Minor dbxout.c changes

[Richard Kenner]

>From geoffk@geoffk.org Thu May 24 05:26:26 2001
Received: from alpha2.ultra.nyu.edu by vlsi1.ultra.nyu.edu (4.1/1.34)
	id AA23787; Thu, 24 May 01 05:26:25 EDT
Received: from geoffk.org (desire.geoffk.org [64.2.60.52])
	by alpha2.ultra.nyu.edu (8.9.3/8.9.3) with ESMTP id FAA25283
	for <kenner@vlsi1.ultra.nyu.edu>; Thu, 24 May 2001 05:20:07 -0400 (EDT)
Received: (from geoffk@localhost)
	by geoffk.org (8.9.3/8.9.3) id CAA01064;
	Thu, 24 May 2001 02:22:54 -0700
To: kenner@vlsi1.ultra.nyu.edu (Richard Kenner)
Cc: gcc-patches@gcc.gnu.org
Subject: Re: Minor dbxout.c changes
References: < 10105182051.AA05548@vlsi1.ultra.nyu.edu >
From: Geoff Keating <geoffk@geoffk.org>
Date: 24 May 2001 02:21:54 -0700
In-Reply-To: kenner@vlsi1.ultra.nyu.edu's message of "Fri, 18 May 01 16:51:49 EDT"
Message-Id: < jmy9rnm79p.fsf@geoffk.org >
Lines: 22
X-Mailer: Gnus v5.5/Emacs 20.3
Status: RO


Hi Richard,

kenner@vlsi1.ultra.nyu.edu (Richard Kenner) writes:

> This mostly fixes a bug where we generate tags on variable-length
> record that causes objdump to crash, but also does some cleanup
> in using host_integerp instead of checking for INTEGER_CST.

How did you test this patch?  I was looking at it, and this line:

> ! 		/* Do not generate a tag for records of variable size,
> ! 		   since this type can not be properly described in the
> ! 		   DBX format, and it confuses some tools such as objdump.  */
here> ! 		&& ! host_integerp (TYPE_SIZE (type), 1))
>   	      {
>   		tree name = TYPE_NAME (type);

seems to do the opposite of what the comment says.

-- 
- Geoffrey Keating <geoffk@geoffk.org>