fix PR8753, eliminating a case where we'd infinitely make a

substitution because it doesn't actually change the IR.  Patch by
Jakub Staszak!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121361 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Transforms/Scalar/MemCpyOptimizer.cpp b/lib/Transforms/Scalar/MemCpyOptimizer.cpp
index 318e212899fe5d61a7f6bba03fe68167168ce32e..6f93e326ba126df9ef960c6cee4502d57d322d50 100644 (file)
--- a/lib/Transforms/Scalar/MemCpyOptimizer.cpp
+++ b/lib/Transforms/Scalar/MemCpyOptimizer.cpp
@@ -675,6 +675,14 @@ bool MemCpyOpt::processMemCpyMemCpyDependence(MemCpyInst *M, MemCpyInst *MDep,
   if (M->getSource() != MDep->getDest() || MDep->isVolatile())
     return false;
   
+  // If dep instruction is reading from our current input, then it is a noop
+  // transfer and substituting the input won't change this instruction.  Just
+  // ignore the input and let someone else zap MDep.  This handles cases like:
+  //    memcpy(a <- a)
+  //    memcpy(b <- a)
+  if (M->getSource() == MDep->getSource())
+    return false;
+  
   // Second, the length of the memcpy's must be the same, or the preceeding one
   // must be larger than the following one.
   ConstantInt *C1 = dyn_cast<ConstantInt>(MDep->getLength());

diff --git a/test/Transforms/MemCpyOpt/crash.ll b/test/Transforms/MemCpyOpt/crash.ll
index bf5b2346757733b99eb56f2d2dab9cb3f4479d8a..cc3a6b05278a466b24669896cad57192fa24723e 100644 (file)
--- a/test/Transforms/MemCpyOpt/crash.ll
+++ b/test/Transforms/MemCpyOpt/crash.ll
@@ -1,5 +1,4 @@
-; RUN: opt < %s -memcpyopt -disable-output
-; PR4882
+; RUN: opt < %s -basicaa -memcpyopt -disable-output
 
 target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64"
 target triple = "armv7-eabi"
@@ -7,7 +6,8 @@ target triple = "armv7-eabi"
 %struct.qw = type { [4 x float] }
 %struct.bar = type { %struct.qw, %struct.qw, %struct.qw, %struct.qw, %struct.qw, float, float}
 
-define arm_aapcs_vfpcc void @test1(%struct.bar* %this) {
+; PR4882
+define void @test1(%struct.bar* %this) {
 entry:
   %0 = getelementptr inbounds %struct.bar* %this, i32 0, i32 0, i32 0, i32 0
   store float 0.000000e+00, float* %0, align 4
@@ -43,3 +43,16 @@ entry:
   store float 0.000000e+00, float* %15, align 4
   unreachable
 }
+
+; PR8753
+
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32,
+i1) nounwind
+
+define void @test2(i32 %cmd) nounwind {
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* undef, i8* undef, i64 20, i32 1, i1
+false) nounwind
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* null, i8* undef, i64 20, i32 1, i1
+false) nounwind
+  ret void
+}