This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

[PATCH] assert gimplifier assumption to prevent silent wrong code

From: Olivier Hainque <hainque at adacore dot com>
To: gcc-patches at gcc dot gnu dot org
Cc: hainque at adacore dot com
Date: Thu, 26 Jun 2008 16:09:26 +0200
Subject: [PATCH] assert gimplifier assumption to prevent silent wrong code

Hello,

The gimplifier has

  /* A subroutine of gimplify_modify_expr. [...]
     In this case we know that the RHS is a CONSTRUCTOR with an empty
     element list.  */
  gimplify_modify_expr_to_memset

On the attached Ada testcase, a front-end bug lead us here with a non
empty contructor, generating wrong code silently.

We have fixed the specific front-end bug involved, and the attached
patch is a suggestion to gcc_assert the assumption so that similar
bugs show up as ICEs instead of silent wrong code.

Bootstrapped and regression tested on x86_64-suse-linux.

Thanks in advance,

Olivier

2008-06-26  Olivier Hainque  <hainque@adacore.com>

	gcc/
	* gimplify.c (gimplify_modify_expr): When turning a varsize
	assignment from constructor into a nullifying memset, assert
	that the constructor is empty.

	testsuite/
	* gnat.dg/aligned_vla.adb: New test.

Attachment: memset_constructor.dif
Description: Text document

Follow-Ups:
- Re: [PATCH] assert gimplifier assumption to prevent silent wrong code
  - From: Richard Guenther

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]