llvm-6502/test/CodeGen/X86/fold-load-vec.ll
Filipe Cabecinhas ca162faee2 Added more insertps optimizations
Summary:
When inserting an element that's coming from a vector load or a broadcast
of a vector (or scalar) load, combine the load into the insertps
instruction.
Added PerformINSERTPSCombine for the case where we need to fix the load
(load of a vector + insertps with a non-zero CountS).
Added patterns for the broadcasts.

Also added tests for SSE4.1, AVX, and AVX2.

Reviewers: delena, nadav, craig.topper

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D3581

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@209156 91177308-0d34-0410-b5e6-96231b3b80d8
2014-05-19 19:45:57 +00:00

40 lines
1.7 KiB
LLVM

; RUN: llc < %s -march=x86-64 -mcpu=corei7 -mattr=+sse4.1 | FileCheck %s
; rdar://12721174
; We should not fold movss into pshufd since pshufd expects m128 while movss
; loads from m32.
define void @sample_test(<4 x float>* %source, <2 x float>* %dest) nounwind {
; CHECK: sample_test
; CHECK-NOT: movaps
; CHECK: insertps
entry:
%source.addr = alloca <4 x float>*, align 8
%dest.addr = alloca <2 x float>*, align 8
%tmp = alloca <2 x float>, align 8
store <4 x float>* %source, <4 x float>** %source.addr, align 8
store <2 x float>* %dest, <2 x float>** %dest.addr, align 8
store <2 x float> zeroinitializer, <2 x float>* %tmp, align 8
%0 = load <4 x float>** %source.addr, align 8
%arrayidx = getelementptr inbounds <4 x float>* %0, i64 0
%1 = load <4 x float>* %arrayidx, align 16
%2 = extractelement <4 x float> %1, i32 0
%3 = load <2 x float>* %tmp, align 8
%4 = insertelement <2 x float> %3, float %2, i32 1
store <2 x float> %4, <2 x float>* %tmp, align 8
%5 = load <2 x float>* %tmp, align 8
%6 = load <2 x float>** %dest.addr, align 8
%arrayidx1 = getelementptr inbounds <2 x float>* %6, i64 0
store <2 x float> %5, <2 x float>* %arrayidx1, align 8
%7 = load <2 x float>** %dest.addr, align 8
%arrayidx2 = getelementptr inbounds <2 x float>* %7, i64 0
%8 = load <2 x float>* %arrayidx2, align 8
%vecext = extractelement <2 x float> %8, i32 0
%9 = load <2 x float>** %dest.addr, align 8
%arrayidx3 = getelementptr inbounds <2 x float>* %9, i64 0
%10 = load <2 x float>* %arrayidx3, align 8
%vecext4 = extractelement <2 x float> %10, i32 1
call void @ext(float %vecext, float %vecext4)
ret void
}
declare void @ext(float, float)