[rfc] Minor cleanup to pascal_val_print

[rfc] Minor cleanup to pascal_val_print

[Ulrich Weigand]

Hello,

p-valprint.c uses "extract_unsigned_integer" to extract the value of a
reference in order to print it, explicitly using gdbarch_ptr_bit as
length.  The corresponding location in the val_print routine of all
other languages (except Ada) simply uses extract_typed_address to get
at that value.

The patch below changes pascal_val_print to do the same thing, getting
rid of a current_gdbarch instance in the process.

Pierre, is there any reason why this would be wrong for Pascal?

Bye,
Ulrich


ChangeLog:

	* p-valprint.c (pascal_val_print): Use extract_typed_address
	to extract reference value.

Index: gdb-head/gdb/p-valprint.c
===================================================================
--- gdb-head.orig/gdb/p-valprint.c
+++ gdb-head/gdb/p-valprint.c
@@ -251,11 +251,10 @@ pascal_val_print (struct type *type, con
       elttype = check_typedef (TYPE_TARGET_TYPE (type));
       if (options->addressprint)
 	{
+	  CORE_ADDR addr
+	    = extract_typed_address (valaddr + embedded_offset, type);
 	  fprintf_filtered (stream, "@");
-	  /* Extract the address, assume that it is unsigned.  */
-	  fputs_filtered (paddress (
-	    extract_unsigned_integer (valaddr + embedded_offset,
-	       gdbarch_ptr_bit (current_gdbarch) / HOST_CHAR_BIT)), stream);
+	  fputs_filtered (paddress (addr), stream);
 	  if (options->deref_ref)
 	    fputs_filtered (": ", stream);
 	}
-- 
  Dr. Ulrich Weigand
  GNU Toolchain for Linux on System z and Cell BE
  Ulrich.Weigand@de.ibm.com

RE: [rfc] Minor cleanup to pascal_val_print

[Pierre Muller]

This patch is OK,
even though I am not sure this part can ever be called...

I am not sure that there is an equivalent of C++ reference type
in pascal language... But I might as well be wrong!
Anyhow, I don't want to remove it either,
as it might be useful.

Patch is approved,
please check it in.


Pierre Muller
Pascal language support maintainer for GDB

Pierre

> -----Message d'origine-----
> De : Ulrich Weigand [mailto:uweigand@de.ibm.com]
> Envoyé : Thursday, May 14, 2009 1:34 PM
> À : gdb-patches@sourceware.org; muller@ics.u-strasbg.fr
> Objet : [rfc] Minor cleanup to pascal_val_print
> 
> Hello,
> 
> p-valprint.c uses "extract_unsigned_integer" to extract the value of a
> reference in order to print it, explicitly using gdbarch_ptr_bit as
> length.  The corresponding location in the val_print routine of all
> other languages (except Ada) simply uses extract_typed_address to get
> at that value.
> 
> The patch below changes pascal_val_print to do the same thing, getting
> rid of a current_gdbarch instance in the process.
> 
> Pierre, is there any reason why this would be wrong for Pascal?
> 
> Bye,
> Ulrich
> 
> 
> ChangeLog:
> 
> 	* p-valprint.c (pascal_val_print): Use extract_typed_address
> 	to extract reference value.
> 
> Index: gdb-head/gdb/p-valprint.c
> ===================================================================
> --- gdb-head.orig/gdb/p-valprint.c
> +++ gdb-head/gdb/p-valprint.c
> @@ -251,11 +251,10 @@ pascal_val_print (struct type *type, con
>        elttype = check_typedef (TYPE_TARGET_TYPE (type));
>        if (options->addressprint)
>  	{
> +	  CORE_ADDR addr
> +	    = extract_typed_address (valaddr + embedded_offset, type);
>  	  fprintf_filtered (stream, "@");
> -	  /* Extract the address, assume that it is unsigned.  */
> -	  fputs_filtered (paddress (
> -	    extract_unsigned_integer (valaddr + embedded_offset,
> -	       gdbarch_ptr_bit (current_gdbarch) / HOST_CHAR_BIT)),
> stream);
> +	  fputs_filtered (paddress (addr), stream);
>  	  if (options->deref_ref)
>  	    fputs_filtered (": ", stream);
>  	}
> --
>   Dr. Ulrich Weigand
>   GNU Toolchain for Linux on System z and Cell BE
>   Ulrich.Weigand@de.ibm.com

Re: [rfc] Minor cleanup to pascal_val_print

[Ulrich Weigand]

Pierre Muller wrote:

> This patch is OK,
> even though I am not sure this part can ever be called...
> 
> I am not sure that there is an equivalent of C++ reference type
> in pascal language... But I might as well be wrong!
> Anyhow, I don't want to remove it either,
> as it might be useful.
> 
> Patch is approved,
> please check it in.

I've checked the patch in now, thanks!

Bye,
Ulrich

-- 
  Dr. Ulrich Weigand
  GNU Toolchain for Linux on System z and Cell BE
  Ulrich.Weigand@de.ibm.com