From: Dmitry Makogon <d.makogon@g.nsu.ru>
Date: Fri, 10 Feb 2023 09:00:16 +0000 (+0700)
Subject: [LVI] Don't traverse uses when calculating range at use
X-Git-Tag: upstream/17.0.6~17965
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=c77c186a647b385c291ddabecd70a2b4f84ae342;p=platform%2Fupstream%2Fllvm.git

[LVI] Don't traverse uses when calculating range at use

This effectively reverts 5c38c6a and 4f772b0.

A recently introduced LazyValueInfo::getConstantRangeAtUse returns incorrect
ranges for values in certain cases. One such example is described in PR60629.
The issue has something to do with traversing PHI uses of a value transitively.
As nikic pointed out, we're effectively reasoning about values from different
loop iterations.

In the faulting test case, CVP made a miscompilation because the calculated
range for a shift argument was incorrect. It returned empty-set, however it is
clearly not a dead code. CVP then erased the shift instruction because
of empty range.
---

diff --git a/llvm/lib/Analysis/LazyValueInfo.cpp b/llvm/lib/Analysis/LazyValueInfo.cpp
index 16d9ad7..efe878d 100644
--- a/llvm/lib/Analysis/LazyValueInfo.cpp
+++ b/llvm/lib/Analysis/LazyValueInfo.cpp
@@ -1662,7 +1662,7 @@ ConstantRange LazyValueInfo::getConstantRangeAtUse(const Use &U,
   // position where V can be constrained by a select or branch condition.
   const Use *CurrU = &U;
   // TODO: Increase limit?
-  const unsigned MaxUsesToInspect = 3;
+  const unsigned MaxUsesToInspect = 0;
   for (unsigned I = 0; I < MaxUsesToInspect; ++I) {
     std::optional<ValueLatticeElement> CondVal;
     auto *CurrI = cast<Instruction>(CurrU->getUser());
diff --git a/llvm/test/Analysis/LazyValueAnalysis/pr60629.ll b/llvm/test/Analysis/LazyValueAnalysis/pr60629.ll
index feae45b..7cc314e 100644
--- a/llvm/test/Analysis/LazyValueAnalysis/pr60629.ll
+++ b/llvm/test/Analysis/LazyValueAnalysis/pr60629.ll
@@ -1,8 +1,6 @@
 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt -passes=correlated-propagation -S %s | FileCheck %s
 
-; XFAIL: *
-
 ; Check that shift is not removed by CVP because of incorrect range returned by LVI::getConstantRangeAtUse.
 ; https://github.com/llvm/llvm-project/issues/60629
 
diff --git a/llvm/test/Transforms/CorrelatedValuePropagation/cond-at-use.ll b/llvm/test/Transforms/CorrelatedValuePropagation/cond-at-use.ll
index 6270ca8..1b09ca0 100644
--- a/llvm/test/Transforms/CorrelatedValuePropagation/cond-at-use.ll
+++ b/llvm/test/Transforms/CorrelatedValuePropagation/cond-at-use.ll
@@ -9,9 +9,9 @@ declare i16 @llvm.abs.i16(i16, i1)
 
 define i16 @sel_true_cond(i16 %x) {
 ; CHECK-LABEL: @sel_true_cond(
-; CHECK-NEXT:    [[SUB1:%.*]] = sub nuw i16 [[X:%.*]], 10
+; CHECK-NEXT:    [[SUB:%.*]] = call i16 @llvm.usub.sat.i16(i16 [[X:%.*]], i16 10)
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp uge i16 [[X]], 10
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[SUB1]], i16 42
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[SUB]], i16 42
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %sub = call i16 @llvm.usub.sat.i16(i16 %x, i16 10)
@@ -76,8 +76,8 @@ define i16 @sel_true_cond_extra_use(i16 %x) {
 
 define i16 @sel_true_cond_chain_speculatable(i16 %x) {
 ; CHECK-LABEL: @sel_true_cond_chain_speculatable(
-; CHECK-NEXT:    [[SUB1:%.*]] = add nuw i16 [[X:%.*]], 1
-; CHECK-NEXT:    [[EXTRA:%.*]] = mul i16 [[SUB1]], 3
+; CHECK-NEXT:    [[SUB:%.*]] = call i16 @llvm.uadd.sat.i16(i16 [[X:%.*]], i16 1)
+; CHECK-NEXT:    [[EXTRA:%.*]] = mul i16 [[SUB]], 3
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ne i16 [[X]], -1
 ; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[EXTRA]], i16 42
 ; CHECK-NEXT:    ret i16 [[SEL]]
@@ -125,9 +125,9 @@ define i16 @sel_true_cond_longer_chain(i16 %x) {
 
 define i16 @sel_false_cond(i16 %x) {
 ; CHECK-LABEL: @sel_false_cond(
-; CHECK-NEXT:    [[SUB1:%.*]] = sub nuw i16 [[X:%.*]], 10
+; CHECK-NEXT:    [[SUB:%.*]] = call i16 @llvm.usub.sat.i16(i16 [[X:%.*]], i16 10)
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X]], 10
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 42, i16 [[SUB1]]
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 42, i16 [[SUB]]
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %sub = call i16 @llvm.usub.sat.i16(i16 %x, i16 10)
@@ -152,13 +152,13 @@ define i16 @sel_false_cond_insufficient(i16 %x) {
 define i16 @phi_true_cond(i16 %x) {
 ; CHECK-LABEL: @phi_true_cond(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[SUB1:%.*]] = sub nuw i16 [[X:%.*]], 10
+; CHECK-NEXT:    [[SUB:%.*]] = call i16 @llvm.usub.sat.i16(i16 [[X:%.*]], i16 10)
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp uge i16 [[X]], 10
 ; CHECK-NEXT:    br i1 [[CMP]], label [[JOIN:%.*]], label [[SPLIT:%.*]]
 ; CHECK:       split:
 ; CHECK-NEXT:    br label [[JOIN]]
 ; CHECK:       join:
-; CHECK-NEXT:    [[PHI:%.*]] = phi i16 [ [[SUB1]], [[ENTRY:%.*]] ], [ 42, [[SPLIT]] ]
+; CHECK-NEXT:    [[PHI:%.*]] = phi i16 [ [[SUB]], [[ENTRY:%.*]] ], [ 42, [[SPLIT]] ]
 ; CHECK-NEXT:    ret i16 [[PHI]]
 ;
 entry:
@@ -229,13 +229,13 @@ join:
 define i16 @phi_false_cond(i16 %x) {
 ; CHECK-LABEL: @phi_false_cond(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[SUB1:%.*]] = sub nuw i16 [[X:%.*]], 10
+; CHECK-NEXT:    [[SUB:%.*]] = call i16 @llvm.usub.sat.i16(i16 [[X:%.*]], i16 10)
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X]], 10
 ; CHECK-NEXT:    br i1 [[CMP]], label [[SPLIT:%.*]], label [[JOIN:%.*]]
 ; CHECK:       split:
 ; CHECK-NEXT:    br label [[JOIN]]
 ; CHECK:       join:
-; CHECK-NEXT:    [[PHI:%.*]] = phi i16 [ [[SUB1]], [[ENTRY:%.*]] ], [ 42, [[SPLIT]] ]
+; CHECK-NEXT:    [[PHI:%.*]] = phi i16 [ [[SUB]], [[ENTRY:%.*]] ], [ 42, [[SPLIT]] ]
 ; CHECK-NEXT:    ret i16 [[PHI]]
 ;
 entry:
@@ -308,10 +308,10 @@ define i16 @loop_cond() {
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    br label [[LOOP:%.*]]
 ; CHECK:       loop:
-; CHECK-NEXT:    [[IV:%.*]] = phi i16 [ 1000, [[ENTRY:%.*]] ], [ [[IV_NEXT1:%.*]], [[LOOP]] ]
+; CHECK-NEXT:    [[IV:%.*]] = phi i16 [ 1000, [[ENTRY:%.*]] ], [ [[IV_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[COUNT:%.*]] = phi i16 [ 0, [[ENTRY]] ], [ [[COUNT_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp eq i16 [[IV]], 0
-; CHECK-NEXT:    [[IV_NEXT1]] = sub nuw i16 [[IV]], 1
+; CHECK-NEXT:    [[IV_NEXT]] = call i16 @llvm.usub.sat.i16(i16 [[IV]], i16 1)
 ; CHECK-NEXT:    [[COUNT_NEXT]] = add i16 [[COUNT]], 1
 ; CHECK-NEXT:    br i1 [[CMP]], label [[EXIT:%.*]], label [[LOOP]]
 ; CHECK:       exit:
@@ -334,8 +334,9 @@ exit:
 
 define i16 @urem_elide(i16 %x) {
 ; CHECK-LABEL: @urem_elide(
-; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X:%.*]], 42
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[X]], i16 24
+; CHECK-NEXT:    [[UREM:%.*]] = urem i16 [[X:%.*]], 42
+; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X]], 42
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[UREM]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %urem = urem i16 %x, 42
@@ -346,10 +347,7 @@ define i16 @urem_elide(i16 %x) {
 
 define i16 @urem_expand(i16 %x) {
 ; CHECK-LABEL: @urem_expand(
-; CHECK-NEXT:    [[X_FROZEN:%.*]] = freeze i16 [[X:%.*]]
-; CHECK-NEXT:    [[UREM_UREM:%.*]] = sub nuw i16 [[X_FROZEN]], 42
-; CHECK-NEXT:    [[UREM_CMP:%.*]] = icmp ult i16 [[X_FROZEN]], 42
-; CHECK-NEXT:    [[UREM:%.*]] = select i1 [[UREM_CMP]], i16 [[X_FROZEN]], i16 [[UREM_UREM]]
+; CHECK-NEXT:    [[UREM:%.*]] = urem i16 [[X:%.*]], 42
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X]], 84
 ; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[UREM]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
@@ -362,11 +360,9 @@ define i16 @urem_expand(i16 %x) {
 
 define i16 @urem_narrow(i16 %x) {
 ; CHECK-LABEL: @urem_narrow(
-; CHECK-NEXT:    [[UREM_LHS_TRUNC:%.*]] = trunc i16 [[X:%.*]] to i8
-; CHECK-NEXT:    [[UREM1:%.*]] = urem i8 [[UREM_LHS_TRUNC]], 42
-; CHECK-NEXT:    [[UREM_ZEXT:%.*]] = zext i8 [[UREM1]] to i16
+; CHECK-NEXT:    [[UREM:%.*]] = urem i16 [[X:%.*]], 42
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X]], 85
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[UREM_ZEXT]], i16 24
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[UREM]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %urem = urem i16 %x, 42
@@ -390,10 +386,11 @@ define i16 @urem_insufficient(i16 %x) {
 
 define i16 @srem_elide(i16 %x) {
 ; CHECK-LABEL: @srem_elide(
-; CHECK-NEXT:    [[CMP1:%.*]] = icmp slt i16 [[X:%.*]], 42
+; CHECK-NEXT:    [[SREM:%.*]] = srem i16 [[X:%.*]], 42
+; CHECK-NEXT:    [[CMP1:%.*]] = icmp slt i16 [[X]], 42
 ; CHECK-NEXT:    [[CMP2:%.*]] = icmp sgt i16 [[X]], -42
 ; CHECK-NEXT:    [[AND:%.*]] = and i1 [[CMP1]], [[CMP2]]
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[AND]], i16 [[X]], i16 24
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[AND]], i16 [[SREM]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %srem = srem i16 %x, 42
@@ -406,13 +403,11 @@ define i16 @srem_elide(i16 %x) {
 
 define i16 @srem_narrow(i16 %x) {
 ; CHECK-LABEL: @srem_narrow(
-; CHECK-NEXT:    [[SREM_LHS_TRUNC:%.*]] = trunc i16 [[X:%.*]] to i8
-; CHECK-NEXT:    [[SREM1:%.*]] = srem i8 [[SREM_LHS_TRUNC]], 42
-; CHECK-NEXT:    [[SREM_SEXT:%.*]] = sext i8 [[SREM1]] to i16
+; CHECK-NEXT:    [[SREM:%.*]] = srem i16 [[X:%.*]], 42
 ; CHECK-NEXT:    [[CMP1:%.*]] = icmp slt i16 [[X]], 43
 ; CHECK-NEXT:    [[CMP2:%.*]] = icmp sgt i16 [[X]], -43
 ; CHECK-NEXT:    [[AND:%.*]] = and i1 [[CMP1]], [[CMP2]]
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[AND]], i16 [[SREM_SEXT]], i16 24
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[AND]], i16 [[SREM]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %srem = srem i16 %x, 42
@@ -425,11 +420,9 @@ define i16 @srem_narrow(i16 %x) {
 
 define i16 @srem_convert(i16 %x) {
 ; CHECK-LABEL: @srem_convert(
-; CHECK-NEXT:    [[X_NONNEG:%.*]] = sub i16 0, [[X:%.*]]
-; CHECK-NEXT:    [[SREM1:%.*]] = urem i16 [[X_NONNEG]], 42
-; CHECK-NEXT:    [[SREM1_NEG:%.*]] = sub i16 0, [[SREM1]]
+; CHECK-NEXT:    [[SREM:%.*]] = srem i16 [[X:%.*]], 42
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i16 [[X]], 0
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[SREM1_NEG]], i16 24
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[SREM]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %srem = srem i16 %x, 42
@@ -440,11 +433,9 @@ define i16 @srem_convert(i16 %x) {
 
 define i16 @sdiv_convert(i16 %x) {
 ; CHECK-LABEL: @sdiv_convert(
-; CHECK-NEXT:    [[X_NONNEG:%.*]] = sub i16 0, [[X:%.*]]
-; CHECK-NEXT:    [[SREM1:%.*]] = udiv i16 [[X_NONNEG]], 42
-; CHECK-NEXT:    [[SREM1_NEG:%.*]] = sub i16 0, [[SREM1]]
+; CHECK-NEXT:    [[SREM:%.*]] = sdiv i16 [[X:%.*]], 42
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i16 [[X]], 0
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[SREM1_NEG]], i16 24
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[SREM]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %srem = sdiv i16 %x, 42
@@ -507,7 +498,7 @@ define i16 @umin_elide(i16 %x) {
 
 define i16 @ashr_convert(i16 %x, i16 %y) {
 ; CHECK-LABEL: @ashr_convert(
-; CHECK-NEXT:    [[ASHR:%.*]] = lshr i16 [[X:%.*]], [[Y:%.*]]
+; CHECK-NEXT:    [[ASHR:%.*]] = ashr i16 [[X:%.*]], [[Y:%.*]]
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp sge i16 [[X]], 0
 ; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[ASHR]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
@@ -520,7 +511,7 @@ define i16 @ashr_convert(i16 %x, i16 %y) {
 
 define i32 @sext_convert(i16 %x) {
 ; CHECK-LABEL: @sext_convert(
-; CHECK-NEXT:    [[EXT:%.*]] = zext i16 [[X:%.*]] to i32
+; CHECK-NEXT:    [[EXT:%.*]] = sext i16 [[X:%.*]] to i32
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp sge i16 [[X]], 0
 ; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i32 [[EXT]], i32 24
 ; CHECK-NEXT:    ret i32 [[SEL]]
@@ -546,8 +537,9 @@ define i16 @infer_flags(i16 %x) {
 
 define i16 @and_elide(i16 %x) {
 ; CHECK-LABEL: @and_elide(
-; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X:%.*]], 8
-; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[X]], i16 24
+; CHECK-NEXT:    [[AND:%.*]] = and i16 [[X:%.*]], 7
+; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i16 [[X]], 8
+; CHECK-NEXT:    [[SEL:%.*]] = select i1 [[CMP]], i16 [[AND]], i16 24
 ; CHECK-NEXT:    ret i16 [[SEL]]
 ;
   %and = and i16 %x, 7