If a conditional branch jumps to the same target, remove the condition

Given that large parts of inst combine is restricted to instructions which have one use, getting rid of a use on the condition can help the effectiveness of the optimizer. Also, it allows the condition to potentially be deleted by instcombine rather than waiting for another pass.

I noticed this completely by accident in another test case. It's not anything that actually came from a real workload.

p.s. We should probably do the same thing for switch instructions.

Differential Revision: http://reviews.llvm.org/D8220

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@231881 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Transforms/InstCombine/InstructionCombining.cpp b/lib/Transforms/InstCombine/InstructionCombining.cpp
index 1fe036b5fbb42dbd5e55e0ca2e6a0ea117cb4a24..434d129bf396b2c5b938b283999a7717fca1b878 100644 (file)
--- a/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -2002,6 +2002,15 @@ Instruction *InstCombiner::visitBranchInst(BranchInst &BI) {
     return &BI;
   }
 
+  // If the condition is irrelevant, remove the use so that other
+  // transforms on the condition become more effective.
+  if (BI.isConditional() &&
+      BI.getSuccessor(0) == BI.getSuccessor(1) &&
+      !isa<UndefValue>(BI.getCondition())) {
+    BI.setCondition(UndefValue::get(BI.getCondition()->getType()));
+    return &BI;
+  }
+
   // Canonicalize fcmp_one -> fcmp_oeq
   FCmpInst::Predicate FPred; Value *Y;
   if (match(&BI, m_Br(m_FCmp(FPred, m_Value(X), m_Value(Y)),

diff --git a/test/Transforms/InstCombine/branch.ll b/test/Transforms/InstCombine/branch.ll
new file mode 100644 (file)
index 0000000..5be93cc
--- /dev/null
+++ b/test/Transforms/InstCombine/branch.ll
@@ -0,0 +1,16 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+define i32 @test(i32 %x) {
+; CHECK-LABEL: @test
+entry:
+; CHECK-NOT: icmp
+; CHECK: br i1 undef, 
+  %cmp = icmp ult i32 %x, 7
+  br i1 %cmp, label %merge, label %merge
+merge:
+; CHECK-LABEL: merge:
+; CHECK: ret i32 %x
+  ret i32 %x
+}
+
+