mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-26 12:43:36 +01:00
db3c93e416
This will remove suboptimal branching from the generated ll/sc loops. The extra simplification pass affects a lot of testcases, which have been modified to accommodate this change: either by modifying the test to become immune to the CFG simplification, or (less preferablt) by adding option -hexagon-initial-cfg-clenaup=0. llvm-svn: 338774
28 lines
1.1 KiB
LLVM
28 lines
1.1 KiB
LLVM
; RUN: llc -march=hexagon < %s | FileCheck %s
|
|
|
|
; This testcase would fail on a bitcast from v64i16 to v32i32. Check that
|
|
; is compiles without errors.
|
|
; CHECK: valign
|
|
; CHECK: vshuff
|
|
|
|
target triple = "hexagon"
|
|
|
|
declare <32 x i32> @llvm.hexagon.V6.lo.128B(<64 x i32>) #0
|
|
declare <64 x i32> @llvm.hexagon.V6.vshuffvdd.128B(<32 x i32>, <32 x i32>, i32) #0
|
|
declare <32 x i32> @llvm.hexagon.V6.valignbi.128B(<32 x i32>, <32 x i32>, i32) #0
|
|
|
|
define void @fred(<64 x i16>* %a0, <32 x i32>* %a1) #1 {
|
|
entry:
|
|
%t0 = bitcast <64 x i16> zeroinitializer to <32 x i32>
|
|
%t1 = tail call <32 x i32> @llvm.hexagon.V6.valignbi.128B(<32 x i32> %t0, <32 x i32> undef, i32 2)
|
|
%t2 = tail call <64 x i32> @llvm.hexagon.V6.vshuffvdd.128B(<32 x i32> undef, <32 x i32> %t1, i32 -2)
|
|
%t3 = tail call <32 x i32> @llvm.hexagon.V6.lo.128B(<64 x i32> %t2)
|
|
store <64 x i16> zeroinitializer, <64 x i16>* %a0, align 128
|
|
store <32 x i32> %t3, <32 x i32>* %a1, align 128
|
|
ret void
|
|
}
|
|
|
|
|
|
attributes #0 = { nounwind readnone }
|
|
attributes #1 = { nounwind "target-cpu"="hexagonv60" "target-features"="+hvxv60,+hvx-length128b" }
|