From e56871b22fb8b75296592e9506f6a0adb277d267 Mon Sep 17 00:00:00 2001
From: Martien de Jong <martien.de-jong@xilinx.com>
Date: Wed, 29 Jan 2025 14:43:45 +0100
Subject: [PATCH] [AIE2P] add end-to-end pipeliner test

---
 .../schedule/postpipeliner/end-to-end.ll      | 71 +++++++++++++++++++
 1 file changed, 71 insertions(+)
 create mode 100644 llvm/test/CodeGen/AIE/aie2p/schedule/postpipeliner/end-to-end.ll

diff --git a/llvm/test/CodeGen/AIE/aie2p/schedule/postpipeliner/end-to-end.ll b/llvm/test/CodeGen/AIE/aie2p/schedule/postpipeliner/end-to-end.ll
new file mode 100644
index 000000000000..514da9cef0e7
--- /dev/null
+++ b/llvm/test/CodeGen/AIE/aie2p/schedule/postpipeliner/end-to-end.ll
@@ -0,0 +1,71 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+
+; This file is licensed under the Apache License v2.0 with LLVM Exceptions.
+; See https://llvm.org/LICENSE.txt for license information.
+; SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+;
+; (c) Copyright 2025 Advanced Micro Devices, Inc. or its affiliates
+
+; This test checks postpipelining for aie2p end-to-end:
+; loop rotation, hwloop generation, legalization, codeselection, hwloop lowering,
+; pipelining, loop distance padding and assembly generation.
+; The result is an eight stage pipeline with seven fully expanded bundles in the prologue
+
+; RUN: llc -O2 -mtriple=aie2p %s -o - | FileCheck %s
+
+define <32 x i16> @zol(i32 %n, ptr %p) {
+; CHECK-LABEL: zol:
+; CHECK:         .p2align 4
+; CHECK-NEXT:  // %bb.0: // %entry
+; CHECK-NEXT:    add.nc lc, r0, #-7
+; CHECK-NEXT:    movxm ls, #.LBB0_1
+; CHECK-NEXT:    movxm le, #.L_LEnd0
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopxm ; nopv
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopxm ; nopv
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopxm ; nopv
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopxm ; nopv
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopxm ; nopv
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopxm ; nopv
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopxm ; nopv
+; CHECK-NEXT:    // implicit-def: $x0
+; CHECK-NEXT:    .p2align 4
+; CHECK-NEXT:  .LBB0_1: // %for.body
+; CHECK-NEXT:    // =>This Inner Loop Header: Depth=1
+; CHECK-NEXT:  .L_LEnd0:
+; CHECK-NEXT:    vlda x2, [p0], #64; nopb ; nops ; nopx ; vadd.16 x0, x2, x0; nopv
+; CHECK-NEXT:  // %bb.2: // %for.cond.cleanup
+; CHECK-NEXT:    nopa ; nopx ; vadd.16 x0, x2, x0
+; CHECK-NEXT:    vadd.16 x0, x2, x0
+; CHECK-NEXT:    vadd.16 x0, x2, x0
+; CHECK-NEXT:    vadd.16 x0, x2, x0
+; CHECK-NEXT:    vadd.16 x0, x2, x0
+; CHECK-NEXT:    vadd.16 x0, x2, x0
+; CHECK-NEXT:    vadd.16 x0, x2, x0
+; CHECK-NEXT:    ret lr
+; CHECK-NEXT:    nop // Delay Slot 5
+; CHECK-NEXT:    nop // Delay Slot 4
+; CHECK-NEXT:    nop // Delay Slot 3
+; CHECK-NEXT:    nop // Delay Slot 2
+; CHECK-NEXT:    nop // Delay Slot 1
+entry:
+  br label %for.body
+
+for.body:                                         ; preds = %entry, %for.body
+  %i.05 = phi i32 [ 0, %entry ], [ %inc, %for.body ]
+  %s.04 = phi <32 x i16> [ undef, %entry ], [ %add, %for.body ]
+  %p.addr.03 = phi ptr [ %p, %entry ], [ %incdec.ptr, %for.body ]
+  %incdec.ptr = getelementptr inbounds i8, ptr %p.addr.03, i20 64
+  %0 = load <32 x i16>, ptr %p.addr.03, align 64
+  %add = add <32 x i16> %0, %s.04
+  %inc = add nuw nsw i32 %i.05, 1
+  %exitcond.not = icmp eq i32 %inc, %n
+  br i1 %exitcond.not, label %for.cond.cleanup, label %for.body, !llvm.loop !0
+
+for.cond.cleanup:
+  ret <32 x i16> %add
+}
+
+!0 = distinct !{!0, !1, !2, !3}
+!1 = !{!"llvm.loop.mustprogress"}
+!2 = !{!"llvm.loop.itercount.range", i64 10}
+!3 = !{!"llvm.loop.unroll.disable"}