llvm/test/Transforms/SLPVectorizer/RISCV/reordered-buildvector-scalars.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
; RUN: opt -S --passes=slp-vectorizer -mtriple=riscv64-unknown-linux-gnu -mcpu=sifive-x280 < %s | FileCheck %s

%struct.ImageParameters = type { i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, ptr, ptr, i32, ptr, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, [9 x [16 x [16 x i16]]], [5 x [16 x [16 x i16]]], [9 x [8 x [8 x i16]]], [2 x [4 x [16 x [16 x i16]]]], [16 x [16 x i16]], [16 x [16 x i32]], ptr, ptr, ptr, ptr, ptr, [1200 x %struct.syntaxelement], ptr, ptr, i32, i32, i32, i32, [4 x [4 x i32]], i32, i32, i32, i32, i32, double, i32, i32, i32, i32, ptr, ptr, ptr, ptr, [15 x i16], i32, i32, i32, i32, i32, i32, i32, i32, [6 x [15 x i32]], i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, [1 x i32], i32, i32, [2 x i32], i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, ptr, i32, i32, i32, i32, i32, double, i32, i32, i32, i32, i32, i32, i32, ptr, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, [2 x i32], i32, i32, i32 }
%struct.syntaxelement = type { i32, i32, i32, i32, i32, i32, i32, i32, ptr, ptr }

@images = external global %struct.ImageParameters

define fastcc i32 @test(i32 %0, i32 %add111.i.i, <4 x i32> %PredPel.i.sroa.86.72.vec.extract, <4 x i32> %1) {
; CHECK-LABEL: define fastcc i32 @test(
; CHECK-SAME: i32 [[TMP0:%.*]], i32 [[ADD111_I_I:%.*]], <4 x i32> [[PREDPEL_I_SROA_86_72_VEC_EXTRACT:%.*]], <4 x i32> [[TMP1:%.*]]) #[[ATTR0:[0-9]+]] {
; CHECK-NEXT:  [[ENTRY:.*:]]
; CHECK-NEXT:    [[LOOPARRAY_SROA_24_0_I_I3:%.*]] = ashr i32 [[TMP0]], 1
; CHECK-NEXT:    [[SHR143_5_I_I9:%.*]] = ashr i32 [[TMP0]], 1
; CHECK-NEXT:    [[ADD1392_I:%.*]] = add i32 [[TMP0]], 1
; CHECK-NEXT:    [[PREDPEL_I_SROA_86_80_VEC_EXTRACT59312:%.*]] = extractelement <4 x i32> [[PREDPEL_I_SROA_86_72_VEC_EXTRACT]], i64 0
; CHECK-NEXT:    [[MUL1445_I:%.*]] = shl i32 [[TMP0]], 1
; CHECK-NEXT:    [[ADD2136_I:%.*]] = or i32 [[LOOPARRAY_SROA_24_0_I_I3]], [[TMP0]]
; CHECK-NEXT:    [[SHR2137_I:%.*]] = lshr i32 [[ADD2136_I]], 1
; CHECK-NEXT:    [[CONV2138_I:%.*]] = trunc i32 [[SHR2137_I]] to i16
; CHECK-NEXT:    [[ADD2157_I:%.*]] = add i32 [[PREDPEL_I_SROA_86_80_VEC_EXTRACT59312]], 1
; CHECK-NEXT:    [[SHR2158_I:%.*]] = lshr i32 [[ADD2157_I]], 1
; CHECK-NEXT:    [[CONV2159_I:%.*]] = trunc i32 [[SHR2158_I]] to i16
; CHECK-NEXT:    [[ADD2174_I:%.*]] = add i32 [[MUL1445_I]], 2
; CHECK-NEXT:    [[SHR2175_I:%.*]] = lshr i32 [[ADD2174_I]], 2
; CHECK-NEXT:    [[CONV2176_I:%.*]] = trunc i32 [[SHR2175_I]] to i16
; CHECK-NEXT:    [[ADD2190_I:%.*]] = or i32 [[ADD1392_I]], 1
; CHECK-NEXT:    [[ADD2191_I:%.*]] = add i32 [[ADD2190_I]], [[TMP0]]
; CHECK-NEXT:    [[CONV2193_I:%.*]] = trunc i32 [[ADD2191_I]] to i16
; CHECK-NEXT:    [[ADD2203_I:%.*]] = or i32 [[TMP0]], 1
; CHECK-NEXT:    [[ADD2204_I:%.*]] = add i32 [[ADD2203_I]], [[TMP0]]
; CHECK-NEXT:    [[CONV2206_I:%.*]] = trunc i32 [[ADD2204_I]] to i16
; CHECK-NEXT:    [[ADD2235_I16:%.*]] = or i32 [[TMP0]], 1
; CHECK-NEXT:    [[ADD2236_I:%.*]] = add i32 [[ADD2235_I16]], 1
; CHECK-NEXT:    [[SHR2237_I:%.*]] = lshr i32 [[ADD2236_I]], 1
; CHECK-NEXT:    [[CONV2238_I:%.*]] = trunc i32 [[SHR2237_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2238_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8196), align 4
; CHECK-NEXT:    store i16 [[CONV2238_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8176), align 8
; CHECK-NEXT:    [[ADD2258_I:%.*]] = or i32 [[ADD111_I_I]], [[TMP0]]
; CHECK-NEXT:    [[SHR2259_I:%.*]] = lshr i32 [[ADD2258_I]], 1
; CHECK-NEXT:    [[CONV2260_I:%.*]] = trunc i32 [[SHR2259_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2260_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8212), align 4
; CHECK-NEXT:    store i16 [[CONV2260_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8192), align 8
; CHECK-NEXT:    store i16 [[CONV2260_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8172), align 4
; CHECK-NEXT:    [[ADD2280_I:%.*]] = add i32 [[ADD111_I_I]], 1
; CHECK-NEXT:    [[SHR2281_I:%.*]] = lshr i32 [[ADD2280_I]], 1
; CHECK-NEXT:    [[CONV2282_I:%.*]] = trunc i32 [[SHR2281_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2282_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8228), align 4
; CHECK-NEXT:    store i16 [[CONV2282_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8208), align 8
; CHECK-NEXT:    store i16 [[CONV2282_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8188), align 4
; CHECK-NEXT:    [[ADD2302_I:%.*]] = add i32 [[TMP0]], 1
; CHECK-NEXT:    [[SHR2303_I:%.*]] = lshr i32 [[ADD2302_I]], 1
; CHECK-NEXT:    [[CONV2304_I:%.*]] = trunc i32 [[SHR2303_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2304_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8224), align 8
; CHECK-NEXT:    store i16 [[CONV2304_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8204), align 4
; CHECK-NEXT:    store i16 [[CONV2304_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8184), align 8
; CHECK-NEXT:    [[ADD2323_I:%.*]] = add i32 [[TMP0]], 1
; CHECK-NEXT:    [[ADD2324_I:%.*]] = or i32 [[ADD2323_I]], [[TMP0]]
; CHECK-NEXT:    [[SHR2325_I:%.*]] = lshr i32 [[ADD2324_I]], 1
; CHECK-NEXT:    [[CONV2326_I:%.*]] = trunc i32 [[SHR2325_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2326_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8220), align 4
; CHECK-NEXT:    store i16 [[CONV2326_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8200), align 8
; CHECK-NEXT:    [[ADD2342_I:%.*]] = add i32 [[SHR143_5_I_I9]], 1
; CHECK-NEXT:    [[SHR2343_I:%.*]] = lshr i32 [[ADD2342_I]], 1
; CHECK-NEXT:    [[CONV2344_I:%.*]] = trunc i32 [[SHR2343_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2344_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8216), align 8
; CHECK-NEXT:    [[ADD2355_I:%.*]] = or i32 [[SHR143_5_I_I9]], 1
; CHECK-NEXT:    [[ADD2356_I:%.*]] = add i32 [[ADD2355_I]], [[TMP0]]
; CHECK-NEXT:    [[CONV2358_I:%.*]] = trunc i32 [[ADD2356_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2358_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8232), align 8
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <4 x i32> [[TMP1]], <4 x i32> poison, <2 x i32> <i32 poison, i32 0>
; CHECK-NEXT:    [[TMP3:%.*]] = insertelement <2 x i32> [[TMP2]], i32 [[LOOPARRAY_SROA_24_0_I_I3]], i32 0
; CHECK-NEXT:    [[TMP4:%.*]] = add <2 x i32> [[TMP3]], splat (i32 1)
; CHECK-NEXT:    [[TMP5:%.*]] = lshr <2 x i32> [[TMP4]], splat (i32 1)
; CHECK-NEXT:    [[TMP6:%.*]] = trunc <2 x i32> [[TMP5]] to <2 x i16>
; CHECK-NEXT:    store <2 x i16> [[TMP6]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8180), align 4
; CHECK-NEXT:    [[ADD2393_I:%.*]] = or i32 [[LOOPARRAY_SROA_24_0_I_I3]], 1
; CHECK-NEXT:    [[ADD2394_I:%.*]] = add i32 [[ADD2393_I]], [[TMP0]]
; CHECK-NEXT:    [[CONV2396_I:%.*]] = trunc i32 [[ADD2394_I]] to i16
; CHECK-NEXT:    store i16 [[CONV2396_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8198), align 2
; CHECK-NEXT:    store i16 [[CONV2396_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8178), align 2
; CHECK-NEXT:    store i16 [[CONV2138_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8214), align 2
; CHECK-NEXT:    store i16 [[CONV2138_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8194), align 2
; CHECK-NEXT:    store i16 [[CONV2138_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8174), align 2
; CHECK-NEXT:    store i16 [[CONV2159_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8230), align 2
; CHECK-NEXT:    store i16 [[CONV2159_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8210), align 2
; CHECK-NEXT:    store i16 [[CONV2159_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8190), align 2
; CHECK-NEXT:    store i16 [[CONV2159_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8170), align 2
; CHECK-NEXT:    store i16 [[CONV2176_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8226), align 2
; CHECK-NEXT:    store i16 [[CONV2176_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8206), align 2
; CHECK-NEXT:    store i16 [[CONV2176_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8186), align 2
; CHECK-NEXT:    store i16 [[CONV2193_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8222), align 2
; CHECK-NEXT:    store i16 [[CONV2193_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8202), align 2
; CHECK-NEXT:    store i16 [[CONV2206_I]], ptr getelementptr inbounds nuw (i8, ptr @images, i64 8218), align 2
; CHECK-NEXT:    ret i32 0
;
entry:
  %LoopArray.sroa.24.0.i.i3 = ashr i32 %0, 1
  %shr143.5.i.i9 = ashr i32 %0, 1
  %add1392.i = add i32 %0, 1
  %PredPel.i.sroa.86.80.vec.extract59312 = extractelement <4 x i32> %PredPel.i.sroa.86.72.vec.extract, i64 0
  %mul1445.i = shl i32 %0, 1
  %PredPel.i.sroa.7.4.vec.extract446 = extractelement <4 x i32> %1, i64 0
  %add1571.i = add i32 %PredPel.i.sroa.7.4.vec.extract446, 1
  %shr1572.i = lshr i32 %add1571.i, 1
  %conv1573.i = trunc i32 %shr1572.i to i16
  %add2136.i = or i32 %LoopArray.sroa.24.0.i.i3, %0
  %shr2137.i = lshr i32 %add2136.i, 1
  %conv2138.i = trunc i32 %shr2137.i to i16
  %add2157.i = add i32 %PredPel.i.sroa.86.80.vec.extract59312, 1
  %shr2158.i = lshr i32 %add2157.i, 1
  %conv2159.i = trunc i32 %shr2158.i to i16
  %add2174.i = add i32 %mul1445.i, 2
  %shr2175.i = lshr i32 %add2174.i, 2
  %conv2176.i = trunc i32 %shr2175.i to i16
  %add2190.i = or i32 %add1392.i, 1
  %add2191.i = add i32 %add2190.i, %0
  %conv2193.i = trunc i32 %add2191.i to i16
  %add2203.i = or i32 %0, 1
  %add2204.i = add i32 %add2203.i, %0
  %conv2206.i = trunc i32 %add2204.i to i16
  %add2214.i = add i32 %LoopArray.sroa.24.0.i.i3, 1
  %shr2215.i = lshr i32 %add2214.i, 1
  %conv2216.i = trunc i32 %shr2215.i to i16
  store i16 %conv2216.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8180), align 4
  %add2235.i16 = or i32 %0, 1
  %add2236.i = add i32 %add2235.i16, 1
  %shr2237.i = lshr i32 %add2236.i, 1
  %conv2238.i = trunc i32 %shr2237.i to i16
  store i16 %conv2238.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8196), align 4
  store i16 %conv2238.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8176), align 8
  %add2258.i = or i32 %add111.i.i, %0
  %shr2259.i = lshr i32 %add2258.i, 1
  %conv2260.i = trunc i32 %shr2259.i to i16
  store i16 %conv2260.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8212), align 4
  store i16 %conv2260.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8192), align 8
  store i16 %conv2260.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8172), align 4
  %add2280.i = add i32 %add111.i.i, 1
  %shr2281.i = lshr i32 %add2280.i, 1
  %conv2282.i = trunc i32 %shr2281.i to i16
  store i16 %conv2282.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8228), align 4
  store i16 %conv2282.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8208), align 8
  store i16 %conv2282.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8188), align 4
  %add2302.i = add i32 %0, 1
  %shr2303.i = lshr i32 %add2302.i, 1
  %conv2304.i = trunc i32 %shr2303.i to i16
  store i16 %conv2304.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8224), align 8
  store i16 %conv2304.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8204), align 4
  store i16 %conv2304.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8184), align 8
  %add2323.i = add i32 %0, 1
  %add2324.i = or i32 %add2323.i, %0
  %shr2325.i = lshr i32 %add2324.i, 1
  %conv2326.i = trunc i32 %shr2325.i to i16
  store i16 %conv2326.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8220), align 4
  store i16 %conv2326.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8200), align 8
  %add2342.i = add i32 %shr143.5.i.i9, 1
  %shr2343.i = lshr i32 %add2342.i, 1
  %conv2344.i = trunc i32 %shr2343.i to i16
  store i16 %conv2344.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8216), align 8
  %add2355.i = or i32 %shr143.5.i.i9, 1
  %add2356.i = add i32 %add2355.i, %0
  %conv2358.i = trunc i32 %add2356.i to i16
  store i16 %conv2358.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8232), align 8
  store i16 %conv1573.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8182), align 2
  %add2393.i = or i32 %LoopArray.sroa.24.0.i.i3, 1
  %add2394.i = add i32 %add2393.i, %0
  %conv2396.i = trunc i32 %add2394.i to i16
  store i16 %conv2396.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8198), align 2
  store i16 %conv2396.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8178), align 2
  store i16 %conv2138.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8214), align 2
  store i16 %conv2138.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8194), align 2
  store i16 %conv2138.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8174), align 2
  store i16 %conv2159.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8230), align 2
  store i16 %conv2159.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8210), align 2
  store i16 %conv2159.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8190), align 2
  store i16 %conv2159.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8170), align 2
  store i16 %conv2176.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8226), align 2
  store i16 %conv2176.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8206), align 2
  store i16 %conv2176.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8186), align 2
  store i16 %conv2193.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8222), align 2
  store i16 %conv2193.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8202), align 2
  store i16 %conv2206.i, ptr getelementptr inbounds nuw (i8, ptr @images, i64 8218), align 2
  ret i32 0
}