RFA: remove redundant code in ada-lang.c

RFA: remove redundant code in ada-lang.c

[Jim Blandy]

I don't have an Ada compiler handy, so I can't test this, but does
this look right, Joel?  It seems to me that since
ada_value_primitive_packed_val is passed arr as its 'obj' parameter,
v's lvalue type should already be set accordingly.

2006-11-22  Jim Blandy  <jimb@codesourcery.com>

	* ada-lang.c (value_subscript_packed): Don't bother to set the
	lvalue type of the returned value here;
	ada_value_primitive_packed_val takes care of that.

Index: gdb-jimb-browse/gdb/ada-lang.c
===================================================================
--- gdb-jimb-browse.orig/gdb/ada-lang.c	2006-11-29 14:59:32.000000000 -0800
+++ gdb-jimb-browse/gdb/ada-lang.c	2006-11-29 15:06:43.393216000 -0800
@@ -1887,10 +1887,6 @@
 
   v = ada_value_primitive_packed_val (arr, NULL, elt_off, bit_off,
                                       bits, elt_type);
-  if (VALUE_LVAL (arr) == lval_internalvar)
-    VALUE_LVAL (v) = lval_internalvar_component;
-  else
-    VALUE_LVAL (v) = VALUE_LVAL (arr);
   return v;
 }
 

Re: RFA: remove redundant code in ada-lang.c

[Joel Brobecker]

Hi Jim,

> I don't have an Ada compiler handy, so I can't test this, but does
> this look right, Joel?  It seems to me that since
> ada_value_primitive_packed_val is passed arr as its 'obj' parameter,
> v's lvalue type should already be set accordingly.
> 
> 2006-11-22  Jim Blandy  <jimb@codesourcery.com>
> 
> 	* ada-lang.c (value_subscript_packed): Don't bother to set the
> 	lvalue type of the returned value here;
> 	ada_value_primitive_packed_val takes care of that.

Paul Hilfinger reviewed the patch for us and said that it looked good.
I tested it on my laptop and showed no regression.

Thanks for the patch!
-- 
Joel

Re: RFA: remove redundant code in ada-lang.c

[Jim Blandy]

Joel Brobecker <brobecker@adacore.com> writes:
>> I don't have an Ada compiler handy, so I can't test this, but does
>> this look right, Joel?  It seems to me that since
>> ada_value_primitive_packed_val is passed arr as its 'obj' parameter,
>> v's lvalue type should already be set accordingly.
>> 
>> 2006-11-22  Jim Blandy  <jimb@codesourcery.com>
>> 
>> 	* ada-lang.c (value_subscript_packed): Don't bother to set the
>> 	lvalue type of the returned value here;
>> 	ada_value_primitive_packed_val takes care of that.
>
> Paul Hilfinger reviewed the patch for us and said that it looked good.
> I tested it on my laptop and showed no regression.

Committed --- thanks!