This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [Patch, fortran] PR40011 - Problems with -fwhole-file


On Mon, May 24, 2010 at 10:52 PM, Thomas Koenig <tkoenig@netcologne.de> wrote:
> Paul Richard Thomas wrote:
>
>> (ii) Modify some of the testcases by fixing their problems and using
>> -fno-whole-file or std=legacy for the rest - or the latter for all of
>> them?
>
> There are test cases where we used a separate subroutine for hiding
> something from the optimizers. ?We should make sure to include at
> least "-O2 -fno-whole-file" (or something similar) in the options for
> the dg-do run test cases.

I can't see a single reason for -fno-whole-file as a way to change
what the Fortran frontend presents to the middle-end.  I have no idea
on whether it is useful to disable the extra warnings or errors the
Frontend produces with -fwhole-file -- for example the C and C++
frontends provide a -fpermissive option that turns some errors into warnings.

Richard.

> ? ? ? ?Thomas
>
>


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]