[Bug fortran/32302] [4.2/4.3 Regression] Incorrect result with -O2

["pault at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>]

------- Comment #5 from pault at gcc dot gnu dot org  2007-06-13 09:01 -------
Dear All,

> > I think that you are probably right - it looks like my patch for PR29786 is to
> > blame, since it is the only thing to touch trans-common.c for a long time.

This is incorrect - it goes back at least as far as gcc-4.2 20061221.  It has
nothing, therefore, to do with my recent patch.  

> If you access to a system, I can give you access to troutmask.
> Send me your ssh public key, and you can play around on 
> sgk@troutmask.  Yeah, it's trans-atlantic.

Thanks but it's not necessary - I am trying to ignore gfortran during the day
(unsuccessfully, I might say...) and normally have an x86_ia64/FC5 to play with
in the evenings.  This latter is unavailable until tomorrow because its mains
protection was blown up by one of last weeeks storms:(

The error creeps in at any level of optimization.  I am trying to reduce the
testcase but I am rather sure that this is not in the front end.

Paul 


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32302