[llvm-branch-commits] [llvm] release/18.x: [ValueTracking] Treat phi as underlying obj when not decomposing further (#84339) (PR #84950)
via llvm-branch-commits
llvm-branch-commits at lists.llvm.org
Tue Mar 12 10:04:18 PDT 2024
https://github.com/llvmbot created https://github.com/llvm/llvm-project/pull/84950
Backport 4cfd4a7896b5fd50274ec8573c259d7ad41741de b274b23665dec30f3ae4fb83ccca8b77e6d3ada3
Requested by: @nikic
>From 78be28fbce89bd342f5f7b0d5402f2d9c0213fe0 Mon Sep 17 00:00:00 2001
From: Florian Hahn <flo at fhahn.com>
Date: Thu, 7 Mar 2024 13:53:02 +0000
Subject: [PATCH 1/2] [LAA] Add test case for #82665.
Test case for https://github.com/llvm/llvm-project/issues/82665.
(cherry picked from commit 4cfd4a7896b5fd50274ec8573c259d7ad41741de)
---
.../underlying-object-loop-varying-phi.ll | 175 ++++++++++++++++++
1 file changed, 175 insertions(+)
create mode 100644 llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll
diff --git a/llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll b/llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll
new file mode 100644
index 00000000000000..1a5a6ac08d4045
--- /dev/null
+++ b/llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll
@@ -0,0 +1,175 @@
+; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py UTC_ARGS: --version 4
+; RUN: opt -passes='print<access-info>' -disable-output %s 2>&1 | FileCheck %s
+
+target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128"
+
+; Test case for https://github.com/llvm/llvm-project/issues/82665.
+define void @indirect_ptr_recurrences_read_write(ptr %A, ptr %B) {
+; CHECK-LABEL: 'indirect_ptr_recurrences_read_write'
+; CHECK-NEXT: loop:
+; CHECK-NEXT: Memory dependences are safe
+; CHECK-NEXT: Dependences:
+; CHECK-NEXT: Run-time memory checks:
+; CHECK-NEXT: Grouped accesses:
+; CHECK-EMPTY:
+; CHECK-NEXT: Non vectorizable stores to invariant address were not found in loop.
+; CHECK-NEXT: SCEV assumptions:
+; CHECK-EMPTY:
+; CHECK-NEXT: Expressions re-written:
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i64 [ 1, %entry ], [ %iv.next, %loop ]
+ %ptr.recur = phi ptr [ %A, %entry ], [ %ptr.next, %loop ]
+ %gep.B = getelementptr inbounds ptr, ptr %B, i64 %iv
+ %ptr.next = load ptr, ptr %gep.B, align 8, !tbaa !6
+ %l = load i32, ptr %ptr.recur, align 4, !tbaa !10
+ %xor = xor i32 %l, 1
+ store i32 %xor, ptr %ptr.recur, align 4, !tbaa !10
+ %iv.next = add nuw nsw i64 %iv, 1
+ %ec = icmp eq i64 %iv.next, 5
+ br i1 %ec, label %exit, label %loop
+
+exit:
+ ret void
+}
+
+define i32 @indirect_ptr_recurrences_read_only_loop(ptr %A, ptr %B) {
+; CHECK-LABEL: 'indirect_ptr_recurrences_read_only_loop'
+; CHECK-NEXT: loop:
+; CHECK-NEXT: Memory dependences are safe
+; CHECK-NEXT: Dependences:
+; CHECK-NEXT: Run-time memory checks:
+; CHECK-NEXT: Grouped accesses:
+; CHECK-EMPTY:
+; CHECK-NEXT: Non vectorizable stores to invariant address were not found in loop.
+; CHECK-NEXT: SCEV assumptions:
+; CHECK-EMPTY:
+; CHECK-NEXT: Expressions re-written:
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i64 [ 1, %entry ], [ %iv.next, %loop ]
+ %ptr.recur = phi ptr [ %A, %entry ], [ %ptr.next, %loop ]
+ %red = phi i32 [ 0, %entry ], [ %xor, %loop ]
+ %gep.B = getelementptr inbounds ptr, ptr %B, i64 %iv
+ %ptr.next = load ptr, ptr %gep.B, align 8, !tbaa !6
+ %l = load i32, ptr %ptr.recur, align 4, !tbaa !10
+ %xor = xor i32 %l, 1
+ %iv.next = add nuw nsw i64 %iv, 1
+ %ec = icmp eq i64 %iv.next, 5
+ br i1 %ec, label %exit, label %loop
+
+exit:
+ ret i32 %xor
+}
+
+define void @indirect_ptr_recurrences_read_write_may_alias_no_tbaa(ptr %A, ptr %B) {
+; CHECK-LABEL: 'indirect_ptr_recurrences_read_write_may_alias_no_tbaa'
+; CHECK-NEXT: loop:
+; CHECK-NEXT: Report: cannot identify array bounds
+; CHECK-NEXT: Dependences:
+; CHECK-NEXT: Run-time memory checks:
+; CHECK-NEXT: Grouped accesses:
+; CHECK-EMPTY:
+; CHECK-NEXT: Non vectorizable stores to invariant address were not found in loop.
+; CHECK-NEXT: SCEV assumptions:
+; CHECK-EMPTY:
+; CHECK-NEXT: Expressions re-written:
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i64 [ 1, %entry ], [ %iv.next, %loop ]
+ %ptr.recur = phi ptr [ %A, %entry ], [ %ptr.next, %loop ]
+ %gep.B = getelementptr inbounds ptr, ptr %B, i64 %iv
+ %ptr.next = load ptr, ptr %gep.B, align 8, !tbaa !6
+ %l = load i32, ptr %ptr.recur, align 4
+ %xor = xor i32 %l, 1
+ store i32 %xor, ptr %ptr.recur, align 4
+ %iv.next = add nuw nsw i64 %iv, 1
+ %ec = icmp eq i64 %iv.next, 5
+ br i1 %ec, label %exit, label %loop
+
+exit:
+ ret void
+}
+
+define void @indirect_ptr_recurrences_read_write_may_alias_different_obj(ptr %A, ptr %B, ptr %C) {
+; CHECK-LABEL: 'indirect_ptr_recurrences_read_write_may_alias_different_obj'
+; CHECK-NEXT: loop:
+; CHECK-NEXT: Report: cannot identify array bounds
+; CHECK-NEXT: Dependences:
+; CHECK-NEXT: Run-time memory checks:
+; CHECK-NEXT: Grouped accesses:
+; CHECK-EMPTY:
+; CHECK-NEXT: Non vectorizable stores to invariant address were not found in loop.
+; CHECK-NEXT: SCEV assumptions:
+; CHECK-EMPTY:
+; CHECK-NEXT: Expressions re-written:
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i64 [ 1, %entry ], [ %iv.next, %loop ]
+ %ptr.recur = phi ptr [ %A, %entry ], [ %ptr.next, %loop ]
+ %gep.B = getelementptr inbounds ptr, ptr %B, i64 %iv
+ %ptr.next = load ptr, ptr %gep.B, align 8, !tbaa !6
+ %l = load i32, ptr %ptr.recur, align 4
+ %xor = xor i32 %l, 1
+ %gep.C = getelementptr inbounds ptr, ptr %C, i64 %iv
+ store i32 %xor, ptr %gep.C, align 4
+ %iv.next = add nuw nsw i64 %iv, 1
+ %ec = icmp eq i64 %iv.next, 5
+ br i1 %ec, label %exit, label %loop
+
+exit:
+ ret void
+}
+
+define void @indirect_ptr_recurrences_read_write_may_noalias_different_obj(ptr %A, ptr %B, ptr noalias %C) {
+; CHECK-LABEL: 'indirect_ptr_recurrences_read_write_may_noalias_different_obj'
+; CHECK-NEXT: loop:
+; CHECK-NEXT: Memory dependences are safe
+; CHECK-NEXT: Dependences:
+; CHECK-NEXT: Run-time memory checks:
+; CHECK-NEXT: Grouped accesses:
+; CHECK-EMPTY:
+; CHECK-NEXT: Non vectorizable stores to invariant address were not found in loop.
+; CHECK-NEXT: SCEV assumptions:
+; CHECK-EMPTY:
+; CHECK-NEXT: Expressions re-written:
+;
+entry:
+ br label %loop
+
+loop:
+ %iv = phi i64 [ 1, %entry ], [ %iv.next, %loop ]
+ %ptr.recur = phi ptr [ %A, %entry ], [ %ptr.next, %loop ]
+ %gep.B = getelementptr inbounds ptr, ptr %B, i64 %iv
+ %ptr.next = load ptr, ptr %gep.B, align 8, !tbaa !6
+ %l = load i32, ptr %ptr.recur, align 4
+ %xor = xor i32 %l, 1
+ %gep.C = getelementptr inbounds ptr, ptr %C, i64 %iv
+ store i32 %xor, ptr %gep.C, align 4
+ %iv.next = add nuw nsw i64 %iv, 1
+ %ec = icmp eq i64 %iv.next, 5
+ br i1 %ec, label %exit, label %loop
+
+exit:
+ ret void
+}
+
+
+!6 = !{!7, !7, i64 0}
+!7 = !{!"any pointer", !8, i64 0}
+!8 = !{!"omnipotent char", !9, i64 0}
+!9 = !{!"Simple C/C++ TBAA"}
+!10 = !{!11, !11, i64 0}
+!11 = !{!"int", !8, i64 0}
>From eca980322cb9d415dd2e9700b031e63c9b136a39 Mon Sep 17 00:00:00 2001
From: Florian Hahn <flo at fhahn.com>
Date: Tue, 12 Mar 2024 08:55:03 +0000
Subject: [PATCH 2/2] [ValueTracking] Treat phi as underlying obj when not
decomposing further (#84339)
At the moment, getUnderlyingObjects simply continues for phis that do
not refer to the same underlying object in loops, without adding them to
the list of underlying objects, effectively ignoring those phis.
Instead of ignoring those phis, add them to the list of underlying
objects. This fixes a miscompile where LoopAccessAnalysis fails to
identify a memory dependence, because no underlying objects can be found
for a set of memory accesses.
Fixes https://github.com/llvm/llvm-project/issues/82665.
PR: https://github.com/llvm/llvm-project/pull/84339
(cherry picked from commit b274b23665dec30f3ae4fb83ccca8b77e6d3ada3)
---
llvm/lib/Analysis/ValueTracking.cpp | 2 ++
.../underlying-object-loop-varying-phi.ll | 7 ++++++-
2 files changed, 8 insertions(+), 1 deletion(-)
diff --git a/llvm/lib/Analysis/ValueTracking.cpp b/llvm/lib/Analysis/ValueTracking.cpp
index 412115eb649c2f..9f9451e4e814ac 100644
--- a/llvm/lib/Analysis/ValueTracking.cpp
+++ b/llvm/lib/Analysis/ValueTracking.cpp
@@ -5986,6 +5986,8 @@ void llvm::getUnderlyingObjects(const Value *V,
if (!LI || !LI->isLoopHeader(PN->getParent()) ||
isSameUnderlyingObjectInLoop(PN, LI))
append_range(Worklist, PN->incoming_values());
+ else
+ Objects.push_back(P);
continue;
}
diff --git a/llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll b/llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll
index 1a5a6ac08d4045..106dc8c13a49fa 100644
--- a/llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll
+++ b/llvm/test/Analysis/LoopAccessAnalysis/underlying-object-loop-varying-phi.ll
@@ -7,8 +7,13 @@ target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128"
define void @indirect_ptr_recurrences_read_write(ptr %A, ptr %B) {
; CHECK-LABEL: 'indirect_ptr_recurrences_read_write'
; CHECK-NEXT: loop:
-; CHECK-NEXT: Memory dependences are safe
+; CHECK-NEXT: Report: unsafe dependent memory operations in loop. Use #pragma clang loop distribute(enable) to allow loop distribution to attempt to isolate the offending operations into a separate loop
+; CHECK-NEXT: Unsafe indirect dependence.
; CHECK-NEXT: Dependences:
+; CHECK-NEXT: IndidrectUnsafe:
+; CHECK-NEXT: %l = load i32, ptr %ptr.recur, align 4, !tbaa !4 ->
+; CHECK-NEXT: store i32 %xor, ptr %ptr.recur, align 4, !tbaa !4
+; CHECK-EMPTY:
; CHECK-NEXT: Run-time memory checks:
; CHECK-NEXT: Grouped accesses:
; CHECK-EMPTY:
More information about the llvm-branch-commits
mailing list