This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: SIMPLE: A language-independent tree IR
On Mon, 2002-01-07 at 05:58, Joseph S. Myers wrote:
> How are you representing the *partial* order constraints represented by
> sequence points? The description of SIMPLE you gave suggests that it
> fully serialises expression evaluation - but there should be a
> language-independent representation in which the partial order constraints
> are explicitly represented.
>
So far, I think we could use sequence or barrier nodes. So that
front-ends could emit something like:
START_SEQ
s1
s2
END_SEQ
> How are you representing the Fortran reassociation constraints? How are
> you representing the various forms of restriction on floating point
> evaluation (C99, Java) discussed?
>
Non-associative operators could be used here. But I haven't thought
about these too much. Suggestions are very much welcome. I suspect
we'll iterate over different options as we convert other front-ends.
I'm hoping somebody from the Java and Fortran FEs will volunteer to work
on those.
> Since you are defining a new format that is supposed to have entirely
> language-independent semantics, I trust you'll be committing full
> documentation for these semantics
>
Absolutely.
Diego.