovaps %xmm1, (%rsi)
-; AVX512F-FAST-NEXT:    vmovaps %xmm0, (%rdx)
-; AVX512F-FAST-NEXT:    vmovaps %xmm2, (%rcx)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i64_stride3_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vpermpd {{.*#+}} zmm0 = mem[0,3,2,3,4,7,6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa 32(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vpalignr {{.*#+}} xmm2 = mem[8,9,10,11,12,13,14,15],xmm1[0,1,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm1 = mem[0,1],xmm1[2,3]
-; AVX512BW-SLOW-NEXT:    vmovaps %xmm0, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovdqa %xmm2, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovdqa %xmm1, (%rcx)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i64_stride3_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovaps {{.*#+}} xmm0 = [1,4]
-; AVX512BW-FAST-NEXT:    vmovaps (%rdi), %zmm1
-; AVX512BW-FAST-NEXT:    vpermpd %zmm1, %zmm0, %zmm0
-; AVX512BW-FAST-NEXT:    vpermpd {{.*#+}} zmm1 = zmm1[0,3,2,3,4,7,6,7]
-; AVX512BW-FAST-NEXT:    vmovaps 16(%rdi), %xmm2
-; AVX512BW-FAST-NEXT:    vblendps {{.*#+}} xmm2 = xmm2[0,1],mem[2,3]
-; AVX512BW-FAST-NEXT:    vmovaps %xmm1, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovaps %xmm0, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovaps %xmm2, (%rcx)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i64_stride3_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovaps {{.*#+}} xmm0 = [1,4]
+; AVX512-FAST-NEXT:    vmovaps (%rdi), %zmm1
+; AVX512-FAST-NEXT:    vpermpd %zmm1, %zmm0, %zmm0
+; AVX512-FAST-NEXT:    vpermpd {{.*#+}} zmm1 = zmm1[0,3,2,3,4,7,6,7]
+; AVX512-FAST-NEXT:    vmovaps 16(%rdi), %xmm2
+; AVX512-FAST-NEXT:    vblendps {{.*#+}} xmm2 = xmm2[0,1],mem[2,3]
+; AVX512-FAST-NEXT:    vmovaps %xmm1, (%rsi)
+; AVX512-FAST-NEXT:    vmovaps %xmm0, (%rdx)
+; AVX512-FAST-NEXT:    vmovaps %xmm2, (%rcx)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <6 x i64>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <6 x i64> %wide.vec, <6 x i64> poison, <2 x i32> <i32 0, i32 3>
   %strided.vec1 = shufflevector <6 x i64> %wide.vec, <6 x i64> poison, <2 x i32> <i32 1, i32 4>
@@ -2571,15 +2545,19 @@ define void @load_i64_stride3_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-4.ll
index 1b508c764cc8cff..e2bf6ef15fdea78 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -67,73 +67,39 @@ define void @load_i64_stride4_vf2(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i64_stride4_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovaps (%rdi), %xmm0
-; AVX512F-SLOW-NEXT:    vmovaps 32(%rdi), %xmm1
-; AVX512F-SLOW-NEXT:    vmovlhps {{.*#+}} xmm2 = xmm0[0],xmm1[0]
-; AVX512F-SLOW-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
-; AVX512F-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
-; AVX512F-SLOW-NEXT:    vmovaps (%rdi), %ymm3
-; AVX512F-SLOW-NEXT:    vunpcklpd {{.*#+}} ymm4 = ymm3[0],ymm1[0],ymm3[2],ymm1[2]
-; AVX512F-SLOW-NEXT:    vunpckhpd {{.*#+}} ymm1 = ymm3[1],ymm1[1],ymm3[3],ymm1[3]
-; AVX512F-SLOW-NEXT:    vmovaps %xmm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovaps %xmm0, (%rdx)
-; AVX512F-SLOW-NEXT:    vextractf128 $1, %ymm4, (%rcx)
-; AVX512F-SLOW-NEXT:    vextractf128 $1, %ymm1, (%r8)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i64_stride4_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovaps (%rdi), %xmm0
+; AVX512-SLOW-NEXT:    vmovaps 32(%rdi), %xmm1
+; AVX512-SLOW-NEXT:    vmovlhps {{.*#+}} xmm2 = xmm0[0],xmm1[0]
+; AVX512-SLOW-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
+; AVX512-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
+; AVX512-SLOW-NEXT:    vmovaps (%rdi), %ymm3
+; AVX512-SLOW-NEXT:    vunpcklpd {{.*#+}} ymm4 = ymm3[0],ymm1[0],ymm3[2],ymm1[2]
+; AVX512-SLOW-NEXT:    vunpckhpd {{.*#+}} ymm1 = ymm3[1],ymm1[1],ymm3[3],ymm1[3]
+; AVX512-SLOW-NEXT:    vmovaps %xmm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovaps %xmm0, (%rdx)
+; AVX512-SLOW-NEXT:    vextractf128 $1, %ymm4, (%rcx)
+; AVX512-SLOW-NEXT:    vextractf128 $1, %ymm1, (%r8)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i64_stride4_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovaps {{.*#+}} xmm0 = [0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,0]
-; AVX512F-FAST-NEXT:    vpermpd (%rdi), %zmm0, %zmm0
-; AVX512F-FAST-NEXT:    vmovaps (%rdi), %xmm1
-; AVX512F-FAST-NEXT:    vunpckhpd {{.*#+}} xmm1 = xmm1[1],mem[1]
-; AVX512F-FAST-NEXT:    vmovaps (%rdi), %ymm2
-; AVX512F-FAST-NEXT:    vmovaps 32(%rdi), %ymm3
-; AVX512F-FAST-NEXT:    vunpcklpd {{.*#+}} ymm4 = ymm2[0],ymm3[0],ymm2[2],ymm3[2]
-; AVX512F-FAST-NEXT:    vunpckhpd {{.*#+}} ymm2 = ymm2[1],ymm3[1],ymm2[3],ymm3[3]
-; AVX512F-FAST-NEXT:    vmovaps %xmm0, (%rsi)
-; AVX512F-FAST-NEXT:    vmovaps %xmm1, (%rdx)
-; AVX512F-FAST-NEXT:    vextractf128 $1, %ymm4, (%rcx)
-; AVX512F-FAST-NEXT:    vextractf128 $1, %ymm2, (%r8)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i64_stride4_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovaps (%rdi), %xmm0
-; AVX512BW-SLOW-NEXT:    vmovaps 32(%rdi), %xmm1
-; AVX512BW-SLOW-NEXT:    vmovlhps {{.*#+}} xmm2 = xmm0[0],xmm1[0]
-; AVX512BW-SLOW-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
-; AVX512BW-SLOW-NEXT:    vmovaps 32(%rdi), %ymm1
-; AVX512BW-SLOW-NEXT:    vmovaps (%rdi), %ymm3
-; AVX512BW-SLOW-NEXT:    vunpcklpd {{.*#+}} ymm4 = ymm3[0],ymm1[0],ymm3[2],ymm1[2]
-; AVX512BW-SLOW-NEXT:    vunpckhpd {{.*#+}} ymm1 = ymm3[1],ymm1[1],ymm3[3],ymm1[3]
-; AVX512BW-SLOW-NEXT:    vmovaps %xmm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovaps %xmm0, (%rdx)
-; AVX512BW-SLOW-NEXT:    vextractf128 $1, %ymm4, (%rcx)
-; AVX512BW-SLOW-NEXT:    vextractf128 $1, %ymm1, (%r8)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i64_stride4_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovaps {{.*#+}} xmm0 = [0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,0]
-; AVX512BW-FAST-NEXT:    vpermpd (%rdi), %zmm0, %zmm0
-; AVX512BW-FAST-NEXT:    vmovaps (%rdi), %xmm1
-; AVX512BW-FAST-NEXT:    vunpckhpd {{.*#+}} xmm1 = xmm1[1],mem[1]
-; AVX512BW-FAST-NEXT:    vmovaps (%rdi), %ymm2
-; AVX512BW-FAST-NEXT:    vmovaps 32(%rdi), %ymm3
-; AVX512BW-FAST-NEXT:    vunpcklpd {{.*#+}} ymm4 = ymm2[0],ymm3[0],ymm2[2],ymm3[2]
-; AVX512BW-FAST-NEXT:    vunpckhpd {{.*#+}} ymm2 = ymm2[1],ymm3[1],ymm2[3],ymm3[3]
-; AVX512BW-FAST-NEXT:    vmovaps %xmm0, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovaps %xmm1, (%rdx)
-; AVX512BW-FAST-NEXT:    vextractf128 $1, %ymm4, (%rcx)
-; AVX512BW-FAST-NEXT:    vextractf128 $1, %ymm2, (%r8)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i64_stride4_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovaps {{.*#+}} xmm0 = [0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,0]
+; AVX512-FAST-NEXT:    vpermpd (%rdi), %zmm0, %zmm0
+; AVX512-FAST-NEXT:    vmovaps (%rdi), %xmm1
+; AVX512-FAST-NEXT:    vunpckhpd {{.*#+}} xmm1 = xmm1[1],mem[1]
+; AVX512-FAST-NEXT:    vmovaps (%rdi), %ymm2
+; AVX512-FAST-NEXT:    vmovaps 32(%rdi), %ymm3
+; AVX512-FAST-NEXT:    vunpcklpd {{.*#+}} ymm4 = ymm2[0],ymm3[0],ymm2[2],ymm3[2]
+; AVX512-FAST-NEXT:    vunpckhpd {{.*#+}} ymm2 = ymm2[1],ymm3[1],ymm2[3],ymm3[3]
+; AVX512-FAST-NEXT:    vmovaps %xmm0, (%rsi)
+; AVX512-FAST-NEXT:    vmovaps %xmm1, (%rdx)
+; AVX512-FAST-NEXT:    vextractf128 $1, %ymm4, (%rcx)
+; AVX512-FAST-NEXT:    vextractf128 $1, %ymm2, (%r8)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <8 x i64>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <8 x i64> %wide.vec, <8 x i64> poison, <2 x i32> <i32 0, i32 4>
   %strided.vec1 = shufflevector <8 x i64> %wide.vec, <8 x i64> poison, <2 x i32> <i32 1, i32 5>
@@ -4179,15 +4145,19 @@ define void @load_i64_stride4_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
 ; AVX512F: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-5.ll
index b677dd7079434f5..08763079d0c68d3 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -220,125 +220,65 @@ define void @load_i64_stride5_vf4(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i64_stride5_vf4:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512F-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm1
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,5,10,15]
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm3 = <1,6,11,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
-; AVX512F-SLOW-NEXT:    vmovdqa 128(%rdi), %xmm4
-; AVX512F-SLOW-NEXT:    vpbroadcastq %xmm4, %ymm5
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3,4,5],ymm5[6,7]
-; AVX512F-SLOW-NEXT:    vinserti128 $1, %xmm4, %ymm0, %ymm4
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <2,7,12,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm5
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm4 = ymm5[0,1,2,3,4,5],ymm4[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <11,0,5,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm0, %zmm1, %zmm5
-; AVX512F-SLOW-NEXT:    vpbroadcastq 144(%rdi), %ymm6
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm5[0,1,2,3,4,5],ymm6[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm6 = <12,1,6,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm0, %zmm1, %zmm6
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm6[0,1,2,3,4,5],mem[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm2, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm3, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm4, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm5, (%r8)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm0, (%r9)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i64_stride5_vf4:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm0
+; AVX512-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm1
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,5,10,15]
+; AVX512-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm3 = <1,6,11,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
+; AVX512-SLOW-NEXT:    vmovdqa 128(%rdi), %xmm4
+; AVX512-SLOW-NEXT:    vpbroadcastq %xmm4, %ymm5
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3,4,5],ymm5[6,7]
+; AVX512-SLOW-NEXT:    vinserti128 $1, %xmm4, %ymm0, %ymm4
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <2,7,12,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm5
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm4 = ymm5[0,1,2,3,4,5],ymm4[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <11,0,5,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm0, %zmm1, %zmm5
+; AVX512-SLOW-NEXT:    vpbroadcastq 144(%rdi), %ymm6
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm5[0,1,2,3,4,5],ymm6[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm6 = <12,1,6,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm0, %zmm1, %zmm6
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm6[0,1,2,3,4,5],mem[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa %ymm2, (%rsi)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm3, (%rdx)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm4, (%rcx)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm5, (%r8)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm0, (%r9)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i64_stride5_vf4:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512F-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm1
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,5,10,15]
-; AVX512F-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm3 = <1,6,11,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,2,4]
-; AVX512F-FAST-NEXT:    vmovdqa 128(%rdi), %ymm5
-; AVX512F-FAST-NEXT:    vpermi2q %ymm5, %ymm3, %ymm4
-; AVX512F-FAST-NEXT:    vinserti128 $1, 128(%rdi), %ymm0, %ymm3
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = <2,7,12,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm6
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm3 = ymm6[0,1,2,3,4,5],ymm3[6,7]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = <11,0,5,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm0, %zmm1, %zmm6
-; AVX512F-FAST-NEXT:    vpbroadcastq 144(%rdi), %ymm7
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm6 = ymm6[0,1,2,3,4,5],ymm7[6,7]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm7 = <12,1,6,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm0, %zmm1, %zmm7
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm7[0,1,2,3,4,5],ymm5[6,7]
-; AVX512F-FAST-NEXT:    vmovdqa %ymm2, (%rsi)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm4, (%rdx)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm3, (%rcx)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm6, (%r8)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm0, (%r9)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i64_stride5_vf4:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512BW-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm1
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,5,10,15]
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm3 = <1,6,11,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
-; AVX512BW-SLOW-NEXT:    vmovdqa 128(%rdi), %xmm4
-; AVX512BW-SLOW-NEXT:    vpbroadcastq %xmm4, %ymm5
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3,4,5],ymm5[6,7]
-; AVX512BW-SLOW-NEXT:    vinserti128 $1, %xmm4, %ymm0, %ymm4
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <2,7,12,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm5
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm4 = ymm5[0,1,2,3,4,5],ymm4[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <11,0,5,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm0, %zmm1, %zmm5
-; AVX512BW-SLOW-NEXT:    vpbroadcastq 144(%rdi), %ymm6
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm5[0,1,2,3,4,5],ymm6[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm6 = <12,1,6,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm0, %zmm1, %zmm6
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm6[0,1,2,3,4,5],mem[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm2, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm3, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm4, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm5, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm0, (%r9)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i64_stride5_vf4:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovdqa64 (%rdi), %zmm0
-; AVX512BW-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm1
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,5,10,15]
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm3 = <1,6,11,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,2,4]
-; AVX512BW-FAST-NEXT:    vmovdqa 128(%rdi), %ymm5
-; AVX512BW-FAST-NEXT:    vpermi2q %ymm5, %ymm3, %ymm4
-; AVX512BW-FAST-NEXT:    vinserti128 $1, 128(%rdi), %ymm0, %ymm3
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = <2,7,12,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm6
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm3 = ymm6[0,1,2,3,4,5],ymm3[6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = <11,0,5,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm0, %zmm1, %zmm6
-; AVX512BW-FAST-NEXT:    vpbroadcastq 144(%rdi), %ymm7
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm6 = ymm6[0,1,2,3,4,5],ymm7[6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm7 = <12,1,6,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm0, %zmm1, %zmm7
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm7[0,1,2,3,4,5],ymm5[6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm2, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm4, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm3, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm6, (%r8)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm0, (%r9)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i64_stride5_vf4:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovdqa64 (%rdi), %zmm0
+; AVX512-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm1
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,5,10,15]
+; AVX512-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm3 = <1,6,11,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,2,4]
+; AVX512-FAST-NEXT:    vmovdqa 128(%rdi), %ymm5
+; AVX512-FAST-NEXT:    vpermi2q %ymm5, %ymm3, %ymm4
+; AVX512-FAST-NEXT:    vinserti128 $1, 128(%rdi), %ymm0, %ymm3
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = <2,7,12,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm1, %zmm0, %zmm6
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm3 = ymm6[0,1,2,3,4,5],ymm3[6,7]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = <11,0,5,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm0, %zmm1, %zmm6
+; AVX512-FAST-NEXT:    vpbroadcastq 144(%rdi), %ymm7
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm6 = ymm6[0,1,2,3,4,5],ymm7[6,7]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm7 = <12,1,6,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm0, %zmm1, %zmm7
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm7[0,1,2,3,4,5],ymm5[6,7]
+; AVX512-FAST-NEXT:    vmovdqa %ymm2, (%rsi)
+; AVX512-FAST-NEXT:    vmovdqa %ymm4, (%rdx)
+; AVX512-FAST-NEXT:    vmovdqa %ymm3, (%rcx)
+; AVX512-FAST-NEXT:    vmovdqa %ymm6, (%r8)
+; AVX512-FAST-NEXT:    vmovdqa %ymm0, (%r9)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <20 x i64>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <20 x i64> %wide.vec, <20 x i64> poison, <4 x i32> <i32 0, i32 5, i32 10, i32 15>
   %strided.vec1 = shufflevector <20 x i64> %wide.vec, <20 x i64> poison, <4 x i32> <i32 1, i32 6, i32 11, i32 16>
@@ -6499,14 +6439,18 @@ define void @load_i64_stride5_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-6.ll
index ffdee1430c71161..6e00c62f404e5a7 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -215,173 +215,89 @@ define void @load_i64_stride6_vf4(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i64_stride6_vf4:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm2
-; AVX512F-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm3
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm0 = <0,6,12,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm0
-; AVX512F-SLOW-NEXT:    vpbroadcastq 144(%rdi), %ymm1
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <1,7,13,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm1
-; AVX512F-SLOW-NEXT:    vmovdqa 128(%rdi), %ymm4
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm4[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <10,0,6,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm2, %zmm3, %zmm5
-; AVX512F-SLOW-NEXT:    vmovdqa 160(%rdi), %xmm6
-; AVX512F-SLOW-NEXT:    vpbroadcastq %xmm6, %ymm7
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm5[0,1,2,3,4,5],ymm7[6,7]
-; AVX512F-SLOW-NEXT:    vinserti128 $1, %xmm6, %ymm0, %ymm6
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm7 = <11,1,7,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm2, %zmm3, %zmm7
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm6 = ymm7[0,1,2,3,4,5],ymm6[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa 160(%rdi), %ymm7
-; AVX512F-SLOW-NEXT:    vpunpcklqdq {{.*#+}} ymm4 = ymm4[0],ymm7[0],ymm4[2],ymm7[2]
-; AVX512F-SLOW-NEXT:    vpermq {{.*#+}} ymm4 = ymm4[0,1,0,3]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} xmm8 = [4,10]
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm4 = ymm8[0,1,2,3],ymm4[4,5,6,7]
-; AVX512F-SLOW-NEXT:    vpbroadcastq 136(%rdi), %ymm8
-; AVX512F-SLOW-NEXT:    vpunpckhqdq {{.*#+}} ymm7 = ymm8[1],ymm7[1],ymm8[3],ymm7[3]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} xmm8 = [5,11]
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm2 = ymm8[0,1,2,3],ymm7[4,5,6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm5, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm6, (%r8)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm4, (%r9)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm2, (%rax)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i64_stride6_vf4:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm2
+; AVX512-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm3
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm0 = <0,6,12,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm0
+; AVX512-SLOW-NEXT:    vpbroadcastq 144(%rdi), %ymm1
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <1,7,13,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm1
+; AVX512-SLOW-NEXT:    vmovdqa 128(%rdi), %ymm4
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm4[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <10,0,6,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm2, %zmm3, %zmm5
+; AVX512-SLOW-NEXT:    vmovdqa 160(%rdi), %xmm6
+; AVX512-SLOW-NEXT:    vpbroadcastq %xmm6, %ymm7
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm5[0,1,2,3,4,5],ymm7[6,7]
+; AVX512-SLOW-NEXT:    vinserti128 $1, %xmm6, %ymm0, %ymm6
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm7 = <11,1,7,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm2, %zmm3, %zmm7
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm6 = ymm7[0,1,2,3,4,5],ymm6[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa 160(%rdi), %ymm7
+; AVX512-SLOW-NEXT:    vpunpcklqdq {{.*#+}} ymm4 = ymm4[0],ymm7[0],ymm4[2],ymm7[2]
+; AVX512-SLOW-NEXT:    vpermq {{.*#+}} ymm4 = ymm4[0,1,0,3]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} xmm8 = [4,10]
+; AVX512-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm4 = ymm8[0,1,2,3],ymm4[4,5,6,7]
+; AVX512-SLOW-NEXT:    vpbroadcastq 136(%rdi), %ymm8
+; AVX512-SLOW-NEXT:    vpunpckhqdq {{.*#+}} ymm7 = ymm8[1],ymm7[1],ymm8[3],ymm7[3]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} xmm8 = [5,11]
+; AVX512-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm2 = ymm8[0,1,2,3],ymm7[4,5,6,7]
+; AVX512-SLOW-NEXT:    vmovdqa %ymm0, (%rsi)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm1, (%rdx)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm5, (%rcx)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm6, (%r8)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm4, (%r9)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm2, (%rax)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i64_stride6_vf4:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-FAST-NEXT:    vmovdqa64 (%rdi), %zmm2
-; AVX512F-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm3
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm0 = <0,6,12,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm0
-; AVX512F-FAST-NEXT:    vpbroadcastq 144(%rdi), %ymm1
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <1,7,13,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm1
-; AVX512F-FAST-NEXT:    vmovdqa 128(%rdi), %ymm4
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm4[6,7]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm5 = <10,0,6,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm2, %zmm3, %zmm5
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = [0,1,2,4]
-; AVX512F-FAST-NEXT:    vmovdqa 160(%rdi), %ymm7
-; AVX512F-FAST-NEXT:    vpermi2q %ymm7, %ymm5, %ymm6
-; AVX512F-FAST-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm5
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm8 = <11,1,7,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm2, %zmm3, %zmm8
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm5 = ymm8[0,1,2,3,4,5],ymm5[6,7]
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm8 = [0,6,0,6]
-; AVX512F-FAST-NEXT:    # ymm8 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vpermi2q %ymm7, %ymm4, %ymm8
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} xmm4 = [4,10]
-; AVX512F-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm4
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm4 = ymm4[0,1,2,3],ymm8[4,5,6,7]
-; AVX512F-FAST-NEXT:    vpbroadcastq 136(%rdi), %ymm8
-; AVX512F-FAST-NEXT:    vpunpckhqdq {{.*#+}} ymm7 = ymm8[1],ymm7[1],ymm8[3],ymm7[3]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} xmm8 = [5,11]
-; AVX512F-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm8[0,1,2,3],ymm7[4,5,6,7]
-; AVX512F-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm6, (%rcx)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm5, (%r8)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm4, (%r9)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm2, (%rax)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i64_stride6_vf4:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm2
-; AVX512BW-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm3
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm0 = <0,6,12,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm0
-; AVX512BW-SLOW-NEXT:    vpbroadcastq 144(%rdi), %ymm1
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <1,7,13,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm1
-; AVX512BW-SLOW-NEXT:    vmovdqa 128(%rdi), %ymm4
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm4[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm5 = <10,0,6,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm2, %zmm3, %zmm5
-; AVX512BW-SLOW-NEXT:    vmovdqa 160(%rdi), %xmm6
-; AVX512BW-SLOW-NEXT:    vpbroadcastq %xmm6, %ymm7
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm5 = ymm5[0,1,2,3,4,5],ymm7[6,7]
-; AVX512BW-SLOW-NEXT:    vinserti128 $1, %xmm6, %ymm0, %ymm6
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm7 = <11,1,7,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm2, %zmm3, %zmm7
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm6 = ymm7[0,1,2,3,4,5],ymm6[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa 160(%rdi), %ymm7
-; AVX512BW-SLOW-NEXT:    vpunpcklqdq {{.*#+}} ymm4 = ymm4[0],ymm7[0],ymm4[2],ymm7[2]
-; AVX512BW-SLOW-NEXT:    vpermq {{.*#+}} ymm4 = ymm4[0,1,0,3]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} xmm8 = [4,10]
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm4 = ymm8[0,1,2,3],ymm4[4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vpbroadcastq 136(%rdi), %ymm8
-; AVX512BW-SLOW-NEXT:    vpunpckhqdq {{.*#+}} ymm7 = ymm8[1],ymm7[1],ymm8[3],ymm7[3]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} xmm8 = [5,11]
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm2 = ymm8[0,1,2,3],ymm7[4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm5, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm6, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm4, (%r9)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm2, (%rax)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i64_stride6_vf4:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-FAST-NEXT:    vmovdqa64 (%rdi), %zmm2
-; AVX512BW-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm3
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm0 = <0,6,12,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm0
-; AVX512BW-FAST-NEXT:    vpbroadcastq 144(%rdi), %ymm1
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <1,7,13,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm1
-; AVX512BW-FAST-NEXT:    vmovdqa 128(%rdi), %ymm4
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm4[6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm5 = <10,0,6,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm2, %zmm3, %zmm5
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = [0,1,2,4]
-; AVX512BW-FAST-NEXT:    vmovdqa 160(%rdi), %ymm7
-; AVX512BW-FAST-NEXT:    vpermi2q %ymm7, %ymm5, %ymm6
-; AVX512BW-FAST-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm5
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm8 = <11,1,7,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm2, %zmm3, %zmm8
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm5 = ymm8[0,1,2,3,4,5],ymm5[6,7]
-; AVX512BW-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm8 = [0,6,0,6]
-; AVX512BW-FAST-NEXT:    # ymm8 = mem[0,1,0,1]
-; AVX512BW-FAST-NEXT:    vpermi2q %ymm7, %ymm4, %ymm8
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} xmm4 = [4,10]
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm4
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm4 = ymm4[0,1,2,3],ymm8[4,5,6,7]
-; AVX512BW-FAST-NEXT:    vpbroadcastq 136(%rdi), %ymm8
-; AVX512BW-FAST-NEXT:    vpunpckhqdq {{.*#+}} ymm7 = ymm8[1],ymm7[1],ymm8[3],ymm7[3]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} xmm8 = [5,11]
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm8[0,1,2,3],ymm7[4,5,6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm6, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm5, (%r8)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm4, (%r9)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm2, (%rax)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i64_stride6_vf4:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-FAST-NEXT:    vmovdqa64 (%rdi), %zmm2
+; AVX512-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm3
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm0 = <0,6,12,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm0
+; AVX512-FAST-NEXT:    vpbroadcastq 144(%rdi), %ymm1
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <1,7,13,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm1
+; AVX512-FAST-NEXT:    vmovdqa 128(%rdi), %ymm4
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm4[6,7]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm5 = <10,0,6,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm2, %zmm3, %zmm5
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm6 = [0,1,2,4]
+; AVX512-FAST-NEXT:    vmovdqa 160(%rdi), %ymm7
+; AVX512-FAST-NEXT:    vpermi2q %ymm7, %ymm5, %ymm6
+; AVX512-FAST-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm5
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm8 = <11,1,7,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm2, %zmm3, %zmm8
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm5 = ymm8[0,1,2,3,4,5],ymm5[6,7]
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm8 = [0,6,0,6]
+; AVX512-FAST-NEXT:    # ymm8 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vpermi2q %ymm7, %ymm4, %ymm8
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} xmm4 = [4,10]
+; AVX512-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm4
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm4 = ymm4[0,1,2,3],ymm8[4,5,6,7]
+; AVX512-FAST-NEXT:    vpbroadcastq 136(%rdi), %ymm8
+; AVX512-FAST-NEXT:    vpunpckhqdq {{.*#+}} ymm7 = ymm8[1],ymm7[1],ymm8[3],ymm7[3]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} xmm8 = [5,11]
+; AVX512-FAST-NEXT:    vpermi2q %zmm3, %zmm2, %zmm8
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm8[0,1,2,3],ymm7[4,5,6,7]
+; AVX512-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
+; AVX512-FAST-NEXT:    vmovdqa %ymm1, (%rdx)
+; AVX512-FAST-NEXT:    vmovdqa %ymm6, (%rcx)
+; AVX512-FAST-NEXT:    vmovdqa %ymm5, (%r8)
+; AVX512-FAST-NEXT:    vmovdqa %ymm4, (%r9)
+; AVX512-FAST-NEXT:    vmovdqa %ymm2, (%rax)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <24 x i64>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <24 x i64> %wide.vec, <24 x i64> poison, <4 x i32> <i32 0, i32 6, i32 12, i32 18>
   %strided.vec1 = shufflevector <24 x i64> %wide.vec, <24 x i64> poison, <4 x i32> <i32 1, i32 7, i32 13, i32 19>
@@ -8731,14 +8647,18 @@ define void @load_i64_stride6_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-7.ll
index 9841686a648a55f..864597f14e320ac 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -294,221 +294,113 @@ define void @load_i64_stride7_vf4(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-ONLY-NEXT:    vzeroupper
 ; AVX2-ONLY-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: load_i64_stride7_vf4:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-SLOW-NEXT:    vmovdqa64 128(%rdi), %zmm3
-; AVX512F-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm4
-; AVX512F-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm5
-; AVX512F-SLOW-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm0
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,7,14,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm1
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm1[0,1,2,3,4,5],ymm0[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <9,0,7,u>
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm5, %zmm4, %zmm1
-; AVX512F-SLOW-NEXT:    vpbroadcastq 176(%rdi), %ymm2
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm2[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa 128(%rdi), %xmm6
-; AVX512F-SLOW-NEXT:    vpbroadcastq %xmm6, %ymm2
-; AVX512F-SLOW-NEXT:    vpunpckhqdq {{.*#+}} ymm2 = ymm2[1],mem[1],ymm2[3],mem[3]
-; AVX512F-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm7
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} xmm7 = xmm7[0,1],mem[2,3]
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm2 = ymm7[0,1,2,3],ymm2[4,5,6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa 64(%rdi), %ymm7
-; AVX512F-SLOW-NEXT:    vpalignr {{.*#+}} ymm7 = mem[8,9,10,11,12,13,14,15],ymm7[0,1,2,3,4,5,6,7],mem[24,25,26,27,28,29,30,31],ymm7[16,17,18,19,20,21,22,23]
-; AVX512F-SLOW-NEXT:    vmovdqa 192(%rdi), %xmm8
-; AVX512F-SLOW-NEXT:    vpalignr {{.*#+}} xmm6 = xmm6[8,9,10,11,12,13,14,15],xmm8[0,1,2,3,4,5,6,7]
-; AVX512F-SLOW-NEXT:    vinserti128 $1, %xmm6, %ymm0, %ymm6
-; AVX512F-SLOW-NEXT:    vperm2i128 {{.*#+}} ymm6 = ymm7[2,3],ymm6[2,3]
-; AVX512F-SLOW-NEXT:    vinserti128 $1, %xmm8, %ymm0, %ymm7
-; AVX512F-SLOW-NEXT:    vmovdqa 128(%rdi), %ymm8
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm8[0,1,2,3,4,5],ymm7[6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} xmm9 = [4,11]
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm9[0,1,2,3],ymm7[4,5,6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa 192(%rdi), %ymm9
-; AVX512F-SLOW-NEXT:    vpalignr {{.*#+}} ymm8 = ymm8[8,9,10,11,12,13,14,15],ymm9[0,1,2,3,4,5,6,7],ymm8[24,25,26,27,28,29,30,31],ymm9[16,17,18,19,20,21,22,23]
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} xmm9 = [5,12]
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm8 = ymm9[0,1,2,3],ymm8[4,5,6,7]
-; AVX512F-SLOW-NEXT:    vbroadcasti128 {{.*#+}} ymm9 = [4,11,4,11]
-; AVX512F-SLOW-NEXT:    # ymm9 = mem[0,1,0,1]
-; AVX512F-SLOW-NEXT:    vpermi2q 192(%rdi), %zmm3, %zmm9
-; AVX512F-SLOW-NEXT:    vmovdqa {{.*#+}} xmm3 = [6,13]
-; AVX512F-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm3
-; AVX512F-SLOW-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3],ymm9[4,5,6,7]
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm2, (%rcx)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm6, (%r8)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm7, (%r9)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm8, (%r10)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm3, (%rax)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: load_i64_stride7_vf4:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-SLOW-NEXT:    vmovdqa64 128(%rdi), %zmm3
+; AVX512-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm4
+; AVX512-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm5
+; AVX512-SLOW-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm0
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,7,14,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm1
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm1[0,1,2,3,4,5],ymm0[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <9,0,7,u>
+; AVX512-SLOW-NEXT:    vpermi2q %zmm5, %zmm4, %zmm1
+; AVX512-SLOW-NEXT:    vpbroadcastq 176(%rdi), %ymm2
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm2[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa 128(%rdi), %xmm6
+; AVX512-SLOW-NEXT:    vpbroadcastq %xmm6, %ymm2
+; AVX512-SLOW-NEXT:    vpunpckhqdq {{.*#+}} ymm2 = ymm2[1],mem[1],ymm2[3],mem[3]
+; AVX512-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm7
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} xmm7 = xmm7[0,1],mem[2,3]
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm2 = ymm7[0,1,2,3],ymm2[4,5,6,7]
+; AVX512-SLOW-NEXT:    vmovdqa 64(%rdi), %ymm7
+; AVX512-SLOW-NEXT:    vpalignr {{.*#+}} ymm7 = mem[8,9,10,11,12,13,14,15],ymm7[0,1,2,3,4,5,6,7],mem[24,25,26,27,28,29,30,31],ymm7[16,17,18,19,20,21,22,23]
+; AVX512-SLOW-NEXT:    vmovdqa 192(%rdi), %xmm8
+; AVX512-SLOW-NEXT:    vpalignr {{.*#+}} xmm6 = xmm6[8,9,10,11,12,13,14,15],xmm8[0,1,2,3,4,5,6,7]
+; AVX512-SLOW-NEXT:    vinserti128 $1, %xmm6, %ymm0, %ymm6
+; AVX512-SLOW-NEXT:    vperm2i128 {{.*#+}} ymm6 = ymm7[2,3],ymm6[2,3]
+; AVX512-SLOW-NEXT:    vinserti128 $1, %xmm8, %ymm0, %ymm7
+; AVX512-SLOW-NEXT:    vmovdqa 128(%rdi), %ymm8
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm8[0,1,2,3,4,5],ymm7[6,7]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} xmm9 = [4,11]
+; AVX512-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm9[0,1,2,3],ymm7[4,5,6,7]
+; AVX512-SLOW-NEXT:    vmovdqa 192(%rdi), %ymm9
+; AVX512-SLOW-NEXT:    vpalignr {{.*#+}} ymm8 = ymm8[8,9,10,11,12,13,14,15],ymm9[0,1,2,3,4,5,6,7],ymm8[24,25,26,27,28,29,30,31],ymm9[16,17,18,19,20,21,22,23]
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} xmm9 = [5,12]
+; AVX512-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm8 = ymm9[0,1,2,3],ymm8[4,5,6,7]
+; AVX512-SLOW-NEXT:    vbroadcasti128 {{.*#+}} ymm9 = [4,11,4,11]
+; AVX512-SLOW-NEXT:    # ymm9 = mem[0,1,0,1]
+; AVX512-SLOW-NEXT:    vpermi2q 192(%rdi), %zmm3, %zmm9
+; AVX512-SLOW-NEXT:    vmovdqa {{.*#+}} xmm3 = [6,13]
+; AVX512-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm3
+; AVX512-SLOW-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3],ymm9[4,5,6,7]
+; AVX512-SLOW-NEXT:    vmovdqa %ymm0, (%rsi)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm1, (%rdx)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm2, (%rcx)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm6, (%r8)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm7, (%r9)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm8, (%r10)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm3, (%rax)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: load_i64_stride7_vf4:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-FAST-NEXT:    vmovdqa64 128(%rdi), %zmm3
-; AVX512F-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm4
-; AVX512F-FAST-NEXT:    vmovdqa64 (%rdi), %zmm5
-; AVX512F-FAST-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm0
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,7,14,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm1
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm1[0,1,2,3,4,5],ymm0[6,7]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <9,0,7,u>
-; AVX512F-FAST-NEXT:    vpermi2q %zmm5, %zmm4, %zmm1
-; AVX512F-FAST-NEXT:    vpbroadcastq 176(%rdi), %ymm2
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm2[6,7]
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm2 = [0,7,0,7]
-; AVX512F-FAST-NEXT:    # ymm2 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vmovdqa 128(%rdi), %ymm6
-; AVX512F-FAST-NEXT:    vpermi2q 160(%rdi), %ymm6, %ymm2
-; AVX512F-FAST-NEXT:    vmovdqa 16(%rdi), %xmm7
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} xmm7 = xmm7[0,1],mem[2,3]
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm7[0,1,2,3],ymm2[4,5,6,7]
-; AVX512F-FAST-NEXT:    vmovdqa 64(%rdi), %ymm7
-; AVX512F-FAST-NEXT:    vpalignr {{.*#+}} ymm7 = mem[8,9,10,11,12,13,14,15],ymm7[0,1,2,3,4,5,6,7],mem[24,25,26,27,28,29,30,31],ymm7[16,17,18,19,20,21,22,23]
-; AVX512F-FAST-NEXT:    vmovdqa 192(%rdi), %xmm8
-; AVX512F-FAST-NEXT:    vpalignr {{.*#+}} xmm9 = mem[8,9,10,11,12,13,14,15],xmm8[0,1,2,3,4,5,6,7]
-; AVX512F-FAST-NEXT:    vinserti128 $1, %xmm9, %ymm0, %ymm9
-; AVX512F-FAST-NEXT:    vperm2i128 {{.*#+}} ymm7 = ymm7[2,3],ymm9[2,3]
-; AVX512F-FAST-NEXT:    vinserti128 $1, %xmm8, %ymm0, %ymm8
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm8 = ymm6[0,1,2,3,4,5],ymm8[6,7]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} xmm9 = [4,11]
-; AVX512F-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm8 = ymm9[0,1,2,3],ymm8[4,5,6,7]
-; AVX512F-FAST-NEXT:    vmovdqa 192(%rdi), %ymm9
-; AVX512F-FAST-NEXT:    vpalignr {{.*#+}} ymm6 = ymm6[8,9,10,11,12,13,14,15],ymm9[0,1,2,3,4,5,6,7],ymm6[24,25,26,27,28,29,30,31],ymm9[16,17,18,19,20,21,22,23]
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} xmm9 = [5,12]
-; AVX512F-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm6 = ymm9[0,1,2,3],ymm6[4,5,6,7]
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm9 = [4,11,4,11]
-; AVX512F-FAST-NEXT:    # ymm9 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vpermi2q 192(%rdi), %zmm3, %zmm9
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} xmm3 = [6,13]
-; AVX512F-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm3
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3],ymm9[4,5,6,7]
-; AVX512F-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm2, (%rcx)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm7, (%r8)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm8, (%r9)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm6, (%r10)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm3, (%rax)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: load_i64_stride7_vf4:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-SLOW-NEXT:    vmovdqa64 128(%rdi), %zmm3
-; AVX512BW-SLOW-NEXT:    vmovdqa64 64(%rdi), %zmm4
-; AVX512BW-SLOW-NEXT:    vmovdqa64 (%rdi), %zmm5
-; AVX512BW-SLOW-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm0
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,7,14,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm1
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm1[0,1,2,3,4,5],ymm0[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} ymm1 = <9,0,7,u>
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm5, %zmm4, %zmm1
-; AVX512BW-SLOW-NEXT:    vpbroadcastq 176(%rdi), %ymm2
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm2[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa 128(%rdi), %xmm6
-; AVX512BW-SLOW-NEXT:    vpbroadcastq %xmm6, %ymm2
-; AVX512BW-SLOW-NEXT:    vpunpckhqdq {{.*#+}} ymm2 = ymm2[1],mem[1],ymm2[3],mem[3]
-; AVX512BW-SLOW-NEXT:    vmovdqa 16(%rdi), %xmm7
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} xmm7 = xmm7[0,1],mem[2,3]
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm2 = ymm7[0,1,2,3],ymm2[4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa 64(%rdi), %ymm7
-; AVX512BW-SLOW-NEXT:    vpalignr {{.*#+}} ymm7 = mem[8,9,10,11,12,13,14,15],ymm7[0,1,2,3,4,5,6,7],mem[24,25,26,27,28,29,30,31],ymm7[16,17,18,19,20,21,22,23]
-; AVX512BW-SLOW-NEXT:    vmovdqa 192(%rdi), %xmm8
-; AVX512BW-SLOW-NEXT:    vpalignr {{.*#+}} xmm6 = xmm6[8,9,10,11,12,13,14,15],xmm8[0,1,2,3,4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vinserti128 $1, %xmm6, %ymm0, %ymm6
-; AVX512BW-SLOW-NEXT:    vperm2i128 {{.*#+}} ymm6 = ymm7[2,3],ymm6[2,3]
-; AVX512BW-SLOW-NEXT:    vinserti128 $1, %xmm8, %ymm0, %ymm7
-; AVX512BW-SLOW-NEXT:    vmovdqa 128(%rdi), %ymm8
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm8[0,1,2,3,4,5],ymm7[6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} xmm9 = [4,11]
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm7 = ymm9[0,1,2,3],ymm7[4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa 192(%rdi), %ymm9
-; AVX512BW-SLOW-NEXT:    vpalignr {{.*#+}} ymm8 = ymm8[8,9,10,11,12,13,14,15],ymm9[0,1,2,3,4,5,6,7],ymm8[24,25,26,27,28,29,30,31],ymm9[16,17,18,19,20,21,22,23]
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} xmm9 = [5,12]
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm8 = ymm9[0,1,2,3],ymm8[4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vbroadcasti128 {{.*#+}} ymm9 = [4,11,4,11]
-; AVX512BW-SLOW-NEXT:    # ymm9 = mem[0,1,0,1]
-; AVX512BW-SLOW-NEXT:    vpermi2q 192(%rdi), %zmm3, %zmm9
-; AVX512BW-SLOW-NEXT:    vmovdqa {{.*#+}} xmm3 = [6,13]
-; AVX512BW-SLOW-NEXT:    vpermi2q %zmm4, %zmm5, %zmm3
-; AVX512BW-SLOW-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3],ymm9[4,5,6,7]
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm2, (%rcx)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm6, (%r8)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm7, (%r9)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm8, (%r10)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm3, (%rax)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: load_i64_stride7_vf4:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-FAST-NEXT:    vmovdqa64 128(%rdi), %zmm3
-; AVX512BW-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm4
-; AVX512BW-FAST-NEXT:    vmovdqa64 (%rdi), %zmm5
-; AVX512BW-FAST-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm0
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,7,14,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm1
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm1[0,1,2,3,4,5],ymm0[6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <9,0,7,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm5, %zmm4, %zmm1
-; AVX512BW-FAST-NEXT:    vpbroadcastq 176(%rdi), %ymm2
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm2[6,7]
-; AVX512BW-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm2 = [0,7,0,7]
-; AVX512BW-FAST-NEXT:    # ymm2 = mem[0,1,0,1]
-; AVX512BW-FAST-NEXT:    vmovdqa 128(%rdi), %ymm6
-; AVX512BW-FAST-NEXT:    vpermi2q 160(%rdi), %ymm6, %ymm2
-; AVX512BW-FAST-NEXT:    vmovdqa 16(%rdi), %xmm7
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} xmm7 = xmm7[0,1],mem[2,3]
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm7[0,1,2,3],ymm2[4,5,6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa 64(%rdi), %ymm7
-; AVX512BW-FAST-NEXT:    vpalignr {{.*#+}} ymm7 = mem[8,9,10,11,12,13,14,15],ymm7[0,1,2,3,4,5,6,7],mem[24,25,26,27,28,29,30,31],ymm7[16,17,18,19,20,21,22,23]
-; AVX512BW-FAST-NEXT:    vmovdqa 192(%rdi), %xmm8
-; AVX512BW-FAST-NEXT:    vpalignr {{.*#+}} xmm9 = mem[8,9,10,11,12,13,14,15],xmm8[0,1,2,3,4,5,6,7]
-; AVX512BW-FAST-NEXT:    vinserti128 $1, %xmm9, %ymm0, %ymm9
-; AVX512BW-FAST-NEXT:    vperm2i128 {{.*#+}} ymm7 = ymm7[2,3],ymm9[2,3]
-; AVX512BW-FAST-NEXT:    vinserti128 $1, %xmm8, %ymm0, %ymm8
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm8 = ymm6[0,1,2,3,4,5],ymm8[6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} xmm9 = [4,11]
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm8 = ymm9[0,1,2,3],ymm8[4,5,6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa 192(%rdi), %ymm9
-; AVX512BW-FAST-NEXT:    vpalignr {{.*#+}} ymm6 = ymm6[8,9,10,11,12,13,14,15],ymm9[0,1,2,3,4,5,6,7],ymm6[24,25,26,27,28,29,30,31],ymm9[16,17,18,19,20,21,22,23]
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} xmm9 = [5,12]
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm6 = ymm9[0,1,2,3],ymm6[4,5,6,7]
-; AVX512BW-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm9 = [4,11,4,11]
-; AVX512BW-FAST-NEXT:    # ymm9 = mem[0,1,0,1]
-; AVX512BW-FAST-NEXT:    vpermi2q 192(%rdi), %zmm3, %zmm9
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} xmm3 = [6,13]
-; AVX512BW-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm3
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3],ymm9[4,5,6,7]
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm1, (%rdx)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm2, (%rcx)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm7, (%r8)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm8, (%r9)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm6, (%r10)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm3, (%rax)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: load_i64_stride7_vf4:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-FAST-NEXT:    vmovdqa64 128(%rdi), %zmm3
+; AVX512-FAST-NEXT:    vmovdqa64 64(%rdi), %zmm4
+; AVX512-FAST-NEXT:    vmovdqa64 (%rdi), %zmm5
+; AVX512-FAST-NEXT:    vinserti128 $1, 160(%rdi), %ymm0, %ymm0
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,7,14,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm1
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm1[0,1,2,3,4,5],ymm0[6,7]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <9,0,7,u>
+; AVX512-FAST-NEXT:    vpermi2q %zmm5, %zmm4, %zmm1
+; AVX512-FAST-NEXT:    vpbroadcastq 176(%rdi), %ymm2
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1,2,3,4,5],ymm2[6,7]
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm2 = [0,7,0,7]
+; AVX512-FAST-NEXT:    # ymm2 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vmovdqa 128(%rdi), %ymm6
+; AVX512-FAST-NEXT:    vpermi2q 160(%rdi), %ymm6, %ymm2
+; AVX512-FAST-NEXT:    vmovdqa 16(%rdi), %xmm7
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} xmm7 = xmm7[0,1],mem[2,3]
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm7[0,1,2,3],ymm2[4,5,6,7]
+; AVX512-FAST-NEXT:    vmovdqa 64(%rdi), %ymm7
+; AVX512-FAST-NEXT:    vpalignr {{.*#+}} ymm7 = mem[8,9,10,11,12,13,14,15],ymm7[0,1,2,3,4,5,6,7],mem[24,25,26,27,28,29,30,31],ymm7[16,17,18,19,20,21,22,23]
+; AVX512-FAST-NEXT:    vmovdqa 192(%rdi), %xmm8
+; AVX512-FAST-NEXT:    vpalignr {{.*#+}} xmm9 = mem[8,9,10,11,12,13,14,15],xmm8[0,1,2,3,4,5,6,7]
+; AVX512-FAST-NEXT:    vinserti128 $1, %xmm9, %ymm0, %ymm9
+; AVX512-FAST-NEXT:    vperm2i128 {{.*#+}} ymm7 = ymm7[2,3],ymm9[2,3]
+; AVX512-FAST-NEXT:    vinserti128 $1, %xmm8, %ymm0, %ymm8
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm8 = ymm6[0,1,2,3,4,5],ymm8[6,7]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} xmm9 = [4,11]
+; AVX512-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm8 = ymm9[0,1,2,3],ymm8[4,5,6,7]
+; AVX512-FAST-NEXT:    vmovdqa 192(%rdi), %ymm9
+; AVX512-FAST-NEXT:    vpalignr {{.*#+}} ymm6 = ymm6[8,9,10,11,12,13,14,15],ymm9[0,1,2,3,4,5,6,7],ymm6[24,25,26,27,28,29,30,31],ymm9[16,17,18,19,20,21,22,23]
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} xmm9 = [5,12]
+; AVX512-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm9
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm6 = ymm9[0,1,2,3],ymm6[4,5,6,7]
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm9 = [4,11,4,11]
+; AVX512-FAST-NEXT:    # ymm9 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vpermi2q 192(%rdi), %zmm3, %zmm9
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} xmm3 = [6,13]
+; AVX512-FAST-NEXT:    vpermi2q %zmm4, %zmm5, %zmm3
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1,2,3],ymm9[4,5,6,7]
+; AVX512-FAST-NEXT:    vmovdqa %ymm0, (%rsi)
+; AVX512-FAST-NEXT:    vmovdqa %ymm1, (%rdx)
+; AVX512-FAST-NEXT:    vmovdqa %ymm2, (%rcx)
+; AVX512-FAST-NEXT:    vmovdqa %ymm7, (%r8)
+; AVX512-FAST-NEXT:    vmovdqa %ymm8, (%r9)
+; AVX512-FAST-NEXT:    vmovdqa %ymm6, (%r10)
+; AVX512-FAST-NEXT:    vmovdqa %ymm3, (%rax)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %wide.vec = load <28 x i64>, ptr %in.vec, align 64
   %strided.vec0 = shufflevector <28 x i64> %wide.vec, <28 x i64> poison, <4 x i32> <i32 0, i32 7, i32 14, i32 21>
   %strided.vec1 = shufflevector <28 x i64> %wide.vec, <28 x i64> poison, <4 x i32> <i32 1, i32 8, i32 15, i32 22>
@@ -12753,6 +12645,10 @@ define void @load_i64_stride7_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512BW-FAST: {{.*}}
+; AVX512BW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-8.ll
index 7d29ed1192bd4df..884a4abfe646c3b 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i64-stride-8.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -11300,6 +11300,8 @@ define void @load_i64_stride8_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, pt
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-2.ll
index f7f6a5e32a2eb5f..06b95c94ce16d49 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -586,6 +586,8 @@ define void @load_i8_stride2_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1) nou
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-3.ll
index 19c97ece8937f77..7d7b43a2234c2ba 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -1629,6 +1629,8 @@ define void @load_i8_stride3_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-4.ll
index 764020b3b48c6d7..df33d3b9d8fd4b1 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -2082,6 +2082,8 @@ define void @load_i8_stride4_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-5.ll
index 1532859fb115a3a..29e3247e1451a5f 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -4260,6 +4260,8 @@ define void @load_i8_stride5_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-6.ll
index 6d38dcbbcf0743f..9ec0e2f036e85c5 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -5042,6 +5042,8 @@ define void @load_i8_stride6_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-7.ll
index 44459e1756009de..3b13741aaf7d5b1 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -12348,6 +12348,8 @@ define void @load_i8_stride7_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-8.ll
index 3ca2aae97def06c..78a2729caf5a4cd 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-load-i8-stride-8.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved loads.
 
@@ -13262,6 +13262,8 @@ define void @load_i8_stride8_vf64(ptr %in.vec, ptr %out.vec0, ptr %out.vec1, ptr
 ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
 ; AVX: {{.*}}
 ; AVX2: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-2.ll
index 31f006340f46890..5cd794b7f1c4dc9 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -571,6 +571,8 @@ define void @store_i16_stride2_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.v
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-3.ll
index 3cc97d69e8b033e..f91dd72bfe3ff5a 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -2534,6 +2534,8 @@ define void @store_i16_stride3_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-4.ll
index 6cd1f13398c6035..0e3ba8789ed1b8c 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -2765,6 +2765,8 @@ define void @store_i16_stride4_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-5.ll
index 2670cc353aa173c..2abbb6805cf41c4 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -6485,6 +6485,8 @@ define void @store_i16_stride5_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-6.ll
index 66fdffccdfddcef..a89f4a4b9d4571f 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -9598,6 +9598,8 @@ define void @store_i16_stride6_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-7.ll
index fe4c2504228a209..64be8677d5a49dc 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -13500,6 +13500,8 @@ define void @store_i16_stride7_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-8.ll
index e5a6eea44b8ad61..1cb76e97601528b 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-8.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -7805,6 +7805,8 @@ define void @store_i16_stride8_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-2.ll
index cabe23547039ba3..0a59406d865e1ed 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -805,6 +805,8 @@ define void @store_i32_stride2_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.v
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-3.ll
index afe43e3c7379d4b..5ff48a44c347a68 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -2799,6 +2799,8 @@ define void @store_i32_stride3_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX: {{.*}}
 ; AVX1: {{.*}}
 ; AVX2-ONLY: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-4.ll
index bf6991e9071774d..31e462a139561bd 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -93,65 +93,35 @@ define void @store_i32_stride4_vf2(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 ; AVX2-FAST-PERLANE-NEXT:    vzeroupper
 ; AVX2-FAST-PERLANE-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: store_i32_stride4_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
-; AVX512F-SLOW-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovsd {{.*#+}} xmm2 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm2[0],xmm1[0]
-; AVX512F-SLOW-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX512F-SLOW-NEXT:    vshufps {{.*#+}} ymm0 = ymm0[0,2,1,3,4,6,5,7]
-; AVX512F-SLOW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512F-SLOW-NEXT:    vmovaps %ymm0, (%r8)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: store_i32_stride4_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; AVX512-SLOW-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovsd {{.*#+}} xmm2 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm2[0],xmm1[0]
+; AVX512-SLOW-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX512-SLOW-NEXT:    vshufps {{.*#+}} ymm0 = ymm0[0,2,1,3,4,6,5,7]
+; AVX512-SLOW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
+; AVX512-SLOW-NEXT:    vmovaps %ymm0, (%r8)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: store_i32_stride4_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
-; AVX512F-FAST-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovsd {{.*#+}} xmm2 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm2[0],xmm1[0]
-; AVX512F-FAST-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX512F-FAST-NEXT:    vmovaps {{.*#+}} ymm1 = [0,2,4,6,1,3,5,7]
-; AVX512F-FAST-NEXT:    vpermps %ymm0, %ymm1, %ymm0
-; AVX512F-FAST-NEXT:    vmovaps %ymm0, (%r8)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: store_i32_stride4_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
-; AVX512BW-SLOW-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovsd {{.*#+}} xmm2 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm2[0],xmm1[0]
-; AVX512BW-SLOW-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX512BW-SLOW-NEXT:    vshufps {{.*#+}} ymm0 = ymm0[0,2,1,3,4,6,5,7]
-; AVX512BW-SLOW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512BW-SLOW-NEXT:    vmovaps %ymm0, (%r8)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: store_i32_stride4_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
-; AVX512BW-FAST-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovsd {{.*#+}} xmm2 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm2[0],xmm1[0]
-; AVX512BW-FAST-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX512BW-FAST-NEXT:    vmovaps {{.*#+}} ymm1 = [0,2,4,6,1,3,5,7]
-; AVX512BW-FAST-NEXT:    vpermps %ymm0, %ymm1, %ymm0
-; AVX512BW-FAST-NEXT:    vmovaps %ymm0, (%r8)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: store_i32_stride4_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; AVX512-FAST-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovsd {{.*#+}} xmm2 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm2[0],xmm1[0]
+; AVX512-FAST-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX512-FAST-NEXT:    vmovaps {{.*#+}} ymm1 = [0,2,4,6,1,3,5,7]
+; AVX512-FAST-NEXT:    vpermps %ymm0, %ymm1, %ymm0
+; AVX512-FAST-NEXT:    vmovaps %ymm0, (%r8)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %in.vec0 = load <2 x i32>, ptr %in.vecptr0, align 64
   %in.vec1 = load <2 x i32>, ptr %in.vecptr1, align 64
   %in.vec2 = load <2 x i32>, ptr %in.vecptr2, align 64
@@ -2977,14 +2947,18 @@ define void @store_i32_stride4_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX: {{.*}}
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-5.ll
index 4bb7f80a795647b..f0c2f3f23a2d038 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -6957,6 +6957,8 @@ define void @store_i32_stride5_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX: {{.*}}
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-6.ll
index 96543b486b35d1c..6e09eba1de1bc9d 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -9539,6 +9539,8 @@ define void @store_i32_stride6_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX: {{.*}}
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-7.ll
index fe216de5231a3cf..33c9d9e182c34c5 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -185,109 +185,57 @@ define void @store_i32_stride7_vf2(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 ; AVX2-FAST-PERLANE-NEXT:    vzeroupper
 ; AVX2-FAST-PERLANE-NEXT:    retq
 ;
-; AVX512F-SLOW-LABEL: store_i32_stride7_vf2:
-; AVX512F-SLOW:       # %bb.0:
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-SLOW-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
-; AVX512F-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
-; AVX512F-SLOW-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm2
-; AVX512F-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-SLOW-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
-; AVX512F-SLOW-NEXT:    vpunpcklqdq {{.*#+}} ymm0 = ymm0[0],ymm3[0],ymm0[2],ymm3[2]
-; AVX512F-SLOW-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,2,4,6,16,20,18,1,3,5,7,17,21,19,u,u>
-; AVX512F-SLOW-NEXT:    vpermi2d %zmm0, %zmm2, %zmm1
-; AVX512F-SLOW-NEXT:    vextracti32x4 $2, %zmm1, 32(%rax)
-; AVX512F-SLOW-NEXT:    vextracti32x4 $3, %zmm1, %xmm0
-; AVX512F-SLOW-NEXT:    vmovq %xmm0, 48(%rax)
-; AVX512F-SLOW-NEXT:    vmovdqa %ymm1, (%rax)
-; AVX512F-SLOW-NEXT:    vzeroupper
-; AVX512F-SLOW-NEXT:    retq
+; AVX512-SLOW-LABEL: store_i32_stride7_vf2:
+; AVX512-SLOW:       # %bb.0:
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-SLOW-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-SLOW-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
+; AVX512-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-SLOW-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
+; AVX512-SLOW-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
+; AVX512-SLOW-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm2
+; AVX512-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-SLOW-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX512-SLOW-NEXT:    vpunpcklqdq {{.*#+}} ymm0 = ymm0[0],ymm3[0],ymm0[2],ymm3[2]
+; AVX512-SLOW-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,2,4,6,16,20,18,1,3,5,7,17,21,19,u,u>
+; AVX512-SLOW-NEXT:    vpermi2d %zmm0, %zmm2, %zmm1
+; AVX512-SLOW-NEXT:    vextracti32x4 $2, %zmm1, 32(%rax)
+; AVX512-SLOW-NEXT:    vextracti32x4 $3, %zmm1, %xmm0
+; AVX512-SLOW-NEXT:    vmovq %xmm0, 48(%rax)
+; AVX512-SLOW-NEXT:    vmovdqa %ymm1, (%rax)
+; AVX512-SLOW-NEXT:    vzeroupper
+; AVX512-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: store_i32_stride7_vf2:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
-; AVX512F-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
-; AVX512F-FAST-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm2
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-FAST-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,2,4,u>
-; AVX512F-FAST-NEXT:    vpermi2q %ymm3, %ymm0, %ymm1
-; AVX512F-FAST-NEXT:    vinserti64x4 $1, %ymm1, %zmm2, %zmm0
-; AVX512F-FAST-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,2,4,6,8,10,12,1,3,5,7,9,11,13,u,u>
-; AVX512F-FAST-NEXT:    vpermd %zmm0, %zmm1, %zmm0
-; AVX512F-FAST-NEXT:    vextracti32x4 $2, %zmm0, 32(%rax)
-; AVX512F-FAST-NEXT:    vextracti32x4 $3, %zmm0, %xmm1
-; AVX512F-FAST-NEXT:    vmovq %xmm1, 48(%rax)
-; AVX512F-FAST-NEXT:    vmovdqa %ymm0, (%rax)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
-;
-; AVX512BW-SLOW-LABEL: store_i32_stride7_vf2:
-; AVX512BW-SLOW:       # %bb.0:
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-SLOW-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-SLOW-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
-; AVX512BW-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
-; AVX512BW-SLOW-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm2
-; AVX512BW-SLOW-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-SLOW-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
-; AVX512BW-SLOW-NEXT:    vpunpcklqdq {{.*#+}} ymm0 = ymm0[0],ymm3[0],ymm0[2],ymm3[2]
-; AVX512BW-SLOW-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,2,4,6,16,20,18,1,3,5,7,17,21,19,u,u>
-; AVX512BW-SLOW-NEXT:    vpermi2d %zmm0, %zmm2, %zmm1
-; AVX512BW-SLOW-NEXT:    vextracti32x4 $2, %zmm1, 32(%rax)
-; AVX512BW-SLOW-NEXT:    vextracti32x4 $3, %zmm1, %xmm0
-; AVX512BW-SLOW-NEXT:    vmovq %xmm0, 48(%rax)
-; AVX512BW-SLOW-NEXT:    vmovdqa %ymm1, (%rax)
-; AVX512BW-SLOW-NEXT:    vzeroupper
-; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: store_i32_stride7_vf2:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
-; AVX512BW-FAST-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm2
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,2,4,u>
-; AVX512BW-FAST-NEXT:    vpermi2q %ymm3, %ymm0, %ymm1
-; AVX512BW-FAST-NEXT:    vinserti64x4 $1, %ymm1, %zmm2, %zmm0
-; AVX512BW-FAST-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,2,4,6,8,10,12,1,3,5,7,9,11,13,u,u>
-; AVX512BW-FAST-NEXT:    vpermd %zmm0, %zmm1, %zmm0
-; AVX512BW-FAST-NEXT:    vextracti32x4 $2, %zmm0, 32(%rax)
-; AVX512BW-FAST-NEXT:    vextracti32x4 $3, %zmm0, %xmm1
-; AVX512BW-FAST-NEXT:    vmovq %xmm1, 48(%rax)
-; AVX512BW-FAST-NEXT:    vmovdqa %ymm0, (%rax)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: store_i32_stride7_vf2:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
+; AVX512-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
+; AVX512-FAST-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm2
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-FAST-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = <0,2,4,u>
+; AVX512-FAST-NEXT:    vpermi2q %ymm3, %ymm0, %ymm1
+; AVX512-FAST-NEXT:    vinserti64x4 $1, %ymm1, %zmm2, %zmm0
+; AVX512-FAST-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,2,4,6,8,10,12,1,3,5,7,9,11,13,u,u>
+; AVX512-FAST-NEXT:    vpermd %zmm0, %zmm1, %zmm0
+; AVX512-FAST-NEXT:    vextracti32x4 $2, %zmm0, 32(%rax)
+; AVX512-FAST-NEXT:    vextracti32x4 $3, %zmm0, %xmm1
+; AVX512-FAST-NEXT:    vmovq %xmm1, 48(%rax)
+; AVX512-FAST-NEXT:    vmovdqa %ymm0, (%rax)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
   %in.vec0 = load <2 x i32>, ptr %in.vecptr0, align 64
   %in.vec1 = load <2 x i32>, ptr %in.vecptr1, align 64
   %in.vec2 = load <2 x i32>, ptr %in.vecptr2, align 64
@@ -12498,14 +12446,18 @@ define void @store_i32_stride7_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX2-ONLY: {{.*}}
+; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
+; AVX512BW-SLOW: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}
 ; AVX512DQBW-SLOW: {{.*}}
+; AVX512F-FAST: {{.*}}
 ; AVX512F-ONLY-FAST: {{.*}}
 ; AVX512F-ONLY-SLOW: {{.*}}
+; AVX512F-SLOW: {{.*}}
 ; FALLBACK0: {{.*}}
 ; FALLBACK1: {{.*}}
 ; FALLBACK10: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-8.ll
index f2b18eb113bbc1d..35d12ffc3a8d306 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i32-stride-8.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -8099,6 +8099,8 @@ define void @store_i32_stride8_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-2.ll
index a9a213c8a590547..57e030ff2a1cee0 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -1578,6 +1578,8 @@ define void @store_i64_stride2_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.v
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-3.ll
index 780c3c55d3d3d0d..55c8cfb9b49566b 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -2700,6 +2700,8 @@ define void @store_i64_stride3_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-4.ll
index 6db64bdc2adac9c..15fa50557067273 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -4245,6 +4245,8 @@ define void @store_i64_stride4_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-5.ll
index c0e1a17243a7e66..fcc1958a46ed497 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -6184,6 +6184,8 @@ define void @store_i64_stride5_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-6.ll
index f5cd64c698d5b1e..12c18c325e5dc25 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -14630,6 +14630,8 @@ define void @store_i64_stride6_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-SLOW: {{.*}}
 ; AVX512F-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-7.ll
index 7f999e06b304cb0..3d88b7b550dacd6 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -20320,6 +20320,8 @@ define void @store_i64_stride7_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-SLOW: {{.*}}
 ; AVX512F-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-8.ll
index af03a07a592e6bf..293a3447577f765 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-8.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -16805,6 +16805,8 @@ define void @store_i64_stride8_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.ve
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-SLOW: {{.*}}
 ; AVX512F-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-2.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-2.ll
index 43d10f2d4f93af2..4aa54baa1ee15ad 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-2.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-2.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -355,6 +355,8 @@ define void @store_i8_stride2_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.ve
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-3.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-3.ll
index 06e94ccb92e6910..035db822be5180b 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-3.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-3.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -1113,6 +1113,8 @@ define void @store_i8_stride3_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 ; AVX2-FAST: {{.*}}
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-4.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-4.ll
index ea7a3a0920c24ac..9e7d970a6abcde6 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-4.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-4.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -727,6 +727,8 @@ define void @store_i8_stride4_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 ; AVX2-FAST-PERLANE: {{.*}}
 ; AVX2-SLOW: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-FAST: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-5.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-5.ll
index 04b9dcb418ceaf1..f75d96b70dcbd03 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-5.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-5.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -4819,6 +4819,8 @@ define void @store_i8_stride5_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-6.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-6.ll
index 0358af331d87ab6..f6fb4837f19cff8 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-6.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-6.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -5301,6 +5301,8 @@ define void @store_i8_stride6_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512DQ-FAST: {{.*}}
 ; AVX512DQ-SLOW: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-7.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-7.ll
index 0909c3f43ea37e0..0ec4e796fcfacaa 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-7.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-7.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -9660,6 +9660,8 @@ define void @store_i8_stride7_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 ; AVX1: {{.*}}
 ; AVX2: {{.*}}
 ; AVX512: {{.*}}
+; AVX512-FAST: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}
 ; AVX512DQBW-FAST: {{.*}}

diff  --git a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-8.ll b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-8.ll
index f8a3ab852b0924a..86843b6c204f425 100644
--- a/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-8.ll
+++ b/llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-8.ll
@@ -4,14 +4,14 @@
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-SLOW,FALLBACK2
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST,FALLBACK3
 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX1,AVX2,AVX2-ONLY,AVX2-FAST-PERLANE,FALLBACK4
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
-; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512F-ONLY-SLOW,FALLBACK5
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512F-ONLY-FAST,FALLBACK6
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-SLOW,AVX512F-SLOW,AVX512DQ-SLOW,FALLBACK7
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512F,AVX512-FAST,AVX512F-FAST,AVX512DQ-FAST,FALLBACK8
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512BW-ONLY-SLOW,FALLBACK9
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512BW-ONLY-FAST,FALLBACK10
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-SLOW,AVX512BW-SLOW,AVX512DQBW-SLOW,FALLBACK11
+; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX512,AVX512BW,AVX512-FAST,AVX512BW-FAST,AVX512DQBW-FAST,FALLBACK12
 
 ; These patterns are produced by LoopVectorizer for interleaved stores.
 
@@ -572,45 +572,45 @@ define void @store_i8_stride8_vf8(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vecp
 ; AVX512F-SLOW-NEXT:    vzeroupper
 ; AVX512F-SLOW-NEXT:    retq
 ;
-; AVX512F-FAST-LABEL: store_i8_stride8_vf8:
-; AVX512F-FAST:       # %bb.0:
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512F-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512F-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512F-FAST-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
-; AVX512F-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
-; AVX512F-FAST-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
-; AVX512F-FAST-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm1
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm2 = [1,3,5,7,1,3,5,7]
-; AVX512F-FAST-NEXT:    # ymm2 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vpermd %ymm1, %ymm2, %ymm3
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm4 = <u,u,u,u,0,4,8,12,u,u,u,u,1,5,9,13,u,u,u,u,2,6,10,14,u,u,u,u,3,7,11,15>
-; AVX512F-FAST-NEXT:    vpshufb %ymm4, %ymm3, %ymm3
-; AVX512F-FAST-NEXT:    vpermd %ymm0, %ymm2, %ymm2
-; AVX512F-FAST-NEXT:    vmovdqa {{.*#+}} ymm5 = <0,4,8,12,u,u,u,u,1,5,9,13,u,u,u,u,2,6,10,14,u,u,u,u,3,7,11,15,u,u,u,u>
-; AVX512F-FAST-NEXT:    vpshufb %ymm5, %ymm2, %ymm2
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm2[0],ymm3[1],ymm2[2],ymm3[3],ymm2[4],ymm3[5],ymm2[6],ymm3[7]
-; AVX512F-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm3 = [0,2,4,6,0,2,4,6]
-; AVX512F-FAST-NEXT:    # ymm3 = mem[0,1,0,1]
-; AVX512F-FAST-NEXT:    vpermd %ymm1, %ymm3, %ymm1
-; AVX512F-FAST-NEXT:    vpshufb %ymm4, %ymm1, %ymm1
-; AVX512F-FAST-NEXT:    vpermd %ymm0, %ymm3, %ymm0
-; AVX512F-FAST-NEXT:    vpshufb %ymm5, %ymm0, %ymm0
-; AVX512F-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
-; AVX512F-FAST-NEXT:    vinserti64x4 $1, %ymm2, %zmm0, %zmm0
-; AVX512F-FAST-NEXT:    vmovdqa64 %zmm0, (%rax)
-; AVX512F-FAST-NEXT:    vzeroupper
-; AVX512F-FAST-NEXT:    retq
+; AVX512-FAST-LABEL: store_i8_stride8_vf8:
+; AVX512-FAST:       # %bb.0:
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
+; AVX512-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r11
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
+; AVX512-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
+; AVX512-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
+; AVX512-FAST-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
+; AVX512-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
+; AVX512-FAST-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX512-FAST-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm1
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm2 = [1,3,5,7,1,3,5,7]
+; AVX512-FAST-NEXT:    # ymm2 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vpermd %ymm1, %ymm2, %ymm3
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm4 = <u,u,u,u,0,4,8,12,u,u,u,u,1,5,9,13,u,u,u,u,2,6,10,14,u,u,u,u,3,7,11,15>
+; AVX512-FAST-NEXT:    vpshufb %ymm4, %ymm3, %ymm3
+; AVX512-FAST-NEXT:    vpermd %ymm0, %ymm2, %ymm2
+; AVX512-FAST-NEXT:    vmovdqa {{.*#+}} ymm5 = <0,4,8,12,u,u,u,u,1,5,9,13,u,u,u,u,2,6,10,14,u,u,u,u,3,7,11,15,u,u,u,u>
+; AVX512-FAST-NEXT:    vpshufb %ymm5, %ymm2, %ymm2
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm2[0],ymm3[1],ymm2[2],ymm3[3],ymm2[4],ymm3[5],ymm2[6],ymm3[7]
+; AVX512-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm3 = [0,2,4,6,0,2,4,6]
+; AVX512-FAST-NEXT:    # ymm3 = mem[0,1,0,1]
+; AVX512-FAST-NEXT:    vpermd %ymm1, %ymm3, %ymm1
+; AVX512-FAST-NEXT:    vpshufb %ymm4, %ymm1, %ymm1
+; AVX512-FAST-NEXT:    vpermd %ymm0, %ymm3, %ymm0
+; AVX512-FAST-NEXT:    vpshufb %ymm5, %ymm0, %ymm0
+; AVX512-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
+; AVX512-FAST-NEXT:    vinserti64x4 $1, %ymm2, %zmm0, %zmm0
+; AVX512-FAST-NEXT:    vmovdqa64 %zmm0, (%rax)
+; AVX512-FAST-NEXT:    vzeroupper
+; AVX512-FAST-NEXT:    retq
 ;
 ; AVX512BW-SLOW-LABEL: store_i8_stride8_vf8:
 ; AVX512BW-SLOW:       # %bb.0:
@@ -655,46 +655,6 @@ define void @store_i8_stride8_vf8(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vecp
 ; AVX512BW-SLOW-NEXT:    vmovdqa64 %zmm0, (%rax)
 ; AVX512BW-SLOW-NEXT:    vzeroupper
 ; AVX512BW-SLOW-NEXT:    retq
-;
-; AVX512BW-FAST-LABEL: store_i8_stride8_vf8:
-; AVX512BW-FAST:       # %bb.0:
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %rax
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512BW-FAST-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm3 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vmovq {{.*#+}} xmm4 = mem[0],zero
-; AVX512BW-FAST-NEXT:    vpunpcklqdq {{.*#+}} xmm3 = xmm4[0],xmm3[0]
-; AVX512BW-FAST-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
-; AVX512BW-FAST-NEXT:    vinserti128 $1, %xmm3, %ymm2, %ymm1
-; AVX512BW-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm2 = [1,3,5,7,1,3,5,7]
-; AVX512BW-FAST-NEXT:    # ymm2 = mem[0,1,0,1]
-; AVX512BW-FAST-NEXT:    vpermd %ymm1, %ymm2, %ymm3
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm4 = <u,u,u,u,0,4,8,12,u,u,u,u,1,5,9,13,u,u,u,u,2,6,10,14,u,u,u,u,3,7,11,15>
-; AVX512BW-FAST-NEXT:    vpshufb %ymm4, %ymm3, %ymm3
-; AVX512BW-FAST-NEXT:    vpermd %ymm0, %ymm2, %ymm2
-; AVX512BW-FAST-NEXT:    vmovdqa {{.*#+}} ymm5 = <0,4,8,12,u,u,u,u,1,5,9,13,u,u,u,u,2,6,10,14,u,u,u,u,3,7,11,15,u,u,u,u>
-; AVX512BW-FAST-NEXT:    vpshufb %ymm5, %ymm2, %ymm2
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm2 = ymm2[0],ymm3[1],ymm2[2],ymm3[3],ymm2[4],ymm3[5],ymm2[6],ymm3[7]
-; AVX512BW-FAST-NEXT:    vbroadcasti128 {{.*#+}} ymm3 = [0,2,4,6,0,2,4,6]
-; AVX512BW-FAST-NEXT:    # ymm3 = mem[0,1,0,1]
-; AVX512BW-FAST-NEXT:    vpermd %ymm1, %ymm3, %ymm1
-; AVX512BW-FAST-NEXT:    vpshufb %ymm4, %ymm1, %ymm1
-; AVX512BW-FAST-NEXT:    vpermd %ymm0, %ymm3, %ymm0
-; AVX512BW-FAST-NEXT:    vpshufb %ymm5, %ymm0, %ymm0
-; AVX512BW-FAST-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
-; AVX512BW-FAST-NEXT:    vinserti64x4 $1, %ymm2, %zmm0, %zmm0
-; AVX512BW-FAST-NEXT:    vmovdqa64 %zmm0, (%rax)
-; AVX512BW-FAST-NEXT:    vzeroupper
-; AVX512BW-FAST-NEXT:    retq
   %in.vec0 = load <8 x i8>, ptr %in.vecptr0, align 64
   %in.vec1 = load <8 x i8>, ptr %in.vecptr1, align 64
   %in.vec2 = load <8 x i8>, ptr %in.vecptr2, align 64
@@ -7028,6 +6988,7 @@ define void @store_i8_stride8_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %in.vec
 }
 ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
 ; AVX1: {{.*}}
+; AVX512-SLOW: {{.*}}
 ; AVX512BW: {{.*}}
 ; AVX512BW-ONLY-FAST: {{.*}}
 ; AVX512BW-ONLY-SLOW: {{.*}}


        
