Re: mips-tdep.c: Fix new-ABI handling of composite return values

["Maciej W. Rozycki" <macro at mips dot com>]

On Tue, 25 Sep 2007, Daniel Jacobowitz wrote:

> 128-bit long double is also handled in this patch - that's not a
> composite type.  I think that's a merge error, though.  If you look
> above the float / double block you'll find another copy of
> 128-bit long double support in HEAD.

 Oops! -- I have missed this bit indeed.

> >    else if (TYPE_CODE (type) == TYPE_CODE_STRUCT
> >  	   && TYPE_NFIELDS (type) <= 2
> >  	   && TYPE_NFIELDS (type) >= 1
> >  	   && ((TYPE_NFIELDS (type) == 1
> > -		&& (TYPE_CODE (TYPE_FIELD_TYPE (type, 0))
> > +     	        && (TYPE_CODE (check_typedef (TYPE_FIELD_TYPE (type, 0)))
> 
> Funny indentation.  Is the only substantive change in this patch the

 Fixed.

> new calls to check_typedef?

 Of course not.  This bit matters much too:

-  if (TYPE_CODE (type) == TYPE_CODE_STRUCT
-      || TYPE_CODE (type) == TYPE_CODE_UNION
-      || TYPE_CODE (type) == TYPE_CODE_ARRAY
+  if (TYPE_CODE (type) == TYPE_CODE_ARRAY

 Here's an updated version.

2007-09-25  David Ung  <davidu@mips.com>
            Maciej W. Rozycki  <macro@mips.com>

	* mips-tdep.c (mips_n32n64_return_value): Per N32/N64 ABI
	rules return composite types in registers as appropriate.

 OK to apply?

  Maciej

12438.diff
Index: binutils-quilt/src/gdb/mips-tdep.c
===================================================================
--- binutils-quilt.orig/src/gdb/mips-tdep.c	2007-09-25 16:42:08.000000000 +0100
+++ binutils-quilt/src/gdb/mips-tdep.c	2007-09-25 16:43:43.000000000 +0100
@@ -3076,9 +3076,30 @@
 			  gdb_byte *readbuf, const gdb_byte *writebuf)
 {
   struct gdbarch_tdep *tdep = gdbarch_tdep (current_gdbarch);
-  if (TYPE_CODE (type) == TYPE_CODE_STRUCT
-      || TYPE_CODE (type) == TYPE_CODE_UNION
-      || TYPE_CODE (type) == TYPE_CODE_ARRAY
+
+  /* From MIPSpro N32 ABI Handbook, Document Number: 007-2816-004
+
+     Function results are returned in $2 (and $3 if needed), or $f0 (and $f2
+     if needed), as appropriate for the type.  Composite results (struct,
+     union, or array) are returned in $2/$f0 and $3/$f2 according to the
+     following rules:
+
+     * A struct with only one or two floating point fields is returned in $f0
+     (and $f2 if necessary).  This is a generalization of the Fortran COMPLEX
+     case.
+
+     * Any other struct or union results of at most 128 bits are returned in
+     $2 (first 64 bits) and $3 (remainder, if necessary).
+
+     * Larger composite results are handled by converting the function to a
+     procedure with an implicit first parameter, which is a pointer to an area
+     reserved by the caller to receive the result.  [The o32-bit ABI requires
+     that all composite results be handled by conversion to implicit first
+     parameters.  The MIPS/SGI Fortran implementation has always made a
+     specific exception to return COMPLEX results in the floating point
+     registers.]  */
+
+  if (TYPE_CODE (type) == TYPE_CODE_ARRAY
       || TYPE_LENGTH (type) > 2 * MIPS64_REGSIZE)
     return RETURN_VALUE_STRUCT_CONVENTION;
   else if (TYPE_CODE (type) == TYPE_CODE_FLT
@@ -3122,12 +3143,12 @@
 	   && TYPE_NFIELDS (type) <= 2
 	   && TYPE_NFIELDS (type) >= 1
 	   && ((TYPE_NFIELDS (type) == 1
-		&& (TYPE_CODE (TYPE_FIELD_TYPE (type, 0))
+		&& (TYPE_CODE (check_typedef (TYPE_FIELD_TYPE (type, 0)))
 		    == TYPE_CODE_FLT))
 	       || (TYPE_NFIELDS (type) == 2
-		   && (TYPE_CODE (TYPE_FIELD_TYPE (type, 0))
+		   && (TYPE_CODE (check_typedef (TYPE_FIELD_TYPE (type, 0)))
 		       == TYPE_CODE_FLT)
-		   && (TYPE_CODE (TYPE_FIELD_TYPE (type, 1))
+		   && (TYPE_CODE (check_typedef (TYPE_FIELD_TYPE (type, 1)))
 		       == TYPE_CODE_FLT)))
 	   && tdep->mips_fpu_type != MIPS_FPU_NONE)
     {