This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] Fix a fallout of PR14179 fix (3.3/3.4/4.0 regression)
- From: Jakub Jelinek <jakub at redhat dot com>
- To: Richard Henderson <rth at redhat dot com>, Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Cc: gcc-patches at gcc dot gnu dot org, giovannibajo at libero dot it
- Date: Mon, 27 Dec 2004 18:41:08 -0500
- Subject: Re: [PATCH] Fix a fallout of PR14179 fix (3.3/3.4/4.0 regression)
- References: <20041227154307.GB10340@devserv.devel.redhat.com> <20041227224652.GA5386@redhat.com>
- Reply-to: Jakub Jelinek <jakub at redhat dot com>
On Mon, Dec 27, 2004 at 02:46:52PM -0800, Richard Henderson wrote:
> On Mon, Dec 27, 2004 at 10:43:07AM -0500, Jakub Jelinek wrote:
> > /* Size types *are* sign extended. */
> > sign_extended_type = (!TYPE_UNSIGNED (TREE_TYPE (t))
> > || (TREE_CODE (TREE_TYPE (t)) == INTEGER_TYPE
> > && TYPE_IS_SIZETYPE (TREE_TYPE (t))));
> Boggle. I wonder what breaks, if anything if this bit
> of bogosity is removed?
I guess the original intent of
was to get overflows detected on signed types AND sizetype and this
sign-extended representation of sizetype was just a side-effect?
I'll see what breaks with this removed. That would be suitable
just for HEAD if it works, right?