Reduce x86 sequence for GP pair to XMM

Added support for punpckldq which is useful for interleaving
32-bit values from two xmm registers.

This new instruction is now used for transfers from GP pairs
to XMM in order to reduce path length.

Change-Id: I70d9b69449dfcfb9a94a628deb74a7cffe96bac7
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
diff --git a/compiler/dex/quick/x86/x86_lir.h b/compiler/dex/quick/x86/x86_lir.h
index d7f61fc..e091a84 100644
--- a/compiler/dex/quick/x86/x86_lir.h
+++ b/compiler/dex/quick/x86/x86_lir.h
@@ -350,6 +350,7 @@
   Binary0fOpCode(kX86Subss),    // float subtract
   Binary0fOpCode(kX86Divsd),    // double divide
   Binary0fOpCode(kX86Divss),    // float divide
+  Binary0fOpCode(kX86Punpckldq),  // Interleave low-order double words
   kX86PsrlqRI,                  // right shift of floating point registers
   kX86PsllqRI,                  // left shift of floating point registers
   kX86SqrtsdRR,                 // sqrt of floating point register