Re: [OMPI users] CUDA supported APIs

[Zhang, Junchao via users]

Leo,
  Thanks for the info. That is interesting.  And yes, Having a CUDA aware MPI 
API list would be very useful.
--Junchao Zhang


On Mon, Aug 19, 2019 at 10:23 AM Fang, Leo 
mailto:leof...@bnl.gov>> wrote:

Hi Junchao,



First, for your second question, the answer is here: 
https://www.mail-archive.com/users@lists.open-mpi.org/msg33279.html. I know 
this because I also asked it earlier  It'd be nice to have this documented in 
the Q though.


As for your first question, I am also interested. It'd be nice for Open MPI 
core devs to keep the supported API list up-to-date. We recently added support 
of CUDA-aware MPI to mpi4py, and such a list is important for us to keep track 
upstream support so that we know whether a test fails due to lack of 
CUDA-awareness or because we messed up (much less likely).


Thanks.



Sincerely,
Leo

---
Yao-Lung Leo Fang
Assistant Computational Scientist
Computational Science Initiative
Brookhaven National Laboratory
Bldg. 725, Room 2-169
P.O. Box 5000, Upton, NY 11973-5000
Office: (631) 344-3265
Email: leof...@bnl.gov
Website: https://leofang.github.io/

寄件者: users 
mailto:users-boun...@lists.open-mpi.org>> 代表 
Zhang, Junchao via users 
mailto:users@lists.open-mpi.org>>
寄件日期: 2019年8月15日 上午 11:52:56
收件者: Open MPI Users mailto:users@lists.open-mpi.org>>
副本: Zhang, Junchao mailto:jczh...@mcs.anl.gov>>
主旨: Re: [OMPI users] CUDA supported APIs

Another question: If MPI_Allgatherv(const void *sendbuf, int sendcount, 
MPI_Datatype sendtype, void *recvbuf, const int recvcounts[],const int 
displs[], MPI_Datatype recvtype, MPI_Comm comm) is cuda aware, are recvcounts, 
displs in CPU memory or GPU memory?

--Junchao Zhang


On Thu, Aug 15, 2019 at 9:55 AM Junchao Zhang 
mailto:jczh...@mcs.anl.gov>> wrote:
Hi,
   Are the APIs at 
https://www.open-mpi.org/faq/?category=runcuda#mpi-apis-cuda
 latest? I could not find MPI_Neighbor_xxx and MPI_Reduce_local.
  Thanks.

--Junchao Zhang
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] **URGENT: Error during testing

[Jeff Squyres (jsquyres) via users]

Note that it is not advisable to run autogen from a distribution tarball.

Specifically: we include the autogen script for advanced users who want to 
tweak their own Open MPI copy (without cloning from git), but 99.99% of users 
can just run "./configure ..." directly (without first running autogen).  If 
you're an advanced user and want / need to run autogen -- cool, go for it.  But 
if you're just downloading an Open MPI tarball and are trying to build / 
install it, there's no need to run autogen.

Put simply: in official release tarballs of Open MPI, we've already run autogen 
for you.


On Aug 19, 2019, at 2:17 PM, Steven Varga via users 
mailto:users@lists.open-mpi.org>> wrote:

Hi
this is steven. I am building custom clusters on AWS Ec2 and had some problems 
in the past. I am getting good result with external pmix 3.1.3
./autogen.sh && ./configure --prefix=/usr/local/ --with-platform=optimized 
--with-hwloc=/usr/local --with-libevent=/usr/local --enable-pmix-binaries  
--enable-pmi-backward-compatibility
make -j4 && sudo make install
and ompenmpi 4.0.1
./autogen.pl && ./configure --prefix=/usr/local 
--with-slurm --with-pmix=/usr/local --enable-mpi1-compatibility 
--with-libevent=/usr/local --with-hwloc=/usr/local

making certain linking against the same libevent
This is on linux most recent custom kernel, and most recent SLURM scheduler.
best:
steve





On Mon, Aug 19, 2019 at 2:07 PM Jeff Squyres (jsquyres) via users 
mailto:users@lists.open-mpi.org>> wrote:
Can you provide some more details?

https://www.open-mpi.org/community/help/


On Aug 19, 2019, at 1:18 PM, Riddhi A Mehta via users 
mailto:users@lists.open-mpi.org>> wrote:

Hello

My name is Riddhi and I am a Graduate Research Assistant in the Dept. of 
Physics & Astronomy at Purdue University. About a month ago I correctly 
configured openmpi on my mac and the ‘mpirun -np 2 ./hello.c’ ran correctly. 
But today, it gave me the following error:


[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file orted/pmix/pmix_server.c at line 262
[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file ess_hnp_module.c at line 667
--
It looks like orte_init failed for some reason; your parallel process is
likely to abort.  There are many reasons that a parallel process can
fail during orte_init; some of which are due to configuration or
environment problems.  This failure appears to be an internal failure;
here's some additional information (which may only be relevant to an
Open MPI developer):

  pmix server init failed
  --> Returned value Bad parameter (-5) instead of ORTE_SUCCESS

I tried to google this error and the only thing I found was to change the 
TMPDIR name to a shorter name. But I don’t think I understand what it’s asking 
me to do. Kindly help since I am not able to understand what went wrong in one 
month and mpirun stopped working.

Thank you

Riddhi Mehta
Research Group: Maxim Lyutikov, Theoretical High Energy Astrophysics
Dept. of Physics & Astronomy
Purdue University

___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users


--
Jeff Squyres
jsquy...@cisco.com




___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users


--
Jeff Squyres
jsquy...@cisco.com




___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] **URGENT: Error during testing

[Jeff Squyres (jsquyres) via users]

Do not use Open MPI v2.0.x -- it's ancient.

You should probably use Open MPI v3.1.x or v4.0.x.



On Aug 19, 2019, at 2:12 PM, Riddhi A Mehta via users 
mailto:users@lists.open-mpi.org>> wrote:

Hi

I followed the exact procedure stated in this link: 
http://www.science.smith.edu/dftwiki/index.php/Install_MPI_on_a_MacBook.


It runs correctly until this line : mpicc -o hello helloWorld.c

After which it gives me the error when I do mpirun.

Thank you

Riddhi

From: users 
mailto:users-boun...@lists.open-mpi.org>> on 
behalf of "Riebs, Andy via users" 
mailto:users@lists.open-mpi.org>>
Reply-To: Open MPI Users 
mailto:users@lists.open-mpi.org>>
Date: Monday, August 19, 2019 at 2:10 PM
To: Open MPI Users mailto:users@lists.open-mpi.org>>
Cc: "Riebs, Andy" mailto:andy.ri...@hpe.com>>
Subject: Re: [OMPI users] **URGENT: Error during testing

Is there any chance that the fact that Riddhi appears to be trying to execute 
an uncompiled hello.c could be the problem here?

From: users [mailto:users-boun...@lists.open-mpi.org] On Behalf Of Jeff Squyres 
(jsquyres) via users
Sent: Monday, August 19, 2019 2:05 PM
To: Open MPI User's List 
mailto:users@lists.open-mpi.org>>
Cc: Jeff Squyres (jsquyres) mailto:jsquy...@cisco.com>>
Subject: Re: [OMPI users] **URGENT: Error during testing

Can you provide some more details?

https://www.open-mpi.org/community/help/




On Aug 19, 2019, at 1:18 PM, Riddhi A Mehta via users 
mailto:users@lists.open-mpi.org>> wrote:

Hello

My name is Riddhi and I am a Graduate Research Assistant in the Dept. of 
Physics & Astronomy at Purdue University. About a month ago I correctly 
configured openmpi on my mac and the ‘mpirun -np 2 ./hello.c’ ran correctly. 
But today, it gave me the following error:


[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file orted/pmix/pmix_server.c at line 262
[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file ess_hnp_module.c at line 667
--
It looks like orte_init failed for some reason; your parallel process is
likely to abort.  There are many reasons that a parallel process can
fail during orte_init; some of which are due to configuration or
environment problems.  This failure appears to be an internal failure;
here's some additional information (which may only be relevant to an
Open MPI developer):

  pmix server init failed
  --> Returned value Bad parameter (-5) instead of ORTE_SUCCESS

I tried to google this error and the only thing I found was to change the 
TMPDIR name to a shorter name. But I don’t think I understand what it’s asking 
me to do. Kindly help since I am not able to understand what went wrong in one 
month and mpirun stopped working.

Thank you

Riddhi Mehta
Research Group: Maxim Lyutikov, Theoretical High Energy Astrophysics
Dept. of Physics & Astronomy
Purdue University

___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users


--
Jeff Squyres
jsquy...@cisco.com



___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users


--
Jeff Squyres
jsquy...@cisco.com




___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] **URGENT: Error during testing

[Steven Varga via users]

Hi
this is steven. I am building custom clusters on AWS Ec2 and had some
problems in the past. I am getting good result with external pmix 3.1.3
./autogen.sh && ./configure --prefix=/usr/local/ --with-platform=optimized
--with-hwloc=/usr/local --with-libevent=/usr/local --enable-pmix-binaries
 --enable-pmi-backward-compatibility
make -j4 && sudo make install
and ompenmpi 4.0.1
./autogen.pl && ./configure --prefix=/usr/local --with-slurm
--with-pmix=/usr/local --enable-mpi1-compatibility
--with-libevent=/usr/local --with-hwloc=/usr/local

making certain linking against the same libevent
This is on linux most recent custom kernel, and most recent SLURM scheduler.
best:
steve





On Mon, Aug 19, 2019 at 2:07 PM Jeff Squyres (jsquyres) via users <
users@lists.open-mpi.org> wrote:

> Can you provide some more details?
>
> https://www.open-mpi.org/community/help/
>
>
> On Aug 19, 2019, at 1:18 PM, Riddhi A Mehta via users <
> users@lists.open-mpi.org> wrote:
>
> Hello
>
> My name is Riddhi and I am a Graduate Research Assistant in the Dept. of
> Physics & Astronomy at Purdue University. About a month ago I correctly
> configured openmpi on my mac and the ‘mpirun -np 2 ./hello.c’ ran
> correctly. But today, it gave me the following error:
>
>
> *[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad
> parameter in file orted/pmix/pmix_server.c at line 262*
> *[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad
> parameter in file ess_hnp_module.c at line 667*
>
> *--*
> *It looks like orte_init failed for some reason; your parallel process is*
> *likely to abort.*  *There are many reasons that a parallel process can*
> *fail during orte_init; some of which are due to configuration or*
> *environment problems.*  *This failure appears to be an internal failure;*
> *here's some additional information (which may only be relevant to an*
> *Open MPI developer):*
>
>   *pmix server init failed*
>   *--> Returned value Bad parameter (-5) instead of ORTE_SUCCESS*
>
> I tried to google this error and the only thing I found was to change the
> TMPDIR name to a shorter name. But I don’t think I understand what it’s
> asking me to do. Kindly help since I am not able to understand what went
> wrong in one month and mpirun stopped working.
>
> Thank you
>
> Riddhi Mehta
> Research Group: Maxim Lyutikov, Theoretical High Energy Astrophysics
> Dept. of Physics & Astronomy
> Purdue University
>
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
>
>
>
> --
> Jeff Squyres
> jsquy...@cisco.com
>
>
>
>
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] **URGENT: Error during testing

[Riddhi A Mehta via users]

Hi

I followed the exact procedure stated in this link: 
http://www.science.smith.edu/dftwiki/index.php/Install_MPI_on_a_MacBook.


It runs correctly until this line : mpicc -o hello helloWorld.c
After which it gives me the error when I do mpirun.

Thank you

Riddhi

From: users  on behalf of "Riebs, Andy via 
users" 
Reply-To: Open MPI Users 
Date: Monday, August 19, 2019 at 2:10 PM
To: Open MPI Users 
Cc: "Riebs, Andy" 
Subject: Re: [OMPI users] **URGENT: Error during testing

Is there any chance that the fact that Riddhi appears to be trying to execute 
an uncompiled hello.c could be the problem here?

From: users [mailto:users-boun...@lists.open-mpi.org] On Behalf Of Jeff Squyres 
(jsquyres) via users
Sent: Monday, August 19, 2019 2:05 PM
To: Open MPI User's List 
Cc: Jeff Squyres (jsquyres) 
Subject: Re: [OMPI users] **URGENT: Error during testing

Can you provide some more details?

https://www.open-mpi.org/community/help/




On Aug 19, 2019, at 1:18 PM, Riddhi A Mehta via users 
mailto:users@lists.open-mpi.org>> wrote:

Hello

My name is Riddhi and I am a Graduate Research Assistant in the Dept. of 
Physics & Astronomy at Purdue University. About a month ago I correctly 
configured openmpi on my mac and the ‘mpirun -np 2 ./hello.c’ ran correctly. 
But today, it gave me the following error:


[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file orted/pmix/pmix_server.c at line 262
[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file ess_hnp_module.c at line 667
--
It looks like orte_init failed for some reason; your parallel process is
likely to abort.  There are many reasons that a parallel process can
fail during orte_init; some of which are due to configuration or
environment problems.  This failure appears to be an internal failure;
here's some additional information (which may only be relevant to an
Open MPI developer):

  pmix server init failed
  --> Returned value Bad parameter (-5) instead of ORTE_SUCCESS

I tried to google this error and the only thing I found was to change the 
TMPDIR name to a shorter name. But I don’t think I understand what it’s asking 
me to do. Kindly help since I am not able to understand what went wrong in one 
month and mpirun stopped working.

Thank you

Riddhi Mehta
Research Group: Maxim Lyutikov, Theoretical High Energy Astrophysics
Dept. of Physics & Astronomy
Purdue University

___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users


--
Jeff Squyres
jsquy...@cisco.com



___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] **URGENT: Error during testing

[Riebs, Andy via users]

Is there any chance that the fact that Riddhi appears to be trying to execute 
an uncompiled hello.c could be the problem here?

From: users [mailto:users-boun...@lists.open-mpi.org] On Behalf Of Jeff Squyres 
(jsquyres) via users
Sent: Monday, August 19, 2019 2:05 PM
To: Open MPI User's List 
Cc: Jeff Squyres (jsquyres) 
Subject: Re: [OMPI users] **URGENT: Error during testing

Can you provide some more details?

https://www.open-mpi.org/community/help/



On Aug 19, 2019, at 1:18 PM, Riddhi A Mehta via users 
mailto:users@lists.open-mpi.org>> wrote:

Hello

My name is Riddhi and I am a Graduate Research Assistant in the Dept. of 
Physics & Astronomy at Purdue University. About a month ago I correctly 
configured openmpi on my mac and the ‘mpirun -np 2 ./hello.c’ ran correctly. 
But today, it gave me the following error:


[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file orted/pmix/pmix_server.c at line 262
[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file ess_hnp_module.c at line 667
--
It looks like orte_init failed for some reason; your parallel process is
likely to abort.  There are many reasons that a parallel process can
fail during orte_init; some of which are due to configuration or
environment problems.  This failure appears to be an internal failure;
here's some additional information (which may only be relevant to an
Open MPI developer):

  pmix server init failed
  --> Returned value Bad parameter (-5) instead of ORTE_SUCCESS

I tried to google this error and the only thing I found was to change the 
TMPDIR name to a shorter name. But I don’t think I understand what it’s asking 
me to do. Kindly help since I am not able to understand what went wrong in one 
month and mpirun stopped working.

Thank you

Riddhi Mehta
Research Group: Maxim Lyutikov, Theoretical High Energy Astrophysics
Dept. of Physics & Astronomy
Purdue University

___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users


--
Jeff Squyres
jsquy...@cisco.com



___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] **URGENT: Error during testing

[Jeff Squyres (jsquyres) via users]

Can you provide some more details?

https://www.open-mpi.org/community/help/


On Aug 19, 2019, at 1:18 PM, Riddhi A Mehta via users 
mailto:users@lists.open-mpi.org>> wrote:

Hello

My name is Riddhi and I am a Graduate Research Assistant in the Dept. of 
Physics & Astronomy at Purdue University. About a month ago I correctly 
configured openmpi on my mac and the ‘mpirun -np 2 ./hello.c’ ran correctly. 
But today, it gave me the following error:


[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file orted/pmix/pmix_server.c at line 262
[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file ess_hnp_module.c at line 667
--
It looks like orte_init failed for some reason; your parallel process is
likely to abort.  There are many reasons that a parallel process can
fail during orte_init; some of which are due to configuration or
environment problems.  This failure appears to be an internal failure;
here's some additional information (which may only be relevant to an
Open MPI developer):

  pmix server init failed
  --> Returned value Bad parameter (-5) instead of ORTE_SUCCESS

I tried to google this error and the only thing I found was to change the 
TMPDIR name to a shorter name. But I don’t think I understand what it’s asking 
me to do. Kindly help since I am not able to understand what went wrong in one 
month and mpirun stopped working.

Thank you

Riddhi Mehta
Research Group: Maxim Lyutikov, Theoretical High Energy Astrophysics
Dept. of Physics & Astronomy
Purdue University

___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users


--
Jeff Squyres
jsquy...@cisco.com




___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

[OMPI users] **URGENT: Error during testing

[Riddhi A Mehta via users]

Hello



My name is Riddhi and I am a Graduate Research Assistant in the Dept. of 
Physics & Astronomy at Purdue University. About a month ago I correctly 
configured openmpi on my mac and the ‘mpirun -np 2 ./hello.c’ ran correctly. 
But today, it gave me the following error:





[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file orted/pmix/pmix_server.c at line 262

[Riddhis-MacBook-Air.local:89582] [[56525,0],0] ORTE_ERROR_LOG: Bad parameter 
in file ess_hnp_module.c at line 667

--

It looks like orte_init failed for some reason; your parallel process is

likely to abort.  There are many reasons that a parallel process can

fail during orte_init; some of which are due to configuration or

environment problems.  This failure appears to be an internal failure;

here's some additional information (which may only be relevant to an

Open MPI developer):



  pmix server init failed

  --> Returned value Bad parameter (-5) instead of ORTE_SUCCESS

I tried to google this error and the only thing I found was to change the 
TMPDIR name to a shorter name. But I don’t think I understand what it’s asking 
me to do. Kindly help since I am not able to understand what went wrong in one 
month and mpirun stopped working.

Thank you

Riddhi Mehta
Research Group: Maxim Lyutikov, Theoretical High Energy Astrophysics
Dept. of Physics & Astronomy
Purdue University

___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] CUDA supported APIs

[Fang, Leo via users]

Hi Junchao,



First, for your second question, the answer is here: 
https://www.mail-archive.com/users@lists.open-mpi.org/msg33279.html. I know 
this because I also asked it earlier  It'd be nice to have this documented in 
the Q though.


As for your first question, I am also interested. It'd be nice for Open MPI 
core devs to keep the supported API list up-to-date. We recently added support 
of CUDA-aware MPI to mpi4py, and such a list is important for us to keep track 
upstream support so that we know whether a test fails due to lack of 
CUDA-awareness or because we messed up (much less likely).


Thanks.



Sincerely,
Leo

---
Yao-Lung Leo Fang
Assistant Computational Scientist
Computational Science Initiative
Brookhaven National Laboratory
Bldg. 725, Room 2-169
P.O. Box 5000, Upton, NY 11973-5000
Office: (631) 344-3265
Email: leof...@bnl.gov
Website: https://leofang.github.io/

寄件者: users  代表 Zhang, Junchao via users 

寄件日期: 2019年8月15日 上午 11:52:56
收件者: Open MPI Users 
副本: Zhang, Junchao 
主旨: Re: [OMPI users] CUDA supported APIs

Another question: If MPI_Allgatherv(const void *sendbuf, int sendcount, 
MPI_Datatype sendtype, void *recvbuf, const int recvcounts[],const int 
displs[], MPI_Datatype recvtype, MPI_Comm comm) is cuda aware, are recvcounts, 
displs in CPU memory or GPU memory?

--Junchao Zhang


On Thu, Aug 15, 2019 at 9:55 AM Junchao Zhang 
mailto:jczh...@mcs.anl.gov>> wrote:
Hi,
   Are the APIs at 
https://www.open-mpi.org/faq/?category=runcuda#mpi-apis-cuda
 latest? I could not find MPI_Neighbor_xxx and MPI_Reduce_local.
  Thanks.

--Junchao Zhang
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Jeff Squyres (jsquyres) via users]

On Aug 19, 2019, at 6:15 AM, Sangam B via users  
wrote:
> 
> subroutine recv(this,lmb)
> class(some__example6), intent(inout) ::  this
> integer, intent(in) :: lmb(2,2)
> 
> integer :: cs3, ierr
> integer(kind=C_LONG) :: size

This ^^ is your problem.  More below.

> ! receive only from buffer at different process
> if(this%is_bf_referred) return
> 
> cs3=this%uspecifier%get_recv_buff_3rd_dim_size(this%xb,this%vwb,lmb)
> if(cs3.eq.0) return ! nothing to recv
> 
> size = this%size_dim(this%gi)*this%size_dim(this%gj)*cs3
> if(this%is_exchange_off) then
>call this%update_stats(size)
>this%bf(:,:,1:cs3) = cmplx(0.,0.)
> else
>call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> this%nrank,this%tag,this%comm_xvw,this%request,ierr)

You are calling MPI_Irecv with "size" as the "COUNT" dummy parameter for 
MPI_IRECV.  According to the MPI spec, this parameter must be a plain INTEGER.

This is why the compiler is erring: Fortran will not automatically cast an 
INTEGER(kind=C_LONG) to INTEGER, and therefore it cannot find a matching 
interface for MPI_IRECV in Open MPI's "mpi" module.

This is valid and correct behavior for the compiler and Open MPI.

The other MPI's may compile if they are not doing strict type checking in their 
"mpi" module -- but depending on the characteristics of the parameter mismatch, 
you may experience run-time problems.  Open MPI's "mpi" module does strict 
compile-time type checking, meaning that what happened is exactly what is 
supposed to happen: you got a compiler error because you tried to pass a 
parameter of the wrong type.  That's actually a good thing!

(Note: I didn't check the other param types; I just stopped when I saw that 
"size" was the wrong type -- you should check the other param types, too)

-- 
Jeff Squyres
jsquy...@cisco.com




___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Sangam B via users]

Hi,

Evan after recompiling OpenMPI with -fdefault-real-8, it fails with same
error.

It seems to me that, it's an issue with OpenMPI itself, because:

Intel MPI + gnu compiler --- works
Intel MPI + intel compiler --- works
Open MPI + gnu compiler ---  fails
Open MPI + AOCC compiler ---  fails

Will send the minimal program shortly

On Mon, Aug 19, 2019 at 4:01 PM Gilles Gouaillardet via users <
users@lists.open-mpi.org> wrote:

> Thanks, but this is not really helping.
>
> Could you please build a Minimal, Reproducible Example as described at
> https://stackoverflow.com/help/minimal-reproducible-example ?
>
> Cheers,
>
> Gilles
>
> On Mon, Aug 19, 2019 at 7:19 PM Sangam B via users
>  wrote:
> >
> > Hi,
> >
> > Here is the sample program snippet:
> >
> > 
> > #include "intrinsic_sizes.h"
> > #include "redef.h"
> >
> > module module1_m
> >
> >   use mod1_m, only:  some__example2
> >   use mod2_m, only:  some__example3
> >   use mod3_m, only:  some__example4
> >
> >   use mpi
> >   use, intrinsic :: iso_c_binding
> >
> > implicit none
> >
> >   private
> >
> >   public :: some__example___memory
> >
> >   type, public, extends(some__example5) :: some__example6
> >  logical, public :: some__example7 = .False.
> >  class(some__example8), private, pointer :: some__example9
> >contains
> >
> > ...
> > ...
> > end type some__example6
> >
> > contains
> > 
> > some_pure_functions here
> > 
> >
> >
> > subroutine recv(this,lmb)
> > class(some__example6), intent(inout) ::  this
> > integer, intent(in) :: lmb(2,2)
> >
> > integer :: cs3, ierr
> > integer(kind=C_LONG) :: size
> >
> > ! receive only from buffer at different process
> > if(this%is_bf_referred) return
> >
> > cs3=this%uspecifier%get_recv_buff_3rd_dim_size(this%xb,this%vwb,lmb)
> > if(cs3.eq.0) return ! nothing to recv
> >
> > size = this%size_dim(this%gi)*this%size_dim(this%gj)*cs3
> > if(this%is_exchange_off) then
> >call this%update_stats(size)
> >this%bf(:,:,1:cs3) = cmplx(0.,0.)
> > else
> >call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> > this%nrank,this%tag,this%comm_xvw,this%request,ierr)
> > end if
> >   end subroutine recv
> >
> >
> > Hope this helps.
> >
> > On Mon, Aug 19, 2019 at 3:21 PM Gilles Gouaillardet via users <
> users@lists.open-mpi.org> wrote:
> >>
> >> Thanks,
> >>
> >> and your reproducer is ?
> >>
> >> Cheers,
> >>
> >> Gilles
> >>
> >> On Mon, Aug 19, 2019 at 6:42 PM Sangam B via users
> >>  wrote:
> >> >
> >> > Hi,
> >> >
> >> > OpenMPI is configured as follows:
> >> >
> >> > export CC=`which clang`
> >> > export CXX=`which clang++`
> >> > export FC=`which flang`
> >> > export F90=`which flang`
> >> >
> >> > ../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort
> --enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2
> --without-knem --without-libfabric --without-lsf --with-verbs=/usr
> --with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm
> >> >
> >> >
> >> > ..
> >> >
> >> > On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:
> >> >>
> >> >> Hi,
> >> >>
> >> >> I get following error if the application is compiled with
> openmpi-3.1.1:
> >> >>
> >> >> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto
> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops
> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form
> -fdefault-real-8 example_program.F90
> >> >> F90-S-0155-Could not resolve generic procedure mpi_irecv (
> example_program.F90  : 97)
> >> >>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
> >> >>
> >> >> Following is the line causing this error:
> >> >>
> >> >> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> >> >> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
> >> >>
> >> >> The program has following module mentioned in the beginning:
> >> >>  use mpi
> >> >>
> >> >> The openmpi has following module files in lib folder:
> >> >> $ ls *.mod
> >> >> mpi_ext.modmpi_f08_ext.mod
>  mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
> >> >> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod
> mpi_f08.mod mpi.mod
> >> >>
> >> >> The same program works with Intel MPI (gcc/intel as base compilers).
> >> >> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base
> compilers. What could be the reason for it?
> >> >>
> >> >> ..
> >> >
> >> > ___
> >> > users mailing list
> >> > users@lists.open-mpi.org
> >> > https://lists.open-mpi.org/mailman/listinfo/users
> >> ___
> >> users mailing list
> >> users@lists.open-mpi.org
> >> https://lists.open-mpi.org/mailman/listinfo/users
> >
> > ___
> > users mailing list
> > users@lists.open-mpi.org
> > https://lists.open-mpi.org/mailman/listinfo/users
> 

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Gilles Gouaillardet via users]

Thanks, but this is not really helping.

Could you please build a Minimal, Reproducible Example as described at
https://stackoverflow.com/help/minimal-reproducible-example ?

Cheers,

Gilles

On Mon, Aug 19, 2019 at 7:19 PM Sangam B via users
 wrote:
>
> Hi,
>
> Here is the sample program snippet:
>
> 
> #include "intrinsic_sizes.h"
> #include "redef.h"
>
> module module1_m
>
>   use mod1_m, only:  some__example2
>   use mod2_m, only:  some__example3
>   use mod3_m, only:  some__example4
>
>   use mpi
>   use, intrinsic :: iso_c_binding
>
> implicit none
>
>   private
>
>   public :: some__example___memory
>
>   type, public, extends(some__example5) :: some__example6
>  logical, public :: some__example7 = .False.
>  class(some__example8), private, pointer :: some__example9
>contains
>
> ...
> ...
> end type some__example6
>
> contains
> 
> some_pure_functions here
> 
>
>
> subroutine recv(this,lmb)
> class(some__example6), intent(inout) ::  this
> integer, intent(in) :: lmb(2,2)
>
> integer :: cs3, ierr
> integer(kind=C_LONG) :: size
>
> ! receive only from buffer at different process
> if(this%is_bf_referred) return
>
> cs3=this%uspecifier%get_recv_buff_3rd_dim_size(this%xb,this%vwb,lmb)
> if(cs3.eq.0) return ! nothing to recv
>
> size = this%size_dim(this%gi)*this%size_dim(this%gj)*cs3
> if(this%is_exchange_off) then
>call this%update_stats(size)
>this%bf(:,:,1:cs3) = cmplx(0.,0.)
> else
>call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
> end if
>   end subroutine recv
>
>
> Hope this helps.
>
> On Mon, Aug 19, 2019 at 3:21 PM Gilles Gouaillardet via users 
>  wrote:
>>
>> Thanks,
>>
>> and your reproducer is ?
>>
>> Cheers,
>>
>> Gilles
>>
>> On Mon, Aug 19, 2019 at 6:42 PM Sangam B via users
>>  wrote:
>> >
>> > Hi,
>> >
>> > OpenMPI is configured as follows:
>> >
>> > export CC=`which clang`
>> > export CXX=`which clang++`
>> > export FC=`which flang`
>> > export F90=`which flang`
>> >
>> > ../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort 
>> > --enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2 
>> > --without-knem --without-libfabric --without-lsf --with-verbs=/usr 
>> > --with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm
>> >
>> >
>> > ..
>> >
>> > On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:
>> >>
>> >> Hi,
>> >>
>> >> I get following error if the application is compiled with openmpi-3.1.1:
>> >>
>> >> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto 
>> >> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops 
>> >> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form 
>> >> -fdefault-real-8 example_program.F90
>> >> F90-S-0155-Could not resolve generic procedure mpi_irecv ( 
>> >> example_program.F90  : 97)
>> >>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
>> >>
>> >> Following is the line causing this error:
>> >>
>> >> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
>> >> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
>> >>
>> >> The program has following module mentioned in the beginning:
>> >>  use mpi
>> >>
>> >> The openmpi has following module files in lib folder:
>> >> $ ls *.mod
>> >> mpi_ext.modmpi_f08_ext.mod   
>> >> mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
>> >> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod  
>> >>mpi.mod
>> >>
>> >> The same program works with Intel MPI (gcc/intel as base compilers).
>> >> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base 
>> >> compilers. What could be the reason for it?
>> >>
>> >> ..
>> >
>> > ___
>> > users mailing list
>> > users@lists.open-mpi.org
>> > https://lists.open-mpi.org/mailman/listinfo/users
>> ___
>> users mailing list
>> users@lists.open-mpi.org
>> https://lists.open-mpi.org/mailman/listinfo/users
>
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Sangam B via users]

Hi,

Here is the sample program snippet:


#include "intrinsic_sizes.h"
#include "redef.h"

module module1_m

  use mod1_m, only:  some__example2
  use mod2_m, only:  some__example3
  use mod3_m, only:  some__example4

  use mpi
  use, intrinsic :: iso_c_binding

implicit none

  private

  public :: some__example___memory

  type, public, extends(some__example5) :: some__example6
 logical, public :: some__example7 = .False.
 class(some__example8), private, pointer :: some__example9
   contains

...
...
end type some__example6

contains

some_pure_functions here



subroutine recv(this,lmb)
class(some__example6), intent(inout) ::  this
integer, intent(in) :: lmb(2,2)

integer :: cs3, ierr
integer(kind=C_LONG) :: size

! receive only from buffer at different process
if(this%is_bf_referred) return

cs3=this%uspecifier%get_recv_buff_3rd_dim_size(this%xb,this%vwb,lmb)
if(cs3.eq.0) return ! nothing to recv

size = this%size_dim(this%gi)*this%size_dim(this%gj)*cs3
if(this%is_exchange_off) then
   call this%update_stats(size)
   this%bf(:,:,1:cs3) = cmplx(0.,0.)
else
   call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
this%nrank,this%tag,this%comm_xvw,this%request,ierr)
end if
  end subroutine recv


Hope this helps.

On Mon, Aug 19, 2019 at 3:21 PM Gilles Gouaillardet via users <
users@lists.open-mpi.org> wrote:

> Thanks,
>
> and your reproducer is ?
>
> Cheers,
>
> Gilles
>
> On Mon, Aug 19, 2019 at 6:42 PM Sangam B via users
>  wrote:
> >
> > Hi,
> >
> > OpenMPI is configured as follows:
> >
> > export CC=`which clang`
> > export CXX=`which clang++`
> > export FC=`which flang`
> > export F90=`which flang`
> >
> > ../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort
> --enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2
> --without-knem --without-libfabric --without-lsf --with-verbs=/usr
> --with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm
> >
> >
> > ..
> >
> > On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:
> >>
> >> Hi,
> >>
> >> I get following error if the application is compiled with openmpi-3.1.1:
> >>
> >> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto
> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops
> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form
> -fdefault-real-8 example_program.F90
> >> F90-S-0155-Could not resolve generic procedure mpi_irecv (
> example_program.F90  : 97)
> >>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
> >>
> >> Following is the line causing this error:
> >>
> >> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> >> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
> >>
> >> The program has following module mentioned in the beginning:
> >>  use mpi
> >>
> >> The openmpi has following module files in lib folder:
> >> $ ls *.mod
> >> mpi_ext.modmpi_f08_ext.mod
>  mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
> >> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod
>  mpi.mod
> >>
> >> The same program works with Intel MPI (gcc/intel as base compilers).
> >> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base
> compilers. What could be the reason for it?
> >>
> >> ..
> >
> > ___
> > users mailing list
> > users@lists.open-mpi.org
> > https://lists.open-mpi.org/mailman/listinfo/users
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
>
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Gilles Gouaillardet via users]

I am not questioning whether you are facing an issue with Open MPI or not.
I am just asking for "the same application" (read minimal source code)
so I can reproduce the issue, investigate it and hopefully help you.

Meanwhile, try rebuilding Open MPI with '-fdefault-real-8' in your
FCFLAGS (since this is what you are using to build your app) and see
whether it helps.

Cheers,

Gilles

Cheers,

Gilles

On Mon, Aug 19, 2019 at 7:06 PM Sangam B via users
 wrote:
>
> Hi,
>
> I've tried both gcc-8.1.0 and AOCC-2.0 compilers with openmpi-3.1.1. It fails 
> for both the compilers.
>
> Posted error message was from OpenMPI-3.1.1 + AOCC-2.0 compiler.
>
> To cross-check whether it is problem with OpenMPI or the base compiler, 
> compiled the same application with Intel MPI using base compiler as (1) intel 
> & (2) gcc. It works for both case.
>
> --
>
>
> On Mon, Aug 19, 2019 at 3:25 PM Gilles Gouaillardet via users 
>  wrote:
>>
>> One more thing ...
>>
>> Your initial message mentioned a failure with gcc 8.2.0, but your
>> follow-up message mentions LLVM compiler.
>>
>> So which compiler did you use to build Open MPI that fails to build your 
>> test ?
>>
>>
>> Cheers,
>>
>> Gilles
>>
>> On Mon, Aug 19, 2019 at 6:49 PM Gilles Gouaillardet
>>  wrote:
>> >
>> > Thanks,
>> >
>> > and your reproducer is ?
>> >
>> > Cheers,
>> >
>> > Gilles
>> >
>> > On Mon, Aug 19, 2019 at 6:42 PM Sangam B via users
>> >  wrote:
>> > >
>> > > Hi,
>> > >
>> > > OpenMPI is configured as follows:
>> > >
>> > > export CC=`which clang`
>> > > export CXX=`which clang++`
>> > > export FC=`which flang`
>> > > export F90=`which flang`
>> > >
>> > > ../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort 
>> > > --enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2 
>> > > --without-knem --without-libfabric --without-lsf --with-verbs=/usr 
>> > > --with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm
>> > >
>> > >
>> > > ..
>> > >
>> > > On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:
>> > >>
>> > >> Hi,
>> > >>
>> > >> I get following error if the application is compiled with openmpi-3.1.1:
>> > >>
>> > >> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto 
>> > >> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops 
>> > >> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form 
>> > >> -fdefault-real-8 example_program.F90
>> > >> F90-S-0155-Could not resolve generic procedure mpi_irecv ( 
>> > >> example_program.F90  : 97)
>> > >>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
>> > >>
>> > >> Following is the line causing this error:
>> > >>
>> > >> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
>> > >> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
>> > >>
>> > >> The program has following module mentioned in the beginning:
>> > >>  use mpi
>> > >>
>> > >> The openmpi has following module files in lib folder:
>> > >> $ ls *.mod
>> > >> mpi_ext.modmpi_f08_ext.mod   
>> > >> mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
>> > >> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod
>> > >>  mpi.mod
>> > >>
>> > >> The same program works with Intel MPI (gcc/intel as base compilers).
>> > >> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base 
>> > >> compilers. What could be the reason for it?
>> > >>
>> > >> ..
>> > >
>> > > ___
>> > > users mailing list
>> > > users@lists.open-mpi.org
>> > > https://lists.open-mpi.org/mailman/listinfo/users
>> ___
>> users mailing list
>> users@lists.open-mpi.org
>> https://lists.open-mpi.org/mailman/listinfo/users
>
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Sangam B via users]

Hi,

I've tried both gcc-8.1.0 and AOCC-2.0 compilers with openmpi-3.1.1. It
fails for both the compilers.

Posted error message was from OpenMPI-3.1.1 + AOCC-2.0 compiler.

To cross-check whether it is problem with OpenMPI or the base compiler,
compiled the same application with Intel MPI using base compiler as (1)
intel & (2) gcc. It works for both case.

--


On Mon, Aug 19, 2019 at 3:25 PM Gilles Gouaillardet via users <
users@lists.open-mpi.org> wrote:

> One more thing ...
>
> Your initial message mentioned a failure with gcc 8.2.0, but your
> follow-up message mentions LLVM compiler.
>
> So which compiler did you use to build Open MPI that fails to build your
> test ?
>
>
> Cheers,
>
> Gilles
>
> On Mon, Aug 19, 2019 at 6:49 PM Gilles Gouaillardet
>  wrote:
> >
> > Thanks,
> >
> > and your reproducer is ?
> >
> > Cheers,
> >
> > Gilles
> >
> > On Mon, Aug 19, 2019 at 6:42 PM Sangam B via users
> >  wrote:
> > >
> > > Hi,
> > >
> > > OpenMPI is configured as follows:
> > >
> > > export CC=`which clang`
> > > export CXX=`which clang++`
> > > export FC=`which flang`
> > > export F90=`which flang`
> > >
> > > ../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort
> --enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2
> --without-knem --without-libfabric --without-lsf --with-verbs=/usr
> --with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm
> > >
> > >
> > > ..
> > >
> > > On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:
> > >>
> > >> Hi,
> > >>
> > >> I get following error if the application is compiled with
> openmpi-3.1.1:
> > >>
> > >> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto
> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops
> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form
> -fdefault-real-8 example_program.F90
> > >> F90-S-0155-Could not resolve generic procedure mpi_irecv (
> example_program.F90  : 97)
> > >>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
> > >>
> > >> Following is the line causing this error:
> > >>
> > >> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> > >> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
> > >>
> > >> The program has following module mentioned in the beginning:
> > >>  use mpi
> > >>
> > >> The openmpi has following module files in lib folder:
> > >> $ ls *.mod
> > >> mpi_ext.modmpi_f08_ext.mod
>  mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
> > >> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod
>mpi.mod
> > >>
> > >> The same program works with Intel MPI (gcc/intel as base compilers).
> > >> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base
> compilers. What could be the reason for it?
> > >>
> > >> ..
> > >
> > > ___
> > > users mailing list
> > > users@lists.open-mpi.org
> > > https://lists.open-mpi.org/mailman/listinfo/users
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
>
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Gilles Gouaillardet via users]

One more thing ...

Your initial message mentioned a failure with gcc 8.2.0, but your
follow-up message mentions LLVM compiler.

So which compiler did you use to build Open MPI that fails to build your test ?


Cheers,

Gilles

On Mon, Aug 19, 2019 at 6:49 PM Gilles Gouaillardet
 wrote:
>
> Thanks,
>
> and your reproducer is ?
>
> Cheers,
>
> Gilles
>
> On Mon, Aug 19, 2019 at 6:42 PM Sangam B via users
>  wrote:
> >
> > Hi,
> >
> > OpenMPI is configured as follows:
> >
> > export CC=`which clang`
> > export CXX=`which clang++`
> > export FC=`which flang`
> > export F90=`which flang`
> >
> > ../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort 
> > --enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2 
> > --without-knem --without-libfabric --without-lsf --with-verbs=/usr 
> > --with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm
> >
> >
> > ..
> >
> > On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:
> >>
> >> Hi,
> >>
> >> I get following error if the application is compiled with openmpi-3.1.1:
> >>
> >> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto 
> >> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops 
> >> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form 
> >> -fdefault-real-8 example_program.F90
> >> F90-S-0155-Could not resolve generic procedure mpi_irecv ( 
> >> example_program.F90  : 97)
> >>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
> >>
> >> Following is the line causing this error:
> >>
> >> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> >> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
> >>
> >> The program has following module mentioned in the beginning:
> >>  use mpi
> >>
> >> The openmpi has following module files in lib folder:
> >> $ ls *.mod
> >> mpi_ext.modmpi_f08_ext.mod   
> >> mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
> >> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod   
> >>   mpi.mod
> >>
> >> The same program works with Intel MPI (gcc/intel as base compilers).
> >> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base 
> >> compilers. What could be the reason for it?
> >>
> >> ..
> >
> > ___
> > users mailing list
> > users@lists.open-mpi.org
> > https://lists.open-mpi.org/mailman/listinfo/users
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Gilles Gouaillardet via users]

Thanks,

and your reproducer is ?

Cheers,

Gilles

On Mon, Aug 19, 2019 at 6:42 PM Sangam B via users
 wrote:
>
> Hi,
>
> OpenMPI is configured as follows:
>
> export CC=`which clang`
> export CXX=`which clang++`
> export FC=`which flang`
> export F90=`which flang`
>
> ../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort 
> --enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2 
> --without-knem --without-libfabric --without-lsf --with-verbs=/usr 
> --with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm
>
>
> ..
>
> On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:
>>
>> Hi,
>>
>> I get following error if the application is compiled with openmpi-3.1.1:
>>
>> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto 
>> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops 
>> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form 
>> -fdefault-real-8 example_program.F90
>> F90-S-0155-Could not resolve generic procedure mpi_irecv ( 
>> example_program.F90  : 97)
>>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
>>
>> Following is the line causing this error:
>>
>> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
>> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
>>
>> The program has following module mentioned in the beginning:
>>  use mpi
>>
>> The openmpi has following module files in lib folder:
>> $ ls *.mod
>> mpi_ext.modmpi_f08_ext.mod   
>> mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
>> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod 
>> mpi.mod
>>
>> The same program works with Intel MPI (gcc/intel as base compilers).
>> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base 
>> compilers. What could be the reason for it?
>>
>> ..
>
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Sangam B via users]

Hi,

OpenMPI is configured as follows:

export CC=`which clang`
export CXX=`which clang++`
export FC=`which flang`
export F90=`which flang`

../configure --prefix=/sw/openmpi/3.1.1/aocc20hpcx210-mpifort
--enable-mpi-fortran --enable-mpi-cxx --without-psm --without-psm2
--without-knem --without-libfabric --without-lsf --with-verbs=/usr
--with-mxm=/sw/hpcx/hpcx-v2.1.0-gcc-MLNX_OFED_LINUX-4.3-1.0.1.0-redhat7.4-x86_64/mxm


..

On Mon, Aug 19, 2019 at 2:43 PM Sangam B  wrote:

> Hi,
>
> I get following error if the application is compiled with openmpi-3.1.1:
>
> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto
> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops
> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form
> -fdefault-real-8 example_program.F90
> F90-S-0155-Could not resolve generic procedure mpi_irecv (
> example_program.F90  : 97)
>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
>
> Following is the line causing this error:
>
> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
>
> The program has following module mentioned in the beginning:
>  use mpi
>
> The openmpi has following module files in lib folder:
> $ ls *.mod
> mpi_ext.modmpi_f08_ext.mod
> mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod
>   mpi.mod
>
> The same program works with Intel MPI (gcc/intel as base compilers).
> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base
> compilers. What could be the reason for it?
>
> ..
>
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Gilles Gouaillardet via users]

Hi,

Can you please post a full but minimal example that evidences the issue?

Also please post your Open MPI configure command line.

Cheers,


Gilles 

Sent from my iPod

> On Aug 19, 2019, at 18:13, Sangam B via users  
> wrote:
> 
> Hi,
> 
> I get following error if the application is compiled with openmpi-3.1.1:
> 
> mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto 
> -J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops 
> -finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form 
> -fdefault-real-8 example_program.F90
> F90-S-0155-Could not resolve generic procedure mpi_irecv ( 
> example_program.F90  : 97)
>   0 inform,   0 warnings,   1 severes, 0 fatal for recv
> 
> Following is the line causing this error:
> 
> call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
> this%nrank,this%tag,this%comm_xvw,this%request,ierr)
> 
> The program has following module mentioned in the beginning:
>  use mpi
> 
> The openmpi has following module files in lib folder:
> $ ls *.mod
> mpi_ext.modmpi_f08_ext.mod   
> mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
> mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod  
>mpi.mod
> 
> The same program works with Intel MPI (gcc/intel as base compilers).
> But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base compilers. 
> What could be the reason for it?
> 
> ..
> ___
> users mailing list
> users@lists.open-mpi.org
> https://lists.open-mpi.org/mailman/listinfo/users
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

[OMPI users] Error with OpenMPI: Could not resolve generic procedure mpi_irecv

[Sangam B via users]

Hi,

I get following error if the application is compiled with openmpi-3.1.1:

mpifort -O3 -march=native -funroll-loops -finline-aggressive -flto
-J./bin/obj_amd64aocc20 -std=f2008 -O3 -march=native -funroll-loops
-finline-aggressive -flto -fallow-fortran-gnu-ext -ffree-form
-fdefault-real-8 example_program.F90
F90-S-0155-Could not resolve generic procedure mpi_irecv (
example_program.F90  : 97)
  0 inform,   0 warnings,   1 severes, 0 fatal for recv

Following is the line causing this error:

call MPI_Irecv(this%bf(:,:,1:cs3),size,MPI_COMPLEX_TYPE,&
this%nrank,this%tag,this%comm_xvw,this%request,ierr)

The program has following module mentioned in the beginning:
 use mpi

The openmpi has following module files in lib folder:
$ ls *.mod
mpi_ext.modmpi_f08_ext.mod
mpi_f08_interfaces.mod  mpi_f08_types.mod  pmpi_f08_interfaces.mod
mpi_f08_callbacks.mod  mpi_f08_interfaces_callbacks.mod  mpi_f08.mod
  mpi.mod

The same program works with Intel MPI (gcc/intel as base compilers).
But fails with OpenMPI, whether gcc-8.1.0 or AOCC are used as base
compilers. What could be the reason for it?

..
___
users mailing list
users@lists.open-mpi.org
https://lists.open-mpi.org/mailman/listinfo/users

Re: [hwloc-users] Netloc feature suggestion

[Brice Goglin]

Hello


Indeed we would like to expose this kind of info but Netloc is
unfornately undermanpowered these days. The code in git master is
outdated. We have a big rework in a branch but it still needs quite a
lot of polishing before being merged


The API is still mostly-scotch-oriented (i.e. for process placement
using communication graphs) because that's pretty-much the only clear
user-request we got in the last years (most people said "we want netloc"
but never gave any idea of what API they actually needed). Of course,
there will be a way to say "I want the entire machine" or "only my
allocated nodes".


The non-scotch API for exposing topology details has been made private
until we understand better what users want. And your request would
definitely help there.


Brice




Le 19/08/2019 à 09:31, Rigel Falcao do Couto Alves a écrit :
>
> Thanks John and Jeff for the replies.
>
>
> Indeed, we are using Slurm here at our cluster; so, for now, I can
> stick with the runtime reading of the network topology's description
> file​, explained here:
>
>
> https://slurm.schedmd.com/topology.conf.html
>
>
> But given the idea of the project is to produce a library that can be
> distributed to anyone in the world, it would still worth it to have a
> way to gather such information on-the-go -- as I can already do
> with /hwloc/​'s topology information. No problem about starting
> simple, i.e. only single-path hierarchies supported in the beginning.
>
>
> The additional /switch/ information (coming from /netloc/) would then
> be added to the graphical output of our tools, allowing users to
> visually analyse how resources placement (both /intra/ and /inter/​
> node) affect their applications.
>
>
>
> 
> *Von:* hwloc-users  im Auftrag
> von John Hearns via hwloc-users 
> *Gesendet:* Freitag, 16. August 2019 07:16
> *An:* Hardware locality user list
> *Cc:* John Hearns
> *Betreff:* Re: [hwloc-users] Netloc feature suggestion
>  
> Hi Rigel. This is very interesting.
> First though I should say - most batch systems have built in node
> grouping utilities.
> PBSPro has bladesets - I think they are called placement groups now.
> I used these when running CFD codes in a Formula 1 team.
> The systems administrator has to set these up manually, using
> knowledge of the switch topology.
> In PBSPro jobs would then 'prefer' to run within the smallest bladeset
> which could accomodate them.
> So you define bladesets for (say) 8/16/24/48 node jobs.
>
> https://pbspro.atlassian.net/wiki/spaces/PD/pages/455180289/Finer+grained+node+grouping
>
> Similarly for Slurm
> https://slurm.schedmd.com/topology.html
>
>
> On Wed, 14 Aug 2019 at 18:53, Rigel Falcao do Couto Alves
> mailto:rigel.al...@tu-dresden.de>> wrote:
>
> Hi,
>
>
> I am doing a PhD in performance analysis of highly parallel CFD
> codes and would like to suggest a feature for Netloc: from
> topic /Build Scotch sub-architectures/
> (at https://www.open-mpi.org/projects/hwloc/doc/v2.0.3/a00329.php),
> create a function-version of /netloc_get_resources/, which could
> retrieve at runtime the network details of the available cluster
> resources (i.e. the nodes allocated to the job). I am mostly
> interested about how many switches (the gray circles in the figure
> below) need to be traversed in order for any pair of
> allocated nodes to communicate with each other:
>
> [removed 200kB image]
>
>
> For example, suppose my job is running within 4 nodes in the
> cluster, illustrated by the numbers above. All I would love to get
> from Netloc - at runtime - is some sort of classification of the
> nodes, like:
>
>
> 1: aa
>
> 2: ab
>
> 3: ba
>
> 4: ca
>
>
> The difference between nodes 1 and 2 is on the last digit, which
> means their MPI communications only need to traverse 1 switch;
> however, between any of them and nodes 3 or 4, the difference
> starts on the second-last digit, which means their communications
> need to traverse two switches. More digits may be left-added to
> the string, per necessity; i.e. if the central gray circle on the
> above figure is connected to another switch, which in turnleads to
> another part of the cluster's structure (with its own switches,
> nodes etc.). For me, it is at the present moment irrelevant
> whether e.g. nodes 1 and 2 are physically - or logically -
> consecutive to each other: /a/, /b/, /c/ etc. would be just
> arbitrary identifiers.
>
>
> I would then use this data to plot the process placement, using
> open-source tools developed here in the University of Dresden
> (Germany); i.e. Scotch is not an option for me. The results of my
> study will be open-source as well and I can gladly share them with
> you once the thesis is finished.
>
>
> I hope I have clearly explained what I have in mind; please let me
>