This is the mail archive of the
mailing list for the GCC project.
Re: Add checks for tree.value, tree.minval, and tree.maxval and others
- From: Richard Henderson <rth at redhat dot com>
- To: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Thu, 18 Mar 2004 15:55:45 -0800
- Subject: Re: Add checks for tree.value, tree.minval, and tree.maxval and others
- References: <10403182300.AA19872@vlsi1.ultra.nyu.edu>
On Thu, Mar 18, 2004 at 06:00:09PM -0500, Richard Kenner wrote:
> Of course. But I always list files just to be safe.
Well, clearly it isn't safe.
If you religiously use one working directory per change, then you're
doing testing as http://gcc.gnu.org/contribute.html#testing requires.
Namely, testing each patch by itself. And then when you're done
testing, you can just say "cvs diff" to get the patch, "cvs commit"
to check it all in. And you won't miss anything.
Really. I mean this. One directory per change. Lack of disk space
is no excuse; 200GB hard drives sell for $149. You are not just a
volunteer working in Botswanaland with zero resources, you are in the
US working for a company that pays you to develop gcc.
To save time and bandwidth, keep an unmodified source tree to the
side, and do "cp -a gcc-unmodified gcc-frobulate". On my machine
this copy takes 9.84 seconds, so unless you've got a truely ancient
machine (see above re working environment), this should not be an