From 9a06696a77c26b4792c8845d7c09cf2459717aaf Mon Sep 17 00:00:00 2001
From: Jakob Stoklund Olesen <stoklund@2pi.dk>
Date: Mon, 29 Oct 2012 17:51:52 +0000
Subject: [PATCH] Completely disallow partial copies in adjustCopiesBackFrom().

Partial copies can show up even when CoalescerPair.isPartial() returns
false. For example:

   %vreg24:dsub_0<def> = COPY %vreg31:dsub_0; QPR:%vreg24,%vreg31

Such a partial-partial copy is not good enough for the transformation
adjustCopiesBackFrom() needs to do.

llvm-svn: 166944
---
 llvm/lib/CodeGen/RegisterCoalescer.cpp    |  3 ++-
 llvm/test/CodeGen/ARM/coalesce-subregs.ll | 28 ++++++++++++++++++++++++++++
 2 files changed, 30 insertions(+), 1 deletion(-)

diff --git a/llvm/lib/CodeGen/RegisterCoalescer.cpp b/llvm/lib/CodeGen/RegisterCoalescer.cpp
index 7592074..ac0ed5f 100644
--- a/llvm/lib/CodeGen/RegisterCoalescer.cpp
+++ b/llvm/lib/CodeGen/RegisterCoalescer.cpp
@@ -430,7 +430,8 @@ bool RegisterCoalescer::adjustCopiesBackFrom(const CoalescerPair &CP,
   // If AValNo is defined as a copy from IntB, we can potentially process this.
   // Get the instruction that defines this value number.
   MachineInstr *ACopyMI = LIS->getInstructionFromIndex(AValNo->def);
-  if (!CP.isCoalescable(ACopyMI))
+  // Don't allow any partial copies, even if isCoalescable() allows them.
+  if (!CP.isCoalescable(ACopyMI) || !ACopyMI->isFullCopy())
     return false;
 
   // Get the LiveRange in IntB that this value number starts with.
diff --git a/llvm/test/CodeGen/ARM/coalesce-subregs.ll b/llvm/test/CodeGen/ARM/coalesce-subregs.ll
index 238ba24..14511ad 100644
--- a/llvm/test/CodeGen/ARM/coalesce-subregs.ll
+++ b/llvm/test/CodeGen/ARM/coalesce-subregs.ll
@@ -289,3 +289,31 @@ bb:
   %tmp18 = insertvalue %struct.wombat.5 %tmp17, <4 x float> undef, 3, 0
   ret %struct.wombat.5 %tmp18
 }
+
+; CHECK: adjustCopiesBackFrom
+; The shuffle in if.else3 must be preserved even though adjustCopiesBackFrom
+; is tempted to remove it.
+; CHECK: %if.else3
+; CHECK: vorr d
+define internal void @adjustCopiesBackFrom(<2 x i64>* noalias nocapture sret %agg.result, <2 x i64> %in) {
+entry:
+  %0 = extractelement <2 x i64> %in, i32 0
+  %cmp = icmp slt i64 %0, 1
+  %.in = select i1 %cmp, <2 x i64> <i64 0, i64 undef>, <2 x i64> %in
+  %1 = extractelement <2 x i64> %in, i32 1
+  %cmp1 = icmp slt i64 %1, 1
+  br i1 %cmp1, label %if.then2, label %if.else3
+
+if.then2:                                         ; preds = %entry
+  %2 = insertelement <2 x i64> %.in, i64 0, i32 1
+  br label %if.end4
+
+if.else3:                                         ; preds = %entry
+  %3 = shufflevector <2 x i64> %.in, <2 x i64> %in, <2 x i32> <i32 0, i32 3>
+  br label %if.end4
+
+if.end4:                                          ; preds = %if.else3, %if.then2
+  %result.2 = phi <2 x i64> [ %2, %if.then2 ], [ %3, %if.else3 ]
+  store <2 x i64> %result.2, <2 x i64>* %agg.result, align 128
+  ret void
+}
-- 
2.7.4