6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-12-21 00:32:23 +00:00
Code Issues Projects Releases Wiki Activity
bc051faae1
llvm-6502/test/CodeGen/PowerPC/unal-altivec.ll

53 lines
2.2 KiB
LLVM
Raw Normal View History

PPC: Initial support for permutation-based unaligned Altivec loads Altivec only directly supports aligned loads, but the loads have a strange property: If given an unaligned address, they truncate the address to the next lower aligned address, and load from there. This property, along with an extra load and some special-purpose permutation-control instructions that generate the appropriate permutations from the original unaligned address, allow efficient lowering of aligned loads. This code uses the trick explained in the Apple Velocity Engine optimization overview document to prevent the needed extra load from possibly causing a page fault if the original address happens to be aligned. As noted in the FIXMEs, there are several additional optimizations that can be performed to reduce the cost of these loads even more. These will be implemented in future commits. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@182691 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-24 23:00:14 +00:00
; RUN: llc < %s -mcpu=g5 | FileCheck %s
target datalayout = "E-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f128:128:128-v128:128:128-n32:64"
target triple = "powerpc64-unknown-linux-gnu"
define void @foo(float* noalias nocapture %a, float* noalias nocapture %b) #0 {
vector.ph:
br label %vector.body
vector.body: ; preds = %vector.body, %vector.ph
%index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ]
%0 = getelementptr inbounds float* %b, i64 %index
%1 = bitcast float* %0 to <4 x float>*
%wide.load = load <4 x float>* %1, align 4
%.sum11 = or i64 %index, 4
%2 = getelementptr float* %b, i64 %.sum11
%3 = bitcast float* %2 to <4 x float>*
%wide.load8 = load <4 x float>* %3, align 4
%4 = fadd <4 x float> %wide.load, <float 1.000000e+00, float 1.000000e+00, float 1.000000e+00, float 1.000000e+00>
%5 = fadd <4 x float> %wide.load8, <float 1.000000e+00, float 1.000000e+00, float 1.000000e+00, float 1.000000e+00>
%6 = getelementptr inbounds float* %a, i64 %index
%7 = bitcast float* %6 to <4 x float>*
store <4 x float> %4, <4 x float>* %7, align 4
%.sum12 = or i64 %index, 4
%8 = getelementptr float* %a, i64 %.sum12
%9 = bitcast float* %8 to <4 x float>*
store <4 x float> %5, <4 x float>* %9, align 4
%index.next = add i64 %index, 8
%10 = icmp eq i64 %index.next, 16000
br i1 %10, label %for.end, label %vector.body
; CHECK: @foo
Prefer to duplicate PPC Altivec loads when expanding unaligned loads When expanding unaligned Altivec loads, we use the decremented offset trick to prevent page faults. Unfortunately, if we have a sequence of consecutive unaligned loads, this leads to suboptimal code generation because the 'extra' load from the first unaligned load can be combined with the base load from the second (but only if the decremented offset trick is not used for the first). Search up and down the chain, through loads and token factors, looking for consecutive loads, and if one is found, don't use the offset reduction trick. These duplicate loads are later combined to yield the desired sequence (in the future, we might want a more-powerful chain search, but that will require some changes to allow the combiner routines to access the AA object). This should complete the initial implementation of the optimized unaligned Altivec load expansion. There is some refactoring that should be done, but that will happen when the unaligned store expansion is added. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@182719 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-26 18:08:30 +00:00
; CHECK-DAG: li [[C0:[0-9]+]], 0
; CHECK-DAG: li [[C16:[0-9]+]], 16
; CHECK-DAG: li [[C31:[0-9]+]], 31
; CHECK-DAG: lvx [[CNST:[0-9]+]],
; CHECK: .LBB0_1:
; CHECK-DAG: lvsl [[PC:[0-9]+]], [[B1:[0-9]+]], [[C0]]
; CHECK-DAG: lvx [[LD1:[0-9]+]], [[B1]], [[C0]]
; CHECK-DAG: add [[B3:[0-9]+]], [[B1]], [[C0]]
; CHECK-DAG: lvx [[LD2:[0-9]+]], [[B3]], [[C16]]
; CHECK-DAG: lvx [[LD3:[0-9]+]], [[B3]], [[C31]]
PPC: Initial support for permutation-based unaligned Altivec loads Altivec only directly supports aligned loads, but the loads have a strange property: If given an unaligned address, they truncate the address to the next lower aligned address, and load from there. This property, along with an extra load and some special-purpose permutation-control instructions that generate the appropriate permutations from the original unaligned address, allow efficient lowering of aligned loads. This code uses the trick explained in the Apple Velocity Engine optimization overview document to prevent the needed extra load from possibly causing a page fault if the original address happens to be aligned. As noted in the FIXMEs, there are several additional optimizations that can be performed to reduce the cost of these loads even more. These will be implemented in future commits. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@182691 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-24 23:00:14 +00:00
; CHECK-DAG: vperm [[R1:[0-9]+]], [[LD1]], [[LD2]], [[PC]]
Prefer to duplicate PPC Altivec loads when expanding unaligned loads When expanding unaligned Altivec loads, we use the decremented offset trick to prevent page faults. Unfortunately, if we have a sequence of consecutive unaligned loads, this leads to suboptimal code generation because the 'extra' load from the first unaligned load can be combined with the base load from the second (but only if the decremented offset trick is not used for the first). Search up and down the chain, through loads and token factors, looking for consecutive loads, and if one is found, don't use the offset reduction trick. These duplicate loads are later combined to yield the desired sequence (in the future, we might want a more-powerful chain search, but that will require some changes to allow the combiner routines to access the AA object). This should complete the initial implementation of the optimized unaligned Altivec load expansion. There is some refactoring that should be done, but that will happen when the unaligned store expansion is added. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@182719 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-26 18:08:30 +00:00
; CHECK-DAG: vperm [[R2:[0-9]+]], [[LD2]], [[LD3]], [[PC]]
; CHECK-DAG: vaddfp {{[0-9]+}}, [[R1]], [[CNST]]
; CHECK-DAG: vaddfp {{[0-9]+}}, [[R2]], [[CNST]]
PPC: Initial support for permutation-based unaligned Altivec loads Altivec only directly supports aligned loads, but the loads have a strange property: If given an unaligned address, they truncate the address to the next lower aligned address, and load from there. This property, along with an extra load and some special-purpose permutation-control instructions that generate the appropriate permutations from the original unaligned address, allow efficient lowering of aligned loads. This code uses the trick explained in the Apple Velocity Engine optimization overview document to prevent the needed extra load from possibly causing a page fault if the original address happens to be aligned. As noted in the FIXMEs, there are several additional optimizations that can be performed to reduce the cost of these loads even more. These will be implemented in future commits. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@182691 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-24 23:00:14 +00:00
; CHECK: blr
for.end: ; preds = %vector.body
ret void
}
attributes #0 = { nounwind }
Reference in New Issue Copy Permalink