48 lines
1.8 KiB
LLVM
48 lines
1.8 KiB
LLVM
; REQUIRES: asserts
|
|
; RUN: llc < %s -mtriple=aarch64 -mcpu=cyclone -mattr=+slow-misaligned-128store -enable-misched -verify-misched -o - | FileCheck %s
|
|
|
|
; Tests to check that the scheduler dependencies derived from alias analysis are
|
|
; correct when we have loads that have been split up so that they can later be
|
|
; merged into STP.
|
|
|
|
; Now that overwritten stores are elided in SelectionDAG, dependencies
|
|
; are resolved and removed before MISCHED. Check that we have
|
|
; equivalent pair of stp calls as a baseline.
|
|
|
|
; CHECK-LABEL: test_splat
|
|
; CHECK: ldr [[REG:w[0-9]+]], [x2]
|
|
; CHECK-DAG: stp w0, [[REG]], [x2, #12]
|
|
; CHECK-DAG: stp [[REG]], w1, [x2, #4]
|
|
define void @test_splat(i32 %x, i32 %y, ptr %p) {
|
|
entry:
|
|
%val = load i32, ptr %p, align 4
|
|
%0 = getelementptr inbounds i32, ptr %p, i64 1
|
|
%1 = getelementptr inbounds i32, ptr %p, i64 2
|
|
%2 = getelementptr inbounds i32, ptr %p, i64 3
|
|
%vec0 = insertelement <4 x i32> undef, i32 %val, i32 0
|
|
%vec1 = insertelement <4 x i32> %vec0, i32 %val, i32 1
|
|
%vec2 = insertelement <4 x i32> %vec1, i32 %val, i32 2
|
|
%vec3 = insertelement <4 x i32> %vec2, i32 %val, i32 3
|
|
store <4 x i32> %vec3, ptr %0, align 4
|
|
store i32 %x, ptr %2, align 4
|
|
store i32 %y, ptr %1, align 4
|
|
ret void
|
|
}
|
|
|
|
declare void @llvm.memset.p0.i64(ptr nocapture, i8, i64, i1)
|
|
%struct.tree_common = type { ptr, ptr, i32 }
|
|
|
|
; CHECK-LABEL: test_zero
|
|
; CHECK-DAG: stp x2, xzr, [x0, #8]
|
|
; CHECK-DAG: str w1, [x0, #16]
|
|
; CHECK-DAG: str xzr, [x0]
|
|
|
|
define void @test_zero(ptr %t, i32 %code, ptr %type) {
|
|
entry:
|
|
tail call void @llvm.memset.p0.i64(ptr align 8 %t, i8 0, i64 24, i1 false)
|
|
%code1 = getelementptr inbounds %struct.tree_common, ptr %t, i64 0, i32 2
|
|
store i32 %code, ptr %code1, align 8
|
|
%type2 = getelementptr inbounds %struct.tree_common, ptr %t, i64 0, i32 1
|
|
store ptr %type, ptr %type2, align 8
|
|
ret void
|
|
}
|