13af1ed8e3
This introduces a new family of combiner helper routines that re-use the target specific cost model from SelectionDAG, and generate inline implementations of the memcpy family of intrinsics. The combines are only enabled at optimization levels higher than -O0, and give very substantial performance improvements. Differential Revision: https://reviews.llvm.org/D65167 llvm-svn: 366951
149 lines
5.8 KiB
YAML
149 lines
5.8 KiB
YAML
# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
|
|
# RUN: llc -march=aarch64 -run-pass=aarch64-prelegalizer-combiner -verify-machineinstrs %s -o - | FileCheck %s
|
|
--- |
|
|
target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
|
|
target triple = "aarch64"
|
|
|
|
define void @test_ms1(i8* nocapture %dst, i32 %c, i32 %len) local_unnamed_addr #0 {
|
|
entry:
|
|
%0 = trunc i32 %c to i8
|
|
%conv = zext i32 %len to i64
|
|
tail call void @llvm.memset.p0i8.i64(i8* align 1 %dst, i8 %0, i64 %conv, i1 false)
|
|
ret void
|
|
}
|
|
|
|
declare void @llvm.memset.p0i8.i64(i8* nocapture writeonly, i8, i64, i1 immarg) #1
|
|
|
|
define void @test_ms2_const(i8* nocapture %dst, i32 %c) local_unnamed_addr #0 {
|
|
entry:
|
|
%0 = trunc i32 %c to i8
|
|
tail call void @llvm.memset.p0i8.i64(i8* align 1 %dst, i8 %0, i64 16, i1 false)
|
|
ret void
|
|
}
|
|
|
|
define void @test_ms3_const_both(i8* nocapture %dst) local_unnamed_addr #0 {
|
|
entry:
|
|
tail call void @llvm.memset.p0i8.i64(i8* align 1 %dst, i8 64, i64 16, i1 false)
|
|
ret void
|
|
}
|
|
|
|
define void @test_ms4_const_both_unaligned(i8* nocapture %dst) local_unnamed_addr #0 {
|
|
entry:
|
|
tail call void @llvm.memset.p0i8.i64(i8* align 1 %dst, i8 64, i64 18, i1 false)
|
|
ret void
|
|
}
|
|
|
|
declare void @llvm.stackprotector(i8*, i8**) #2
|
|
|
|
attributes #0 = { nounwind ssp uwtable "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="cyclone" "target-features"="+aes,+crypto,+fp-armv8,+neon,+sha2,+zcm,+zcz" "unsafe-fp-math"="false" "use-soft-float"="false" }
|
|
attributes #1 = { argmemonly nounwind }
|
|
|
|
...
|
|
---
|
|
name: test_ms1
|
|
alignment: 2
|
|
tracksRegLiveness: true
|
|
body: |
|
|
bb.1.entry:
|
|
liveins: $w1, $w2, $x0
|
|
|
|
; CHECK-LABEL: name: test_ms1
|
|
; CHECK: liveins: $w1, $w2, $x0
|
|
; CHECK: [[COPY:%[0-9]+]]:_(p0) = COPY $x0
|
|
; CHECK: [[COPY1:%[0-9]+]]:_(s32) = COPY $w1
|
|
; CHECK: [[COPY2:%[0-9]+]]:_(s32) = COPY $w2
|
|
; CHECK: [[TRUNC:%[0-9]+]]:_(s8) = G_TRUNC [[COPY1]](s32)
|
|
; CHECK: [[ZEXT:%[0-9]+]]:_(s64) = G_ZEXT [[COPY2]](s32)
|
|
; CHECK: G_INTRINSIC_W_SIDE_EFFECTS intrinsic(@llvm.memset), [[COPY]](p0), [[TRUNC]](s8), [[ZEXT]](s64) :: (store 1 into %ir.dst)
|
|
; CHECK: RET_ReallyLR
|
|
%0:_(p0) = COPY $x0
|
|
%1:_(s32) = COPY $w1
|
|
%2:_(s32) = COPY $w2
|
|
%3:_(s8) = G_TRUNC %1(s32)
|
|
%4:_(s64) = G_ZEXT %2(s32)
|
|
G_INTRINSIC_W_SIDE_EFFECTS intrinsic(@llvm.memset), %0(p0), %3(s8), %4(s64) :: (store 1 into %ir.dst)
|
|
RET_ReallyLR
|
|
|
|
...
|
|
---
|
|
name: test_ms2_const
|
|
alignment: 2
|
|
tracksRegLiveness: true
|
|
body: |
|
|
bb.1.entry:
|
|
liveins: $w1, $x0
|
|
|
|
; CHECK-LABEL: name: test_ms2_const
|
|
; CHECK: liveins: $w1, $x0
|
|
; CHECK: [[COPY:%[0-9]+]]:_(p0) = COPY $x0
|
|
; CHECK: [[COPY1:%[0-9]+]]:_(s32) = COPY $w1
|
|
; CHECK: [[TRUNC:%[0-9]+]]:_(s8) = G_TRUNC [[COPY1]](s32)
|
|
; CHECK: [[ZEXT:%[0-9]+]]:_(s64) = G_ZEXT [[TRUNC]](s8)
|
|
; CHECK: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 72340172838076673
|
|
; CHECK: [[MUL:%[0-9]+]]:_(s64) = G_MUL [[ZEXT]], [[C]]
|
|
; CHECK: G_STORE [[MUL]](s64), [[COPY]](p0) :: (store 8 into %ir.dst, align 1)
|
|
; CHECK: [[C1:%[0-9]+]]:_(s64) = G_CONSTANT i64 8
|
|
; CHECK: [[GEP:%[0-9]+]]:_(p0) = G_GEP [[COPY]], [[C1]](s64)
|
|
; CHECK: G_STORE [[MUL]](s64), [[GEP]](p0) :: (store 8 into %ir.dst + 8, align 1)
|
|
; CHECK: RET_ReallyLR
|
|
%0:_(p0) = COPY $x0
|
|
%1:_(s32) = COPY $w1
|
|
%3:_(s64) = G_CONSTANT i64 16
|
|
%2:_(s8) = G_TRUNC %1(s32)
|
|
G_INTRINSIC_W_SIDE_EFFECTS intrinsic(@llvm.memset), %0(p0), %2(s8), %3(s64) :: (store 1 into %ir.dst)
|
|
RET_ReallyLR
|
|
|
|
...
|
|
---
|
|
name: test_ms3_const_both
|
|
alignment: 2
|
|
tracksRegLiveness: true
|
|
body: |
|
|
bb.1.entry:
|
|
liveins: $x0
|
|
|
|
; CHECK-LABEL: name: test_ms3_const_both
|
|
; CHECK: liveins: $x0
|
|
; CHECK: [[COPY:%[0-9]+]]:_(p0) = COPY $x0
|
|
; CHECK: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 4629771061636907072
|
|
; CHECK: G_STORE [[C]](s64), [[COPY]](p0) :: (store 8 into %ir.dst, align 1)
|
|
; CHECK: [[C1:%[0-9]+]]:_(s64) = G_CONSTANT i64 8
|
|
; CHECK: [[GEP:%[0-9]+]]:_(p0) = G_GEP [[COPY]], [[C1]](s64)
|
|
; CHECK: G_STORE [[C]](s64), [[GEP]](p0) :: (store 8 into %ir.dst + 8, align 1)
|
|
; CHECK: RET_ReallyLR
|
|
%0:_(p0) = COPY $x0
|
|
%1:_(s8) = G_CONSTANT i8 64
|
|
%2:_(s64) = G_CONSTANT i64 16
|
|
G_INTRINSIC_W_SIDE_EFFECTS intrinsic(@llvm.memset), %0(p0), %1(s8), %2(s64) :: (store 1 into %ir.dst)
|
|
RET_ReallyLR
|
|
|
|
...
|
|
---
|
|
name: test_ms4_const_both_unaligned
|
|
alignment: 2
|
|
tracksRegLiveness: true
|
|
body: |
|
|
bb.1.entry:
|
|
liveins: $x0
|
|
|
|
; CHECK-LABEL: name: test_ms4_const_both_unaligned
|
|
; CHECK: liveins: $x0
|
|
; CHECK: [[COPY:%[0-9]+]]:_(p0) = COPY $x0
|
|
; CHECK: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 4629771061636907072
|
|
; CHECK: G_STORE [[C]](s64), [[COPY]](p0) :: (store 8 into %ir.dst, align 1)
|
|
; CHECK: [[C1:%[0-9]+]]:_(s64) = G_CONSTANT i64 8
|
|
; CHECK: [[GEP:%[0-9]+]]:_(p0) = G_GEP [[COPY]], [[C1]](s64)
|
|
; CHECK: G_STORE [[C]](s64), [[GEP]](p0) :: (store 8 into %ir.dst + 8, align 1)
|
|
; CHECK: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[C]](s64)
|
|
; CHECK: [[C2:%[0-9]+]]:_(s64) = G_CONSTANT i64 16
|
|
; CHECK: [[GEP1:%[0-9]+]]:_(p0) = G_GEP [[COPY]], [[C2]](s64)
|
|
; CHECK: G_STORE [[TRUNC]](s16), [[GEP1]](p0) :: (store 2 into %ir.dst + 16, align 1)
|
|
; CHECK: RET_ReallyLR
|
|
%0:_(p0) = COPY $x0
|
|
%1:_(s8) = G_CONSTANT i8 64
|
|
%2:_(s64) = G_CONSTANT i64 18
|
|
G_INTRINSIC_W_SIDE_EFFECTS intrinsic(@llvm.memset), %0(p0), %1(s8), %2(s64) :: (store 1 into %ir.dst)
|
|
RET_ReallyLR
|
|
|
|
...
|