From 68623a0e9fbebfb2aa2edab710d021b26f3f1edf Mon Sep 17 00:00:00 2001
From: David Majnemer <david.majnemer@gmail.com>
Date: Mon, 25 Jul 2016 02:21:25 +0000
Subject: [PATCH] [GVNHoist] Merge metadata on hoisted instructions less
 conservatively

We can combine metadata from multiple instructions intelligently for
certain metadata nodes.

llvm-svn: 276602
---
 llvm/lib/Transforms/Scalar/GVNHoist.cpp | 24 +++++++++++++++++-------
 llvm/test/Transforms/GVN/hoist-md.ll    | 25 +++++++++++++++++++++++++
 2 files changed, 42 insertions(+), 7 deletions(-)

diff --git a/llvm/lib/Transforms/Scalar/GVNHoist.cpp b/llvm/lib/Transforms/Scalar/GVNHoist.cpp
index 0b5e94f..194405d 100644
--- a/llvm/lib/Transforms/Scalar/GVNHoist.cpp
+++ b/llvm/lib/Transforms/Scalar/GVNHoist.cpp
@@ -23,6 +23,7 @@
 #include "llvm/Analysis/ValueTracking.h"
 #include "llvm/Transforms/Scalar.h"
 #include "llvm/Transforms/Scalar/GVN.h"
+#include "llvm/Transforms/Utils/Local.h"
 #include "llvm/Transforms/Utils/MemorySSA.h"
 
 using namespace llvm;
@@ -172,6 +173,15 @@ typedef DenseMap<const BasicBlock *, bool> BBSideEffectsSet;
 typedef SmallVector<Instruction *, 4> SmallVecInsn;
 typedef SmallVectorImpl<Instruction *> SmallVecImplInsn;
 
+static void combineKnownMetadata(Instruction *ReplInst, Instruction *I) {
+  static const unsigned KnownIDs[] = {
+      LLVMContext::MD_tbaa,           LLVMContext::MD_alias_scope,
+      LLVMContext::MD_noalias,        LLVMContext::MD_range,
+      LLVMContext::MD_fpmath,         LLVMContext::MD_invariant_load,
+      LLVMContext::MD_invariant_group};
+  combineMetadata(ReplInst, I, KnownIDs);
+}
+
 // This pass hoists common computations across branches sharing common
 // dominator. The primary goal is to reduce the code size, and in some
 // cases reduce critical path (by exposing more ILP).
@@ -604,15 +614,15 @@ public:
     ClonedGep->insertBefore(HoistPt->getTerminator());
     // Conservatively discard any optimization hints, they may differ on the
     // other paths.
-    ClonedGep->dropUnknownNonDebugMetadata();
-    for (const Instruction *OtherInst : InstructionsToHoist) {
-      const GetElementPtrInst *OtherGep;
+    for (Instruction *OtherInst : InstructionsToHoist) {
+      GetElementPtrInst *OtherGep;
       if (auto *OtherLd = dyn_cast<LoadInst>(OtherInst))
         OtherGep = cast<GetElementPtrInst>(OtherLd->getPointerOperand());
       else
         OtherGep = cast<GetElementPtrInst>(
             cast<StoreInst>(OtherInst)->getPointerOperand());
       ClonedGep->intersectOptionalDataWith(OtherGep);
+      combineKnownMetadata(ClonedGep, OtherGep);
     }
     Repl->replaceUsesOfWith(Gep, ClonedGep);
 
@@ -622,11 +632,11 @@ public:
       ClonedVal->insertBefore(HoistPt->getTerminator());
       // Conservatively discard any optimization hints, they may differ on the
       // other paths.
-      ClonedVal->dropUnknownNonDebugMetadata();
-      for (const Instruction *OtherInst : InstructionsToHoist) {
-        const auto *OtherVal =
+      for (Instruction *OtherInst : InstructionsToHoist) {
+        auto *OtherVal =
             cast<Instruction>(cast<StoreInst>(OtherInst)->getValueOperand());
         ClonedVal->intersectOptionalDataWith(OtherVal);
+        combineKnownMetadata(ClonedVal, OtherVal);
       }
       Repl->replaceUsesOfWith(Val, ClonedVal);
     }
@@ -668,7 +678,6 @@ public:
         Repl->moveBefore(HoistPt->getTerminator());
         // TBAA may differ on one of the other paths, we need to get rid of
         // anything which might conflict.
-        Repl->dropUnknownNonDebugMetadata();
       }
 
       if (isa<LoadInst>(Repl))
@@ -702,6 +711,7 @@ public:
             ++NumCallsRemoved;
           }
           Repl->intersectOptionalDataWith(I);
+          combineKnownMetadata(Repl, I);
           I->replaceAllUsesWith(Repl);
           I->eraseFromParent();
         }
diff --git a/llvm/test/Transforms/GVN/hoist-md.ll b/llvm/test/Transforms/GVN/hoist-md.ll
index 8661542..902bbb7 100644
--- a/llvm/test/Transforms/GVN/hoist-md.ll
+++ b/llvm/test/Transforms/GVN/hoist-md.ll
@@ -71,3 +71,28 @@ if.end:                                           ; preds = %if.else, %if.then
 !4 = !{!"Simple C++ TBAA"}
 !5 = !{!6, !6, i64 0}
 !6 = !{!"_ZTS1e", !3, i64 0}
+
+define i32 @test4(i1 %b, i32* %y) {
+entry:
+  br i1 %b, label %if.then, label %if.end
+
+if.then:                                          ; preds = %entry
+  %0 = load i32, i32* %y, align 4, !range !7
+  br label %return
+
+if.end:                                           ; preds = %entry
+  %1 = load i32, i32* %y, align 4, !range !8
+  br label %return
+
+return:                                           ; preds = %if.end, %if.then
+  %retval.0 = phi i32 [ %0, %if.then ], [ %1, %if.end ]
+  ret i32 %retval.0
+}
+; CHECK-LABEL: define i32 @test4(
+; CHECK: %[[load:.*]] = load i32, i32* %y, align 4, !range ![[range_md:.*]]
+; CHECK: %[[phi:.*]] = phi i32 [ %[[load]], %{{.*}} ], [ %[[load]], %{{.*}} ]
+; CHECK: ret i32 %[[phi]]
+
+!7 = !{i32 0, i32 2}
+!8 = !{i32 3, i32 4}
+; CHECK: ![[range_md]] = !{i32 0, i32 2, i32 3, i32 4}
-- 
2.7.4