[llvm] [LAA] Pre-commit tests exercising different types (PR #151091)

Ramkumar Ramachandra via llvm-commits llvm-commits at lists.llvm.org
Wed Aug 6 05:39:26 PDT 2025


https://github.com/artagnon updated https://github.com/llvm/llvm-project/pull/151091

>From be3db448f8f0acdd9f75f87fecf0059abb7a9153 Mon Sep 17 00:00:00 2001
From: Ramkumar Ramachandra <ramkumar.ramachandra at codasip.com>
Date: Tue, 29 Jul 2025 07:54:08 +0100
Subject: [PATCH 1/3] [LAA] Pre-commit tests exercising different types

Pre-commit tests exercising different types of source/sink in
depend_diff_types.ll, in preparation to weaken the HasSameSize check in
LoopAccessAnalysis.
---
 .../LoopAccessAnalysis/depend_diff_types.ll   | 240 ++++++++++++++++++
 1 file changed, 240 insertions(+)

diff --git a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
index 1bf80489b1ea0..548a7c0fd17ed 100644
--- a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
+++ b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
@@ -262,3 +262,243 @@ loop:
 exit:
   ret void
 }
+
+; The type-size of the source is different from that of the sink,
+; and the dependence distance is 10.
+; TODO: Relax the HasSameSize check; the strided accesses are
+; independent, as determined by both the source size and the sink size.
+; This test should report no dependencies.
+define void @different_type_sizes_strided_accesses_independent(ptr %dst) {
+; CHECK-LABEL: 'different_type_sizes_strided_accesses_independent'
+; CHECK-NEXT:    loop:
+; 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:  Unknown data dependence.
+; CHECK-NEXT:      Dependences:
+; CHECK-NEXT:        Unknown:
+; CHECK-NEXT:            store i16 0, ptr %gep.iv, align 2 ->
+; CHECK-NEXT:            store i8 1, ptr %gep.10.iv, align 1
+; CHECK-EMPTY:
+; 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:
+  %gep.10 = getelementptr nuw i8, ptr %dst, i64 10
+  br label %loop
+
+loop:
+  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
+  %gep.iv = getelementptr i8, ptr %dst, i64 %iv
+  store i16 0, ptr %gep.iv
+  %gep.10.iv = getelementptr i8, ptr %gep.10, i64 %iv
+  store i8 1, ptr %gep.10.iv
+  %iv.next = add i64 %iv, 8
+  %ec = icmp eq i64 %iv.next, 64
+  br i1 %ec, label %exit, label %loop
+
+exit:
+  ret void
+}
+
+; Variant of the above, where the source size forbids strided access
+; independence.
+; TODO: Relax the HasSameSize check; this test should report a backward
+; loop-carried dependence.
+define void @different_type_sizes_strided_accesses_dependent(ptr %dst) {
+; CHECK-LABEL: 'different_type_sizes_strided_accesses_dependent'
+; CHECK-NEXT:    loop:
+; 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:  Unknown data dependence.
+; CHECK-NEXT:      Dependences:
+; CHECK-NEXT:        Unknown:
+; CHECK-NEXT:            store i64 0, ptr %gep.iv, align 4 ->
+; CHECK-NEXT:            store i8 1, ptr %gep.10.iv, align 1
+; CHECK-EMPTY:
+; 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:
+  %gep.10 = getelementptr nuw i8, ptr %dst, i64 10
+  br label %loop
+
+loop:
+  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
+  %gep.iv = getelementptr i8, ptr %dst, i64 %iv
+  store i64 0, ptr %gep.iv
+  %gep.10.iv = getelementptr i8, ptr %gep.10, i64 %iv
+  store i8 1, ptr %gep.10.iv
+  %iv.next = add i64 %iv, 8
+  %ec = icmp eq i64 %iv.next, 64
+  br i1 %ec, label %exit, label %loop
+
+exit:
+  ret void
+}
+
+; Source type-size differs from that of the sink, but when
+; determining backward dependence, only the source size
+; is relevant.
+; TODO: Relax the HasSameSize check; this test should report
+; BackwardVectorizable.
+define void @different_type_sizes_source_size_backwardvectorizible(ptr %dst) {
+; CHECK-LABEL: 'different_type_sizes_source_size_backwardvectorizible'
+; CHECK-NEXT:    loop:
+; 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:  Unknown data dependence.
+; CHECK-NEXT:      Dependences:
+; CHECK-NEXT:        Unknown:
+; CHECK-NEXT:            store i16 0, ptr %gep.iv, align 2 ->
+; CHECK-NEXT:            store i32 1, ptr %gep.10.iv, align 4
+; CHECK-EMPTY:
+; 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:
+  %gep.10 = getelementptr nuw i8, ptr %dst, i64 10
+  br label %loop
+
+loop:
+  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
+  %gep.iv = getelementptr i8, ptr %dst, i64 %iv
+  store i16 0, ptr %gep.iv
+  %gep.10.iv = getelementptr i8, ptr %gep.10, i64 %iv
+  store i32 1, ptr %gep.10.iv
+  %iv.next = add i64 %iv, 8
+  %ec = icmp eq i64 %iv.next, 64
+  br i1 %ec, label %exit, label %loop
+
+exit:
+  ret void
+}
+
+; Source type-size differs from that of the sink, and when
+; determining forward dependence, the source size can
+; prevent forwarding.
+define void @different_type_sizes_forward(ptr %dst) {
+; CHECK-LABEL: 'different_type_sizes_forward'
+; CHECK-NEXT:    loop:
+; CHECK-NEXT:      Memory dependences are safe
+; CHECK-NEXT:      Dependences:
+; CHECK-NEXT:        Forward:
+; CHECK-NEXT:            store i32 0, ptr %gep.10.iv, align 4 ->
+; CHECK-NEXT:            store i16 1, ptr %gep.iv, align 2
+; CHECK-EMPTY:
+; 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:
+  %gep.10 = getelementptr nuw i8, ptr %dst, i64 10
+  br label %loop
+
+loop:
+  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
+  %gep.10.iv = getelementptr i8, ptr %gep.10, i64 %iv
+  store i32 0, ptr %gep.10.iv
+  %gep.iv = getelementptr i8, ptr %dst, i64 %iv
+  store i16 1, ptr %gep.iv
+  %iv.next = add i64 %iv, 8
+  %ec = icmp eq i64 %iv.next, 64
+  br i1 %ec, label %exit, label %loop
+
+exit:
+  ret void
+}
+
+; Same as the above, but here, the store size should not prevent
+; ld->st forwarding.
+; TODO: Relax the HasSameSize check; this test should report a
+; forward dependence.
+define void @different_type_sizes_store_size_cannot_prevent_forwarding(ptr %A, ptr noalias %B) {
+; CHECK-LABEL: 'different_type_sizes_store_size_cannot_prevent_forwarding'
+; CHECK-NEXT:    loop:
+; 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:  Forward loop carried data dependence that prevents store-to-load forwarding.
+; CHECK-NEXT:      Dependences:
+; CHECK-NEXT:        ForwardButPreventsForwarding:
+; CHECK-NEXT:            store i32 0, ptr %gep.A, align 4 ->
+; CHECK-NEXT:            %l = load i16, ptr %gep.A.1, align 2
+; CHECK-EMPTY:
+; 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:
+  %A.1 = getelementptr i32, ptr %A, i64 1
+  br label %loop
+
+loop:
+  %iv = phi i64 [ 1022, %entry ], [ %iv.next, %loop ]
+  %gep.A = getelementptr inbounds i32, ptr %A, i64 %iv
+  store i32 0, ptr %gep.A
+  %gep.A.1 = getelementptr i32, ptr %A.1, i64 %iv
+  %l = load i16, ptr %gep.A.1
+  store i16 %l, ptr %B
+  %iv.next = add nsw i64 %iv, -1
+  %cmp = icmp eq i64 %iv, 0
+  br i1 %cmp, label %exit, label %loop
+
+exit:
+  ret void
+}
+
+; Same as the above, but here, the load size prevents
+; ld->st forwarding.
+define void @different_type_sizes_load_size_prevents_forwarding(ptr %A, ptr noalias %B) {
+; CHECK-LABEL: 'different_type_sizes_load_size_prevents_forwarding'
+; CHECK-NEXT:    loop:
+; 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:  Forward loop carried data dependence that prevents store-to-load forwarding.
+; CHECK-NEXT:      Dependences:
+; CHECK-NEXT:        ForwardButPreventsForwarding:
+; CHECK-NEXT:            store i16 0, ptr %gep.A, align 2 ->
+; CHECK-NEXT:            %l = load i32, ptr %gep.A.1, align 4
+; CHECK-EMPTY:
+; 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:
+  %A.1 = getelementptr i32, ptr %A, i64 1
+  br label %loop
+
+loop:
+  %iv = phi i64 [ 1022, %entry ], [ %iv.next, %loop ]
+  %gep.A = getelementptr inbounds i32, ptr %A, i64 %iv
+  store i16 0, ptr %gep.A
+  %gep.A.1 = getelementptr i32, ptr %A.1, i64 %iv
+  %l = load i32, ptr %gep.A.1
+  store i32 %l, ptr %B
+  %iv.next = add nsw i64 %iv, -1
+  %cmp = icmp eq i64 %iv, 0
+  br i1 %cmp, label %exit, label %loop
+
+exit:
+  ret void
+}

>From 5c253085f1b3d164e58ae14bb0dc7ebfb6da482c Mon Sep 17 00:00:00 2001
From: Ramkumar Ramachandra <ramkumar.ramachandra at codasip.com>
Date: Mon, 4 Aug 2025 11:57:29 +0100
Subject: [PATCH 2/3] [LAA] Add test for store size > dep dist

---
 .../LoopAccessAnalysis/depend_diff_types.ll   | 39 +++++++++++++++++++
 1 file changed, 39 insertions(+)

diff --git a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
index 548a7c0fd17ed..af8219dad72d0 100644
--- a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
+++ b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
@@ -344,6 +344,45 @@ exit:
   ret void
 }
 
+; Variant of the above, where the store size exceeds the dependence
+; distance.
+define void @different_type_sizes_strided_accesses_store_size_exceeds_depdist(ptr %dst) {
+; CHECK-LABEL: 'different_type_sizes_strided_accesses_store_size_exceeds_depdist'
+; CHECK-NEXT:    loop:
+; 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:        IndirectUnsafe:
+; CHECK-NEXT:            store i16 0, ptr %gep.iv, align 2 ->
+; CHECK-NEXT:            store i128 1, ptr %gep.10.iv, align 4
+; CHECK-EMPTY:
+; 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:
+  %gep.10 = getelementptr nuw i8, ptr %dst, i64 10
+  br label %loop
+
+loop:
+  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
+  %gep.iv = getelementptr i8, ptr %dst, i64 %iv
+  store i16 0, ptr %gep.iv
+  %gep.10.iv = getelementptr i8, ptr %gep.10, i64 %iv
+  store i128 1, ptr %gep.10.iv
+  %iv.next = add i64 %iv, 8
+  %ec = icmp eq i64 %iv.next, 64
+  br i1 %ec, label %exit, label %loop
+
+exit:
+  ret void
+}
+
+
 ; Source type-size differs from that of the sink, but when
 ; determining backward dependence, only the source size
 ; is relevant.

>From e7e09917b5a27df36600da2abefeabb0c4a8357f Mon Sep 17 00:00:00 2001
From: Ramkumar Ramachandra <ramkumar.ramachandra at codasip.com>
Date: Wed, 6 Aug 2025 13:19:40 +0100
Subject: [PATCH 3/3] [LAA] Improve strided-accesses tests

Co-authored-by: Igor Kirillov <igor.kirillov at arm.com>
---
 .../LoopAccessAnalysis/depend_diff_types.ll   | 55 +++++++++++++------
 1 file changed, 37 insertions(+), 18 deletions(-)

diff --git a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
index af8219dad72d0..023a8c056968f 100644
--- a/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
+++ b/llvm/test/Analysis/LoopAccessAnalysis/depend_diff_types.ll
@@ -263,8 +263,17 @@ exit:
   ret void
 }
 
-; The type-size of the source is different from that of the sink,
-; and the dependence distance is 10.
+;       i16       i32
+; [ . . 0 0 . . 1 1] [ 1 1 0 0 . . 1 1 ]
+;  ^~~^ gep i8 = 1
+;  ^ ~~ ^ iv.2 = iv + 2
+;       ^ ~~~~~ ^ dependence distance = 4
+;              ^ ~~~~~~~~~~~~~~~~~ ^ 8
+;       ^ ~~~~~~~~~~~~~~~~ ^ 8
+;   ^ ~~~~~~~~~~~~~~~~ ^ iv.next = iv + 8
+;
+; Measurements are in bytes.
+;
 ; TODO: Relax the HasSameSize check; the strided accesses are
 ; independent, as determined by both the source size and the sink size.
 ; This test should report no dependencies.
@@ -276,7 +285,7 @@ define void @different_type_sizes_strided_accesses_independent(ptr %dst) {
 ; CHECK-NEXT:      Dependences:
 ; CHECK-NEXT:        Unknown:
 ; CHECK-NEXT:            store i16 0, ptr %gep.iv, align 2 ->
-; CHECK-NEXT:            store i8 1, ptr %gep.10.iv, align 1
+; CHECK-NEXT:            store i32 1, ptr %gep.4.iv, align 4
 ; CHECK-EMPTY:
 ; CHECK-NEXT:      Run-time memory checks:
 ; CHECK-NEXT:      Grouped accesses:
@@ -287,16 +296,17 @@ define void @different_type_sizes_strided_accesses_independent(ptr %dst) {
 ; CHECK-NEXT:      Expressions re-written:
 ;
 entry:
-  %gep.10 = getelementptr nuw i8, ptr %dst, i64 10
+  %gep.4 = getelementptr nuw i8, ptr %dst, i64 4
   br label %loop
 
 loop:
   %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
-  %gep.iv = getelementptr i8, ptr %dst, i64 %iv
+  %iv.2 = add nuw nsw i64 %iv, 2
+  %gep.iv = getelementptr i8, ptr %dst, i64 %iv.2
   store i16 0, ptr %gep.iv
-  %gep.10.iv = getelementptr i8, ptr %gep.10, i64 %iv
-  store i8 1, ptr %gep.10.iv
-  %iv.next = add i64 %iv, 8
+  %gep.4.iv = getelementptr i8, ptr %gep.4, i64 %iv.2
+  store i32 1, ptr %gep.4.iv
+  %iv.next = add nuw nsw i64 %iv, 8
   %ec = icmp eq i64 %iv.next, 64
   br i1 %ec, label %exit, label %loop
 
@@ -304,8 +314,16 @@ exit:
   ret void
 }
 
-; Variant of the above, where the source size forbids strided access
-; independence.
+
+;     i16      i64
+; [ . 0 0 . 1 1 1 1] [ 1 x x 1 1 1 1 1 ]
+;  ^~~^ gep i8 = 1
+;  ^~~^ iv.1 = iv + 1
+;     ^ ~~ ^ dependence distance = 3
+;     ^ ~~~~~~~~~~~~~~~~ ^ 8
+;           ^ ~~~~~~~~~~~~~~~~ ^ 8
+;   ^ ~~~~~~~~~~~~~~~~ ^ iv.next = iv + 8
+;
 ; TODO: Relax the HasSameSize check; this test should report a backward
 ; loop-carried dependence.
 define void @different_type_sizes_strided_accesses_dependent(ptr %dst) {
@@ -315,8 +333,8 @@ define void @different_type_sizes_strided_accesses_dependent(ptr %dst) {
 ; CHECK-NEXT:  Unknown data dependence.
 ; CHECK-NEXT:      Dependences:
 ; CHECK-NEXT:        Unknown:
-; CHECK-NEXT:            store i64 0, ptr %gep.iv, align 4 ->
-; CHECK-NEXT:            store i8 1, ptr %gep.10.iv, align 1
+; CHECK-NEXT:            store i16 0, ptr %gep.iv, align 2 ->
+; CHECK-NEXT:            store i64 1, ptr %gep.3.iv, align 4
 ; CHECK-EMPTY:
 ; CHECK-NEXT:      Run-time memory checks:
 ; CHECK-NEXT:      Grouped accesses:
@@ -327,16 +345,17 @@ define void @different_type_sizes_strided_accesses_dependent(ptr %dst) {
 ; CHECK-NEXT:      Expressions re-written:
 ;
 entry:
-  %gep.10 = getelementptr nuw i8, ptr %dst, i64 10
+  %gep.3 = getelementptr nuw i8, ptr %dst, i64 3
   br label %loop
 
 loop:
   %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
-  %gep.iv = getelementptr i8, ptr %dst, i64 %iv
-  store i64 0, ptr %gep.iv
-  %gep.10.iv = getelementptr i8, ptr %gep.10, i64 %iv
-  store i8 1, ptr %gep.10.iv
-  %iv.next = add i64 %iv, 8
+  %iv.1 = add nuw nsw i64 %iv, 1
+  %gep.iv = getelementptr i8, ptr %dst, i64 %iv.1
+  store i16 0, ptr %gep.iv
+  %gep.3.iv = getelementptr i8, ptr %gep.3, i64 %iv.1
+  store i64 1, ptr %gep.3.iv
+  %iv.next = add nuw nsw i64 %iv, 8
   %ec = icmp eq i64 %iv.next, 64
   br i1 %ec, label %exit, label %loop
 



More information about the llvm-commits mailing list