You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

123 lines
6.5 KiB

; RUN: llc -O3 -march=hexagon < %s | FileCheck %s
; CHECK: v{{[0-9]+}} = vmem(r{{[0-9]+}}+#0)
target triple = "hexagon"
; Function Attrs: nounwind
define void @f0(i16* nocapture %a0) #0 {
b0:
br i1 undef, label %b1, label %b5
b1: ; preds = %b0
%v0 = bitcast i16* %a0 to <16 x i32>*
br label %b2
b2: ; preds = %b4, %b1
%v1 = phi i32 [ 0, %b1 ], [ %v50, %b4 ]
%v2 = phi <16 x i32>* [ %v0, %b1 ], [ undef, %b4 ]
br label %b3
b3: ; preds = %b3, %b2
%v3 = phi i32 [ -4, %b2 ], [ %v40, %b3 ]
%v4 = add i32 0, -64
%v5 = getelementptr inbounds i8, i8* null, i32 %v4
%v6 = bitcast i8* %v5 to <16 x i32>*
%v7 = load <16 x i32>, <16 x i32>* %v6, align 64, !tbaa !0
%v8 = tail call <16 x i32> @llvm.hexagon.V6.vlalignbi(<16 x i32> undef, <16 x i32> %v7, i32 4)
%v9 = tail call <16 x i32> @llvm.hexagon.V6.vabsdiffub(<16 x i32> %v8, <16 x i32> zeroinitializer)
%v10 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> %v9, <16 x i32> undef)
%v11 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v10, <16 x i32> undef, <16 x i32> undef)
%v12 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> zeroinitializer, <16 x i32> %v11, <16 x i32> undef)
%v13 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> undef, <16 x i32> %v12, <16 x i32> undef)
%v14 = tail call <16 x i32> @llvm.hexagon.V6.valignbi(<16 x i32> undef, <16 x i32> undef, i32 1)
%v15 = tail call <16 x i32> @llvm.hexagon.V6.vabsdiffub(<16 x i32> %v14, <16 x i32> zeroinitializer)
%v16 = tail call <16 x i32> @llvm.hexagon.V6.vabsdiffub(<16 x i32> zeroinitializer, <16 x i32> zeroinitializer)
%v17 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> zeroinitializer, <16 x i32> undef)
%v18 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> %v15, <16 x i32> undef)
%v19 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> zeroinitializer, <16 x i32> undef)
%v20 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> %v16, <16 x i32> undef)
%v21 = tail call <16 x i32> @llvm.hexagon.V6.vmux(<64 x i1> %v19, <16 x i32> undef, <16 x i32> zeroinitializer)
%v22 = tail call <16 x i32> @llvm.hexagon.V6.vmux(<64 x i1> %v20, <16 x i32> undef, <16 x i32> zeroinitializer)
%v23 = tail call <32 x i32> @llvm.hexagon.V6.vcombine(<16 x i32> %v22, <16 x i32> %v21)
%v24 = tail call <32 x i32> @llvm.hexagon.V6.vmpabus.acc(<32 x i32> zeroinitializer, <32 x i32> %v23, i32 16843009)
%v25 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v17, <16 x i32> %v13, <16 x i32> undef)
%v26 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v18, <16 x i32> %v25, <16 x i32> undef)
%v27 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v19, <16 x i32> %v26, <16 x i32> undef)
%v28 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v20, <16 x i32> %v27, <16 x i32> undef)
%v29 = tail call <16 x i32> @llvm.hexagon.V6.vabsdiffub(<16 x i32> undef, <16 x i32> zeroinitializer)
%v30 = tail call <16 x i32> @llvm.hexagon.V6.vabsdiffub(<16 x i32> zeroinitializer, <16 x i32> zeroinitializer)
%v31 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> undef, <16 x i32> undef)
%v32 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> %v29, <16 x i32> undef)
%v33 = tail call <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32> %v30, <16 x i32> undef)
%v34 = tail call <32 x i32> @llvm.hexagon.V6.vmpabus.acc(<32 x i32> %v24, <32 x i32> zeroinitializer, i32 16843009)
%v35 = tail call <32 x i32> @llvm.hexagon.V6.vmpabus.acc(<32 x i32> %v34, <32 x i32> undef, i32 16843009)
%v36 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> zeroinitializer, <16 x i32> %v28, <16 x i32> undef)
%v37 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v31, <16 x i32> %v36, <16 x i32> undef)
%v38 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v32, <16 x i32> %v37, <16 x i32> undef)
%v39 = tail call <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1> %v33, <16 x i32> %v38, <16 x i32> undef)
%v40 = add nsw i32 %v3, 3
%v41 = icmp eq i32 %v40, 5
br i1 %v41, label %b4, label %b3
b4: ; preds = %b3
%v42 = phi <16 x i32> [ %v39, %b3 ]
%v43 = phi <32 x i32> [ %v35, %b3 ]
%v44 = tail call <16 x i32> @llvm.hexagon.V6.lo(<32 x i32> %v43)
%v45 = tail call <32 x i32> @llvm.hexagon.V6.vshuffvdd(<16 x i32> zeroinitializer, <16 x i32> %v44, i32 -2)
%v46 = tail call <32 x i32> @llvm.hexagon.V6.vunpackub(<16 x i32> %v42)
%v47 = tail call <16 x i32> @llvm.hexagon.V6.lo(<32 x i32> %v45)
store <16 x i32> %v47, <16 x i32>* %v2, align 64, !tbaa !0
%v48 = getelementptr inbounds <16 x i32>, <16 x i32>* null, i32 1
%v49 = tail call <16 x i32> @llvm.hexagon.V6.hi(<32 x i32> %v46)
store <16 x i32> %v49, <16 x i32>* %v48, align 64, !tbaa !0
%v50 = add nsw i32 %v1, 1
%v51 = icmp slt i32 %v50, 0
br i1 %v51, label %b2, label %b5
b5: ; preds = %b4, %b0
ret void
}
; Function Attrs: nounwind readnone
declare <32 x i32> @llvm.hexagon.V6.vcombine(<16 x i32>, <16 x i32>) #1
; Function Attrs: nounwind readnone
declare <16 x i32> @llvm.hexagon.V6.vabsdiffub(<16 x i32>, <16 x i32>) #1
; Function Attrs: nounwind readnone
declare <64 x i1> @llvm.hexagon.V6.vgtub(<16 x i32>, <16 x i32>) #1
; Function Attrs: nounwind readnone
declare <16 x i32> @llvm.hexagon.V6.vmux(<64 x i1>, <16 x i32>, <16 x i32>) #1
; Function Attrs: nounwind readnone
declare <16 x i32> @llvm.hexagon.V6.vaddbnq(<64 x i1>, <16 x i32>, <16 x i32>) #1
; Function Attrs: nounwind readnone
declare <16 x i32> @llvm.hexagon.V6.vlalignbi(<16 x i32>, <16 x i32>, i32) #1
; Function Attrs: nounwind readnone
declare <16 x i32> @llvm.hexagon.V6.valignbi(<16 x i32>, <16 x i32>, i32) #1
; Function Attrs: nounwind readnone
declare <32 x i32> @llvm.hexagon.V6.vmpabus.acc(<32 x i32>, <32 x i32>, i32) #1
; Function Attrs: nounwind readnone
declare <32 x i32> @llvm.hexagon.V6.vshuffvdd(<16 x i32>, <16 x i32>, i32) #1
; Function Attrs: nounwind readnone
declare <16 x i32> @llvm.hexagon.V6.hi(<32 x i32>) #1
; Function Attrs: nounwind readnone
declare <16 x i32> @llvm.hexagon.V6.lo(<32 x i32>) #1
; Function Attrs: nounwind readnone
declare <32 x i32> @llvm.hexagon.V6.vunpackub(<16 x i32>) #1
attributes #0 = { nounwind "target-cpu"="hexagonv60" "target-features"="+hvxv60,+hvx-length64b" }
attributes #1 = { nounwind readnone }
!0 = !{!1, !1, i64 0}
!1 = !{!"omnipotent char", !2, i64 0}
!2 = !{!"Simple C/C++ TBAA"}