From c99e17fef5f34ac536192fa7b915641f1962c7b9 Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev@redking.me.uk>
Date: Tue, 13 Jul 2021 13:48:04 +0100
Subject: [PATCH] [InstCombine] Pre-commit ashr(or(neg(x),x),bw-1) -->
 sext(icmp_ne(x,0)) tests from D105764

Added 'thwart complexity-based canonicalization' hacks and the lshr(or(neg(x),x),bw-1) --> zext(icmp_ne(x,0)) variants suggested by Sanjay.
---
 .../InstCombine/sub-ashr-or-to-icmp-select.ll | 86 ++++++++++++++++
 .../InstCombine/sub-lshr-or-to-icmp-select.ll | 98 +++++++++++++++++++
 2 files changed, 184 insertions(+)
 create mode 100644 llvm/test/Transforms/InstCombine/sub-lshr-or-to-icmp-select.ll

diff --git a/llvm/test/Transforms/InstCombine/sub-ashr-or-to-icmp-select.ll b/llvm/test/Transforms/InstCombine/sub-ashr-or-to-icmp-select.ll
index fa5a8f693819..21d7c8a62119 100644
--- a/llvm/test/Transforms/InstCombine/sub-ashr-or-to-icmp-select.ll
+++ b/llvm/test/Transforms/InstCombine/sub-ashr-or-to-icmp-select.ll
@@ -72,6 +72,19 @@ define i64 @sub_ashr_or_i64(i64 %x, i64 %y) {
   ret i64 %or
 }
 
+define i32 @neg_or_ashr_i32(i32 %x) {
+; CHECK-LABEL: @neg_or_ashr_i32(
+; CHECK-NEXT:    [[NEG:%.*]] = sub i32 0, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = ashr i32 [[OR]], 31
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %neg = sub i32 0, %x
+  %or = or i32 %neg, %x
+  %shr = ashr i32 %or, 31
+  ret i32 %shr
+}
+
 ; nuw nsw
 
 define i32 @sub_ashr_or_i32_nuw_nsw(i32 %x, i32 %y) {
@@ -100,6 +113,21 @@ define i32 @sub_ashr_or_i32_commute(i32 %x, i32 %y) {
   ret i32 %or
 }
 
+define i32 @neg_or_ashr_i32_commute(i32 %x0) {
+; CHECK-LABEL: @neg_or_ashr_i32_commute(
+; CHECK-NEXT:    [[X:%.*]] = sdiv i32 42, [[X0:%.*]]
+; CHECK-NEXT:    [[NEG:%.*]] = sub nsw i32 0, [[X]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[X]], [[NEG]]
+; CHECK-NEXT:    [[SHR:%.*]] = ashr i32 [[OR]], 31
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %x = sdiv i32 42, %x0 ; thwart complexity-based canonicalization
+  %neg = sub i32 0, %x
+  %or = or i32 %x, %neg
+  %shr = ashr i32 %or, 31
+  ret i32 %shr
+}
+
 ; Vector Types
 
 define <4 x i32> @sub_ashr_or_i32_vec(<4 x i32> %x, <4 x i32> %y) {
@@ -126,6 +154,19 @@ define <4 x i32> @sub_ashr_or_i32_vec_nuw_nsw(<4 x i32> %x, <4 x i32> %y) {
   ret <4 x i32> %or
 }
 
+define <4 x i32> @neg_or_ashr_i32_vec(<4 x i32> %x) {
+; CHECK-LABEL: @neg_or_ashr_i32_vec(
+; CHECK-NEXT:    [[NEG:%.*]] = sub <4 x i32> zeroinitializer, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or <4 x i32> [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = ashr <4 x i32> [[OR]], <i32 31, i32 31, i32 31, i32 31>
+; CHECK-NEXT:    ret <4 x i32> [[SHR]]
+;
+  %neg = sub <4 x i32> zeroinitializer, %x
+  %or = or <4 x i32> %neg, %x
+  %shr = ashr <4 x i32> %or, <i32 31, i32 31, i32 31, i32 31>
+  ret <4 x i32> %shr
+}
+
 define <4 x i32> @sub_ashr_or_i32_vec_commute(<4 x i32> %x, <4 x i32> %y) {
 ; CHECK-LABEL: @sub_ashr_or_i32_vec_commute(
 ; CHECK-NEXT:    [[TMP1:%.*]] = icmp slt <4 x i32> [[Y:%.*]], [[X:%.*]]
@@ -138,6 +179,21 @@ define <4 x i32> @sub_ashr_or_i32_vec_commute(<4 x i32> %x, <4 x i32> %y) {
   ret <4 x i32> %or
 }
 
+define <4 x i32> @neg_or_ashr_i32_vec_commute(<4 x i32> %x0) {
+; CHECK-LABEL: @neg_or_ashr_i32_vec_commute(
+; CHECK-NEXT:    [[X:%.*]] = sdiv <4 x i32> <i32 42, i32 42, i32 42, i32 42>, [[X0:%.*]]
+; CHECK-NEXT:    [[NEG:%.*]] = sub nsw <4 x i32> zeroinitializer, [[X]]
+; CHECK-NEXT:    [[OR:%.*]] = or <4 x i32> [[X]], [[NEG]]
+; CHECK-NEXT:    [[SHR:%.*]] = ashr <4 x i32> [[OR]], <i32 31, i32 31, i32 31, i32 31>
+; CHECK-NEXT:    ret <4 x i32> [[SHR]]
+;
+  %x = sdiv <4 x i32> <i32 42, i32 42, i32 42, i32 42>, %x0 ; thwart complexity-based canonicalization
+  %neg = sub <4 x i32> zeroinitializer, %x
+  %or = or <4 x i32> %x, %neg
+  %shr = ashr <4 x i32> %or, <i32 31, i32 31, i32 31, i32 31>
+  ret <4 x i32> %shr
+}
+
 ; Extra uses
 
 define i32 @sub_ashr_or_i32_extra_use_sub(i32 %x, i32 %y, i32* %p) {
@@ -169,6 +225,21 @@ define i32 @sub_ashr_or_i32_extra_use_or(i32 %x, i32 %y, i32* %p) {
   ret i32 %or
 }
 
+define i32 @neg_extra_use_or_ashr_i32(i32 %x, i32* %p) {
+; CHECK-LABEL: @neg_extra_use_or_ashr_i32(
+; CHECK-NEXT:    [[NEG:%.*]] = sub i32 0, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = ashr i32 [[OR]], 31
+; CHECK-NEXT:    store i32 [[NEG]], i32* [[P:%.*]], align 4
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %neg = sub i32 0, %x
+  %or = or i32 %neg, %x
+  %shr = ashr i32 %or, 31
+  store i32 %neg, i32* %p
+  ret i32 %shr
+}
+
 ; Negative Tests
 
 define i32 @sub_ashr_or_i32_extra_use_ashr(i32 %x, i32 %y, i32* %p) {
@@ -199,6 +270,21 @@ define i32 @sub_ashr_or_i32_no_nsw_nuw(i32 %x, i32 %y) {
   ret i32 %or
 }
 
+define i32 @neg_or_extra_use_ashr_i32(i32 %x, i32* %p) {
+; CHECK-LABEL: @neg_or_extra_use_ashr_i32(
+; CHECK-NEXT:    [[NEG:%.*]] = sub i32 0, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = ashr i32 [[OR]], 31
+; CHECK-NEXT:    store i32 [[OR]], i32* [[P:%.*]], align 4
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %neg = sub i32 0, %x
+  %or = or i32 %neg, %x
+  %shr = ashr i32 %or, 31
+  store i32 %or, i32* %p
+  ret i32 %shr
+}
+
 define <4 x i32> @sub_ashr_or_i32_vec_undef1(<4 x i32> %x) {
 ; CHECK-LABEL: @sub_ashr_or_i32_vec_undef1(
 ; CHECK-NEXT:    [[SUB:%.*]] = sub <4 x i32> <i32 255, i32 255, i32 undef, i32 255>, [[X:%.*]]
diff --git a/llvm/test/Transforms/InstCombine/sub-lshr-or-to-icmp-select.ll b/llvm/test/Transforms/InstCombine/sub-lshr-or-to-icmp-select.ll
new file mode 100644
index 000000000000..830b9f99b8eb
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/sub-lshr-or-to-icmp-select.ll
@@ -0,0 +1,98 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -instcombine %s -S -o - | FileCheck %s
+
+; Scalar Types
+
+define i32 @neg_or_lshr_i32(i32 %x) {
+; CHECK-LABEL: @neg_or_lshr_i32(
+; CHECK-NEXT:    [[NEG:%.*]] = sub i32 0, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = lshr i32 [[OR]], 31
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %neg = sub i32 0, %x
+  %or = or i32 %neg, %x
+  %shr = lshr i32 %or, 31
+  ret i32 %shr
+}
+
+; Commute
+
+define i32 @neg_or_lshr_i32_commute(i32 %x0) {
+; CHECK-LABEL: @neg_or_lshr_i32_commute(
+; CHECK-NEXT:    [[X:%.*]] = sdiv i32 42, [[X0:%.*]]
+; CHECK-NEXT:    [[NEG:%.*]] = sub nsw i32 0, [[X]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[X]], [[NEG]]
+; CHECK-NEXT:    [[SHR:%.*]] = lshr i32 [[OR]], 31
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %x = sdiv i32 42, %x0 ; thwart complexity-based canonicalization
+  %neg = sub i32 0, %x
+  %or = or i32 %x, %neg
+  %shr = lshr i32 %or, 31
+  ret i32 %shr
+}
+
+; Vector Types
+
+define <4 x i32> @neg_or_lshr_i32_vec(<4 x i32> %x) {
+; CHECK-LABEL: @neg_or_lshr_i32_vec(
+; CHECK-NEXT:    [[NEG:%.*]] = sub <4 x i32> zeroinitializer, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or <4 x i32> [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = lshr <4 x i32> [[OR]], <i32 31, i32 31, i32 31, i32 31>
+; CHECK-NEXT:    ret <4 x i32> [[SHR]]
+;
+  %neg = sub <4 x i32> zeroinitializer, %x
+  %or = or <4 x i32> %neg, %x
+  %shr = lshr <4 x i32> %or, <i32 31, i32 31, i32 31, i32 31>
+  ret <4 x i32> %shr
+}
+
+define <4 x i32> @neg_or_lshr_i32_vec_commute(<4 x i32> %x0) {
+; CHECK-LABEL: @neg_or_lshr_i32_vec_commute(
+; CHECK-NEXT:    [[X:%.*]] = sdiv <4 x i32> <i32 42, i32 42, i32 42, i32 42>, [[X0:%.*]]
+; CHECK-NEXT:    [[NEG:%.*]] = sub nsw <4 x i32> zeroinitializer, [[X]]
+; CHECK-NEXT:    [[OR:%.*]] = or <4 x i32> [[X]], [[NEG]]
+; CHECK-NEXT:    [[SHR:%.*]] = lshr <4 x i32> [[OR]], <i32 31, i32 31, i32 31, i32 31>
+; CHECK-NEXT:    ret <4 x i32> [[SHR]]
+;
+  %x = sdiv <4 x i32> <i32 42, i32 42, i32 42, i32 42>, %x0 ; thwart complexity-based canonicalization
+  %neg = sub <4 x i32> zeroinitializer, %x
+  %or = or <4 x i32> %x, %neg
+  %shr = lshr <4 x i32> %or, <i32 31, i32 31, i32 31, i32 31>
+  ret <4 x i32> %shr
+}
+
+; Extra uses
+
+define i32 @neg_extra_use_or_lshr_i32(i32 %x, i32* %p) {
+; CHECK-LABEL: @neg_extra_use_or_lshr_i32(
+; CHECK-NEXT:    [[NEG:%.*]] = sub i32 0, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = lshr i32 [[OR]], 31
+; CHECK-NEXT:    store i32 [[NEG]], i32* [[P:%.*]], align 4
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %neg = sub i32 0, %x
+  %or = or i32 %neg, %x
+  %shr = lshr i32 %or, 31
+  store i32 %neg, i32* %p
+  ret i32 %shr
+}
+
+; Negative Tests
+
+define i32 @neg_or_extra_use_lshr_i32(i32 %x, i32* %p) {
+; CHECK-LABEL: @neg_or_extra_use_lshr_i32(
+; CHECK-NEXT:    [[NEG:%.*]] = sub i32 0, [[X:%.*]]
+; CHECK-NEXT:    [[OR:%.*]] = or i32 [[NEG]], [[X]]
+; CHECK-NEXT:    [[SHR:%.*]] = lshr i32 [[OR]], 31
+; CHECK-NEXT:    store i32 [[OR]], i32* [[P:%.*]], align 4
+; CHECK-NEXT:    ret i32 [[SHR]]
+;
+  %neg = sub i32 0, %x
+  %or = or i32 %neg, %x
+  %shr = lshr i32 %or, 31
+  store i32 %or, i32* %p
+  ret i32 %shr
+}