This is the mail archive of the
mailing list for the GCC project.
Re: std::pow implementation
- From: Segher Boessenkool <segher at koffie dot nl>
- To: Robert Dewar <dewar at gnat dot com>
- Cc: Richard dot Earnshaw at arm dot com, coyote at coyotegulch dot com,aoliva at redhat dot com, gcc at gcc dot gnu dot org, gdr at integrable-solutions dot net,kgardas at objectsecurity dot com, rguenth at tat dot physik dot uni-tuebingen dot de
- Date: Thu, 31 Jul 2003 23:31:51 +0200
- Subject: Re: std::pow implementation
- References: <20030730160945.5C81AF2E03@nile.gnat.com>
Robert Dewar wrote:
I strongly disagree with this. Profiling is simply not an effective way of
identifying functions that benefit from inlining. Profiling is a very crude
tool that is helpful in eliminating hot spots but that's all.
Sort of. The important weakness of profiling is that
profiling only shows how a certain translation of the
program is performing; it doesn't show much about the
possible performance of other translations.
In fact, the liberal use of inline tends to *prevent* premature optimization
(of the kind I mentioned above, avoidance of abstraction), and certainly does
not distort the code.