forked from OSchip/llvm-project
27 lines
924 B
LLVM
27 lines
924 B
LLVM
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
|
|
|
|
; There was an infinite loop in DAGCombiner from a target build_vector
|
|
; combine and a generic insert_vector_elt combine.
|
|
|
|
; GCN-LABEL: {{^}}combine_loop:
|
|
; GCN: flat_load_short_d16_hi
|
|
; GCN: flat_store_short
|
|
define amdgpu_kernel void @combine_loop(i16* %arg) #0 {
|
|
bb:
|
|
br label %bb1
|
|
|
|
bb1:
|
|
%tmp = phi <2 x i16> [ <i16 15360, i16 15360>, %bb ], [ %tmp5, %bb1 ]
|
|
%tmp2 = phi half [ 0xH0000, %bb ], [ %tmp8, %bb1 ]
|
|
%tmp3 = load volatile half, half* null, align 536870912
|
|
%tmp4 = bitcast half %tmp3 to i16
|
|
%tmp5 = insertelement <2 x i16> <i16 0, i16 undef>, i16 %tmp4, i32 1
|
|
%tmp6 = bitcast i16* %arg to half*
|
|
store volatile half %tmp2, half* %tmp6, align 2
|
|
%tmp7 = bitcast <2 x i16> %tmp to <2 x half>
|
|
%tmp8 = extractelement <2 x half> %tmp7, i32 0
|
|
br label %bb1
|
|
}
|
|
|
|
attributes #0 = { nounwind }
|