[Bug fortran/95397] New: [Fortran/OpenACC] Wrong results with 'loop vector' inside 'routine'

[Bug fortran/95397] New: [Fortran/OpenACC] Wrong results with 'loop vector' inside 'routine'

[tschwinge at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95397

            Bug ID: 95397
           Summary: [Fortran/OpenACC] Wrong results with 'loop vector'
                    inside 'routine'
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Keywords: openacc
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: tschwinge at gcc dot gnu.org
                CC: burnus at gcc dot gnu.org
  Target Milestone: ---

Created attachment 48631
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48631&action=edit
'openacc-github-302-2_.f90'

See attached.  This is a simplified version of
<https://github.com/OpenACC/openacc-spec/issues/302#issuecomment-635348932>,
'openacc-github-302-2.f90.txt'.  With nvptx offloading:

    $ ./a.out
               0           0           0           0           0           0   
       0           0           0           0

Expected:

    $ ACC_DEVICE_TYPE=host ./a.out
               1           2           3           4           5           6   
       7           8           9          10

Is something wrong with my Fortran/OpenACC, or with GCC's Fortran/OpenACC?

[Bug fortran/95397] [Fortran/OpenACC] Wrong results with 'loop vector' inside 'routine'

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95397

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2020-07-25
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |WAITING

--- Comment #1 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
Is it still failing? I cannot reproduce it.

[Bug fortran/95397] [Fortran/OpenACC] Wrong results with 'loop vector' inside 'routine'

[tschwinge at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95397

Thomas Schwinge <tschwinge at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW
   Last reconfirmed|2020-07-25 00:00:00         |2020-7-27

--- Comment #2 from Thomas Schwinge <tschwinge at gcc dot gnu.org> ---
(In reply to Dominique d'Humieres from comment #1)
> Is it still failing?

Yes, I still see this fail in the same way as originally reported.

> I cannot reproduce it.

:-| Hmm, strange.  What GPU hardware and CUDA version is that?

[Bug fortran/95397] [Fortran/OpenACC] Wrong results with 'loop vector' inside 'routine'

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95397

--- Comment #3 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
> :-| Hmm, strange.  What GPU hardware and CUDA version is that?

AMD Radeon Pro 5500M, hence no CUDA AFAIK. I also don't know what I should with
AMD.