; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
; RUN: llc < %s -mtriple=x86_64-unknown-unknown | FileCheck %s

declare void @llvm.lifetime.start.p0(i64 immarg, ptr nocapture) #0
declare void @llvm.lifetime.end.p0(i64 immarg, ptr nocapture) #0

define void @SHA256_Compress_Generic(ptr noundef %ctx) #1 {
; CHECK-LABEL: SHA256_Compress_Generic:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    movbel 0, %eax
; CHECK-NEXT:    movbel 12(%rdi), %ecx
; CHECK-NEXT:    vmovd %eax, %xmm0
; CHECK-NEXT:    vmovdqa {{.*#+}} xmm1 = [128,128,128,128,0,1,2,3,128,128,128,128,128,128,128,128]
; CHECK-NEXT:    vpshufb %xmm1, %xmm0, %xmm2
; CHECK-NEXT:    vpsrld $17, %xmm2, %xmm0
; CHECK-NEXT:    vpslld $15, %xmm2, %xmm3
; CHECK-NEXT:    vpor %xmm0, %xmm3, %xmm0
; CHECK-NEXT:    vpsrld $19, %xmm2, %xmm3
; CHECK-NEXT:    vpslld $13, %xmm2, %xmm4
; CHECK-NEXT:    vpor %xmm3, %xmm4, %xmm3
; CHECK-NEXT:    vpxor %xmm3, %xmm0, %xmm0
; CHECK-NEXT:    vpxor %xmm2, %xmm0, %xmm0
; CHECK-NEXT:    vmovd %ecx, %xmm3
; CHECK-NEXT:    vpshufb %xmm1, %xmm3, %xmm1
; CHECK-NEXT:    vpaddd %xmm0, %xmm1, %xmm1
; CHECK-NEXT:    vpsrld $17, %xmm1, %xmm0
; CHECK-NEXT:    vpslld $15, %xmm1, %xmm3
; CHECK-NEXT:    vpor %xmm0, %xmm3, %xmm0
; CHECK-NEXT:    vpsrld $19, %xmm1, %xmm3
; CHECK-NEXT:    vpslld $13, %xmm1, %xmm4
; CHECK-NEXT:    vpor %xmm3, %xmm4, %xmm3
; CHECK-NEXT:    vpxor %xmm3, %xmm0, %xmm0
; CHECK-NEXT:    vpxor %xmm1, %xmm0, %xmm0
; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
; CHECK-NEXT:    vpsrld $17, %xmm0, %xmm3
; CHECK-NEXT:    vpslld $15, %xmm0, %xmm4
; CHECK-NEXT:    vpor %xmm3, %xmm4, %xmm3
; CHECK-NEXT:    vpsrld $19, %xmm0, %xmm4
; CHECK-NEXT:    vpslld $13, %xmm0, %xmm5
; CHECK-NEXT:    vpor %xmm4, %xmm5, %xmm4
; CHECK-NEXT:    vpxor %xmm4, %xmm3, %xmm3
; CHECK-NEXT:    vpsrld $10, %xmm0, %xmm0
; CHECK-NEXT:    vpxor %xmm0, %xmm3, %xmm0
; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
; CHECK-NEXT:    vpsrld $17, %xmm0, %xmm3
; CHECK-NEXT:    vpslld $15, %xmm0, %xmm4
; CHECK-NEXT:    vpor %xmm3, %xmm4, %xmm3
; CHECK-NEXT:    vpsrld $19, %xmm0, %xmm4
; CHECK-NEXT:    vpslld $13, %xmm0, %xmm5
; CHECK-NEXT:    vpor %xmm4, %xmm5, %xmm4
; CHECK-NEXT:    vpxor %xmm4, %xmm3, %xmm3
; CHECK-NEXT:    vpsrld $10, %xmm0, %xmm4
; CHECK-NEXT:    vpxor %xmm4, %xmm3, %xmm3
; CHECK-NEXT:    vpblendd {{.*#+}} xmm2 = xmm1[0],xmm2[1],xmm1[2,3]
; CHECK-NEXT:    vpshufd {{.*#+}} xmm2 = xmm2[1,0,2,3]
; CHECK-NEXT:    vpaddd %xmm3, %xmm2, %xmm2
; CHECK-NEXT:    vpsrld $17, %xmm2, %xmm3
; CHECK-NEXT:    vpslld $15, %xmm2, %xmm4
; CHECK-NEXT:    vpor %xmm3, %xmm4, %xmm3
; CHECK-NEXT:    vpsrld $19, %xmm2, %xmm4
; CHECK-NEXT:    vpslld $13, %xmm2, %xmm5
; CHECK-NEXT:    vpor %xmm4, %xmm5, %xmm4
; CHECK-NEXT:    vpxor %xmm4, %xmm3, %xmm3
; CHECK-NEXT:    vpsrld $10, %xmm2, %xmm2
; CHECK-NEXT:    vpxor %xmm2, %xmm3, %xmm2
; CHECK-NEXT:    vpsrlq $32, %xmm1, %xmm3
; CHECK-NEXT:    vpaddd %xmm2, %xmm3, %xmm1
; CHECK-NEXT:    vpsrld $17, %xmm1, %xmm2
; CHECK-NEXT:    vpslld $15, %xmm1, %xmm4
; CHECK-NEXT:    vpor %xmm2, %xmm4, %xmm2
; CHECK-NEXT:    vpsrld $19, %xmm1, %xmm4
; CHECK-NEXT:    vpslld $13, %xmm1, %xmm5
; CHECK-NEXT:    vpor %xmm4, %xmm5, %xmm4
; CHECK-NEXT:    vpxor %xmm4, %xmm2, %xmm2
; CHECK-NEXT:    vpsrld $10, %xmm1, %xmm4
; CHECK-NEXT:    vpxor %xmm4, %xmm2, %xmm2
; CHECK-NEXT:    vpaddd %xmm2, %xmm3, %xmm2
; CHECK-NEXT:    vpsrld $17, %xmm2, %xmm3
; CHECK-NEXT:    vpslld $15, %xmm2, %xmm4
; CHECK-NEXT:    vpor %xmm3, %xmm4, %xmm3
; CHECK-NEXT:    vpsrld $19, %xmm2, %xmm4
; CHECK-NEXT:    vpslld $13, %xmm2, %xmm5
; CHECK-NEXT:    vpor %xmm4, %xmm5, %xmm4
; CHECK-NEXT:    vpxor %xmm4, %xmm3, %xmm3
; CHECK-NEXT:    vpsrld $10, %xmm2, %xmm2
; CHECK-NEXT:    vpxor %xmm2, %xmm3, %xmm2
; CHECK-NEXT:    vpaddd %xmm2, %xmm0, %xmm0
; CHECK-NEXT:    vpsrld $17, %xmm0, %xmm2
; CHECK-NEXT:    vpslld $15, %xmm0, %xmm3
; CHECK-NEXT:    vpor %xmm2, %xmm3, %xmm2
; CHECK-NEXT:    vpsrld $19, %xmm0, %xmm3
; CHECK-NEXT:    vpslld $13, %xmm0, %xmm4
; CHECK-NEXT:    vpor %xmm3, %xmm4, %xmm3
; CHECK-NEXT:    vpxor %xmm3, %xmm2, %xmm2
; CHECK-NEXT:    vpsrld $10, %xmm0, %xmm3
; CHECK-NEXT:    vpxor %xmm3, %xmm2, %xmm2
; CHECK-NEXT:    vpsllq $32, %xmm1, %xmm3
; CHECK-NEXT:    vpaddd %xmm2, %xmm3, %xmm2
; CHECK-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT:    vpunpcklqdq {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[2],ymm0[2]
; CHECK-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
; CHECK-NEXT:    vmovdqu %ymm0, 132(%rdi)
; CHECK-NEXT:    vzeroupper
; CHECK-NEXT:    retq
entry:
  %0 = load i32, ptr null, align 4
  %1 = tail call i32 asm "bswap $0", "=r,0,~{dirflag},~{fpsr},~{flags}"(i32 %0) #3
  %arrayidx14 = getelementptr inbounds [64 x i32], ptr %ctx, i64 0, i64 3
  %2 = load i32, ptr %arrayidx14, align 4
  %3 = tail call i32 asm "bswap $0", "=r,0,~{dirflag},~{fpsr},~{flags}"(i32 %2) #3
  %4 = insertelement <2 x i32> zeroinitializer, i32 %1, i64 1
  %5 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %4, <2 x i32> %4, <2 x i32> <i32 15, i32 15>)
  %6 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %4, <2 x i32> %4, <2 x i32> <i32 13, i32 13>)
  %7 = xor <2 x i32> %5, %6
  %8 = lshr <2 x i32> %4, zeroinitializer
  %9 = xor <2 x i32> %7, %8
  %10 = insertelement <2 x i32> zeroinitializer, i32 %3, i64 0
  %11 = shufflevector <2 x i32> zeroinitializer, <2 x i32> %10, <2 x i32> <i32 1, i32 2>
  %12 = add <2 x i32> %11, %9
  %13 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %12, <2 x i32> %12, <2 x i32> <i32 15, i32 15>)
  %14 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %12, <2 x i32> %12, <2 x i32> <i32 13, i32 13>)
  %15 = xor <2 x i32> %13, %14
  %16 = lshr <2 x i32> %12, zeroinitializer
  %17 = xor <2 x i32> %15, %16
  %18 = add <2 x i32> %4, %17
  %19 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %18, <2 x i32> %18, <2 x i32> <i32 15, i32 15>)
  %20 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %18, <2 x i32> %18, <2 x i32> <i32 13, i32 13>)
  %21 = xor <2 x i32> %19, %20
  %22 = lshr <2 x i32> %18, <i32 10, i32 10>
  %23 = xor <2 x i32> %21, %22
  %24 = add <2 x i32> %4, %23
  %25 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %24, <2 x i32> %24, <2 x i32> <i32 15, i32 15>)
  %26 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %24, <2 x i32> %24, <2 x i32> <i32 13, i32 13>)
  %27 = xor <2 x i32> %25, %26
  %28 = lshr <2 x i32> %24, <i32 10, i32 10>
  %29 = xor <2 x i32> %27, %28
  %30 = shufflevector <2 x i32> %4, <2 x i32> %12, <2 x i32> <i32 1, i32 2>
  %31 = add <2 x i32> %30, %29
  %32 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %31, <2 x i32> %31, <2 x i32> <i32 15, i32 15>)
  %33 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %31, <2 x i32> %31, <2 x i32> <i32 13, i32 13>)
  %34 = xor <2 x i32> %32, %33
  %35 = lshr <2 x i32> %31, <i32 10, i32 10>
  %36 = xor <2 x i32> %34, %35
  %37 = shufflevector <2 x i32> %12, <2 x i32> zeroinitializer, <2 x i32> <i32 1, i32 2>
  %38 = add <2 x i32> %37, %36
  %arrayidx918 = getelementptr inbounds [64 x i32], ptr %ctx, i64 0, i64 33
  store <2 x i32> %38, ptr %arrayidx918, align 4
  %arrayidx1012 = getelementptr inbounds [64 x i32], ptr %ctx, i64 0, i64 35
  %39 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %38, <2 x i32> %38, <2 x i32> <i32 15, i32 15>)
  %40 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %38, <2 x i32> %38, <2 x i32> <i32 13, i32 13>)
  %41 = xor <2 x i32> %39, %40
  %42 = lshr <2 x i32> %38, <i32 10, i32 10>
  %43 = xor <2 x i32> %41, %42
  %44 = add <2 x i32> %37, %43
  store <2 x i32> zeroinitializer, ptr %arrayidx1012, align 4
  %arrayidx1106 = getelementptr inbounds [64 x i32], ptr %ctx, i64 0, i64 37
  %45 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %44, <2 x i32> %44, <2 x i32> <i32 15, i32 15>)
  %46 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %44, <2 x i32> %44, <2 x i32> <i32 13, i32 13>)
  %47 = xor <2 x i32> %45, %46
  %48 = lshr <2 x i32> %44, <i32 10, i32 10>
  %49 = xor <2 x i32> %47, %48
  %50 = lshr <2 x i32> %24, zeroinitializer
  %51 = add <2 x i32> %50, %49
  store <2 x i32> %51, ptr %arrayidx1106, align 4
  %arrayidx1200 = getelementptr inbounds [64 x i32], ptr %ctx, i64 0, i64 39
  %52 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %51, <2 x i32> %51, <2 x i32> <i32 15, i32 15>)
  %53 = tail call <2 x i32> @llvm.fshl.v2i32(<2 x i32> %51, <2 x i32> %51, <2 x i32> <i32 13, i32 13>)
  %54 = xor <2 x i32> %52, %53
  %55 = lshr <2 x i32> %51, <i32 10, i32 10>
  %56 = xor <2 x i32> %54, %55
  %57 = shufflevector <2 x i32> %38, <2 x i32> zeroinitializer, <2 x i32> <i32 poison, i32 0>
  %58 = insertelement <2 x i32> %57, i32 0, i64 0
  %59 = add <2 x i32> %58, %56
  store <2 x i32> %59, ptr %arrayidx1200, align 4
  ret void

; uselistorder directives
  uselistorder <2 x i32> %4, { 7, 0, 1, 6, 5, 4, 3, 2 }
  uselistorder <2 x i32> %38, { 6, 5, 4, 3, 2, 1, 0 }
}

declare <2 x i32> @llvm.fshl.v2i32(<2 x i32>, <2 x i32>, <2 x i32>) #2

; uselistorder directives
uselistorder ptr @llvm.fshl.v2i32, { 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0 }

attributes #0 = { nocallback nofree nosync nounwind willreturn memory(argmem: readwrite) }
attributes #1 = { nounwind sspstrong memory(argmem: readwrite) uwtable "min-legal-vector-width"="0" "no-trapping-math"="true" "probe-stack"="inline-asm" "stack-protector-buffer-size"="8" "target-cpu"="skylake" "target-features"="+adx,+aes,+avx,+avx2,+bmi,+bmi2,+clflushopt,+cmov,+crc32,+cx16,+cx8,+f16c,+fma,+fsgsbase,+fxsr,+invpcid,+lzcnt,+mmx,+movbe,+pclmul,+popcnt,+prfchw,+rdrnd,+rdseed,+sahf,+sgx,+sse,+sse2,+sse3,+sse4.1,+sse4.2,+ssse3,+x87,+xsave,+xsavec,+xsaveopt,+xsaves" }
attributes #2 = { nocallback nofree nosync nounwind speculatable willreturn memory(none) }
attributes #3 = { nounwind memory(none) }