This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: fix -Werror regressions with gcc 4.8 on x86_64
- From: Chris Metcalf <cmetcalf at ezchip dot com>
- To: Joseph Myers <joseph at codesourcery dot com>, Alexandre Oliva <aoliva at redhat dot com>
- Cc: <libc-alpha at sourceware dot org>, Andreas Schwab <schwab at linux-m68k dot org>
- Date: Tue, 23 Dec 2014 16:21:01 -0500
- Subject: Re: fix -Werror regressions with gcc 4.8 on x86_64
- Authentication-results: sourceware.org; auth=none
- Authentication-results: spf=none (sender IP is ) smtp dot mailfrom=cmetcalf at ezchip dot com;
- References: <orr3vr3pa1 dot fsf at free dot home> <oregrq414g dot fsf at free dot home> <alpine dot DEB dot 2 dot 10 dot 1412231952260 dot 31777 at digraph dot polyomino dot org dot uk>
On 12/23/2014 2:56 PM, Joseph Myers wrote:
Are you building with -DNDEBUG? That's the only way I can see these
variables being unused (set-but-not-used). NDEBUG is definitely a broken
configuration for running the testsuite, as it means tests aren't checking
what they should if the verification uses assert - the warnings are
showing up a problem with your environment, not with glibc (see what
Roland said in
<https://sourceware.org/ml/libc-alpha/2014-12/msg00589.html>). And I'm
doubtful NDEBUG is a sensible configuration for building glibc itself
either; at least, it's not a well-tested one.
Thus, this patch seems wrong, and I'm dubious of the previous patch
without a clean consensus to support NDEBUG builds of glibc.
NDEBUG is added by default to RedHat/Fedora builds (other than prereleases).
This has been true since 2009; Andreas can perhaps comment on this, since the
changelog shows he enabled it.
--
Chris Metcalf, EZChip Semiconductor
http://www.ezchip.com