GCC Bugzilla Bug 5679

When the call is expanded, there does not seem to be a target tree for the
returned object, although GCC cannot determine its size and thus needs one.

The problem disappears if the type conversion is removed, so the test case is
probably minimal.

This bugs is also triggered when compiling GtkAda 1.2.12.

Release:
GCC 3.1 CVS from 2002-02-12

Environment:
Debian GNU/Linux unstable, x86

How-To-Repeat:
Compile the attached code.

Fix:
Unknown.  The bug might be caused by the incomplete merge from the ACT tree.

From: Dara Hazeghi <dhazeghi@yahoo.com>
To: gcc-gnats@gcc.gnu.org, adi@drcomp.erfurt.thur.de, fw@deneb.enyo.de
Cc:  
Subject: Re: ada/5679: ICE in assign_stack_temp_for_type, returning unconstrained type conversion ("GtkAda bug")
Date: Sat, 10 May 2003 14:29:01 -0700

 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit- 
 trail&database=gcc&pr=5679
 
 Hello,
 
 gcc 3.1 is quite old. Would it be possible for you to determine if this  
 problem still occurs on a more current version (ie 3.2.3)? Thanks,
 
 Dara
 

From: Dara Hazeghi <dhazeghi@yahoo.com>
To: gcc-gnats@gcc.gnu.org, fw@deneb.enyo.de, adi@drcomp.erfurt.thur.de
Cc:  
Subject: Re: ada/5679: ICE in assign_stack_temp_for_type, returning unconstrained type conversion ("GtkAda bug")
Date: Mon, 12 May 2003 01:47:24 -0700

 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit- 
 trail&database=gcc&pr=5679
 
 Hello,
 
 sorry for the earlier message. I can confirm that this bug still occurs  
 on gcc 3.2, 3.3 branch and mainline (20030510) on i686-linux. No  
 further confirmation is needed!
 
 Dara
 

State-Changed-From-To: open->analyzed
State-Changed-Why: Confirmed in 3.3 and mainline

This is a dup of 9596.

*** This bug has been marked as a duplicate of 9596 ***

I closed this bug the wrong way.

*** Bug 9596 has been marked as a duplicate of this bug. ***

*** Bug 6452 has been marked as a duplicate of this bug. ***

Geert, this appears to be a rather serious (and often encountered) bug. Do you
have any insight 
into what it's cause might be? Thanks,

Dara

Confirmed still present with gcc 3.3 branch and mainline (20030620).

*** Bug 8368 has been marked as a duplicate of this bug. ***

*** Bug 9132 has been marked as a duplicate of this bug. ***

*** Bug 11541 has been marked as a duplicate of this bug. ***

It also happens on powerpc-apple-darwin6.6 on the mainline (20030719).

The original (GtkAda) code was wrong, and has been fixed a while
ago.

It does not make sense for a function to return such a giant array.
There may still be a remaining "hole" in the gcc back-end, but this
is neither important, nor because of "an incomplete merge", so I am
closing this report.

Arno

Subject: Re:  ICE in assign_stack_temp_for_type, returning unconstrained type
conversion ("GtkAda bug")

charlet at gcc dot gnu dot org wrote:

> The original (GtkAda) code was wrong, and has been fixed a while
> ago.
> 
> It does not make sense for a function to return such a giant array.

I was under the impression that the code uses a documented GNAT
extension.  I'm pretty sure that this the example is not valid Ada code.

Subject: Re:  ICE in assign_stack_temp_for_type, returning unconstrained type
conversion ("GtkAda bug")

> I was under the impression that the code uses a documented GNAT
> extension.

No, that's not the case.

Arno