This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Should we continue to hold up 2.20 for -Wundef fixes?
- From: Allan McRae <allan at archlinux dot org>
- To: Roland McGrath <roland at hack dot frob dot com>, "Carlos O'Donell" <carlos at redhat dot com>
- Cc: GNU C Library <libc-alpha at sourceware dot org>, Siddhesh Poyarekar <siddhesh at redhat dot com>, "Joseph S. Myers" <joseph at codesourcery dot com>
- Date: Thu, 04 Sep 2014 11:49:37 +1000
- Subject: Re: Should we continue to hold up 2.20 for -Wundef fixes?
- Authentication-results: sourceware.org; auth=none
- References: <54073A95 dot 7000105 at redhat dot com> <20140903211832 dot F39382C39C9 at topped-with-meat dot com>
On 04/09/14 07:18, Roland McGrath wrote:
> Unlike Joseph, I don't think any -Wundef fixes are especially
> risky--largely because of the ease of comparing .o files to convince
> ourselves each change is safe. But I don't have a strong opinion about
> which side of the release cut they land on. My previous recommendations
> about this were based on the notion of people taking the released source
> tree (perhaps some time in the future), seeing a zillion noisy warnings,
> and coming back to us thinking something Must Be Wrong (or, similarly, not
> noticing the messages that really did matter to their configuration when
> they were lost in a sea of distracting -Wundef warnings). But that is just
> one consideration, and a fairly weak one, compared to continued delay and
> other people's differing perceptions of the risk of further changes going
> in soon.
>
Most of these I would deem safe. However, given some of them needed
fixes due to issues on particular architectures), I'd prefer the ones
that touch ports that they are not specifically tested on to wait.
Saying that, I think Siddhesh has tested on many architectures since the
initial posting.
I think they should wait for 2.21 and we should get 2.20 out the door
now. It is not too last minute for me to be happy.
Allan