This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: [ping] unreviewed patches for 4.4
- From: FX <fxcoudert at gmail dot com>
- To: "Daniel Franke" <franke dot daniel at gmail dot com>
- Cc: fortran at gcc dot gnu dot org
- Date: Tue, 19 Feb 2008 10:06:48 +0000
- Subject: Re: [ping] unreviewed patches for 4.4
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; bh=3sAmybzgR99sKtQrPtSx5YQ4RAvuvVGDJwIaKa0LNCA=; b=MyP9EAd/OXMwCcVlxdJdVqk4KdkeC/C0sISRqIwZULoJq0E82341LZDK2Aemkl5IP6eAFYQMqz17Ub1x0GxuWSLZwHaMxMhX6C8P2A08XXvSG4/PEjZPJOCgeX+JjYY1LiyaqCEnQg+d15MKZ8CbX9fqJ9v28AZnzmXhuoTP+RY=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=MtPmGMIJzqR2rnITWPo0rDApq5a8cB2aA8QWTb0PWFZubPGOBm2YiDifJHvN5VGPfE0EPvzX4E4ioJzZGdU+kwh/2AesRJZhayVpNI1hjdEk2zl2MwxkUSPlejAfBMU8fs3gEWya+ScZKfpnTMFzIXzN3CnW2tzTV3JNnKvTvCY=
- References: <640ad44b0802190127s47137b87y19da36336329ad01@mail.gmail.com>
> PRs 31463, 33950, 34296 - fix inconsistent warnings if function
> return value is not set
> http://gcc.gnu.org/ml/fortran/2008-01/msg00355.html
I think it would be good to hide them behind a warning option.
-Wreturn-type is probably the best choice among existing options,
although clearly not perfect. Maybe we should just bind it to -W or
-Wall?
My second concern is your use of TREE_NO_WARNING: if another,
unrelated warning is to be emitted about that function, my understand
is that TREE_NO_WARNING will kill it altogether. I'm trying to see if
there is a really-life situation where that might happen, but I'm
pretty confident there is. I can't access SVN from where I am, so I
can't check, but isn't there any other type of TREE_ macro that might
be more subtle? TREE_USED, maybe?
> Although neither is a regression, how about backport to 4.3.1?
Like Tobias, I'd prefer we don't. Less risky (in this case, admittedly
not much) and we keep developer time for mainline development.
--
FX Coudert
http://www.homepages.ucl.ac.uk/~uccafco/