1 files changed, 48 insertions, 27 deletions

diff --git a/llvm/lib/CodeGen/RegisterCoalescer.cpp b/llvm/lib/CodeGen/RegisterCoalescer.cpp
index 20ad6445344d..8313927dd2aa 100644
--- a/llvm/lib/CodeGen/RegisterCoalescer.cpp
+++ b/llvm/lib/CodeGen/RegisterCoalescer.cpp
@@ -113,7 +113,7 @@ static cl::opt<unsigned> LargeIntervalSizeThreshold(
 
 static cl::opt<unsigned> LargeIntervalFreqThreshold(
     "large-interval-freq-threshold", cl::Hidden,
-    cl::desc("For a large interval, if it is coalesed with other live "
+    cl::desc("For a large interval, if it is coalesced with other live "
              "intervals many times more than the threshold, stop its "
              "coalescing to control the compile time. "),
     cl::init(256));
@@ -1325,11 +1325,6 @@ bool RegisterCoalescer::reMaterializeTrivialDef(const CoalescerPair &CP,
   const MCInstrDesc &MCID = DefMI->getDesc();
   if (MCID.getNumDefs() != 1)
     return false;
-  // Only support subregister destinations when the def is read-undef.
-  MachineOperand &DstOperand = CopyMI->getOperand(0);
-  Register CopyDstReg = DstOperand.getReg();
-  if (DstOperand.getSubReg() && !DstOperand.isUndef())
-    return false;
 
   // If both SrcIdx and DstIdx are set, correct rematerialization would widen
   // the register substantially (beyond both source and dest size). This is bad
@@ -1339,6 +1334,32 @@ bool RegisterCoalescer::reMaterializeTrivialDef(const CoalescerPair &CP,
   if (SrcIdx && DstIdx)
     return false;
 
+  // Only support subregister destinations when the def is read-undef.
+  MachineOperand &DstOperand = CopyMI->getOperand(0);
+  Register CopyDstReg = DstOperand.getReg();
+  if (DstOperand.getSubReg() && !DstOperand.isUndef())
+    return false;
+
+  // In the physical register case, checking that the def is read-undef is not
+  // enough. We're widening the def and need to avoid clobbering other live
+  // values in the unused register pieces.
+  //
+  // TODO: Targets may support rewriting the rematerialized instruction to only
+  // touch relevant lanes, in which case we don't need any liveness check.
+  if (CopyDstReg.isPhysical() && CP.isPartial()) {
+    for (MCRegUnit Unit : TRI->regunits(DstReg)) {
+      // Ignore the register units we are writing anyway.
+      if (is_contained(TRI->regunits(CopyDstReg), Unit))
+        continue;
+
+      // Check if the other lanes we are defining are live at the
+      // rematerialization point.
+      LiveRange &LR = LIS->getRegUnit(Unit);
+      if (LR.liveAt(CopyIdx))
+        return false;
+    }
+  }
+
   const unsigned DefSubIdx = DefMI->getOperand(0).getSubReg();
   const TargetRegisterClass *DefRC = TII->getRegClass(MCID, 0, TRI, *MF);
   if (!DefMI->isImplicitDef()) {
@@ -1375,27 +1396,6 @@ bool RegisterCoalescer::reMaterializeTrivialDef(const CoalescerPair &CP,
   NewMI.setDebugLoc(DL);
 
   // In a situation like the following:
-  //
-  //    undef %2.subreg:reg = INST %1:reg         ; DefMI (rematerializable),
-  //                                              ; DefSubIdx = subreg
-  //    %3:reg = COPY %2                          ; SrcIdx = DstIdx = 0
-  //    .... = SOMEINSTR %3:reg
-  //
-  // there are no subranges for %3 so after rematerialization we need
-  // to explicitly create them. Undefined subranges are removed later on.
-  if (DstReg.isVirtual() && DefSubIdx && !CP.getSrcIdx() && !CP.getDstIdx() &&
-      MRI->shouldTrackSubRegLiveness(DstReg)) {
-    LiveInterval &DstInt = LIS->getInterval(DstReg);
-    if (!DstInt.hasSubRanges()) {
-      LaneBitmask FullMask = MRI->getMaxLaneMaskForVReg(DstReg);
-      LaneBitmask UsedLanes = TRI->getSubRegIndexLaneMask(DefSubIdx);
-      LaneBitmask UnusedLanes = FullMask & ~UsedLanes;
-      DstInt.createSubRangeFrom(LIS->getVNInfoAllocator(), UsedLanes, DstInt);
-      DstInt.createSubRangeFrom(LIS->getVNInfoAllocator(), UnusedLanes, DstInt);
-    }
-  }
-
-  // In a situation like the following:
   //     %0:subreg = instr              ; DefMI, subreg = DstIdx
   //     %1        = copy %0:subreg ; CopyMI, SrcIdx = 0
   // instead of widening %1 to the register class of %0 simply do:
@@ -1523,6 +1523,27 @@ bool RegisterCoalescer::reMaterializeTrivialDef(const CoalescerPair &CP,
     // sure that "undef" is not set.
     if (NewIdx == 0)
       NewMI.getOperand(0).setIsUndef(false);
+
+    // In a situation like the following:
+    //
+    //    undef %2.subreg:reg = INST %1:reg    ; DefMI (rematerializable),
+    //                                         ; Defines only some of lanes,
+    //                                         ; so DefSubIdx = NewIdx = subreg
+    //    %3:reg = COPY %2                     ; Copy full reg
+    //    .... = SOMEINSTR %3:reg              ; Use full reg
+    //
+    // there are no subranges for %3 so after rematerialization we need
+    // to explicitly create them. Undefined subranges are removed later on.
+    if (NewIdx && !DstInt.hasSubRanges() &&
+        MRI->shouldTrackSubRegLiveness(DstReg)) {
+      LaneBitmask FullMask = MRI->getMaxLaneMaskForVReg(DstReg);
+      LaneBitmask UsedLanes = TRI->getSubRegIndexLaneMask(NewIdx);
+      LaneBitmask UnusedLanes = FullMask & ~UsedLanes;
+      VNInfo::Allocator &Alloc = LIS->getVNInfoAllocator();
+      DstInt.createSubRangeFrom(Alloc, UsedLanes, DstInt);
+      DstInt.createSubRangeFrom(Alloc, UnusedLanes, DstInt);
+    }
+
     // Add dead subregister definitions if we are defining the whole register
     // but only part of it is live.
     // This could happen if the rematerialization instruction is rematerializing