A64: Fix Fmov with signalling NaN literals.

test/cctest/test-assembler-a64.cc

 // Copyright 2013 the V8 project authors. All rights reserved.
 // Redistribution and use in source and binary forms, with or without
 // modification, are permitted provided that the following conditions are
 // met:
 //
 //     * Redistributions of source code must retain the above copyright
 //       notice, this list of conditions and the following disclaimer.
 //     * Redistributions in binary form must reproduce the above
 //       copyright notice, this list of conditions and the following
 //       disclaimer in the documentation and/or other materials provided
@@ skipping 5455 matching lines @@
     }
   }
 }
 
 
 static void FminFmaxFloatHelper(float n, float m, float min, float max,
                                 float minnm, float maxnm) {
   SETUP();
 
   START();
-  // TODO(all): Signalling NaNs are sometimes converted by the C compiler to
-  // quiet NaNs on implicit casts from float to double. Here, we move the raw
-  // bits into a W register first, so we get the correct value. Fix Fmov so this
-  // additional step is no longer needed.
-  __ Mov(w0, float_to_rawbits(n));
-  __ Fmov(s0, w0);
-  __ Mov(w0, float_to_rawbits(m));
-  __ Fmov(s1, w0);
+  __ Fmov(s0, n);
+  __ Fmov(s1, m);
   __ Fmin(s28, s0, s1);
   __ Fmax(s29, s0, s1);
   __ Fminnm(s30, s0, s1);
   __ Fmaxnm(s31, s0, s1);
   END();
 
   RUN();
 
   ASSERT_EQUAL_FP32(min, s28);
   ASSERT_EQUAL_FP32(max, s29);
@@ skipping 4529 matching lines @@
   AbsHelperX(-42);
   AbsHelperX(kXMinInt);
   AbsHelperX(kXMaxInt);
 
   AbsHelperW(0);
   AbsHelperW(42);
   AbsHelperW(-42);
   AbsHelperW(kWMinInt);
   AbsHelperW(kWMaxInt);
 }