On Mon, 2008-10-13 at 11:52 +0200, Bernd Jendrissek wrote:
> On Sun, Oct 12, 2008 at 8:46 PM, Peter Clifton <[EMAIL PROTECTED]> wrote:
> > For example, with some macros defined to pick out the syntax understood
> > by any specific compilers, we could add branch prediction information to
> > certain hot-spot routines.
> 
> I'd rather just use -fprofile-generate and -fprofile-use instead of
> adding more tokens to the source.

Doesn't this rely on running the program to generate profiling data,
ensuring the run fully exercises the various code-paths, then
re-compiling?

This is the kind of information, for the _really_ hot routines, which
the developers could profile and add.

FWIW, I tried - and it didn't seem to make much difference. Perhaps GCC
had made the right decision anyway - or the difference was so small to
be lost in the noise.


Best wishes,
-- 
Peter Clifton

Electrical Engineering Division,
Engineering Department,
University of Cambridge,
9, JJ Thomson Avenue,
Cambridge
CB3 0FA

Tel: +44 (0)7729 980173 - (No signal in the lab!)



_______________________________________________
geda-dev mailing list
[email protected]
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev

Reply via email to