[Gomp-discuss] Re: OpenMP, HPC, and the future of GCC

[Steven Bosscher]

Per Bothner wrote:
> Lars Segerlund wrote:
> > Perhaps you should take the time to get informed, before wasting
> > other peoples time
> 
> My point is that it is that the Gcc community is not responsible
> for implementing OpenMP, and we are not required to merge in
> an extension in to the main sources.  It is up to the OpenMP
> community to produce an implementation, and if they want it
> merged in, that it be is minimally disruptive to the cleanliness
> (such as it is!) of Gcc.

Nobody has said you're responsible for that.  Nobody has said you're
required to merge it.  On the contrary, there's a group of people
offering you to develop it and consider it for merging.  Yet even before
you see a patch you dismiss the whole plan.

Questions you don't seem to be able to answer correctly:
a) Define the "OpenMP community"
b) Define the "Gcc community"

Answers:

a) There is such community.  There is just a large group of people
supporting and using it.  What *does* exist is a group of volunteers,
with respected GCC hackers among them, who *offer* to implement OpenMP
support, and who wish to discuss the design of this implementation on
this list.  Yet all you do is diss OpenMP and tell us to go play
somewhere else.

b) That is whoever is willing and capable (by the judgement of whoever
rejects or accepts patches) of helping with the development of GCC. 
Clearly you don't understand this, instead you seem to be a purist with
no sense of what is asked for in a compiler in the real world.

Is this your idea of an "open development environment"?
Is this your idea of encouraging new initiatives?


> I am making a point about *process*, primarily.  The *design* of
> OpenMP also raises concerns, but that is a different issue.

Oh is it?  I don't think so.  If you already know you would reject it
because *you* don't like the desing, then why don't you say that now. 
That way we would not start an effort that is doomed from the beginning.
Our time is not free either, you know.


> > also conisder that here are people willing to make an effort
> > to support 'multi threading' or 'native parallallism' in gcc,
> > which if you had a look at most modern systems might be of
> > major importance to gcc and the free community.
> 
> It *might* be.  However, there have been hundreds of designs
> for language-based 'multi threading' or 'native parallallism'.
> That suggests caution before making any major changes to our
> code-base to support it.

BS.

I'll buy you a beer (if I ever meet you:) if you can name me just one
such designs that works out-of-the-box for C/C++ and Fortran and that is
as populair as OpenMP.

Also, we've said before that we don't design concurrency support around
OpenMP, but rather we want to develop a framework to support
concurrency.  OpenMP would be just one thing that can be built on that.


> > Perhaps something as simple as reading openmp's homepage
> > should give a hint as to which institutions have contributed
> > to the standard,
> 
> So what?  All that means it's a standard with a lot of backing;
> it means nothing about its quality.  For that the *people*
> who contributed would be more important.

Ah come on, of course the quality of the standard has its influence on
the implementation!  And so does the quality of the implementation of
GCC itself.  Did you even read the earlier thread about OpenMP pragmas?
Did you even *look* at who's willing to contribute?

Also, you said this morning that C++ isn't so well designed.  But I
think that is just your purist point of view again, and that the C++
user base has a very different opinion.  Why do you think more people
use C++ than Java?  You probably don't understand that, because
"well-designed" by your standards and "usable" for your target audience
are two different things.

Why do you think OpenMP became the de facto standard for MP programming?
Exactly, because it is usable.

If you want to have a compiler that only implements "quality standards"
by your measures, then clearly GCC is just an academic project to you.


> To emphasize:  efforts to make Gcc better for "computational"
> domains is very welcome.  It is just that any design needs to
> avoid further complicating and slowing down Gcc.

You don't make it sound like it's welcome.  On the contrary.


> Right now, there is a lot of concern and focus about the
> compilation speed of Gcc, especially when not opimizing.
> So it is very important that any new vectorization support
> not slow the compiler down further, at least by default.

Vectorizing is orthogonal to this discussion.  It has nothing to do with
OpenMP.  Never has, never will.

And, nobody was even talking about vectorization in this thread, let
alone "new vectorization" whatever that may be.


> GCC is Free Software.  You're free to use it as you will.  You're
> free to create a fork.  We're free to reject your changes.

Oh hear, is this a SC member is speaking?  Nobody is talking about a
fork, so why you think you should say this is unclear to me.  And there
is nothing to reject yet but somehow you seem to reject it already.

I'm sorry, but your whole mail just sucks, and your attitude makes me
sick.

Greetz
Steven