diff mbox

Dissociate store_expr's temp from exp so that it is not marked as addressable

Message ID 20120404131401.GA19344@virgil.arch.suse.de
State New
Headers show

Commit Message

Martin Jambor April 4, 2012, 1:14 p.m. UTC 
Hi,

On Tue, Apr 03, 2012 at 11:02:11AM +0200, Eric Botcazou wrote:
> > Yeah, that sounds reasonable.
> 
> There is a further subtlety in the second temp allocation when the expression 
> doesn't use the alias set of its type.  In that case, we cannot pass the type 
> to set_mem_attributes.  In fact, since assign_stack_temp_for_type already 
> calls the appropriate set_mem_* routines, the best thing to do might be to 
> remove the call to set_mem_attributes altogether in that case.
> 

So, something like this?  Bootstrapped and tested on x86_64-linux and
ia64-linux, I'm currently having problems bootsrapping sparc64 which
is what I need this mainly for but those are unelated and this should
help.

Thanks,

Martin



2012-04-03  Martin Jambor  <mjambor@suse.cz>

	* expr.c (expand_expr_real_1): Pass type, not the expression, to
	set_mem_attributes for a memory temporary.  Do not call the
	function for temporaries with a different alias set.

Comments

Eric Botcazou April 6, 2012, 4:13 p.m. UTC | #1 
> 2012-04-03  Martin Jambor  <mjambor@suse.cz>
>
> 	* expr.c (expand_expr_real_1): Pass type, not the expression, to
> 	set_mem_attributes for a memory temporary.  Do not call the
> 	function for temporaries with a different alias set.

The last sentence is unprecise, this would rather be: "Do not call the function 
for the memory temporary created for a bitfield".

I wonder whether we should simplify the bitfield case in the process.  Once we 
remove the call to set_mem_attributes, I think the

		/* If the reference doesn't use the alias set of its type,
		   we cannot create the temporary using that type.  */

is useless, so we could try to revert r122014 in the process.
Eric Botcazou April 6, 2012, 4:21 p.m. UTC | #2 
> @@ -9870,7 +9871,14 @@ expand_expr_real_1 (tree exp, rtx target
>  	if (op0 == orig_op0)
>  	  op0 = copy_rtx (op0);
>
> -	set_mem_attributes (op0, exp, 0);
> +	/* If op0 is a temporary because of forcing to memory, pass only the
> +	   type to set_mem_attributes so that the original expression is never
> +	   marked as ADDRESSABLE through MEM_EXPR of the temporary.  */
> +	if (mem_attrs_from_type)
> +	  set_mem_attributes (op0, TREE_TYPE (exp), 0);

set_mem_attributes (op0, type, 0); is equivalent.
diff mbox

Patch

Index: src/gcc/expr.c
===================================================================
--- src.orig/gcc/expr.c
+++ src/gcc/expr.c
@@ -9572,6 +9572,7 @@  expand_expr_real_1 (tree exp, rtx target
 	tree tem = get_inner_reference (exp, &bitsize, &bitpos, &offset,
 					&mode1, &unsignedp, &volatilep, true);
 	rtx orig_op0, memloc;
+	bool mem_attrs_from_type = false;
 
 	/* If we got back the original object, something is wrong.  Perhaps
 	   we are evaluating an expression too early.  In any event, don't
@@ -9677,6 +9678,7 @@  expand_expr_real_1 (tree exp, rtx target
 	    memloc = assign_temp (nt, 1, 1, 1);
 	    emit_move_insn (memloc, op0);
 	    op0 = memloc;
+	    mem_attrs_from_type = true;
 	  }
 
 	if (offset)
@@ -9849,7 +9851,6 @@  expand_expr_real_1 (tree exp, rtx target
 		emit_move_insn (new_rtx, op0);
 		op0 = copy_rtx (new_rtx);
 		PUT_MODE (op0, BLKmode);
-		set_mem_attributes (op0, exp, 1);
 	      }
 
 	    return op0;
@@ -9870,7 +9871,14 @@  expand_expr_real_1 (tree exp, rtx target
 	if (op0 == orig_op0)
 	  op0 = copy_rtx (op0);
 
-	set_mem_attributes (op0, exp, 0);
+	/* If op0 is a temporary because of forcing to memory, pass only the
+	   type to set_mem_attributes so that the original expression is never
+	   marked as ADDRESSABLE through MEM_EXPR of the temporary.  */
+	if (mem_attrs_from_type)
+	  set_mem_attributes (op0, TREE_TYPE (exp), 0);
+	else
+	  set_mem_attributes (op0, exp, 0);
+
 	if (REG_P (XEXP (op0, 0)))
 	  mark_reg_pointer (XEXP (op0, 0), MEM_ALIGN (op0));