Default fortran changed to +gcc8

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Default fortran changed to +gcc8

Mark Brethen-2
I followed the note below after a selfupdate:

NOTE: Default fortran changed to +gcc8; consider switching variants to enable
pre-built packages for mpich-default by running:
  "sudo port clean mpich-default && sudo port upgrade mpich-default +gcc8-gcc<N>” (should be ggc7)

However it resulted in:

Warning: Skipping upgrade since mpich-default 3.2.1_5 >= mpich-default 3.2.1_5, even though installed variants "+gcc7" do not match "+gcc8". Use 'upgrade --enforce-variants' to switch to the requested variants.
--->  Computing dependencies for gcc8
--->  Fetching archive for gcc8
--->  Attempting to fetch gcc8-8.2.0_3.darwin_16.x86_64.tbz2 from https://packages.macports.org/gcc8
--->  Attempting to fetch gcc8-8.2.0_3.darwin_16.x86_64.tbz2.rmd160 from https://packages.macports.org/gcc8
--->  Installing gcc8 @8.2.0_3
--->  Activating gcc8 @8.2.0_3
--->  Cleaning gcc8
--->  Updating database of binaries
--->  Updating database of C++ stdlib usage
--->  Scanning binaries for linking errors
--->  Found 33 broken files, matching files to ports    
--->  Found 2 broken ports, determining rebuild order
You can always run 'port rev-upgrade' again to fix errors.
The following ports will be rebuilt:
 py36-tables @3.4.4
 octave @4.4.1+accelerate+app+docs+gfortran+graphicsmagick+qt5+sound+sundials
Continue? [Y/n]:

an hour later it finished rebuilding those ports.

should I now issue port 'rev-upgrade' or 'upgrade --enforce-variants’ to rebuild mpich-default with gcc8?


Mark Brethen
[hidden email]



Reply | Threaded
Open this post in threaded view
|

Re: Default fortran changed to +gcc8

Ryan Schmidt-24


On Nov 14, 2018, at 18:53, Mark Brethen wrote:

> I followed the note below after a selfupdate:
>
> NOTE: Default fortran changed to +gcc8; consider switching variants to enable
> pre-built packages for mpich-default by running:
>  "sudo port clean mpich-default && sudo port upgrade mpich-default +gcc8-gcc<N>” (should be ggc7)
>
> However it resulted in:
>
> Warning: Skipping upgrade since mpich-default 3.2.1_5 >= mpich-default 3.2.1_5, even though installed variants "+gcc7" do not match "+gcc8". Use 'upgrade --enforce-variants' to switch to the requested variants.
> --->  Computing dependencies for gcc8
> --->  Fetching archive for gcc8
> --->  Attempting to fetch gcc8-8.2.0_3.darwin_16.x86_64.tbz2 from https://packages.macports.org/gcc8
> --->  Attempting to fetch gcc8-8.2.0_3.darwin_16.x86_64.tbz2.rmd160 from https://packages.macports.org/gcc8
> --->  Installing gcc8 @8.2.0_3
> --->  Activating gcc8 @8.2.0_3
> --->  Cleaning gcc8
> --->  Updating database of binaries
> --->  Updating database of C++ stdlib usage
> --->  Scanning binaries for linking errors
> --->  Found 33 broken files, matching files to ports    
> --->  Found 2 broken ports, determining rebuild order
> You can always run 'port rev-upgrade' again to fix errors.
> The following ports will be rebuilt:
> py36-tables @3.4.4
> octave @4.4.1+accelerate+app+docs+gfortran+graphicsmagick+qt5+sound+sundials
> Continue? [Y/n]:
>
> an hour later it finished rebuilding those ports.
>
> should I now issue port 'rev-upgrade'

MacPorts just ran rev-upgrade for you.

There's not enough information in the above output to know why py36-tables and octave were considered broken; we would need to have seen debug output (-d) to have made that determination.

But if MacPorts no longer shows these ports as broken, then the problem has been solved for your computer at least (and since the broken ports have been rebuilt and are no longer there, we can't make any further analysis of that the problem with them might have been).

> or 'upgrade --enforce-variants’ to rebuild mpich-default with gcc8?

If you want to switch mpich-default to that variant, yes. I'm not sure why the note the port displayed did not tell you to use "--enforce-variants"; that would seem to be necessary to achieve the intended result. Eric, Sean?

Reply | Threaded
Open this post in threaded view
|

Re: Default fortran changed to +gcc8

Mark Brethen-2
BTW, py36-tables also has those reinplace warnings.

--->  Extracting py36-tables
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5ARRAY-opt.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5ATTR.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5TB-opt.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5Zbzip2.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5Zlzo.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/idx-opt.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/typeconv.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/tables/_comp_bzip2.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/tables/_comp_lzo.c
Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/tables/lrucacheextension.c


Mark Brethen



On Nov 14, 2018, at 7:21 PM, Ryan Schmidt <[hidden email]> wrote:



On Nov 14, 2018, at 18:53, Mark Brethen wrote:

I followed the note below after a selfupdate:

NOTE: Default fortran changed to +gcc8; consider switching variants to enable
pre-built packages for mpich-default by running:
"sudo port clean mpich-default && sudo port upgrade mpich-default +gcc8-gcc<N>” (should be ggc7)

However it resulted in:

Warning: Skipping upgrade since mpich-default 3.2.1_5 >= mpich-default 3.2.1_5, even though installed variants "+gcc7" do not match "+gcc8". Use 'upgrade --enforce-variants' to switch to the requested variants.
--->  Computing dependencies for gcc8
--->  Fetching archive for gcc8
--->  Attempting to fetch gcc8-8.2.0_3.darwin_16.x86_64.tbz2 from https://packages.macports.org/gcc8
--->  Attempting to fetch gcc8-8.2.0_3.darwin_16.x86_64.tbz2.rmd160 from https://packages.macports.org/gcc8
--->  Installing gcc8 @8.2.0_3
--->  Activating gcc8 @8.2.0_3
--->  Cleaning gcc8
--->  Updating database of binaries
--->  Updating database of C++ stdlib usage
--->  Scanning binaries for linking errors
--->  Found 33 broken files, matching files to ports     
--->  Found 2 broken ports, determining rebuild order
You can always run 'port rev-upgrade' again to fix errors.
The following ports will be rebuilt:
py36-tables @3.4.4
octave @4.4.1+accelerate+app+docs+gfortran+graphicsmagick+qt5+sound+sundials
Continue? [Y/n]: 

an hour later it finished rebuilding those ports. 

should I now issue port 'rev-upgrade'

MacPorts just ran rev-upgrade for you.

There's not enough information in the above output to know why py36-tables and octave were considered broken; we would need to have seen debug output (-d) to have made that determination.

But if MacPorts no longer shows these ports as broken, then the problem has been solved for your computer at least (and since the broken ports have been rebuilt and are no longer there, we can't make any further analysis of that the problem with them might have been).

or 'upgrade --enforce-variants’ to rebuild mpich-default with gcc8?

If you want to switch mpich-default to that variant, yes. I'm not sure why the note the port displayed did not tell you to use "--enforce-variants"; that would seem to be necessary to achieve the intended result. Eric, Sean?

Reply | Threaded
Open this post in threaded view
|

Re: Default fortran changed to +gcc8

Ryan Schmidt-24


On Nov 14, 2018, at 19:33, Mark Brethen wrote:

> BTW, py36-tables also has those reinplace warnings.
>
> --->  Extracting py36-tables
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5ARRAY-opt.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5ATTR.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5TB-opt.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5Zbzip2.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/H5Zlzo.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/idx-opt.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/src/typeconv.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/tables/_comp_bzip2.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/tables/_comp_lzo.c
> Warning: reinplace s:utils.h:xxx_utils.h: didn't change anything in /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_python_py-tables/py36-tables/work/tables-3.4.4/tables/lrucacheextension.c

Thanks, I've fixed that now:

https://github.com/macports/macports-ports/commit/c5d42e383a79710059f27d6b558aeb396cc6cc8b

If you could file any other problems you notice as tickets in our issue tracker, that will let us assign them to the appropriate maintainers and track them properly.