re PR java/22578 (should inline floatToIntBits et al)

gcc/java
	PR java/22578:
	* check-init.c (check_init): Handle VIEW_CONVERT_EXPR.
	* builtins.c (convert_real): New function.
	(java_builtins): Handle Float.intBitsToFloat,
	Float.floatToRawIntBits, Double.longBitsToDouble,
	Double.doubleToRawLongBits.
libjava
	PR java/22578:
	* gcj/javaprims.h: Updated.
	* sources.am, Makefile.in: Rebuilt.
	* java/lang/natDouble.cc (doubleToLongBits): Moved to VMDouble.
	(doubleToRawLongBits): Likewise.
	(longBitsToDouble): Likewise.
	(toString): Likewise.
	(parseDouble): Likewise.
	* java/lang/natFloat.cc (floatToIntBits): Moved to VMFloat.
	(floatToRawIntBits): Likewise.
	(intBitsToFloat): Likewise.
	* java/lang/VMDouble.java: New file.
	* java/lang/VMFloat.java: New file.
	* java/lang/Float.java, java/lang/Double.java: Removed.

From-SVN: r110759

libjava/java/lang/VMDouble.java

@@ -0,0 +1,111 @@
+/* VMDouble.java -- VM Specific Double methods
+   Copyright (C) 2003, 2005, 2006  Free Software Foundation
+
+This file is part of GNU Classpath.
+
+GNU Classpath is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2, or (at your option)
+any later version.
+
+GNU Classpath is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with GNU Classpath; see the file COPYING.  If not, write to the
+Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+02110-1301 USA.
+
+Linking this library statically or dynamically with other modules is
+making a combined work based on this library.  Thus, the terms and
+conditions of the GNU General Public License cover the whole
+combination.
+
+As a special exception, the copyright holders of this library give you
+permission to link this library with independent modules to produce an
+executable, regardless of the license terms of these independent
+modules, and to copy and distribute the resulting executable under
+terms of your choice, provided that you also meet, for each linked
+independent module, the terms and conditions of the license of that
+module.  An independent module is a module which is not derived from
+or based on this library.  If you modify this library, you may extend
+this exception to your version of the library, but you are not
+obligated to do so.  If you do not wish to do so, delete this
+exception statement from your version. */
+
+package java.lang;
+
+import gnu.classpath.Configuration;
+
+/*
+ * This class is a reference version, mainly for compiling a class library
+ * jar.  It is likely that VM implementers replace this with their own
+ * version that can communicate effectively with the VM.
+ */
+
+/**
+ * Code relocated from java.lang.Double by 
+ * @author Dave Grove (groved@us.ibm.com)
+ */
+final class VMDouble
+{
+  /**
+   * Convert the double to the IEEE 754 floating-point "double format" bit
+   * layout. Bit 63 (the most significant) is the sign bit, bits 62-52
+   * (masked by 0x7ff0000000000000L) represent the exponent, and bits 51-0
+   * (masked by 0x000fffffffffffffL) are the mantissa. This function
+   * collapses all versions of NaN to 0x7ff8000000000000L. The result of this
+   * function can be used as the argument to
+   * <code>Double.longBitsToDouble(long)</code> to obtain the original
+   * <code>double</code> value.
+   *
+   * @param value the <code>double</code> to convert
+   * @return the bits of the <code>double</code>
+   * @see #longBitsToDouble(long)
+   */
+  public static native long doubleToLongBits(double value);
+
+  /**
+   * Convert the double to the IEEE 754 floating-point "double format" bit
+   * layout. Bit 63 (the most significant) is the sign bit, bits 62-52
+   * (masked by 0x7ff0000000000000L) represent the exponent, and bits 51-0
+   * (masked by 0x000fffffffffffffL) are the mantissa. This function
+   * leaves NaN alone, rather than collapsing to a canonical value. The
+   * result of this function can be used as the argument to
+   * <code>Double.longBitsToDouble(long)</code> to obtain the original
+   * <code>double</code> value.
+   *
+   * @param value the <code>double</code> to convert
+   * @return the bits of the <code>double</code>
+   * @see #longBitsToDouble(long)
+   */
+  public static native long doubleToRawLongBits(double value);
+
+  /**
+   * Convert the argument in IEEE 754 floating-point "double format" bit
+   * layout to the corresponding float. Bit 63 (the most significant) is the
+   * sign bit, bits 62-52 (masked by 0x7ff0000000000000L) represent the
+   * exponent, and bits 51-0 (masked by 0x000fffffffffffffL) are the mantissa.
+   * This function leaves NaN alone, so that you can recover the bit pattern
+   * with <code>Double.doubleToRawLongBits(double)</code>.
+   *
+   * @param bits the bits to convert
+   * @return the <code>double</code> represented by the bits
+   * @see #doubleToLongBits(double)
+   * @see #doubleToRawLongBits(double)
+   */
+  public static native double longBitsToDouble(long bits);
+
+  /**
+   * Helper method to convert to string.
+   *
+   * @param d the double to convert
+   * @param isFloat true if the conversion is requested by Float (results in
+   *        fewer digits)
+   */
+  public static native String toString(double d, boolean isFloat);
+
+  public static native double parseDouble(String str);
+}