This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: egcs build report
- To: Joe Buck <jbuck at synopsys dot com>
- Subject: Re: egcs build report
- From: Gerald Pfeifer <pfeifer at dbai dot tuwien dot ac dot at>
- Date: Mon, 22 Jun 1998 18:14:46 +0200 (MET DST)
- cc: egcs at cygnus dot com, law at cygnus dot com, tomh at taz dot ccs dot fau dot edu
On Thu, 18 Jun 1998, Joe Buck wrote:
> But those users will install binary distributions, which generally will
> be built without -g.
I seriously doubt this. Currently the documentation is very weak in
this area:
markab[189]:/sw/swtest/egcs/.../wwwdocs/htdocs/install% grep FLAG *
markab[190]:/sw/swtest/egcs/.../wwwdocs/htdocs/install%
> Considering that disk prices are now about 1/100th of what they were
> when this policy began, I don't think it's a problem.
As Mark Mitchell has mentioned before, disk space _is_ a problem. Not
everyone installing egcs is administrator and has full access, we also
should think about ``plain'' users interested in egcs.
> A compromise might be to include in the installation instructions "If
> you're really short on disk space and not competent to help us debug,
> do xyz".
I guess that's a good idea. Anyone willing to write that piece of
documentation?
(After having caused the longest thread in the history of egcs just
recently, I'm somewhat hesitant... :-\ )
> Proper instruction should be given! As g++ FAQ maintainer, I don't want
> to suddenly get lots of mail from folks who can't debug their code
> (because someone installed a libstdc++ with no debug symbols).
What I would like to see -- preferrably by default -- is that the
executables are built without debugging information, while the libraries
are made with -g.
Is there any problem with this approach resp. making it the default?
Gerald
--
Gerald Pfeifer (Jerry) Vienna University of Technology
pfeifer@dbai.tuwien.ac.at http://www.dbai.tuwien.ac.at/~pfeifer/