oshmem/fortran: fix warning mesages

[karasevb]

fix warning:

pshmem_put_f.c: In function 'shmem_put_f':
pshmem_put_f.c:36:5: warning: passing argument 2 of 'mca_spml.spml_put' makes integer from pointer without a cast [enabled by default]
     MCA_SPML_CALL(put(FPTR_2_VOID_PTR(target),
     ^
pshmem_put_f.c:36:5: note: expected 'size_t' but argument is of type 'void *'
pshmem_put_f.c:36:5: warning: passing argument 3 of 'mca_spml.spml_put' makes pointer from integer
without a cast [enabled by default]
pshmem_put_f.c:36:5: note: expected 'void *' but argument is of type 'int'

Please review this PR

[jsquyres]

@karasevb Github pro tip: enclose multi-line verbatim blocks with 3 tick marks. E.g.:

pshmem_put_f.c: In function 'shmem_put_f':
pshmem_put_f.c:36:5: warning: passing argument 2 of 'mca_spml.spml_put' makes integer from pointer without a cast [enabled by default]
MCA_SPML_CALL(put(FPTR_2_VOID_PTR(target),
^
pshmem_put_f.c:36:5: note: expected 'size_t' but argument is of type 'void *'
pshmem_put_f.c:36:5: warning: passing argument 3 of 'mca_spml.spml_put' makes pointer from integer without a cast [enabled by default]
pshmem_put_f.c:36:5: note: expected 'void *' but argument is of type 'int'

If you're putting code in the verbatim block, you can even tell Github to do syntax hilighting -- see https://github.com/adam-p/markdown-here/wiki/Markdown-Cheatsheet#code-and-syntax-highlighting.

[artpol84]

@rhc54 please check

[karasevb]

Problem was introduced in #1867

[karasevb]

@jladd-mlnx @miked-mellanox @hpcraink please, review.

[lanl-ompi]

Test FAILed.

[jsquyres]

What is the *8 for?

[Di0gen]

bot:mellanox:retest

[Di0gen]

bot:mellanox:retest

[lanl-ompi]

Test FAILed.

[karasevb]

@jsquyres
shmem_put considers 64-bit values and the size calculation is performed transparently.
MCA_SPML_CALL assumes that you provide the size in bytes so you need to x8 here like in shmem_put64_f:
https://github.com/open-mpi/ompi/blob/master/oshmem/shmem/fortran/shmem_put64_f.c#L38

[lanl-ompi]

Test FAILed.

[hppritcha]

Is this targeting 2.0.1 release?

[artpol84]

Depends on whether #1867 was PR'd to release or not. I'll check later.

[jsquyres]

@karasevb Cool; thanks for the explanation.

[artpol84]

@hppritcha as it is for now #1867 wasn't cherry-picked to release:
https://github.com/open-mpi/ompi-release/blob/v2.x/oshmem/shmem/fortran/shmem_put_f.c#L35

I didn't found the PR pending. If we do want that @karasevb can take care of that for both PR's. I guess we want that for consistency. @jladd-mlnx @miked-mellanox what do you think?

[artpol84]

bot:mellanox:retest

[mike-dubman]

@artpol84 - agree, better have in sync for consistency.

[artpol84]

Ok @karasevb will PR both of them once this is merged.

[artpol84]

bot:mellanox:retest

[hppritcha]

checking the lanl bot
bot:lanl:retest

[artpol84]

bot:mellanox:retest

[artpol84]

Bot:retest

[karasevb]

@igor-ivanov please review

[igor-ivanov]

shmem_put() and shmem_get() are part of old API that is supported to avoid issue with old applications.
All old function can be found at https://github.com/open-mpi/ompi/blob/master/oshmem/include/shmem-compat.h

both functions should operate with long type.

Change looks reasonable. Please consider correct formatting (you need to use spaces instead of tab).

[artpol84]

@igor-ivanov thank you!
@karasevb fix tabs please!

[artpol84]

@karasevb thank you!
Please PR to release along with @hpcraink PR