mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2025-08-10 02:25:47 +00:00
6683efb4cd
and was actually very wrong, fix it and make it simpler. Also remove the ConcatVectors function, which is unused now. - Fix a introduction of useless nodes in r126664 and r126264. The VUNPCKL* should never be introduced cause we don't want duplicate nodes for 128 AVX and non-AVX modes, the actual instruction difference only exists during isel, but not for target specific DAG nodes. We only introduce V* target nodes when there is no 128-bit version already there. - Fix a fragile test and make it more useful. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135729 91177308-0d34-0410-b5e6-96231b3b80d8
21 lines
986 B
LLVM
21 lines
986 B
LLVM
; RUN: llc < %s -mattr=+avx | 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-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
|
|
target triple = "x86_64-unknown-linux-gnu"
|
|
|
|
define void @try_([2 x <4 x double>]* noalias %incarray, [2 x <4 x double>]* noalias %incarrayb ) {
|
|
entry:
|
|
%incarray1 = alloca [2 x <4 x double>]*, align 8
|
|
%incarrayb1 = alloca [2 x <4 x double>]*, align 8
|
|
%carray = alloca [2 x <4 x double>], align 16
|
|
%r = getelementptr [2 x <4 x double>]* %incarray, i32 0, i32 0
|
|
%rb = getelementptr [2 x <4 x double>]* %incarrayb, i32 0, i32 0
|
|
%r3 = load <4 x double>* %r, align 8
|
|
%r4 = load <4 x double>* %rb, align 8
|
|
%r11 = shufflevector <4 x double> %r3, <4 x double> %r4, <4 x i32> < i32 0, i32 4, i32 1, i32 5 > ; <<4 x double>> [#uses=1]
|
|
; CHECK-NOT: vunpcklpd %ymm
|
|
%r12 = getelementptr [2 x <4 x double>]* %carray, i32 0, i32 1
|
|
store <4 x double> %r11, <4 x double>* %r12, align 4
|
|
ret void
|
|
}
|