Preserve load alignment and dereferenceable metadata during some transformations

Reviewed By: hfinkel

Differential Revision: http://reviews.llvm.org/D13953

llvm-svn: 251809

16 files changed, 327 insertions, 9 deletions

diff --git a/llvm/include/llvm/IR/Metadata.h b/llvm/include/llvm/IR/Metadata.h
index 26302109ff0..8805cec1471 100644
--- a/llvm/include/llvm/IR/Metadata.h
+++ b/llvm/include/llvm/IR/Metadata.h
@@ -964,6 +964,8 @@ public:
   static MDNode *getMostGenericFPMath(MDNode *A, MDNode *B);
   static MDNode *getMostGenericRange(MDNode *A, MDNode *B);
   static MDNode *getMostGenericAliasScope(MDNode *A, MDNode *B);
+  static MDNode *getMostGenericAlignmentOrDereferenceable(MDNode *A, MDNode *B);
+
 };
 
 /// \brief Tuple of metadata.
diff --git a/llvm/lib/IR/Metadata.cpp b/llvm/lib/IR/Metadata.cpp
index 7d8c3523743..80f18daa79b 100644
--- a/llvm/lib/IR/Metadata.cpp
+++ b/llvm/lib/IR/Metadata.cpp
@@ -953,6 +953,17 @@ MDNode *MDNode::getMostGenericRange(MDNode *A, MDNode *B) {
   return MDNode::get(A->getContext(), MDs);
 }
 
+MDNode *MDNode::getMostGenericAlignmentOrDereferenceable(MDNode *A, MDNode *B) {
+  if (!A || !B)
+    return nullptr;
+
+  ConstantInt *AVal = mdconst::extract<ConstantInt>(A->getOperand(0));
+  ConstantInt *BVal = mdconst::extract<ConstantInt>(B->getOperand(0));
+  if (AVal->getZExtValue() < BVal->getZExtValue())
+    return A;
+  return B;
+}
+
 //===----------------------------------------------------------------------===//
 // NamedMDNode implementation.
 //
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp b/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
index 4fba7f95797..933849f460b 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
@@ -367,7 +367,13 @@ static LoadInst *combineLoadToNewType(InstCombiner &IC, LoadInst &LI, Type *NewT
                              MDB.createRange(NonNullInt, NullInt));
       }
       break;
-
+    case LLVMContext::MD_align:
+    case LLVMContext::MD_dereferenceable:
+    case LLVMContext::MD_dereferenceable_or_null:
+      // These only directly apply if the new type is also a pointer.
+      if (NewTy->isPointerTy())
+        NewLoad->setMetadata(ID, N);
+      break;
     case LLVMContext::MD_range:
       // FIXME: It would be nice to propagate this in some way, but the type
       // conversions make it hard. If the new type is a pointer, we could
@@ -418,6 +424,9 @@ static StoreInst *combineStoreToNewValue(InstCombiner &IC, StoreInst &SI, Value
     case LLVMContext::MD_invariant_load:
     case LLVMContext::MD_nonnull:
     case LLVMContext::MD_range:
+    case LLVMContext::MD_align:
+    case LLVMContext::MD_dereferenceable:
+    case LLVMContext::MD_dereferenceable_or_null:
       // These don't apply for stores.
       break;
     }
@@ -755,10 +764,12 @@ Instruction *InstCombiner::visitLoadInst(LoadInst &LI) {
                                DefMaxInstsToScan, AA, &AATags)) {
     if (LoadInst *NLI = dyn_cast<LoadInst>(AvailableVal)) {
       unsigned KnownIDs[] = {
-          LLVMContext::MD_tbaa,           LLVMContext::MD_alias_scope,
-          LLVMContext::MD_noalias,        LLVMContext::MD_range,
-          LLVMContext::MD_invariant_load, LLVMContext::MD_nonnull,
-          LLVMContext::MD_invariant_group};
+          LLVMContext::MD_tbaa,            LLVMContext::MD_alias_scope,
+          LLVMContext::MD_noalias,         LLVMContext::MD_range,
+          LLVMContext::MD_invariant_load,  LLVMContext::MD_nonnull,
+          LLVMContext::MD_invariant_group, LLVMContext::MD_align,
+          LLVMContext::MD_dereferenceable,
+          LLVMContext::MD_dereferenceable_or_null};
       combineMetadata(NLI, &LI, KnownIDs);
     };
 
diff --git a/llvm/lib/Transforms/InstCombine/InstCombinePHI.cpp b/llvm/lib/Transforms/InstCombine/InstCombinePHI.cpp
index 6c67fd38882..1d3b837f124 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombinePHI.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombinePHI.cpp
@@ -358,7 +358,10 @@ Instruction *InstCombiner::FoldPHIArgLoadIntoPHI(PHINode &PN) {
     LLVMContext::MD_invariant_load,
     LLVMContext::MD_alias_scope,
     LLVMContext::MD_noalias,
-    LLVMContext::MD_nonnull
+    LLVMContext::MD_nonnull,
+    LLVMContext::MD_align,
+    LLVMContext::MD_dereferenceable,
+    LLVMContext::MD_dereferenceable_or_null,
   };
 
   for (unsigned ID : KnownIDs)
diff --git a/llvm/lib/Transforms/Utils/Local.cpp b/llvm/lib/Transforms/Utils/Local.cpp
index b6080f29faa..0bd5fa9f877 100644
--- a/llvm/lib/Transforms/Utils/Local.cpp
+++ b/llvm/lib/Transforms/Utils/Local.cpp
@@ -1430,6 +1430,15 @@ void llvm::combineMetadata(Instruction *K, const Instruction *J,
       case LLVMContext::MD_invariant_group:
         // Preserve !invariant.group in K.
         break;
+      case LLVMContext::MD_align:
+        K->setMetadata(Kind, 
+          MDNode::getMostGenericAlignmentOrDereferenceable(JMD, KMD));
+        break;
+      case LLVMContext::MD_dereferenceable:
+      case LLVMContext::MD_dereferenceable_or_null:
+        K->setMetadata(Kind, 
+          MDNode::getMostGenericAlignmentOrDereferenceable(JMD, KMD));
+        break;
     }
   }
   // Set !invariant.group from J if J has it. If both instructions have it
diff --git a/llvm/lib/Transforms/Utils/SimplifyCFG.cpp b/llvm/lib/Transforms/Utils/SimplifyCFG.cpp
index e44a04aaba6..f9d5d2d4d36 100644
--- a/llvm/lib/Transforms/Utils/SimplifyCFG.cpp
+++ b/llvm/lib/Transforms/Utils/SimplifyCFG.cpp
@@ -1097,7 +1097,9 @@ static bool HoistThenElseCodeToIf(BranchInst *BI,
     unsigned KnownIDs[] = {
         LLVMContext::MD_tbaa,    LLVMContext::MD_range,
         LLVMContext::MD_fpmath,  LLVMContext::MD_invariant_load,
-        LLVMContext::MD_nonnull, LLVMContext::MD_invariant_group};
+        LLVMContext::MD_nonnull, LLVMContext::MD_invariant_group,
+        LLVMContext::MD_align,   LLVMContext::MD_dereferenceable,
+        LLVMContext::MD_dereferenceable_or_null};
     combineMetadata(I1, I2, KnownIDs);
     I2->eraseFromParent();
     Changed = true;
diff --git a/llvm/test/Transforms/InstCombine/load-combine-metadata-2.ll b/llvm/test/Transforms/InstCombine/load-combine-metadata-2.ll
new file mode 100644
index 00000000000..bec0d7d2c36
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/load-combine-metadata-2.ll
@@ -0,0 +1,20 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+target datalayout = "e-m:e-p:64:64:64-i64:64-f80:128-n8:16:32:64-S128"
+
+; CHECK-LABEL: @test_load_load_combine_metadata(
+; Check that align metadata is combined
+; CHECK: load i32*, i32** %0
+; CHECK-SAME: !align ![[ALIGN:[0-9]+]]
+define void @test_load_load_combine_metadata(i32**, i32**, i32**) {
+  %a = load i32*, i32** %0, !align !0
+  %b = load i32*, i32** %0, !align !1
+  store i32 0, i32* %a
+  store i32 0, i32* %b
+  ret void
+}
+
+; CHECK: ![[ALIGN]] = !{i64 4}
+
+!0 = !{i64 4}
+!1 = !{i64 8}
\ No newline at end of file
diff --git a/llvm/test/Transforms/InstCombine/load-combine-metadata-3.ll b/llvm/test/Transforms/InstCombine/load-combine-metadata-3.ll
new file mode 100644
index 00000000000..bad4bb24059
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/load-combine-metadata-3.ll
@@ -0,0 +1,20 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+target datalayout = "e-m:e-p:64:64:64-i64:64-f80:128-n8:16:32:64-S128"
+
+; CHECK-LABEL: @test_load_load_combine_metadata(
+; Check that dereferenceable metadata is combined
+; CHECK: load i32*, i32** %0
+; CHECK-SAME: !dereferenceable ![[DEREF:[0-9]+]]
+define void @test_load_load_combine_metadata(i32**, i32**, i32**) {
+  %a = load i32*, i32** %0, !dereferenceable !0
+  %b = load i32*, i32** %0, !dereferenceable !1
+  store i32 0, i32* %a
+  store i32 0, i32* %b
+  ret void
+}
+
+; CHECK: ![[DEREF]] = !{i64 4}
+
+!0 = !{i64 4}
+!1 = !{i64 8}
\ No newline at end of file
diff --git a/llvm/test/Transforms/InstCombine/load-combine-metadata-4.ll b/llvm/test/Transforms/InstCombine/load-combine-metadata-4.ll
new file mode 100644
index 00000000000..2a1ffcd0605
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/load-combine-metadata-4.ll
@@ -0,0 +1,20 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+target datalayout = "e-m:e-p:64:64:64-i64:64-f80:128-n8:16:32:64-S128"
+
+; CHECK-LABEL: @test_load_load_combine_metadata(
+; Check that dereferenceable_or_null metadata is combined
+; CHECK: load i32*, i32** %0
+; CHECK-SAME: !dereferenceable_or_null ![[DEREF:[0-9]+]]
+define void @test_load_load_combine_metadata(i32**, i32**, i32**) {
+  %a = load i32*, i32** %0, !dereferenceable_or_null !0
+  %b = load i32*, i32** %0, !dereferenceable_or_null !1
+  store i32 0, i32* %a
+  store i32 0, i32* %b
+  ret void
+}
+
+; CHECK: ![[DEREF]] = !{i64 4}
+
+!0 = !{i64 4}
+!1 = !{i64 8}
diff --git a/llvm/test/Transforms/InstCombine/loadstore-metadata.ll b/llvm/test/Transforms/InstCombine/loadstore-metadata.ll
index a30c0bc852e..f72e36a7ea3 100644
--- a/llvm/test/Transforms/InstCombine/loadstore-metadata.ll
+++ b/llvm/test/Transforms/InstCombine/loadstore-metadata.ll
@@ -31,7 +31,7 @@ define float @test_load_cast_combine_range(i32* %ptr) {
 ; CHECK-NOT: !range
 ; CHECK: ret float
 entry:
-  %l = load i32, i32* %ptr, !range !5
+  %l = load i32, i32* %ptr, !range !6
   %c = bitcast i32 %l to float
   ret float %c
 }
@@ -57,6 +57,39 @@ entry:
   ret i32 %c
 }
 
+define i8* @test_load_cast_combine_align(i32** %ptr) {
+; Ensure (cast (load (...))) -> (load (cast (...))) preserves align
+; metadata.
+; CHECK-LABEL: @test_load_cast_combine_align(
+; CHECK: load i8*, i8** %{{.*}}, !align !5
+entry:
+  %l = load i32*, i32** %ptr, !align !5
+  %c = bitcast i32* %l to i8*
+  ret i8* %c
+}
+
+define i8* @test_load_cast_combine_deref(i32** %ptr) {
+; Ensure (cast (load (...))) -> (load (cast (...))) preserves dereferenceable
+; metadata.
+; CHECK-LABEL: @test_load_cast_combine_deref(
+; CHECK: load i8*, i8** %{{.*}}, !dereferenceable !5
+entry:
+  %l = load i32*, i32** %ptr, !dereferenceable !5
+  %c = bitcast i32* %l to i8*
+  ret i8* %c
+}
+
+define i8* @test_load_cast_combine_deref_or_null(i32** %ptr) {
+; Ensure (cast (load (...))) -> (load (cast (...))) preserves
+; dereferenceable_or_null metadata.
+; CHECK-LABEL: @test_load_cast_combine_deref_or_null(
+; CHECK: load i8*, i8** %{{.*}}, !dereferenceable_or_null !5
+entry:
+  %l = load i32*, i32** %ptr, !dereferenceable_or_null !5
+  %c = bitcast i32* %l to i8*
+  ret i8* %c
+}
+
 define void @test_load_cast_combine_loop(float* %src, i32* %dst, i32 %n) {
 ; Ensure (cast (load (...))) -> (load (cast (...))) preserves loop access
 ; metadata.
@@ -110,4 +143,5 @@ entry:
 !2 = !{ !2, !1 }
 !3 = !{ }
 !4 = !{ i32 1 }
-!5 = !{ i32 0, i32 42 }
+!5 = !{ i64 8 }
+!6 = !{ i32 0, i32 42 }
diff --git a/llvm/test/Transforms/InstCombine/phi-load-metadata-2.ll b/llvm/test/Transforms/InstCombine/phi-load-metadata-2.ll
new file mode 100644
index 00000000000..cfbf2dea8a7
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/phi-load-metadata-2.ll
@@ -0,0 +1,30 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+declare void @bar()
+declare void @baz()
+
+; Check that dereferenceable metadata is combined
+; CHECK-LABEL: cont:
+; CHECK: load i32*, i32**
+; CHECK-SAME: !dereferenceable ![[DEREF:[0-9]+]]
+define i32* @test_phi_combine_load_metadata(i1 %c, i32** dereferenceable(8) %p1, i32** dereferenceable(8) %p2) {
+  br i1 %c, label %t, label %f
+t:
+  call void @bar()
+  %v1 = load i32*, i32** %p1, align 8, !dereferenceable !0
+  br label %cont
+
+f:
+  call void @baz()
+  %v2 = load i32*, i32** %p2, align 8, !dereferenceable !1
+  br label %cont
+
+cont:
+  %res = phi i32* [ %v1, %t ], [ %v2, %f ]
+  ret i32* %res
+}
+
+; CHECK: ![[DEREF]] = !{i64 8}
+
+!0 = !{i64 8}
+!1 = !{i64 16}
diff --git a/llvm/test/Transforms/InstCombine/phi-load-metadata-3.ll b/llvm/test/Transforms/InstCombine/phi-load-metadata-3.ll
new file mode 100644
index 00000000000..39049c9c718
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/phi-load-metadata-3.ll
@@ -0,0 +1,30 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+declare void @bar()
+declare void @baz()
+
+; Check that dereferenceable_or_null metadata is combined
+; CHECK-LABEL: cont:
+; CHECK: load i32*, i32**
+; CHECK-SAME: !dereferenceable_or_null ![[DEREF:[0-9]+]]
+define i32* @test_phi_combine_load_metadata(i1 %c, i32** dereferenceable(8) %p1, i32** dereferenceable(8) %p2) {
+  br i1 %c, label %t, label %f
+t:
+  call void @bar()
+  %v1 = load i32*, i32** %p1, align 8, !dereferenceable_or_null !0
+  br label %cont
+
+f:
+  call void @baz()
+  %v2 = load i32*, i32** %p2, align 8, !dereferenceable_or_null !1
+  br label %cont
+
+cont:
+  %res = phi i32* [ %v1, %t ], [ %v2, %f ]
+  ret i32* %res
+}
+
+; CHECK: ![[DEREF]] = !{i64 8}
+
+!0 = !{i64 8}
+!1 = !{i64 16}
diff --git a/llvm/test/Transforms/InstCombine/phi-load-metadata.ll b/llvm/test/Transforms/InstCombine/phi-load-metadata.ll
new file mode 100644
index 00000000000..004a355ca44
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/phi-load-metadata.ll
@@ -0,0 +1,30 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+declare void @bar()
+declare void @baz()
+
+; Check that align metadata is combined
+; CHECK-LABEL: cont:
+; CHECK: load i32*, i32**
+; CHECK-SAME: !align ![[ALIGN:[0-9]+]]
+define i32* @test_phi_combine_load_metadata(i1 %c, i32** dereferenceable(8) %p1, i32** dereferenceable(8) %p2) {
+  br i1 %c, label %t, label %f
+t:
+  call void @bar()
+  %v1 = load i32*, i32** %p1, align 8, !align !0
+  br label %cont
+
+f:
+  call void @baz()
+  %v2 = load i32*, i32** %p2, align 8, !align !1
+  br label %cont
+
+cont:
+  %res = phi i32* [ %v1, %t ], [ %v2, %f ]
+  ret i32* %res
+}
+
+; CHECK: ![[ALIGN]] = !{i64 8}
+
+!0 = !{i64 8}
+!1 = !{i64 16}
diff --git a/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata-2.ll b/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata-2.ll
new file mode 100644
index 00000000000..94d3565ce98
--- /dev/null
+++ b/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata-2.ll
@@ -0,0 +1,32 @@
+; RUN: opt < %s -simplifycfg -S | FileCheck %s
+
+declare void @bar(i32*)
+declare void @baz(i32*)
+
+; CHECK-LABEL: @test_load_combine_metadata(
+; Check that dereferenceable metadata is combined
+; CHECK: load i32*, i32** %p
+; CHECK-SAME: !dereferenceable ![[DEREF:[0-9]+]]
+; CHECK: t:
+; CHECK: f:
+define void @test_load_combine_metadata(i1 %c, i32** %p) {
+  br i1 %c, label %t, label %f
+
+t:
+  %v1 = load i32*, i32** %p, !dereferenceable !0
+  call void @bar(i32* %v1)
+  br label %cont
+
+f:
+  %v2 = load i32*, i32** %p, !dereferenceable !1
+  call void @baz(i32* %v2)
+  br label %cont
+
+cont:
+  ret void
+}
+
+; CHECK: ![[DEREF]] = !{i64 8}
+
+!0 = !{i64 8}
+!1 = !{i64 16}
diff --git a/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata-3.ll b/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata-3.ll
new file mode 100644
index 00000000000..92bdf6ec5c1
--- /dev/null
+++ b/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata-3.ll
@@ -0,0 +1,32 @@
+; RUN: opt < %s -simplifycfg -S | FileCheck %s
+
+declare void @bar(i32*)
+declare void @baz(i32*)
+
+; CHECK-LABEL: @test_load_combine_metadata(
+; Check that dereferenceable_or_null metadata is combined
+; CHECK: load i32*, i32** %p
+; CHECK-SAME: !dereferenceable_or_null ![[DEREF:[0-9]+]]
+; CHECK: t:
+; CHECK: f:
+define void @test_load_combine_metadata(i1 %c, i32** %p) {
+  br i1 %c, label %t, label %f
+
+t:
+  %v1 = load i32*, i32** %p, !dereferenceable_or_null !0
+  call void @bar(i32* %v1)
+  br label %cont
+
+f:
+  %v2 = load i32*, i32** %p, !dereferenceable_or_null !1
+  call void @baz(i32* %v2)
+  br label %cont
+
+cont:
+  ret void
+}
+
+; CHECK: ![[DEREF]] = !{i64 8}
+
+!0 = !{i64 8}
+!1 = !{i64 16}
diff --git a/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata.ll b/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata.ll
new file mode 100644
index 00000000000..89815c84315
--- /dev/null
+++ b/llvm/test/Transforms/SimplifyCFG/preserve-load-metadata.ll
@@ -0,0 +1,32 @@
+; RUN: opt < %s -simplifycfg -S | FileCheck %s
+
+declare void @bar(i32*)
+declare void @baz(i32*)
+
+; CHECK-LABEL: @test_load_combine_metadata(
+; Check that align metadata is combined
+; CHECK: load i32*, i32** %p
+; CHECK-SAME: !align ![[ALIGN:[0-9]+]]
+; CHECK: t:
+; CHECK: f:
+define void @test_load_combine_metadata(i1 %c, i32** %p) {
+  br i1 %c, label %t, label %f
+
+t:
+  %v1 = load i32*, i32** %p, !align !0
+  call void @bar(i32* %v1)
+  br label %cont
+
+f:
+  %v2 = load i32*, i32** %p, !align !1
+  call void @baz(i32* %v2)
+  br label %cont
+
+cont:
+  ret void
+}
+
+; CHECK: ![[ALIGN]] = !{i64 8}
+
+!0 = !{i64 8}
+!1 = !{i64 16}