[Fwd: GCC 4.3.2 Status Report (2008-08-18)]

[Paul Richard Thomas]

Dear All,

On Mon, Aug 18, 2008 at 5:24 PM, Tobias Burnus <burnus@net-b.de> wrote:
> Forwarded for completeness.
> To my knowledge no one is currently working on gfortran regressions.
>
> Currently, the following PRs are marked as gfortran regressions:
>
> a) 36599 major execution regression for induct.f90 polyhedron benc...
> b) 36371 [4.3 Regression] Wrong locus for errors in DATA statement

This one can easily be fixed by anybody but me - The patch was applied
to trunk a long while back; apparently with no destructive effects.  I
will not be in a position to fix it for some weeks, so feel free:-)

> c) 36700 [4.3/4.4 Regression] ICE on calling a function
> d) 35680 [4.3/4.4 regression] ICE on invalid transfer in variable ...
> e) 36463 [4.4 Regression] ICE in expand_expr_real_1, at expr.c:726...
>
> (a) is a bit unclear, seems to be some specific problem, more ME related
> (b) is diagnostic and a patch is available (checked in into 4.4/trunk)
>    Close or backport? Or is something missing?
> (c) and (d) are ice-on-invalid-code
>    Nothing to say here, needs to be fixed, but is not urgent
> (e) is an ice-on-valid-code PR
>    The ICE is in so far new that there was before a rejects-valid
>    problem. One patch fixed a parser/resolve problem and now this
>    program ICEs later in the code path.
>    In my opinion this is not a real regression.

I agree with the rest - ie; we can live with them until stage 3.

Cheers

Paul