maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/Transforms/MemCpyOpt/fca2memcpy.ll

130 lines
4.2 KiB
LLVM
Raw Normal View History

Implement load to store => memcpy in MemCpyOpt for aggregates Summary: Most of the tool chain is able to optimize scalar and memcpy like operation effisciently while it isn't that good with aggregates. In order to improve the support of aggregate, we try to change aggregate manipulation into either scalar or memcpy like ones whenever possible without loosing informations. This is one such opportunity. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15894 llvm-svn: 256868
2016-01-06 04:17:48 +08:00
; RUN: opt -memcpyopt -S < %s | FileCheck %s
target datalayout = "e-i64:64-f80:128-n8:16:32:64"
target triple = "x86_64-unknown-linux-gnu"
Promote aggregate store to memset when possible Summary: As per title. This will allow the optimizer to pick up on it. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15923 llvm-svn: 256969
2016-01-07 03:47:24 +08:00
%S = type { i8*, i8, i32 }
Implement load to store => memcpy in MemCpyOpt for aggregates Summary: Most of the tool chain is able to optimize scalar and memcpy like operation effisciently while it isn't that good with aggregates. In order to improve the support of aggregate, we try to change aggregate manipulation into either scalar or memcpy like ones whenever possible without loosing informations. This is one such opportunity. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15894 llvm-svn: 256868
2016-01-06 04:17:48 +08:00
define void @copy(%S* %src, %S* %dst) {
; CHECK-LABEL: copy
Improve load/store to memcpy for aggregate Summary: It turns out that if we don't try to do it at the store location, we can do it before any operation that alias the load, as long as no operation alias the store. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15903 llvm-svn: 256923
2016-01-06 17:30:39 +08:00
; CHECK-NOT: load
Implement load to store => memcpy in MemCpyOpt for aggregates Summary: Most of the tool chain is able to optimize scalar and memcpy like operation effisciently while it isn't that good with aggregates. In order to improve the support of aggregate, we try to change aggregate manipulation into either scalar or memcpy like ones whenever possible without loosing informations. This is one such opportunity. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15894 llvm-svn: 256868
2016-01-06 04:17:48 +08:00
; CHECK: call void @llvm.memmove.p0i8.p0i8.i64
; CHECK-NEXT: ret void
%1 = load %S, %S* %src
store %S %1, %S* %dst
ret void
}
define void @noaliassrc(%S* noalias %src, %S* %dst) {
; CHECK-LABEL: noaliassrc
Improve load/store to memcpy for aggregate Summary: It turns out that if we don't try to do it at the store location, we can do it before any operation that alias the load, as long as no operation alias the store. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15903 llvm-svn: 256923
2016-01-06 17:30:39 +08:00
; CHECK-NOT: load
Implement load to store => memcpy in MemCpyOpt for aggregates Summary: Most of the tool chain is able to optimize scalar and memcpy like operation effisciently while it isn't that good with aggregates. In order to improve the support of aggregate, we try to change aggregate manipulation into either scalar or memcpy like ones whenever possible without loosing informations. This is one such opportunity. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15894 llvm-svn: 256868
2016-01-06 04:17:48 +08:00
; CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
; CHECK-NEXT: ret void
%1 = load %S, %S* %src
store %S %1, %S* %dst
ret void
}
define void @noaliasdst(%S* %src, %S* noalias %dst) {
; CHECK-LABEL: noaliasdst
Improve load/store to memcpy for aggregate Summary: It turns out that if we don't try to do it at the store location, we can do it before any operation that alias the load, as long as no operation alias the store. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15903 llvm-svn: 256923
2016-01-06 17:30:39 +08:00
; CHECK-NOT: load
; CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
; CHECK-NEXT: ret void
%1 = load %S, %S* %src
store %S %1, %S* %dst
ret void
}
define void @destroysrc(%S* %src, %S* %dst) {
; CHECK-LABEL: destroysrc
Promote aggregate store to memset when possible Summary: As per title. This will allow the optimizer to pick up on it. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15923 llvm-svn: 256969
2016-01-07 03:47:24 +08:00
; CHECK: load %S, %S* %src
; CHECK: call void @llvm.memset.p0i8.i64
; CHECK-NEXT: store %S %1, %S* %dst
; CHECK-NEXT: ret void
Improve load/store to memcpy for aggregate Summary: It turns out that if we don't try to do it at the store location, we can do it before any operation that alias the load, as long as no operation alias the store. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15903 llvm-svn: 256923
2016-01-06 17:30:39 +08:00
%1 = load %S, %S* %src
store %S zeroinitializer, %S* %src
store %S %1, %S* %dst
ret void
}
define void @destroynoaliassrc(%S* noalias %src, %S* %dst) {
; CHECK-LABEL: destroynoaliassrc
; CHECK-NOT: load
Implement load to store => memcpy in MemCpyOpt for aggregates Summary: Most of the tool chain is able to optimize scalar and memcpy like operation effisciently while it isn't that good with aggregates. In order to improve the support of aggregate, we try to change aggregate manipulation into either scalar or memcpy like ones whenever possible without loosing informations. This is one such opportunity. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15894 llvm-svn: 256868
2016-01-06 04:17:48 +08:00
; CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
Promote aggregate store to memset when possible Summary: As per title. This will allow the optimizer to pick up on it. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15923 llvm-svn: 256969
2016-01-07 03:47:24 +08:00
; CHECK-NEXT: call void @llvm.memset.p0i8.i64
Implement load to store => memcpy in MemCpyOpt for aggregates Summary: Most of the tool chain is able to optimize scalar and memcpy like operation effisciently while it isn't that good with aggregates. In order to improve the support of aggregate, we try to change aggregate manipulation into either scalar or memcpy like ones whenever possible without loosing informations. This is one such opportunity. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15894 llvm-svn: 256868
2016-01-06 04:17:48 +08:00
; CHECK-NEXT: ret void
%1 = load %S, %S* %src
Improve load/store to memcpy for aggregate Summary: It turns out that if we don't try to do it at the store location, we can do it before any operation that alias the load, as long as no operation alias the store. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15903 llvm-svn: 256923
2016-01-06 17:30:39 +08:00
store %S zeroinitializer, %S* %src
Implement load to store => memcpy in MemCpyOpt for aggregates Summary: Most of the tool chain is able to optimize scalar and memcpy like operation effisciently while it isn't that good with aggregates. In order to improve the support of aggregate, we try to change aggregate manipulation into either scalar or memcpy like ones whenever possible without loosing informations. This is one such opportunity. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15894 llvm-svn: 256868
2016-01-06 04:17:48 +08:00
store %S %1, %S* %dst
ret void
}
define void @copyalias(%S* %src, %S* %dst) {
; CHECK-LABEL: copyalias
; CHECK-NEXT: [[LOAD:%[a-z0-9\.]+]] = load %S, %S* %src
; CHECK-NOT: load
; CHECK: call void @llvm.memmove.p0i8.p0i8.i64
; CHECK-NEXT: store %S [[LOAD]], %S* %dst
; CHECK-NEXT: ret void
%1 = load %S, %S* %src
%2 = load %S, %S* %src
store %S %1, %S* %dst
store %S %2, %S* %dst
ret void
}
Fix PR26051: Memcpy optimization should introduce a call to memcpy before the store destination position This is a conservative fix, I expect Amaury to relax this. Follow-up for r256923 From: Mehdi Amini <mehdi.amini@apple.com> llvm-svn: 256999
2016-01-07 07:50:22 +08:00
Merge clang's isRepeatedBytePattern with LLVM's isBytewiseValue Summary: his code was in CGDecl.cpp and really belongs in LLVM's isBytewiseValue. Teach isBytewiseValue the tricks clang's isRepeatedBytePattern had, including merging undef properly, and recursing on more types. clang part of this patch: D51752 Subscribers: dexonsmith, llvm-commits Differential Revision: https://reviews.llvm.org/D51751 llvm-svn: 342709
2018-09-21 13:17:42 +08:00
; If the store address is computed in a complex manner, make
Imporove load to store => memcpy Summary: This now try to reorder instructions in order to help create the optimizable pattern. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Differential Revision: http://reviews.llvm.org/D16523 llvm-svn: 263503
2016-03-15 06:52:27 +08:00
; sure we lift the computation as well if needed and possible.
define void @addrproducer(%S* %src, %S* %dst) {
Merge clang's isRepeatedBytePattern with LLVM's isBytewiseValue Summary: his code was in CGDecl.cpp and really belongs in LLVM's isBytewiseValue. Teach isBytewiseValue the tricks clang's isRepeatedBytePattern had, including merging undef properly, and recursing on more types. clang part of this patch: D51752 Subscribers: dexonsmith, llvm-commits Differential Revision: https://reviews.llvm.org/D51751 llvm-svn: 342709
2018-09-21 13:17:42 +08:00
; CHECK-LABEL: addrproducer(
; CHECK-NEXT: %[[DSTCAST:[0-9]+]] = bitcast %S* %dst to i8*
; CHECK-NEXT: %dst2 = getelementptr %S, %S* %dst, i64 1
; CHECK-NEXT: %[[DST2CAST:[0-9]+]] = bitcast %S* %dst2 to i8*
; CHECK-NEXT: %[[SRCCAST:[0-9]+]] = bitcast %S* %src to i8*
; CHECK-NEXT: call void @llvm.memmove.p0i8.p0i8.i64(i8* align 8 %[[DST2CAST]], i8* align 8 %[[SRCCAST]], i64 16, i1 false)
; CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* align 8 %[[DSTCAST]], i8 undef, i64 16, i1 false)
Imporove load to store => memcpy Summary: This now try to reorder instructions in order to help create the optimizable pattern. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Differential Revision: http://reviews.llvm.org/D16523 llvm-svn: 263503
2016-03-15 06:52:27 +08:00
; CHECK-NEXT: ret void
Fix PR26051: Memcpy optimization should introduce a call to memcpy before the store destination position This is a conservative fix, I expect Amaury to relax this. Follow-up for r256923 From: Mehdi Amini <mehdi.amini@apple.com> llvm-svn: 256999
2016-01-07 07:50:22 +08:00
%1 = load %S, %S* %src
store %S undef, %S* %dst
%dst2 = getelementptr %S , %S* %dst, i64 1
store %S %1, %S* %dst2
ret void
}
Imporove load to store => memcpy Summary: This now try to reorder instructions in order to help create the optimizable pattern. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Differential Revision: http://reviews.llvm.org/D16523 llvm-svn: 263503
2016-03-15 06:52:27 +08:00
define void @aliasaddrproducer(%S* %src, %S* %dst, i32* %dstidptr) {
Merge clang's isRepeatedBytePattern with LLVM's isBytewiseValue Summary: his code was in CGDecl.cpp and really belongs in LLVM's isBytewiseValue. Teach isBytewiseValue the tricks clang's isRepeatedBytePattern had, including merging undef properly, and recursing on more types. clang part of this patch: D51752 Subscribers: dexonsmith, llvm-commits Differential Revision: https://reviews.llvm.org/D51751 llvm-svn: 342709
2018-09-21 13:17:42 +08:00
; CHECK-LABEL: aliasaddrproducer(
; CHECK-NEXT: %[[SRC:[0-9]+]] = load %S, %S* %src
; CHECK-NEXT: %[[DSTCAST:[0-9]+]] = bitcast %S* %dst to i8*
; CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* align 8 %[[DSTCAST]], i8 undef, i64 16, i1 false)
; CHECK-NEXT: %dstindex = load i32, i32* %dstidptr
; CHECK-NEXT: %dst2 = getelementptr %S, %S* %dst, i32 %dstindex
; CHECK-NEXT: store %S %[[SRC]], %S* %dst2
; CHECK-NEXT: ret void
Imporove load to store => memcpy Summary: This now try to reorder instructions in order to help create the optimizable pattern. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Differential Revision: http://reviews.llvm.org/D16523 llvm-svn: 263503
2016-03-15 06:52:27 +08:00
%1 = load %S, %S* %src
store %S undef, %S* %dst
%dstindex = load i32, i32* %dstidptr
%dst2 = getelementptr %S , %S* %dst, i32 %dstindex
store %S %1, %S* %dst2
ret void
}
define void @noaliasaddrproducer(%S* %src, %S* noalias %dst, i32* noalias %dstidptr) {
Merge clang's isRepeatedBytePattern with LLVM's isBytewiseValue Summary: his code was in CGDecl.cpp and really belongs in LLVM's isBytewiseValue. Teach isBytewiseValue the tricks clang's isRepeatedBytePattern had, including merging undef properly, and recursing on more types. clang part of this patch: D51752 Subscribers: dexonsmith, llvm-commits Differential Revision: https://reviews.llvm.org/D51751 llvm-svn: 342709
2018-09-21 13:17:42 +08:00
; CHECK-LABEL: noaliasaddrproducer(
; CHECK-NEXT: %[[SRCCAST:[0-9]+]] = bitcast %S* %src to i8*
; CHECK-NEXT: %[[LOADED:[0-9]+]] = load i32, i32* %dstidptr
; CHECK-NEXT: %dstindex = or i32 %[[LOADED]], 1
; CHECK-NEXT: %dst2 = getelementptr %S, %S* %dst, i32 %dstindex
; CHECK-NEXT: %[[DST2CAST:[0-9]+]] = bitcast %S* %dst2 to i8*
; CHECK-NEXT: %[[SRCCAST2:[0-9]+]] = bitcast %S* %src to i8*
; CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[DST2CAST]], i8* align 8 %[[SRCCAST2]], i64 16, i1 false)
; CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* align 8 %[[SRCCAST]], i8 undef, i64 16, i1 false)
; CHECK-NEXT: ret void
Imporove load to store => memcpy Summary: This now try to reorder instructions in order to help create the optimizable pattern. Reviewers: craig.topper, spatel, dexonsmith, Prazek, chandlerc, joker.eph, majnemer Differential Revision: http://reviews.llvm.org/D16523 llvm-svn: 263503
2016-03-15 06:52:27 +08:00
%1 = load %S, %S* %src
store %S undef, %S* %src
%2 = load i32, i32* %dstidptr
%dstindex = or i32 %2, 1
%dst2 = getelementptr %S , %S* %dst, i32 %dstindex
store %S %1, %S* %dst2
ret void
}