This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: RFC: Do not install specs file
- From: Daniel Jacobowitz <drow at false dot org>
- To: Mark Mitchell <mark at codesourcery dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Sun, 19 Dec 2004 22:16:39 -0500
- Subject: Re: RFC: Do not install specs file
- References: <200412200229.iBK2TwXF020264@sirius.codesourcery.com>
On Sun, Dec 19, 2004 at 06:29:58PM -0800, Mark Mitchell wrote:
>
> Zack suggested that we might win by not installing the specs file,
> and, in fact, this turns out to be surprisingly true. I measured a
> 0.3% improvement in Qt builds simply by making that change. Some
> combination of reduced work on the part of the driver, and reduced
> disk access, makes this worthwhile. Furthermore, in internal
> discussions, we couldn't see any reason that people need the specs
> file to be installed, especially since it can be recreated with
> -dumpspecs.
>
> Does anyone object to this change (which I have tested on
> i686-pc-linux-gnu, but not installed)?
I don't object to the concept, but I don't like the implementation...
:-) Is it still necessary to create the specs file in the build
directory? I would rather not if we can avoid it. This invocation of
CC_FOR_TARGET has caused me headaches before.
It gives me the creeps to leave just the rm -f there. Since this is in
$(libsubdir), can we skip that? It would only burn people who
frequently 'make install' into the same directory, and only this once.
--
Daniel Jacobowitz