This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: PATCH: libstdc++/5697
- From: Gerald Pfeifer <pfeifer at dbai dot tuwien dot ac dot at>
- To: Loren James Rittle <rittle at labs dot mot dot com>
- Cc: gcc-patches at gcc dot gnu dot org, <libstdc++ at gcc dot gnu dot org>
- Date: Thu, 21 Feb 2002 08:16:10 +0100 (CET)
- Subject: Re: PATCH: libstdc++/5697
On Tue, 19 Feb 2002, Loren James Rittle wrote:
>> I am getting kind of nervous when I see new bootstrap failures introduced
>> directly before the branch date for a release.
> Let us not overstate the problem please.
As it was due to the colon-in-objdir issue, it really wasn't that huge;
however, in general I have been noticing that libstdc++ has been receiving
relatively invasive changes relatively late in the release processes.
> The patch below makes them consistent and uses the proper automake way
> to obtain the CONFIG_HEADER instead of attempting to construct its
> name using ad hoc methods
Thanks!
> BTW, are you aware that if you have a colon in your srcdir, many
> Makefiles in gcc will break? Also, are you aware that if you have a
> colon in your target triple, many Makefiles in gcc will break? I know
> these things by trivial visual inspection. I am beginning to wonder...
Target triples we have full control over, so this isn't a problem. I am
mostly testing colon-in-objdir as I this had worked without problems for
years, and I simply want to avoid regressions showing up only after formal
releases. (Else I would have eliminated the weird colon ages ago. ;) )
Gerald
--
Gerald "Jerry" pfeifer@dbai.tuwien.ac.at http://www.dbai.tuwien.ac.at/~pfeifer/