mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2025-02-01 13:11:39 +01:00
e757c4f305
Instead of inserting everything after the 'root' of the reduction, insert all instructions as close to their operands as possible. This can help reduce register pressure. Differential Revision: https://reviews.llvm.org/D67392 llvm-svn: 374981
151 lines
6.7 KiB
LLVM
151 lines
6.7 KiB
LLVM
; RUN: llc -mtriple=thumbv7em -mcpu=cortex-m4 -O3 %s -o - | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-LLC
|
|
; RUN: opt -S -mtriple=armv7-a -arm-parallel-dsp -dce %s -o - | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-OPT
|
|
|
|
; TODO: Think we should be able to use smlsdx/smlsldx here.
|
|
|
|
; CHECK-LABEL: complex_dot_prod
|
|
|
|
; CHECK-LLC: smlaldx
|
|
; CHECK-LLC: smulbb
|
|
; CHECK-LLC: smultt
|
|
; CHECK-LLC: smlaldx
|
|
; CHECK-LLC: smlalbb
|
|
; CHECK-LLC: smultt
|
|
; CHECK-LLC: smlalbb
|
|
; CHECK-LLC: smultt
|
|
; CHECK-LLC: smlaldx
|
|
; CHECK-LLC: smlalbb
|
|
; CHECK-LLC: smultt
|
|
; CHECK-LLC: smlaldx
|
|
; CHECK-LCC: pop.w {r4, r5, r6, r7, r8, r9, r10, pc}
|
|
|
|
; CHECK-OPT: [[ADDR_A:%[^ ]+]] = bitcast i16* %pSrcA to i32*
|
|
; CHECK-OPT: [[A:%[^ ]+]] = load i32, i32* [[ADDR_A]], align 2
|
|
; CHECK-OPT: [[ADDR_A_2:%[^ ]+]] = getelementptr inbounds i16, i16* %pSrcA, i32 2
|
|
; CHECK-OPT: [[ADDR_B:%[^ ]+]] = bitcast i16* %pSrcB to i32*
|
|
; CHECK-OPT: [[B:%[^ ]+]] = load i32, i32* [[ADDR_B]], align 2
|
|
; CHECK-OPT: [[ACC0:%[^ ]+]] = call i64 @llvm.arm.smlaldx(i32 [[A]], i32 [[B]], i64 0)
|
|
; CHECK-OPT: [[ADDR_B_2:%[^ ]+]] = getelementptr inbounds i16, i16* %pSrcB, i32 2
|
|
; CHECK-OPT: [[CAST_ADDR_A_2:%[^ ]+]] = bitcast i16* [[ADDR_A_2]] to i32*
|
|
; CHECK-OPT: [[A_2:%[^ ]+]] = load i32, i32* [[CAST_ADDR_A_2]], align 2
|
|
; CHECK-OPT: [[ADDR_A_4:%[^ ]+]] = getelementptr inbounds i16, i16* %pSrcA, i32 4
|
|
; CHECK-OPT: [[CAST_ADDR_B_2:%[^ ]+]] = bitcast i16* [[ADDR_B_2]] to i32*
|
|
; CHECK-OPT: [[B_2:%[^ ]+]] = load i32, i32* [[CAST_ADDR_B_2]], align 2
|
|
; CHECK-OPT: [[ACC1:%[^ ]+]] = call i64 @llvm.arm.smlaldx(i32 [[A_2]], i32 [[B_2]], i64 [[ACC0]])
|
|
; CHECK-OPT: [[ADDR_B_4:%[^ ]+]] = getelementptr inbounds i16, i16* %pSrcB, i32 4
|
|
; CHECK-OPT: [[CAST_ADDR_A_4:%[^ ]+]] = bitcast i16* [[ADDR_A_4]] to i32*
|
|
; CHECK-OPT: [[A_4:%[^ ]+]] = load i32, i32* [[CAST_ADDR_A_4]], align 2
|
|
; CHECK-OPT: [[ADDR_A_6:%[^ ]+]] = getelementptr inbounds i16, i16* %pSrcA, i32 6
|
|
; CHECK-OPT: [[CAST_ADDR_B_4:%[^ ]+]] = bitcast i16* [[ADDR_B_4]] to i32*
|
|
; CHECK-OPT: [[B_4:%[^ ]+]] = load i32, i32* [[CAST_ADDR_B_4]], align 2
|
|
; CHECK-OPT: [[ACC2:%[^ ]+]] = call i64 @llvm.arm.smlaldx(i32 [[A_4]], i32 [[B_4]], i64 [[ACC1]])
|
|
; CHECK-OPT: [[ADDR_B_6:%[^ ]+]] = getelementptr inbounds i16, i16* %pSrcB, i32 6
|
|
; CHECK-OPT: [[CAST_ADDR_A_6:%[^ ]+]] = bitcast i16* [[ADDR_A_6]] to i32*
|
|
; CHECK-OPT: [[A_6:%[^ ]+]] = load i32, i32* [[CAST_ADDR_A_6]], align 2
|
|
; CHECK-OPT: [[CAST_ADDR_B_6:%[^ ]+]] = bitcast i16* [[ADDR_B_6]] to i32*
|
|
; CHECK-OPT: [[B_6:%[^ ]+]] = load i32, i32* [[CAST_ADDR_B_6]], align 2
|
|
; CHECK-OPT: call i64 @llvm.arm.smlaldx(i32 [[A_6]], i32 [[B_6]], i64 [[ACC2]])
|
|
|
|
define dso_local arm_aapcscc void @complex_dot_prod(i16* nocapture readonly %pSrcA, i16* nocapture readonly %pSrcB, i32* nocapture %realResult, i32* nocapture %imagResult) {
|
|
entry:
|
|
%incdec.ptr = getelementptr inbounds i16, i16* %pSrcA, i32 1
|
|
%0 = load i16, i16* %pSrcA, align 2
|
|
%incdec.ptr1 = getelementptr inbounds i16, i16* %pSrcA, i32 2
|
|
%1 = load i16, i16* %incdec.ptr, align 2
|
|
%incdec.ptr2 = getelementptr inbounds i16, i16* %pSrcB, i32 1
|
|
%2 = load i16, i16* %pSrcB, align 2
|
|
%incdec.ptr3 = getelementptr inbounds i16, i16* %pSrcB, i32 2
|
|
%3 = load i16, i16* %incdec.ptr2, align 2
|
|
%conv = sext i16 %0 to i32
|
|
%conv4 = sext i16 %2 to i32
|
|
%mul = mul nsw i32 %conv4, %conv
|
|
%conv5 = sext i32 %mul to i64
|
|
%conv7 = sext i16 %3 to i32
|
|
%mul8 = mul nsw i32 %conv7, %conv
|
|
%conv9 = sext i32 %mul8 to i64
|
|
%conv11 = sext i16 %1 to i32
|
|
%mul13 = mul nsw i32 %conv7, %conv11
|
|
%conv14 = sext i32 %mul13 to i64
|
|
%sub = sub nsw i64 %conv5, %conv14
|
|
%mul17 = mul nsw i32 %conv4, %conv11
|
|
%conv18 = sext i32 %mul17 to i64
|
|
%add19 = add nsw i64 %conv9, %conv18
|
|
%incdec.ptr20 = getelementptr inbounds i16, i16* %pSrcA, i32 3
|
|
%4 = load i16, i16* %incdec.ptr1, align 2
|
|
%incdec.ptr21 = getelementptr inbounds i16, i16* %pSrcA, i32 4
|
|
%5 = load i16, i16* %incdec.ptr20, align 2
|
|
%incdec.ptr22 = getelementptr inbounds i16, i16* %pSrcB, i32 3
|
|
%6 = load i16, i16* %incdec.ptr3, align 2
|
|
%incdec.ptr23 = getelementptr inbounds i16, i16* %pSrcB, i32 4
|
|
%7 = load i16, i16* %incdec.ptr22, align 2
|
|
%conv24 = sext i16 %4 to i32
|
|
%conv25 = sext i16 %6 to i32
|
|
%mul26 = mul nsw i32 %conv25, %conv24
|
|
%conv27 = sext i32 %mul26 to i64
|
|
%add28 = add nsw i64 %sub, %conv27
|
|
%conv30 = sext i16 %7 to i32
|
|
%mul31 = mul nsw i32 %conv30, %conv24
|
|
%conv32 = sext i32 %mul31 to i64
|
|
%conv34 = sext i16 %5 to i32
|
|
%mul36 = mul nsw i32 %conv30, %conv34
|
|
%conv37 = sext i32 %mul36 to i64
|
|
%sub38 = sub nsw i64 %add28, %conv37
|
|
%mul41 = mul nsw i32 %conv25, %conv34
|
|
%conv42 = sext i32 %mul41 to i64
|
|
%add33 = add nsw i64 %add19, %conv42
|
|
%add43 = add nsw i64 %add33, %conv32
|
|
%incdec.ptr44 = getelementptr inbounds i16, i16* %pSrcA, i32 5
|
|
%8 = load i16, i16* %incdec.ptr21, align 2
|
|
%incdec.ptr45 = getelementptr inbounds i16, i16* %pSrcA, i32 6
|
|
%9 = load i16, i16* %incdec.ptr44, align 2
|
|
%incdec.ptr46 = getelementptr inbounds i16, i16* %pSrcB, i32 5
|
|
%10 = load i16, i16* %incdec.ptr23, align 2
|
|
%incdec.ptr47 = getelementptr inbounds i16, i16* %pSrcB, i32 6
|
|
%11 = load i16, i16* %incdec.ptr46, align 2
|
|
%conv48 = sext i16 %8 to i32
|
|
%conv49 = sext i16 %10 to i32
|
|
%mul50 = mul nsw i32 %conv49, %conv48
|
|
%conv51 = sext i32 %mul50 to i64
|
|
%add52 = add nsw i64 %sub38, %conv51
|
|
%conv54 = sext i16 %11 to i32
|
|
%mul55 = mul nsw i32 %conv54, %conv48
|
|
%conv56 = sext i32 %mul55 to i64
|
|
%conv58 = sext i16 %9 to i32
|
|
%mul60 = mul nsw i32 %conv54, %conv58
|
|
%conv61 = sext i32 %mul60 to i64
|
|
%sub62 = sub nsw i64 %add52, %conv61
|
|
%mul65 = mul nsw i32 %conv49, %conv58
|
|
%conv66 = sext i32 %mul65 to i64
|
|
%add57 = add nsw i64 %add43, %conv66
|
|
%add67 = add nsw i64 %add57, %conv56
|
|
%incdec.ptr68 = getelementptr inbounds i16, i16* %pSrcA, i32 7
|
|
%12 = load i16, i16* %incdec.ptr45, align 2
|
|
%13 = load i16, i16* %incdec.ptr68, align 2
|
|
%incdec.ptr70 = getelementptr inbounds i16, i16* %pSrcB, i32 7
|
|
%14 = load i16, i16* %incdec.ptr47, align 2
|
|
%15 = load i16, i16* %incdec.ptr70, align 2
|
|
%conv72 = sext i16 %12 to i32
|
|
%conv73 = sext i16 %14 to i32
|
|
%mul74 = mul nsw i32 %conv73, %conv72
|
|
%conv75 = sext i32 %mul74 to i64
|
|
%add76 = add nsw i64 %sub62, %conv75
|
|
%conv78 = sext i16 %15 to i32
|
|
%mul79 = mul nsw i32 %conv78, %conv72
|
|
%conv80 = sext i32 %mul79 to i64
|
|
%conv82 = sext i16 %13 to i32
|
|
%mul84 = mul nsw i32 %conv78, %conv82
|
|
%conv85 = sext i32 %mul84 to i64
|
|
%sub86 = sub nsw i64 %add76, %conv85
|
|
%mul89 = mul nsw i32 %conv73, %conv82
|
|
%conv90 = sext i32 %mul89 to i64
|
|
%add81 = add nsw i64 %add67, %conv90
|
|
%add91 = add nsw i64 %add81, %conv80
|
|
%16 = lshr i64 %sub86, 6
|
|
%conv92 = trunc i64 %16 to i32
|
|
store i32 %conv92, i32* %realResult, align 4
|
|
%17 = lshr i64 %add91, 6
|
|
%conv94 = trunc i64 %17 to i32
|
|
store i32 %conv94, i32* %imagResult, align 4
|
|
ret void
|
|
}
|