Re: [OMPI users] Openmpi compilation errors

[Bruno Queiros]

Thank you for the help Luis

I did as you told me, i created a .pfg90.rc on the home directory of my
user with:

switch -pthread is replace(-lpthread) positional(linker)

i tried to compile openmpi with make, but i still got the same error.

Am i doing something wrong? How can i overcome this -pthread error?


Em qui, 28 de mai de 2015 às 18:36, Luis Kornblueh <
luis.kornbl...@mpimet.mpg.de> escreveu:

> Hi Bruno,
>
> once more: it might be the case that the file needs to be named .pgf90rc
>
> Sorry,
> Luis
>
> On 28/05/15 19:08, Bruno Queiros wrote:
> > Thomas
> >
> > I did get a configure successfull, but make fails with
> >
> > pgf90-Error-Unknown switch: -pthread
> > make[2]: *** [libmpi_usempi.la ] Error 1
> >
> > I have tried and did a fake pgf90 has mentioned on the mailing list
> >
> > #! /bin/bash
> > newargs='echo $@ | sed s/-pthread//g -'
> > #echo "/opt/pgi/linux86-64/10.4/bin/pgf90  $newargs"
> > /opt/pgi/linux86-64/10.4/bin/pgf90  $newargs
> > exit
> >
> > But for some reason this doesn't work, the make just stops.
> >
> > Can you point me in the right direction?
> >
> > Em qui, 28 de mai de 2015 às 14:10, Bruno Queiros  > > escreveu:
> >
> > Thank you very much Thomas
> >
> > I will try this
> >
> > Em qui, 28 de mai de 2015 às 12:25, Thomas Jahns  > > escreveu:
> >
> > On 05/28/15 12:20, Jeff Squyres (jsquyres) wrote:
> >  > Sounds like your pgcc compiler installation is busted.
> > You'll need to get that fixed to compile/install Open MPI.
> >
> > An alternative might be to use a system compiler (which is
> > probably gcc or
> > clang) for C and C++ and only use pgfortran for the Fortran
> > parts of OpenMPI,
> > i.e. try something like
> >
> > ./configure CC=gcc CXX=g++ FC=pgf90 F77=pgf77 F90=pgf90
> > --prefix=/opt/openmpi-1.8.5_pgf90
> >
> > Regards, Thomas Jahns
> >
> >
> > ___
> > users mailing list
> > us...@open-mpi.org 
> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> > Link to this post:
> > http://www.open-mpi.org/community/lists/users/2015/05/26963.php
> >
> >
> >
> > ___
> > users mailing list
> > us...@open-mpi.org
> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> > Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26967.php
> >
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26969.php
>

Re: [OMPI users] Openmpi compilation errors

[Jeff Squyres (jsquyres)]

Just to clarify: the underlying issue here is that Open MPI's configure script 
is assuming that whatever linker flags it discovers for the C compiler are also 
valid for the Fortran compiler.  This is apparently not true when you mix the 
GCC and Portland compilers -- bummer (i.e., Open MPI's configure script is 
determining that -pthread works with gcc and assumes that it works with 
pgf90... but it apparently doesn't, which is why Luis's solution is 
particularly clever).

If the portland-config-script-in-your-$HOME solution isn't work, I'd double 
check that you have the file named correctly (e.g., I thought the Portland 
Fortran compiler's name was pgf90, not pfg90 -- I would assume that the config 
filename is representative of the compiler's executable name...?), that it is 
actually being seen by the Portland compiler, etc.

You can easily test whether this Portland config trick is working: just invoke 
"pgf90 -pthread somefile.f90" and see if it compiles that somefile.f90 file or 
if it complains about the -pthread option.  If the Portland config trick is 
working, (assuming somefile.f90 is a valid Fortran file), it'll compile without 
complaining about -pthread. The Open MPI will likely be able to build, too.


> On May 29, 2015, at 6:14 AM, Bruno Queiros  wrote:
> 
> Thank you for the help Luis
> 
> I did as you told me, i created a .pfg90.rc on the home directory of my user 
> with:
> 
> switch -pthread is replace(-lpthread) positional(linker)
> 
> i tried to compile openmpi with make, but i still got the same error.
> 
> Am i doing something wrong? How can i overcome this -pthread error?
> 
> 
> Em qui, 28 de mai de 2015 às 18:36, Luis Kornblueh 
>  escreveu:
> Hi Bruno,
> 
> once more: it might be the case that the file needs to be named .pgf90rc
> 
> Sorry,
> Luis
> 
> On 28/05/15 19:08, Bruno Queiros wrote:
> > Thomas
> >
> > I did get a configure successfull, but make fails with
> >
> > pgf90-Error-Unknown switch: -pthread
> > make[2]: *** [libmpi_usempi.la ] Error 1
> >
> > I have tried and did a fake pgf90 has mentioned on the mailing list
> >
> > #! /bin/bash
> > newargs='echo $@ | sed s/-pthread//g -'
> > #echo "/opt/pgi/linux86-64/10.4/bin/pgf90  $newargs"
> > /opt/pgi/linux86-64/10.4/bin/pgf90  $newargs
> > exit
> >
> > But for some reason this doesn't work, the make just stops.
> >
> > Can you point me in the right direction?
> >
> > Em qui, 28 de mai de 2015 às 14:10, Bruno Queiros  > > escreveu:
> >
> > Thank you very much Thomas
> >
> > I will try this
> >
> > Em qui, 28 de mai de 2015 às 12:25, Thomas Jahns  > > escreveu:
> >
> > On 05/28/15 12:20, Jeff Squyres (jsquyres) wrote:
> >  > Sounds like your pgcc compiler installation is busted.
> > You'll need to get that fixed to compile/install Open MPI.
> >
> > An alternative might be to use a system compiler (which is
> > probably gcc or
> > clang) for C and C++ and only use pgfortran for the Fortran
> > parts of OpenMPI,
> > i.e. try something like
> >
> > ./configure CC=gcc CXX=g++ FC=pgf90 F77=pgf77 F90=pgf90
> > --prefix=/opt/openmpi-1.8.5_pgf90
> >
> > Regards, Thomas Jahns
> >
> >
> > ___
> > users mailing list
> > us...@open-mpi.org 
> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> > Link to this post:
> > http://www.open-mpi.org/community/lists/users/2015/05/26963.php
> >
> >
> >
> > ___
> > users mailing list
> > us...@open-mpi.org
> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> > Link to this post: 
> > http://www.open-mpi.org/community/lists/users/2015/05/26967.php
> >
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/05/26969.php
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/05/26971.php


-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to: 
http://www.cisco.com/web/about/doing_business/legal/cri/

Re: [OMPI users] Openmpi compilation errors

[Bruno Queiros]

Hello Jeff

I understand that using Portland compiler isn't "advised" by Open Mpi, i
was just wondering if there's a way of doing it, since i need Open Mpi
compiled with PG fortran and not gfortran for example.

The name of the binary is correct: pgf90 the name of the file is also
correct .pgf90.rc i do have some doubts about the content of the file. Is
this enough?

switch -pthread is replace(-lpthread) positional(linker)

If i do a source .pgf90.rc i do get errors:

-bash: ./.pgf90.rc: line 1: syntax error near unexpected token `('
-bash: ./.pgf90.rc: line 1: `switch -pthread is replace(-lpthread)
positional(linker)'

Besides that i tried what you mentioned: pgf90 -pthread somefile.f90 , and
that gives the expected error:

pgf90 -pthread helloworld.f90
pgf90-Error-Unknown switch: -pthread


My real question is, can i change Open Mpi configuration, to not use
-pthread flag on fortran compiler?



Em sex, 29 de mai de 2015 às 11:36, Jeff Squyres (jsquyres) <
jsquy...@cisco.com> escreveu:

> Just to clarify: the underlying issue here is that Open MPI's configure
> script is assuming that whatever linker flags it discovers for the C
> compiler are also valid for the Fortran compiler.  This is apparently not
> true when you mix the GCC and Portland compilers -- bummer (i.e., Open
> MPI's configure script is determining that -pthread works with gcc and
> assumes that it works with pgf90... but it apparently doesn't, which is why
> Luis's solution is particularly clever).
>
> If the portland-config-script-in-your-$HOME solution isn't work, I'd
> double check that you have the file named correctly (e.g., I thought the
> Portland Fortran compiler's name was pgf90, not pfg90 -- I would assume
> that the config filename is representative of the compiler's executable
> name...?), that it is actually being seen by the Portland compiler, etc.
>
> You can easily test whether this Portland config trick is working: just
> invoke "pgf90 -pthread somefile.f90" and see if it compiles that
> somefile.f90 file or if it complains about the -pthread option.  If the
> Portland config trick is working, (assuming somefile.f90 is a valid Fortran
> file), it'll compile without complaining about -pthread. The Open MPI will
> likely be able to build, too.
>
>
> > On May 29, 2015, at 6:14 AM, Bruno Queiros  wrote:
> >
> > Thank you for the help Luis
> >
> > I did as you told me, i created a .pfg90.rc on the home directory of my
> user with:
> >
> > switch -pthread is replace(-lpthread) positional(linker)
> >
> > i tried to compile openmpi with make, but i still got the same error.
> >
> > Am i doing something wrong? How can i overcome this -pthread error?
> >
> >
> > Em qui, 28 de mai de 2015 às 18:36, Luis Kornblueh <
> luis.kornbl...@mpimet.mpg.de> escreveu:
> > Hi Bruno,
> >
> > once more: it might be the case that the file needs to be named .pgf90rc
> >
> > Sorry,
> > Luis
> >
> > On 28/05/15 19:08, Bruno Queiros wrote:
> > > Thomas
> > >
> > > I did get a configure successfull, but make fails with
> > >
> > > pgf90-Error-Unknown switch: -pthread
> > > make[2]: *** [libmpi_usempi.la ] Error 1
> > >
> > > I have tried and did a fake pgf90 has mentioned on the mailing list
> > >
> > > #! /bin/bash
> > > newargs='echo $@ | sed s/-pthread//g -'
> > > #echo "/opt/pgi/linux86-64/10.4/bin/pgf90  $newargs"
> > > /opt/pgi/linux86-64/10.4/bin/pgf90  $newargs
> > > exit
> > >
> > > But for some reason this doesn't work, the make just stops.
> > >
> > > Can you point me in the right direction?
> > >
> > > Em qui, 28 de mai de 2015 às 14:10, Bruno Queiros  > > > escreveu:
> > >
> > > Thank you very much Thomas
> > >
> > > I will try this
> > >
> > > Em qui, 28 de mai de 2015 às 12:25, Thomas Jahns  > > > escreveu:
> > >
> > > On 05/28/15 12:20, Jeff Squyres (jsquyres) wrote:
> > >  > Sounds like your pgcc compiler installation is busted.
> > > You'll need to get that fixed to compile/install Open MPI.
> > >
> > > An alternative might be to use a system compiler (which is
> > > probably gcc or
> > > clang) for C and C++ and only use pgfortran for the Fortran
> > > parts of OpenMPI,
> > > i.e. try something like
> > >
> > > ./configure CC=gcc CXX=g++ FC=pgf90 F77=pgf77 F90=pgf90
> > > --prefix=/opt/openmpi-1.8.5_pgf90
> > >
> > > Regards, Thomas Jahns
> > >
> > >
> > > ___
> > > users mailing list
> > > us...@open-mpi.org 
> > > Subscription:
> http://www.open-mpi.org/mailman/listinfo.cgi/users
> > > Link to this post:
> > >
> http://www.open-mpi.org/community/lists/users/2015/05/26963.php
> > >
> > >
> > >
> > > ___
> > > users mailing list
> > > us...@open-mpi.org
> > > Subscription: http://www.open-mpi.org/mailman/li

Re: [OMPI users] Openmpi compilation errors

[Jeff Squyres (jsquyres)]

On May 29, 2015, at 6:54 AM, Bruno Queiros  wrote:
> 
> I understand that using Portland compiler isn't "advised" by Open Mpi, i was 
> just wondering if there's a way of doing it, since i need Open Mpi compiled 
> with PG fortran and not gfortran for example.

A further clarification: the Portland compiler is a fine compiler; we certainly 
support it.  The only thing we recommend is avoiding mixing compiler suites 
whenever possible (e.g., using one compiler for C and another for Fortran).  It 
leads to complications like this.

That being said, it looks like you have a Portland Fortran license, and no 
corresponding Portland C license, and therefore you're somewhat forced into 
this situation.

> The name of the binary is correct: pgf90 the name of the file is also correct 
> .pgf90.rc i do have some doubts about the content of the file. Is this enough?
> 
> switch -pthread is replace(-lpthread) positional(linker)

I'm not a Portland customer -- I don't know.  You'll need to check their 
documentation.

> If i do a source .pgf90.rc i do get errors:
> 
> -bash: ./.pgf90.rc: line 1: syntax error near unexpected token `('
> -bash: ./.pgf90.rc: line 1: `switch -pthread is replace(-lpthread) 
> positional(linker)'

I'm guessing that this file is not intended to be sourced by the shell, but 
rather noticed and read/used by the pgf90 compiler when it is invoked.

Sidenote: isn't there a pgfortran compiler executable that is supposed to be 
preferred over "pgf90" these days?  (remember my disclaimer: I'm not a Portland 
customer, so I could be totally off base here...)  Have you tried pgfortran to 
see if it accepts the -pthread option?  Sometimes the different compiler 
executable entry points behave slightly differently...

> Besides that i tried what you mentioned: pgf90 -pthread somefile.f90 , and 
> that gives the expected error:
> 
> pgf90 -pthread helloworld.f90 
> pgf90-Error-Unknown switch: -pthread

Sounds like this file is somehow not yet correct, or you have a version of the 
Portland compiler that does not use this config file.  You should check the 
Portland Fortran compiler documentation.

If you can't get this Portland config file workaround to work, you can also 
make your own "my_fortran_compiler" script that simply strips out the -pthread 
argv token and then invokes pgf90 (or pgfortran) with the rest of the argv.  
Then you can:

./configure FC=my_fortran_compiler ...

I.e., you're just intercepting Open MPI's invocation of the Fortran compiler, 
removing the CLI option that pgf90(pgfortran) won't understand, and then 
invoking pgf90(pgfortran).  Make sense?

> My real question is, can i change Open Mpi configuration, to not use -pthread 
> flag on fortran compiler?

I'm sorry, no -- Open MPI requires threading support.

-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to: 
http://www.cisco.com/web/about/doing_business/legal/cri/

Re: [OMPI users] Openmpi compilation errors

[Luis Kornblueh]

Hi,

Bruno is using 10.4 as I ready only a second ago in one of his emails. I 
have no idea if the compiler resource file trick works with this old one.


By the way it has to be .mypgf90rc not .pgf90rc at least following the 
current documentation.


Cheerio,
Luis

On 29/05/15 13:07, Jeff Squyres (jsquyres) wrote:

On May 29, 2015, at 6:54 AM, Bruno Queiros  wrote:


I understand that using Portland compiler isn't "advised" by Open Mpi, i was 
just wondering if there's a way of doing it, since i need Open Mpi compiled with PG 
fortran and not gfortran for example.


A further clarification: the Portland compiler is a fine compiler; we certainly 
support it.  The only thing we recommend is avoiding mixing compiler suites 
whenever possible (e.g., using one compiler for C and another for Fortran).  It 
leads to complications like this.

That being said, it looks like you have a Portland Fortran license, and no 
corresponding Portland C license, and therefore you're somewhat forced into 
this situation.


The name of the binary is correct: pgf90 the name of the file is also correct 
.pgf90.rc i do have some doubts about the content of the file. Is this enough?

switch -pthread is replace(-lpthread) positional(linker)


I'm not a Portland customer -- I don't know.  You'll need to check their 
documentation.


If i do a source .pgf90.rc i do get errors:

-bash: ./.pgf90.rc: line 1: syntax error near unexpected token `('
-bash: ./.pgf90.rc: line 1: `switch -pthread is replace(-lpthread) 
positional(linker)'


I'm guessing that this file is not intended to be sourced by the shell, but 
rather noticed and read/used by the pgf90 compiler when it is invoked.

Sidenote: isn't there a pgfortran compiler executable that is supposed to be preferred 
over "pgf90" these days?  (remember my disclaimer: I'm not a Portland customer, 
so I could be totally off base here...)  Have you tried pgfortran to see if it accepts 
the -pthread option?  Sometimes the different compiler executable entry points behave 
slightly differently...


Besides that i tried what you mentioned: pgf90 -pthread somefile.f90 , and that 
gives the expected error:

pgf90 -pthread helloworld.f90
pgf90-Error-Unknown switch: -pthread


Sounds like this file is somehow not yet correct, or you have a version of the 
Portland compiler that does not use this config file.  You should check the 
Portland Fortran compiler documentation.

If you can't get this Portland config file workaround to work, you can also make your own 
"my_fortran_compiler" script that simply strips out the -pthread argv token and 
then invokes pgf90 (or pgfortran) with the rest of the argv.  Then you can:

./configure FC=my_fortran_compiler ...

I.e., you're just intercepting Open MPI's invocation of the Fortran compiler, 
removing the CLI option that pgf90(pgfortran) won't understand, and then 
invoking pgf90(pgfortran).  Make sense?


My real question is, can i change Open Mpi configuration, to not use -pthread 
flag on fortran compiler?


I'm sorry, no -- Open MPI requires threading support.

[OMPI users] Building OpenMPI on Raspberry Pi 2

[Jeff Layton]

Good morning,

I'm building OpenMPI from source on a Raspberry Pi 2 and
I've hit an error. The error is:

make[2]: Entering directory '/work/pi/src/openmpi-1.8.5/opal/asm'
  CPPASatomic-asm.lo
atomic-asm.S: Assembler messages:
atomic-asm.S:7: Error: selected processor does not support ARM mode `dmb'
atomic-asm.S:15: Error: selected processor does not support ARM mode `dmb'
atomic-asm.S:23: Error: selected processor does not support ARM mode `dmb'
atomic-asm.S:55: Error: selected processor does not support ARM mode `dmb'
atomic-asm.S:70: Error: selected processor does not support ARM mode `dmb'
atomic-asm.S:86: Error: selected processor does not support ARM mode 
`ldrexd r4,r5,[r0]'
atomic-asm.S:91: Error: selected processor does not support ARM mode 
`strexd r1,r6,r7,[r0]'
atomic-asm.S:107: Error: selected processor does not support ARM mode 
`ldrexd r4,r5,[r0]'
atomic-asm.S:112: Error: selected processor does not support ARM mode 
`strexd r1,r6,r7,[r0]'

atomic-asm.S:115: Error: selected processor does not support ARM mode `dmb'
atomic-asm.S:130: Error: selected processor does not support ARM mode 
`ldrexd r4,r5,[r0]'

atomic-asm.S:135: Error: selected processor does not support ARM mode `dmb'
atomic-asm.S:136: Error: selected processor does not support ARM mode 
`strexd r1,r6,r7,[r0]'

Makefile:1608: recipe for target 'atomic-asm.lo' failed
make[2]: *** [atomic-asm.lo] Error 1
make[2]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal/asm'
Makefile:2149: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
Makefile:1698: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1


I was doing some googling and I saw where I need to specify
CCFLAGS when I run configure but I want to make 100% sure
I have the right arguments. Can anyone help?

Thanks!

Jeff

P.S. I'd rather build OpenMPI from source rather than use the pre-built
packages in the Raspbian repository.

Re: [OMPI users] Building OpenMPI on Raspberry Pi 2

[Gilles Gouaillardet]

Jeff,

shall I assume you made a typo and wrote CCFLAGS instead of CFLAGS ?

also, can you double check the flags are correctly passed to the assembler
with
cd opal/asm
make -n atomic-asm.lo

Cheers,

Gilles

On Friday, May 29, 2015, Jeff Layton  wrote:

> Good morning,
>
> I'm building OpenMPI from source on a Raspberry Pi 2 and
> I've hit an error. The error is:
>
> make[2]: Entering directory '/work/pi/src/openmpi-1.8.5/opal/asm'
>   CPPASatomic-asm.lo
> atomic-asm.S: Assembler messages:
> atomic-asm.S:7: Error: selected processor does not support ARM mode `dmb'
> atomic-asm.S:15: Error: selected processor does not support ARM mode `dmb'
> atomic-asm.S:23: Error: selected processor does not support ARM mode `dmb'
> atomic-asm.S:55: Error: selected processor does not support ARM mode `dmb'
> atomic-asm.S:70: Error: selected processor does not support ARM mode `dmb'
> atomic-asm.S:86: Error: selected processor does not support ARM mode
> `ldrexd r4,r5,[r0]'
> atomic-asm.S:91: Error: selected processor does not support ARM mode
> `strexd r1,r6,r7,[r0]'
> atomic-asm.S:107: Error: selected processor does not support ARM mode
> `ldrexd r4,r5,[r0]'
> atomic-asm.S:112: Error: selected processor does not support ARM mode
> `strexd r1,r6,r7,[r0]'
> atomic-asm.S:115: Error: selected processor does not support ARM mode `dmb'
> atomic-asm.S:130: Error: selected processor does not support ARM mode
> `ldrexd r4,r5,[r0]'
> atomic-asm.S:135: Error: selected processor does not support ARM mode `dmb'
> atomic-asm.S:136: Error: selected processor does not support ARM mode
> `strexd r1,r6,r7,[r0]'
> Makefile:1608: recipe for target 'atomic-asm.lo' failed
> make[2]: *** [atomic-asm.lo] Error 1
> make[2]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal/asm'
> Makefile:2149: recipe for target 'all-recursive' failed
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
> Makefile:1698: recipe for target 'all-recursive' failed
> make: *** [all-recursive] Error 1
>
>
> I was doing some googling and I saw where I need to specify
> CCFLAGS when I run configure but I want to make 100% sure
> I have the right arguments. Can anyone help?
>
> Thanks!
>
> Jeff
>
> P.S. I'd rather build OpenMPI from source rather than use the pre-built
> packages in the Raspbian repository.
>
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26976.php
>

Re: [OMPI users] Building OpenMPI on Raspberry Pi 2

[Jeff Layton]

Gilles,

oops - yes, CFLAGS. But I also saw this posting:

https://www.open-mpi.org/community/lists/users/2013/01/2.php

where CCASFLAGS is used (I assume because for asm). I'm trying
this flag when I configure Open MPI.

Thanks!

Jeff



Jeff,

shall I assume you made a typo and wrote CCFLAGS instead of CFLAGS ?

also, can you double check the flags are correctly passed to the 
assembler with

cd opal/asm
make -n atomic-asm.lo

Cheers,

Gilles

On Friday, May 29, 2015, Jeff Layton > wrote:


Good morning,

I'm building OpenMPI from source on a Raspberry Pi 2 and
I've hit an error. The error is:

make[2]: Entering directory '/work/pi/src/openmpi-1.8.5/opal/asm'
  CPPASatomic-asm.lo
atomic-asm.S: Assembler messages:
atomic-asm.S:7: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:15: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:23: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:55: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:70: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:86: Error: selected processor does not support ARM
mode `ldrexd r4,r5,[r0]'
atomic-asm.S:91: Error: selected processor does not support ARM
mode `strexd r1,r6,r7,[r0]'
atomic-asm.S:107: Error: selected processor does not support ARM
mode `ldrexd r4,r5,[r0]'
atomic-asm.S:112: Error: selected processor does not support ARM
mode `strexd r1,r6,r7,[r0]'
atomic-asm.S:115: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:130: Error: selected processor does not support ARM
mode `ldrexd r4,r5,[r0]'
atomic-asm.S:135: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:136: Error: selected processor does not support ARM
mode `strexd r1,r6,r7,[r0]'
Makefile:1608: recipe for target 'atomic-asm.lo' failed
make[2]: *** [atomic-asm.lo] Error 1
make[2]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal/asm'
Makefile:2149: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
Makefile:1698: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1


I was doing some googling and I saw where I need to specify
CCFLAGS when I run configure but I want to make 100% sure
I have the right arguments. Can anyone help?

Thanks!

Jeff

P.S. I'd rather build OpenMPI from source rather than use the
pre-built
packages in the Raspbian repository.


___
users mailing list
us...@open-mpi.org
Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
Link to this post:
http://www.open-mpi.org/community/lists/users/2015/05/26976.php



___
users mailing list
us...@open-mpi.org
Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
Link to this post: 
http://www.open-mpi.org/community/lists/users/2015/05/26977.php

Re: [OMPI users] Openmpi compilation errors

[Bruno Queiros]

Thank you Jeff for the clarification and help.

Em sex, 29 de mai de 2015 às 12:08, Jeff Squyres (jsquyres) <
jsquy...@cisco.com> escreveu:

> On May 29, 2015, at 6:54 AM, Bruno Queiros  wrote:
> >
> > I understand that using Portland compiler isn't "advised" by Open Mpi, i
> was just wondering if there's a way of doing it, since i need Open Mpi
> compiled with PG fortran and not gfortran for example.
>
> A further clarification: the Portland compiler is a fine compiler; we
> certainly support it.  The only thing we recommend is avoiding mixing
> compiler suites whenever possible (e.g., using one compiler for C and
> another for Fortran).  It leads to complications like this.
>
> That being said, it looks like you have a Portland Fortran license, and no
> corresponding Portland C license, and therefore you're somewhat forced into
> this situation.
>
> > The name of the binary is correct: pgf90 the name of the file is also
> correct .pgf90.rc i do have some doubts about the content of the file. Is
> this enough?
> >
> > switch -pthread is replace(-lpthread) positional(linker)
>
> I'm not a Portland customer -- I don't know.  You'll need to check their
> documentation.
>
> > If i do a source .pgf90.rc i do get errors:
> >
> > -bash: ./.pgf90.rc: line 1: syntax error near unexpected token `('
> > -bash: ./.pgf90.rc: line 1: `switch -pthread is replace(-lpthread)
> positional(linker)'
>
> I'm guessing that this file is not intended to be sourced by the shell,
> but rather noticed and read/used by the pgf90 compiler when it is invoked.
>
> Sidenote: isn't there a pgfortran compiler executable that is supposed to
> be preferred over "pgf90" these days?  (remember my disclaimer: I'm not a
> Portland customer, so I could be totally off base here...)  Have you tried
> pgfortran to see if it accepts the -pthread option?  Sometimes the
> different compiler executable entry points behave slightly differently...
>
> > Besides that i tried what you mentioned: pgf90 -pthread somefile.f90 ,
> and that gives the expected error:
> >
> > pgf90 -pthread helloworld.f90
> > pgf90-Error-Unknown switch: -pthread
>
> Sounds like this file is somehow not yet correct, or you have a version of
> the Portland compiler that does not use this config file.  You should check
> the Portland Fortran compiler documentation.
>
> If you can't get this Portland config file workaround to work, you can
> also make your own "my_fortran_compiler" script that simply strips out the
> -pthread argv token and then invokes pgf90 (or pgfortran) with the rest of
> the argv.  Then you can:
>
> ./configure FC=my_fortran_compiler ...
>
> I.e., you're just intercepting Open MPI's invocation of the Fortran
> compiler, removing the CLI option that pgf90(pgfortran) won't understand,
> and then invoking pgf90(pgfortran).  Make sense?
>
> > My real question is, can i change Open Mpi configuration, to not use
> -pthread flag on fortran compiler?
>
> I'm sorry, no -- Open MPI requires threading support.
>
> --
> Jeff Squyres
> jsquy...@cisco.com
> For corporate legal information go to:
> http://www.cisco.com/web/about/doing_business/legal/cri/
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26974.php
>

Re: [OMPI users] Openmpi compilation errors

[Bruno Queiros]

Luis

Can you point me to that documentation you're reading?

Em sex, 29 de mai de 2015 às 13:04, Luis Kornblueh <
luis.kornbl...@mpimet.mpg.de> escreveu:

> Hi,
>
> Bruno is using 10.4 as I ready only a second ago in one of his emails. I
> have no idea if the compiler resource file trick works with this old one.
>
> By the way it has to be .mypgf90rc not .pgf90rc at least following the
> current documentation.
>
> Cheerio,
> Luis
>
> On 29/05/15 13:07, Jeff Squyres (jsquyres) wrote:
> > On May 29, 2015, at 6:54 AM, Bruno Queiros  wrote:
> >>
> >> I understand that using Portland compiler isn't "advised" by Open Mpi,
> i was just wondering if there's a way of doing it, since i need Open Mpi
> compiled with PG fortran and not gfortran for example.
> >
> > A further clarification: the Portland compiler is a fine compiler; we
> certainly support it.  The only thing we recommend is avoiding mixing
> compiler suites whenever possible (e.g., using one compiler for C and
> another for Fortran).  It leads to complications like this.
> >
> > That being said, it looks like you have a Portland Fortran license, and
> no corresponding Portland C license, and therefore you're somewhat forced
> into this situation.
> >
> >> The name of the binary is correct: pgf90 the name of the file is also
> correct .pgf90.rc i do have some doubts about the content of the file. Is
> this enough?
> >>
> >> switch -pthread is replace(-lpthread) positional(linker)
> >
> > I'm not a Portland customer -- I don't know.  You'll need to check their
> documentation.
> >
> >> If i do a source .pgf90.rc i do get errors:
> >>
> >> -bash: ./.pgf90.rc: line 1: syntax error near unexpected token `('
> >> -bash: ./.pgf90.rc: line 1: `switch -pthread is replace(-lpthread)
> positional(linker)'
> >
> > I'm guessing that this file is not intended to be sourced by the shell,
> but rather noticed and read/used by the pgf90 compiler when it is invoked.
> >
> > Sidenote: isn't there a pgfortran compiler executable that is supposed
> to be preferred over "pgf90" these days?  (remember my disclaimer: I'm not
> a Portland customer, so I could be totally off base here...)  Have you
> tried pgfortran to see if it accepts the -pthread option?  Sometimes the
> different compiler executable entry points behave slightly differently...
> >
> >> Besides that i tried what you mentioned: pgf90 -pthread somefile.f90 ,
> and that gives the expected error:
> >>
> >> pgf90 -pthread helloworld.f90
> >> pgf90-Error-Unknown switch: -pthread
> >
> > Sounds like this file is somehow not yet correct, or you have a version
> of the Portland compiler that does not use this config file.  You should
> check the Portland Fortran compiler documentation.
> >
> > If you can't get this Portland config file workaround to work, you can
> also make your own "my_fortran_compiler" script that simply strips out the
> -pthread argv token and then invokes pgf90 (or pgfortran) with the rest of
> the argv.  Then you can:
> >
> > ./configure FC=my_fortran_compiler ...
> >
> > I.e., you're just intercepting Open MPI's invocation of the Fortran
> compiler, removing the CLI option that pgf90(pgfortran) won't understand,
> and then invoking pgf90(pgfortran).  Make sense?
> >
> >> My real question is, can i change Open Mpi configuration, to not use
> -pthread flag on fortran compiler?
> >
> > I'm sorry, no -- Open MPI requires threading support.
> >
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26975.php
>

Re: [OMPI users] Building OpenMPI on Raspberry Pi 2

[Jeff Layton]

On 05/29/2015 09:35 AM, Jeff Layton wrote:

Gilles,

oops - yes, CFLAGS. But I also saw this posting:

https://www.open-mpi.org/community/lists/users/2013/01/2.php

where CCASFLAGS is used (I assume because for asm). I'm trying
this flag when I configure Open MPI.


I tried using the CCASFLAGS flag from the above link and it didn't work. 
The error

now reads:

Making all in mca/memory/linux
make[2]: Entering directory 
'/work/pi/src/openmpi-1.8.5/opal/mca/memory/linux'

  CC   memory_linux_component.lo
  CC   memory_linux_ptmalloc2.lo
  CC   memory_linux_munmap.lo
  CC   malloc.lo
/tmp/cc7g4mWi.s: Assembler messages:
/tmp/cc7g4mWi.s:948: Error: selected processor does not support ARM mode 
`dmb'

Makefile:1694: recipe for target 'malloc.lo' failed
make[2]: *** [malloc.lo] Error 1
make[2]: Leaving directory 
'/work/pi/src/openmpi-1.8.5/opal/mca/memory/linux'

Makefile:2149: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
Makefile:1698: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1


I used the configure flag CCASFLAGS=-march=armv7-a

Not sure if that is correct or not. I'm guessing I'm using the wrong
architecture for the Pi 2. Suggestions?

Thanks!

Jeff




Thanks!

Jeff



Jeff,

shall I assume you made a typo and wrote CCFLAGS instead of CFLAGS ?

also, can you double check the flags are correctly passed to the 
assembler with

cd opal/asm
make -n atomic-asm.lo

Cheers,

Gilles

On Friday, May 29, 2015, Jeff Layton > wrote:


Good morning,

I'm building OpenMPI from source on a Raspberry Pi 2 and
I've hit an error. The error is:

make[2]: Entering directory '/work/pi/src/openmpi-1.8.5/opal/asm'
  CPPASatomic-asm.lo
atomic-asm.S: Assembler messages:
atomic-asm.S:7: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:15: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:23: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:55: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:70: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:86: Error: selected processor does not support ARM
mode `ldrexd r4,r5,[r0]'
atomic-asm.S:91: Error: selected processor does not support ARM
mode `strexd r1,r6,r7,[r0]'
atomic-asm.S:107: Error: selected processor does not support ARM
mode `ldrexd r4,r5,[r0]'
atomic-asm.S:112: Error: selected processor does not support ARM
mode `strexd r1,r6,r7,[r0]'
atomic-asm.S:115: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:130: Error: selected processor does not support ARM
mode `ldrexd r4,r5,[r0]'
atomic-asm.S:135: Error: selected processor does not support ARM
mode `dmb'
atomic-asm.S:136: Error: selected processor does not support ARM
mode `strexd r1,r6,r7,[r0]'
Makefile:1608: recipe for target 'atomic-asm.lo' failed
make[2]: *** [atomic-asm.lo] Error 1
make[2]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal/asm'
Makefile:2149: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
Makefile:1698: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1


I was doing some googling and I saw where I need to specify
CCFLAGS when I run configure but I want to make 100% sure
I have the right arguments. Can anyone help?

Thanks!

Jeff

P.S. I'd rather build OpenMPI from source rather than use the
pre-built
packages in the Raspbian repository.

Re: [OMPI users] Openmpi compilation errors

[Timothy Brown]

> On May 29, 2015, at 5:07 AM, Jeff Squyres (jsquyres)  
> wrote:
> 
> On May 29, 2015, at 6:54 AM, Bruno Queiros  wrote:
> 
>> The name of the binary is correct: pgf90 the name of the file is also 
>> correct .pgf90.rc i do have some doubts about the content of the file. Is 
>> this enough?
>> 
>> switch -pthread is replace(-lpthread) positional(linker)
> 
> I'm not a Portland customer -- I don't know.  You'll need to check their 
> documentation.
> 

Here I have a siterc file within the PGI bin directory, for example:

/some/long/path/pgi/15.3/bin/siterc

I have exactly the same line as you have specified. 

If you are unable to put it in the PGI installation bin directory you can put 
it in a file ${HOME}/.mypgf90rc, as is described in section 1.8.2 (page 14) of 
the PGI Compilers Users Guide ( http://www.pgroup.com/doc/pgiug.pdf ).
 

>> If i do a source .pgf90.rc i do get errors:
>> 
>> -bash: ./.pgf90.rc: line 1: syntax error near unexpected token `('
>> -bash: ./.pgf90.rc: line 1: `switch -pthread is replace(-lpthread) 
>> positional(linker)'
> 
> I'm guessing that this file is not intended to be sourced by the shell, but 
> rather noticed and read/used by the pgf90 compiler when it is invoked.
> 

Jeff, your right. It's not for your shell to source it'd for the compiler to 
read.


> Sidenote: isn't there a pgfortran compiler executable that is supposed to be 
> preferred over "pgf90" these days?  (remember my disclaimer: I'm not a 
> Portland customer, so I could be totally off base here...)  Have you tried 
> pgfortran to see if it accepts the -pthread option?  Sometimes the different 
> compiler executable entry points behave slightly differently...
> 

I've built Openmpi 1.8.5 with the following configure line:

./configure  \
  --prefix=/curc/tools/x86_64/rh6/software/openmpi/1.8.5/pgi/15.3 \
  --with-threads=posix \
  --enable-mpi-thread-multiple \
  --with-slurm \
  --with-pmi=/curc/slurm/slurm/current/

Please note, I am using the following environment variables:
CC=pgcc
FC=pgfortran
F90=pgf90
F77=pgf77
CXX=pgc++

I do not use pgprepro for CPP as I found it to be flaky.

Hope this helps.
Timothy

[OMPI users] How can I discover valid values for MCA parameters?

[Blosch, Edwin L]

Sometimes I want to use one of the option flags, for example today it is 
mtl_mxm_verbose.  How do I discover the valid possible values of various MCA 
parameters?

I've tried ompi_info --all but it does not show the possible values, only the 
current value

I've tried ompi_info --param  all  but no matter what string I 
give for framework, I get no output at all.

Thanks

Re: [OMPI users] How can I discover valid values for MCA parameters?

[Nathan Hjelm]

At the moment we only provide the list of possible values for emumerated
integers. If an integer is enumerated you will see something like this
in the ompi_info -a output:

Valid values: 0:"static", 1:"sequential", 2:"unknown-root"

I am hoping to expand this support to more variables in the future.

BTW, verbosities generally range from 0 (none) to 100 (all). If I can
get agreement on standard values I may make an enumerator for those
parameters.

-Nathan

On Fri, May 29, 2015 at 04:05:57PM +, Blosch, Edwin L wrote:
>Sometimes I want to use one of the option flags, for example today it is
>mtl_mxm_verbose.  How do I discover the valid possible values of various
>MCA parameters?
> 
> 
> 
>I've tried ompi_info --all but it does not show the possible values, only
>the current value
> 
> 
> 
>I've tried ompi_info --param  all  but no matter what
>string I give for framework, I get no output at all.
> 
> 
> 
>Thanks
> 
> 
> 
> 

> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/05/26983.php



pgp9yn9DNrTTz.pgp
Description: PGP signature

Re: [OMPI users] How can I discover valid values for MCA parameters?

[Blosch, Edwin L]

Follow-up to the 2nd question which I now realize is something else.

I can see output when I do: ompi_info --param coll fca  with a version 
of OpenMPI that was built with --prefix set to its installed location.

I cannot get the output when I use a relocated version, i.e. built in one place 
and installed in another, even after I set OPAL_PREFIX to reflect the installed 
location.

From: users [mailto:users-boun...@open-mpi.org] On Behalf Of Blosch, Edwin L
Sent: Friday, May 29, 2015 11:06 AM
To: Open MPI Users (us...@open-mpi.org)
Subject: EXTERNAL: [OMPI users] How can I discover valid values for MCA 
parameters?

Sometimes I want to use one of the option flags, for example today it is 
mtl_mxm_verbose.  How do I discover the valid possible values of various MCA 
parameters?

I've tried ompi_info --all but it does not show the possible values, only the 
current value

I've tried ompi_info --param  all  but no matter what string I 
give for framework, I get no output at all.

Thanks

Re: [OMPI users] Building OpenMPI on Raspberry Pi 2

[George Bosilca]

As you are not cross-compiling I would expect gcc to use the right assembly
by default. What is happening is you force the native mode (-march=native) ?

  George.


On Fri, May 29, 2015 at 10:09 AM, Jeff Layton  wrote:

>  On 05/29/2015 09:35 AM, Jeff Layton wrote:
>
> Gilles,
>
> oops - yes, CFLAGS. But I also saw this posting:
>
> https://www.open-mpi.org/community/lists/users/2013/01/2.php
>
> where CCASFLAGS is used (I assume because for asm). I'm trying
> this flag when I configure Open MPI.
>
>
> I tried using the CCASFLAGS flag from the above link and it didn't work.
> The error
> now reads:
>
> Making all in mca/memory/linux
> make[2]: Entering directory
> '/work/pi/src/openmpi-1.8.5/opal/mca/memory/linux'
>   CC   memory_linux_component.lo
>   CC   memory_linux_ptmalloc2.lo
>   CC   memory_linux_munmap.lo
>   CC   malloc.lo
> /tmp/cc7g4mWi.s: Assembler messages:
> /tmp/cc7g4mWi.s:948: Error: selected processor does not support ARM mode
> `dmb'
> Makefile:1694: recipe for target 'malloc.lo' failed
> make[2]: *** [malloc.lo] Error 1
> make[2]: Leaving directory
> '/work/pi/src/openmpi-1.8.5/opal/mca/memory/linux'
> Makefile:2149: recipe for target 'all-recursive' failed
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
> Makefile:1698: recipe for target 'all-recursive' failed
> make: *** [all-recursive] Error 1
>
>
> I used the configure flag CCASFLAGS=-march=armv7-a
>
> Not sure if that is correct or not. I'm guessing I'm using the wrong
> architecture for the Pi 2. Suggestions?
>
> Thanks!
>
> Jeff
>
>
>
>
> Thanks!
>
> Jeff
>
>
>  Jeff,
>
>  shall I assume you made a typo and wrote CCFLAGS instead of CFLAGS ?
>
>  also, can you double check the flags are correctly passed to the
> assembler with
> cd opal/asm
> make -n atomic-asm.lo
>
>  Cheers,
>
>  Gilles
>
> On Friday, May 29, 2015, Jeff Layton  wrote:
>
>> Good morning,
>>
>> I'm building OpenMPI from source on a Raspberry Pi 2 and
>> I've hit an error. The error is:
>>
>> make[2]: Entering directory '/work/pi/src/openmpi-1.8.5/opal/asm'
>>   CPPASatomic-asm.lo
>> atomic-asm.S: Assembler messages:
>> atomic-asm.S:7: Error: selected processor does not support ARM mode `dmb'
>> atomic-asm.S:15: Error: selected processor does not support ARM mode `dmb'
>> atomic-asm.S:23: Error: selected processor does not support ARM mode `dmb'
>> atomic-asm.S:55: Error: selected processor does not support ARM mode `dmb'
>> atomic-asm.S:70: Error: selected processor does not support ARM mode `dmb'
>> atomic-asm.S:86: Error: selected processor does not support ARM mode
>> `ldrexd r4,r5,[r0]'
>> atomic-asm.S:91: Error: selected processor does not support ARM mode
>> `strexd r1,r6,r7,[r0]'
>> atomic-asm.S:107: Error: selected processor does not support ARM mode
>> `ldrexd r4,r5,[r0]'
>> atomic-asm.S:112: Error: selected processor does not support ARM mode
>> `strexd r1,r6,r7,[r0]'
>> atomic-asm.S:115: Error: selected processor does not support ARM mode
>> `dmb'
>> atomic-asm.S:130: Error: selected processor does not support ARM mode
>> `ldrexd r4,r5,[r0]'
>> atomic-asm.S:135: Error: selected processor does not support ARM mode
>> `dmb'
>> atomic-asm.S:136: Error: selected processor does not support ARM mode
>> `strexd r1,r6,r7,[r0]'
>> Makefile:1608: recipe for target 'atomic-asm.lo' failed
>> make[2]: *** [atomic-asm.lo] Error 1
>> make[2]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal/asm'
>> Makefile:2149: recipe for target 'all-recursive' failed
>> make[1]: *** [all-recursive] Error 1
>> make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
>> Makefile:1698: recipe for target 'all-recursive' failed
>> make: *** [all-recursive] Error 1
>>
>>
>> I was doing some googling and I saw where I need to specify
>> CCFLAGS when I run configure but I want to make 100% sure
>> I have the right arguments. Can anyone help?
>>
>> Thanks!
>>
>> Jeff
>>
>> P.S. I'd rather build OpenMPI from source rather than use the pre-built
>> packages in the Raspbian repository.
>>
>>
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26981.php
>

Re: [OMPI users] Building OpenMPI on Raspberry Pi 2

[Jeff Layton]

George,

I changed my configure command to be:

./configure CCASFLAGS=-march=native

and I get an error while running configure:

...
*** Assembler
checking dependency style of gcc -std=gnu99... gcc3
checking for BSD- or MS-compatible name lister (nm)... /usr/bin/nm -B
checking the name lister (/usr/bin/nm -B) interface... BSD nm
checking for fgrep... /bin/grep -F
checking if need to remove -g from CCASFLAGS... no
checking whether to enable smp locks... yes
checking if .proc/endp is needed... no
checking directive for setting text section... .text
checking directive for exporting symbols... .globl
checking for objdump... objdump
checking if .note.GNU-stack is needed... yes
checking suffix for labels... :
checking prefix for global symbol labels... none
configure: error: Could not determine global symbol label prefix


Not sure why configure failed at this point.

Thanks!

Jeff


As you are not cross-compiling I would expect gcc to use the right 
assembly by default. What is happening is you force the native mode 
(-march=native) ?


  George.


On Fri, May 29, 2015 at 10:09 AM, Jeff Layton > wrote:


On 05/29/2015 09:35 AM, Jeff Layton wrote:

Gilles,

oops - yes, CFLAGS. But I also saw this posting:

https://www.open-mpi.org/community/lists/users/2013/01/2.php

where CCASFLAGS is used (I assume because for asm). I'm trying
this flag when I configure Open MPI.


I tried using the CCASFLAGS flag from the above link and it didn't
work. The error
now reads:

Making all in mca/memory/linux
make[2]: Entering directory
'/work/pi/src/openmpi-1.8.5/opal/mca/memory/linux'
  CC   memory_linux_component.lo
  CC   memory_linux_ptmalloc2.lo
  CC   memory_linux_munmap.lo
  CC   malloc.lo
/tmp/cc7g4mWi.s: Assembler messages:
/tmp/cc7g4mWi.s:948: Error: selected processor does not support
ARM mode `dmb'
Makefile:1694: recipe for target 'malloc.lo' failed
make[2]: *** [malloc.lo] Error 1
make[2]: Leaving directory
'/work/pi/src/openmpi-1.8.5/opal/mca/memory/linux'
Makefile:2149: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
Makefile:1698: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1


I used the configure flag CCASFLAGS=-march=armv7-a

Not sure if that is correct or not. I'm guessing I'm using the wrong
architecture for the Pi 2. Suggestions?

Thanks!

Jeff





Thanks!

Jeff



Jeff,

shall I assume you made a typo and wrote CCFLAGS instead of CFLAGS ?

also, can you double check the flags are correctly passed to the
assembler with
cd opal/asm
make -n atomic-asm.lo

Cheers,

Gilles

On Friday, May 29, 2015, Jeff Layton mailto:layto...@att.net>> wrote:

Good morning,

I'm building OpenMPI from source on a Raspberry Pi 2 and
I've hit an error. The error is:

make[2]: Entering directory
'/work/pi/src/openmpi-1.8.5/opal/asm'
  CPPASatomic-asm.lo
atomic-asm.S: Assembler messages:
atomic-asm.S:7: Error: selected processor does not support
ARM mode `dmb'
atomic-asm.S:15: Error: selected processor does not support
ARM mode `dmb'
atomic-asm.S:23: Error: selected processor does not support
ARM mode `dmb'
atomic-asm.S:55: Error: selected processor does not support
ARM mode `dmb'
atomic-asm.S:70: Error: selected processor does not support
ARM mode `dmb'
atomic-asm.S:86: Error: selected processor does not support
ARM mode `ldrexd r4,r5,[r0]'
atomic-asm.S:91: Error: selected processor does not support
ARM mode `strexd r1,r6,r7,[r0]'
atomic-asm.S:107: Error: selected processor does not support
ARM mode `ldrexd r4,r5,[r0]'
atomic-asm.S:112: Error: selected processor does not support
ARM mode `strexd r1,r6,r7,[r0]'
atomic-asm.S:115: Error: selected processor does not support
ARM mode `dmb'
atomic-asm.S:130: Error: selected processor does not support
ARM mode `ldrexd r4,r5,[r0]'
atomic-asm.S:135: Error: selected processor does not support
ARM mode `dmb'
atomic-asm.S:136: Error: selected processor does not support
ARM mode `strexd r1,r6,r7,[r0]'
Makefile:1608: recipe for target 'atomic-asm.lo' failed
make[2]: *** [atomic-asm.lo] Error 1
make[2]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal/asm'
Makefile:2149: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/work/pi/src/openmpi-1.8.5/opal'
Makefile:1698: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1


I was doing some googling a

[OMPI users] mpirun

[Walt Brainerd]

On Windows, mpirun appears to take about 5 seconds
to start. I can't try it on Linux. Intel takes no time to
start executing its version.

Is this expected?

-- 
Walt Brainerd

Re: [OMPI users] mpirun

[Ralph Castain]

I assume you mean on cygwin? Or is this an older version that supported native 
Windows?

> On May 29, 2015, at 12:34 PM, Walt Brainerd  wrote:
> 
> On Windows, mpirun appears to take about 5 seconds
> to start. I can't try it on Linux. Intel takes no time to
> start executing its version.
> 
> Is this expected?
> 
> -- 
> Walt Brainerd
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/05/26988.php

Re: [OMPI users] mpirun

[Walt Brainerd]

It behaved this way with the Cygwin version (very recent update)
and with 1.8.5 that I built from source.

On Fri, May 29, 2015 at 12:35 PM, Ralph Castain  wrote:

> I assume you mean on cygwin? Or is this an older version that supported
> native Windows?
>
> > On May 29, 2015, at 12:34 PM, Walt Brainerd 
> wrote:
> >
> > On Windows, mpirun appears to take about 5 seconds
> > to start. I can't try it on Linux. Intel takes no time to
> > start executing its version.
> >
> > Is this expected?
> >
> > --
> > Walt Brainerd
> > ___
> > users mailing list
> > us...@open-mpi.org
> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> > Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26988.php
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26989.php
>



-- 
Walt Brainerd

Re: [OMPI users] mpirun

[Walt Brainerd]

I should explain a little more. Sorry.
I built OpenMPI and Opencoarrays and Gfortran 5.1.
It is with this  combination that I am running a
program using coarrays compiled by gfortran.

The command "mpirun -np 4 a.exe" appears to take
about 5 seconds to start up. This is true of another
person using the same software. It is the same
whether using mpirun that I built from source or the
mpirun in cygwin.

After the 5 seconds, the program appears to run fine.

Thanks for your comments.

On Fri, May 29, 2015 at 12:35 PM, Ralph Castain  wrote:

> I assume you mean on cygwin? Or is this an older version that supported
> native Windows?
>
> > On May 29, 2015, at 12:34 PM, Walt Brainerd 
> wrote:
> >
> > On Windows, mpirun appears to take about 5 seconds
> > to start. I can't try it on Linux. Intel takes no time to
> > start executing its version.
> >
> > Is this expected?
> >
> > --
> > Walt Brainerd
> > ___
> > users mailing list
> > us...@open-mpi.org
> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> > Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26988.php
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/05/26989.php
>



-- 
Walt Brainerd

Re: [OMPI users] mpirun

[Ralph Castain]

Since you are able to build OMPI from scratch, can you build it with 
—enable-debug? If so, then you could set an MCA param or two for debug purposes 
that might help us narrow down where the delay is coming from. I’d start with 
"—mca state_base_verbose 5” and see if there is some noticeable hang during the 
output.


> On May 29, 2015, at 12:58 PM, Walt Brainerd  wrote:
> 
> I should explain a little more. Sorry.
> I built OpenMPI and Opencoarrays and Gfortran 5.1.
> It is with this  combination that I am running a
> program using coarrays compiled by gfortran.
> 
> The command "mpirun -np 4 a.exe" appears to take
> about 5 seconds to start up. This is true of another
> person using the same software. It is the same
> whether using mpirun that I built from source or the
> mpirun in cygwin.
> 
> After the 5 seconds, the program appears to run fine.
> 
> Thanks for your comments.
> 
> On Fri, May 29, 2015 at 12:35 PM, Ralph Castain  > wrote:
> I assume you mean on cygwin? Or is this an older version that supported 
> native Windows?
> 
> > On May 29, 2015, at 12:34 PM, Walt Brainerd  > > wrote:
> >
> > On Windows, mpirun appears to take about 5 seconds
> > to start. I can't try it on Linux. Intel takes no time to
> > start executing its version.
> >
> > Is this expected?
> >
> > --
> > Walt Brainerd
> > ___
> > users mailing list
> > us...@open-mpi.org 
> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
> > 
> > Link to this post: 
> > http://www.open-mpi.org/community/lists/users/2015/05/26988.php 
> > 
> 
> ___
> users mailing list
> us...@open-mpi.org 
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
> 
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/05/26989.php 
> 
> 
> 
> 
> -- 
> Walt Brainerd
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/05/26991.php

Re: [OMPI users] mpirun

[Ralph Castain]

Another thing you might want to try: see how long it takes to run a simple MPI 
“hello” - see our “examples” directory. This will tell us if there is something 
about this app that is causing the delay, or something in the launch support


> On May 29, 2015, at 1:13 PM, Ralph Castain  wrote:
> 
> Since you are able to build OMPI from scratch, can you build it with 
> —enable-debug? If so, then you could set an MCA param or two for debug 
> purposes that might help us narrow down where the delay is coming from. I’d 
> start with "—mca state_base_verbose 5” and see if there is some noticeable 
> hang during the output.
> 
> 
>> On May 29, 2015, at 12:58 PM, Walt Brainerd > > wrote:
>> 
>> I should explain a little more. Sorry.
>> I built OpenMPI and Opencoarrays and Gfortran 5.1.
>> It is with this  combination that I am running a
>> program using coarrays compiled by gfortran.
>> 
>> The command "mpirun -np 4 a.exe" appears to take
>> about 5 seconds to start up. This is true of another
>> person using the same software. It is the same
>> whether using mpirun that I built from source or the
>> mpirun in cygwin.
>> 
>> After the 5 seconds, the program appears to run fine.
>> 
>> Thanks for your comments.
>> 
>> On Fri, May 29, 2015 at 12:35 PM, Ralph Castain > > wrote:
>> I assume you mean on cygwin? Or is this an older version that supported 
>> native Windows?
>> 
>> > On May 29, 2015, at 12:34 PM, Walt Brainerd > > > wrote:
>> >
>> > On Windows, mpirun appears to take about 5 seconds
>> > to start. I can't try it on Linux. Intel takes no time to
>> > start executing its version.
>> >
>> > Is this expected?
>> >
>> > --
>> > Walt Brainerd
>> > ___
>> > users mailing list
>> > us...@open-mpi.org 
>> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>> > 
>> > Link to this post: 
>> > http://www.open-mpi.org/community/lists/users/2015/05/26988.php 
>> > 
>> 
>> ___
>> users mailing list
>> us...@open-mpi.org 
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>> 
>> Link to this post: 
>> http://www.open-mpi.org/community/lists/users/2015/05/26989.php 
>> 
>> 
>> 
>> 
>> -- 
>> Walt Brainerd
>> ___
>> users mailing list
>> us...@open-mpi.org 
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>> Link to this post: 
>> http://www.open-mpi.org/community/lists/users/2015/05/26991.php
> 

Re: [OMPI users] mpirun

[Gilles Gouaillardet]

Walt,

can you disable firewall and network if possible  and give it an other try ?

Cheers,

Gilles



On Saturday, May 30, 2015, Walt Brainerd  wrote:

> It behaved this way with the Cygwin version (very recent update)
> and with 1.8.5 that I built from source.
>
> On Fri, May 29, 2015 at 12:35 PM, Ralph Castain  > wrote:
>
>> I assume you mean on cygwin? Or is this an older version that supported
>> native Windows?
>>
>> > On May 29, 2015, at 12:34 PM, Walt Brainerd > > wrote:
>> >
>> > On Windows, mpirun appears to take about 5 seconds
>> > to start. I can't try it on Linux. Intel takes no time to
>> > start executing its version.
>> >
>> > Is this expected?
>> >
>> > --
>> > Walt Brainerd
>> > ___
>> > users mailing list
>> > us...@open-mpi.org 
>> > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>> > Link to this post:
>> http://www.open-mpi.org/community/lists/users/2015/05/26988.php
>>
>> ___
>> users mailing list
>> us...@open-mpi.org 
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>> Link to this post:
>> http://www.open-mpi.org/community/lists/users/2015/05/26989.php
>>
>
>
>
> --
> Walt Brainerd
>

Re: [OMPI users] mpirun

[Tim Prince]

I don't recall Walt 's cases taking all of 5 seconds to start. More annoying is 
the hang after completion.

Sent via the ASUS PadFone X mini, an AT&T 4G LTE smartphone

 Original Message 
From:Ralph Castain 
Sent:Fri, 29 May 2015 15:35:15 -0400
To:Open MPI Users 
Subject:Re: [OMPI users] mpirun

>I assume you mean on cygwin? Or is this an older version that supported native 
>Windows?
>
>> On May 29, 2015, at 12:34 PM, Walt Brainerd  wrote:
>> 
>> On Windows, mpirun appears to take about 5 seconds
>> to start. I can't try it on Linux. Intel takes no time to
>> start executing its version.
>> 
>> Is this expected?
>> 
>> -- 
>> Walt Brainerd
>> ___
>> users mailing list
>> us...@open-mpi.org
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>> Link to this post: 
>> http://www.open-mpi.org/community/lists/users/2015/05/26988.php
>
>___
>users mailing list
>us...@open-mpi.org
>Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>Link to this post: 
>http://www.open-mpi.org/community/lists/users/2015/05/26989.php