What is acceptable for -ffast-math? A numerical viewpoint

[Laurent Guerby]

<<
The point is that the rules have to be coherent and well documented and
understood. You are NOT going to achieve that unless you have input from
people who understand :-)
>>

I was merely proposing to have a flag (one of the name I proposed
was -funsafe-fast-math) where the rules would be:

1. anyone can go ahead, code and document an optimization

2. volunteers field test it and report on performance and precision

3. unless 2 showed up unfixable practical problems, the optimization
is enabled under -funsafe-fast-math

Do you actively oppose to this proposal?

I have no opinion on other flags, since I do not consider myself an
expert. (My only feeling is that no GCC FP optimization work will be
done at all since all proposals will be discussed endlessly - even
with a "clear set of criteria", whatever that means - and there will
always someone to show up with something that will be badly broken by
the optimization.)

-- 
Laurent Guerby <guerby@acm.org>