RE: i386 backend

["David Rasmussen" <pinkfloydhomer at yahoo dot com>]

> <<I like gcc, but I'm not using it anymore, since I need all the
> performance I
> can get in my domain, and MSVC++ generated code is about 30% faster than
> that of gcc, while Intel's is 40-50% faster.
>
> What is being done to address these issues, if anything?
> >>
>
> It is not easy to address vague reports like this. What would be helpful
> is some detailed examples with behcnmark results and analyses of
> why the differences are there. It is also important to know exactly what
> optimization options you used/tried on both compilers.
>
> Generally when you compare two compilers, there will always be cases
> whre one compiler is faster than another (in both directions). To
> known whether a given instance is something that is useful to study
> further, you need the example to look at!

I know that I am being vague. I have no other choice, as my knowledge on the
internals of a compiler backend is rather small.
I do know, however, that what I am experiencing is not isolated. Many people
from many different places, working with many different projects, with a lot
of different code, has reported this to me. Professional people as well as
amateurs. So the problem is not unknown.

Even if I can't be very precise in pointing out _where_ the problems is, it
should still be recognized as a problem. A very simplistic way of putting it
is: "Why is gcc producing slower code than MSVC++ or Intel C++?". I know
that it is not always the case, but from all the different types of programs
I've compiled, and all the people who I have talked to, it is almost always
the case. So there is room for improvement.