7005241: C1: SEGV in java.util.concurrent.LinkedTransferQueue.xfer() with compressed oops

Summary: Implementation of the CAS primitive for x64 compressed oops was incorrect. It kills rscratch2 register (r11), which is allocatable in C1. Also, we don't need to restore cmpval as it's never used after that, so we need only one temporary register, which can be scratch1.
Reviewed-by: kvn, never

src/cpu/x86/vm/c1_LIRAssembler_x86.cpp

@@ -1993,15 +1993,14 @@ void LIR_Assembler::emit_compare_and_swap(LIR_OpCompareAndSwap* op) {
     if ( op->code() == lir_cas_obj) {
 #ifdef _LP64
       if (UseCompressedOops) {
-        __ mov(rscratch1, cmpval);
         __ encode_heap_oop(cmpval);
-        __ mov(rscratch2, newval);
-        __ encode_heap_oop(rscratch2);
+        __ mov(rscratch1, newval);
+        __ encode_heap_oop(rscratch1);
         if (os::is_MP()) {
           __ lock();
         }
-        __ cmpxchgl(rscratch2, Address(addr, 0));
-        __ mov(cmpval, rscratch1);
+        // cmpval (rax) is implicitly used by this instruction
+        __ cmpxchgl(rscratch1, Address(addr, 0));
       } else
 #endif
       {