Re: [RFA] powerpc - extract a float return value

[Jerome Guitton <guitton@act-europe.fr>]

[-- Attachment #1: Type: text/plain, Size: 1087 bytes --]

Kevin Buettner (kevinb@redhat.com):

> The rs6000 return value code will need to be revamped to not refer to
> DEPRECATED_REGISTER_BYTE.  And, in fact, it probably ought to be
> restructured to use the mechanisms found in ppc-linux.c.  See
> ppc_linux_return_value() and ppc_sysv_abi_return_value().  This
> is not a barrier to getting your current patch in; I only mention
> it in case you wish to tackle this bit of work too.

Sorry, I would not really easy for me, I do not have a convenient
setup for PPC (that is actually the reason why it took me 3 days
to be able to test against the testsuite :-)

> > I have not yet tested it against the testsuite, I will do that
> > tomorrow. In the meantime, if you have comments I would be happy to
> > address them!
> 
> Please let me know the results of your tests.  If they look good,
> please commit your patch.

OK, thanks. Tested on the ppc-elf simulator, on a x86-linux host.
No regression.

I have slightly modified the patch, I do not see any reason to have a special
case for doubles now. OK to apply with this change?

-- 
Jerome

[-- Attachment #2: diff.3 --]
[-- Type: text/plain, Size: 1489 bytes --]

2004-04-23  Jerome Guitton  <guitton@gnat.fr>

	* rs6000-tdep.c (rs6000_extract_return_value): When extracting a float,
	use convert_typed_floating to get the appropriate format.

Index: rs6000-tdep.c
===================================================================
RCS file: /cvs/src/src/gdb/rs6000-tdep.c,v
retrieving revision 1.191
diff -u -p -r1.191 rs6000-tdep.c
--- rs6000-tdep.c	1 Apr 2004 21:00:59 -0000	1.191
+++ rs6000-tdep.c	21 Apr 2004 10:14:03 -0000
@@ -1251,22 +1251,15 @@ rs6000_extract_return_value (struct type
   if (TYPE_CODE (valtype) == TYPE_CODE_FLT)
     {
 
-      double dd;
-      float ff;
       /* floats and doubles are returned in fpr1. fpr's have a size of 8 bytes.
          We need to truncate the return value into float size (4 byte) if
          necessary.  */
 
-      if (TYPE_LENGTH (valtype) > 4)	/* this is a double */
-	memcpy (valbuf,
-		&regbuf[DEPRECATED_REGISTER_BYTE (FP0_REGNUM + 1)],
-		TYPE_LENGTH (valtype));
-      else
-	{			/* float */
-	  memcpy (&dd, &regbuf[DEPRECATED_REGISTER_BYTE (FP0_REGNUM + 1)], 8);
-	  ff = (float) dd;
-	  memcpy (valbuf, &ff, sizeof (float));
-	}
+      convert_typed_floating (&regbuf[DEPRECATED_REGISTER_BYTE
+                                      (FP0_REGNUM + 1)],
+                              builtin_type_double,
+                              valbuf,
+                              valtype);
     }
   else if (TYPE_CODE (valtype) == TYPE_CODE_ARRAY
            && TYPE_LENGTH (valtype) == 16