[mpich-discuss] Nemesis engine query by Viswanath

Halim Amer aamer at anl.gov
Fri Jul 31 09:43:06 CDT 2015


You can learn more about multithreading support in MPI from the standard 
(www.mpi-forum.org/docs/mpi-3.1/mpi31-report.pdf) in section 12.4.

If you need us to help, we need more information such as configure line, 
target platform, compiler version, and most importantly, a toy program 
that reproduces the bug.

--Halim


On 7/31/15 4:41 AM, Viswanath Krishnamurthy wrote:
> I did try initializing Multithreading support
> #include<mpi.h>
> int main()
> {
>
>          int provided;
>          int returnvalue =
> MPI_Init_thread(&argc,&argv,MPI_THREAD_MULTIPLE,&provided);
> if(provided < MPI_THREAD_MULTIPLE)
>          {
>                  printf("\n THREAD LIBRARY DOESN'T HAVE MULTITHREADING
> SUPPORT:");
>                  exit(1);
>          }
> }
>
> The code compiles but throws an error -
> Assertion failed in file
> /home/viswa/libraries/mpich-3.1.4/src/include/mpiimplthreadpost.h at
> line 163: depth > 0 && depth < 10
> internal ABORT - process 1
> internal ABORT - process 0
>
> Could you please refer me to some documentation for mpi_init_thread/
> MPICH-multithreading documentation as I am relatively new to it.
>
> Thanks,
> Viswanath
>
>
>
> On Fri, Jul 31, 2015 at 2:41 AM, <discuss-request at mpich.org
> <mailto:discuss-request at mpich.org>> wrote:
>
>     Send discuss mailing list submissions to
>     discuss at mpich.org <mailto:discuss at mpich.org>
>
>     To subscribe or unsubscribe via the World Wide Web, visit
>     https://lists.mpich.org/mailman/listinfo/discuss
>     or, via email, send a message with subject or body 'help' to
>     discuss-request at mpich.org <mailto:discuss-request at mpich.org>
>
>     You can reach the person managing the list at
>     discuss-owner at mpich.org <mailto:discuss-owner at mpich.org>
>
>     When replying, please edit your Subject line so it is more specific
>     than "Re: Contents of discuss digest..."
>
>
>     Today's Topics:
>
>         1.  PANFS Remove RAID0 and add RAIDN to MPICH 3.2 (Victorelli, Ron)
>         2. Re:  PANFS Remove RAID0 and add RAIDN to MPICH 3.2 (Rob Latham)
>         3. Re:  hydra, stdin close(), and SLURM (Aaron Knister)
>         4. Re:  Nemesis engine (Viswanath Krishnamurthy)
>         5. Re:  Nemesis engine (Halim Amer)
>         6.  Active loop in MPI_Waitany? (Dorier, Matthieu)
>         7. Re:  Active loop in MPI_Waitany? (Jeff Hammond)
>
>
>     ----------------------------------------------------------------------
>
>     Message: 1
>     Date: Wed, 29 Jul 2015 12:07:40 +0000
>     From: "Victorelli, Ron" <rvictorelli at panasas.com
>     <mailto:rvictorelli at panasas.com>>
>     To: "discuss at mpich.org <mailto:discuss at mpich.org>"
>     <discuss at mpich.org <mailto:discuss at mpich.org>>
>     Subject: [mpich-discuss] PANFS Remove RAID0 and add RAIDN to MPICH 3.2
>     Message-ID:
>
>     <BN3PR08MB12888AACA7907A8ACED7531AA18C0 at BN3PR08MB1288.namprd08.prod.outlook.com
>     <mailto:BN3PR08MB12888AACA7907A8ACED7531AA18C0 at BN3PR08MB1288.namprd08.prod.outlook.com>>
>
>     Content-Type: text/plain; charset="us-ascii"
>
>     I am a developer at Panasas, and we would like to provide a patch that
>     removes RAID0 support and adds RAIDN support to romio (MPICH 3.2):
>
>     src/mpi/romio/adio/ad_panfs/ad_panfs_open.c
>
>     I currently do not have an MCS or trac account.
>
>     Thank You
>
>     Ron Victorelli
>     Software Engineer
>     Panasas, Inc
>     Email: rvictorelli at panasas.com
>     <mailto:rvictorelli at panasas.com><mailto:rvictorelli at panasas.com
>     <mailto:rvictorelli at panasas.com>>
>     Tel: 412 -323-6422 <tel:412%20-323-6422>
>     www.panasas.com <http://www.panasas.com><http://www.panasas.com>
>     [Panasas_Logo_LR]
>     -------------- next part --------------
>     An HTML attachment was scrubbed...
>     URL:
>     <http://lists.mpich.org/pipermail/discuss/attachments/20150729/00b240fc/attachment-0001.html>
>     -------------- next part --------------
>     A non-text attachment was scrubbed...
>     Name: image001.jpg
>     Type: image/jpeg
>     Size: 3610 bytes
>     Desc: image001.jpg
>     URL:
>     <http://lists.mpich.org/pipermail/discuss/attachments/20150729/00b240fc/attachment-0001.jpg>
>
>     ------------------------------
>
>     Message: 2
>     Date: Wed, 29 Jul 2015 15:52:31 -0500
>     From: Rob Latham <robl at mcs.anl.gov <mailto:robl at mcs.anl.gov>>
>     To: <discuss at mpich.org <mailto:discuss at mpich.org>>,
>     <rvictorelli at panasas.com <mailto:rvictorelli at panasas.com>>
>     Subject: Re: [mpich-discuss] PANFS Remove RAID0 and add RAIDN to MPICH
>              3.2
>     Message-ID: <55B93D0F.7040703 at mcs.anl.gov
>     <mailto:55B93D0F.7040703 at mcs.anl.gov>>
>     Content-Type: text/plain; charset="windows-1252"; format=flowed
>
>
>
>     On 07/29/2015 07:07 AM, Victorelli, Ron wrote:
>      > I am a developer at Panasas, and we would like to provide a patch
>     that
>      >
>      > removes RAID0 support and adds RAIDN support to romio (MPICH 3.2):
>      >
>      > src/mpi/romio/adio/ad_panfs/ad_panfs_open.c
>      >
>      > I currently do not have an MCS or trac account.
>
>     Hi Ron.  I'm pleased to have contributions from Panasas.   It's your
>     first since 2007!
>
>     If you've got a lot of patches in the works, maybe we should go ahead
>     and set you up with a trac account and/or a git tree.
>
>     If you're just looking to get this patch into the tree, that's fine too:
>     it's definitely easier and you will just need to 'git format-patch' your
>     changes and email them to me.
>
>     ==rob
>
>      >
>      > Thank You
>      >
>      > Ron Victorelli
>      >
>      > Software Engineer
>      >
>      > Panasas, Inc
>      >
>      > Email: rvictorelli at panasas.com <mailto:rvictorelli at panasas.com>
>     <mailto:rvictorelli at panasas.com <mailto:rvictorelli at panasas.com>>
>      >
>      > Tel: 412 -323-6422 <tel:412%20-323-6422>
>      >
>      > www.panasas.com <http://www.panasas.com> <http://www.panasas.com>
>      >
>      > Panasas_Logo_LR
>      >
>      >
>      >
>      > _______________________________________________
>      > discuss mailing list discuss at mpich.org <mailto:discuss at mpich.org>
>      > To manage subscription options or unsubscribe:
>      > https://lists.mpich.org/mailman/listinfo/discuss
>      >
>
>     --
>     Rob Latham
>     Mathematics and Computer Science Division
>     Argonne National Lab, IL USA
>
>
>     ------------------------------
>
>     Message: 3
>     Date: Wed, 29 Jul 2015 17:14:53 -0400
>     From: Aaron Knister <aaron.s.knister at nasa.gov
>     <mailto:aaron.s.knister at nasa.gov>>
>     To: <discuss at mpich.org <mailto:discuss at mpich.org>>
>     Subject: Re: [mpich-discuss] hydra, stdin close(), and SLURM
>     Message-ID: <55B9424D.6030104 at nasa.gov
>     <mailto:55B9424D.6030104 at nasa.gov>>
>     Content-Type: text/plain; charset="windows-1252"; Format="flowed"
>
>     Thanks Pavan!
>
>     -Aaron
>
>     On 7/28/15 3:23 PM, Balaji, Pavan wrote:
>      > Hi Aaron,
>      >
>      > I've committed it to mpich/master:
>      >
>      >
>     http://git.mpich.org/mpich.git/commitdiff/6b41775b2056ff18b3c28aab71764e35904c00fa
>      >
>      > Thanks for the contribution.
>      >
>      > This should be in tonight's nightlies:
>      >
>      > http://www.mpich.org/static/downloads/nightly/master/mpich/
>      >
>      > ... and in the upcoming mpich-3.2rc1 release.
>      >
>      >    -- Pavan
>      >
>      >
>      >
>      >
>      > On 7/27/15, 1:40 PM, "Balaji, Pavan" <balaji at anl.gov
>     <mailto:balaji at anl.gov>> wrote:
>      >
>      >> Hi Aaron,
>      >>
>      >>
>      >>
>      >> Please send the patch to me directly.
>      >>
>      >> General guidelines as to the kind of patches we ask for:
>      >>
>      >> https://wiki.mpich.org/mpich/index.php/Version_Control_Systems_101
>      >>
>      >> You can ignore the git workflow related text, which is for our
>     internal testing.  I'll take care of that for you.
>      >>
>      >> Thanks,
>      >>
>      >>   -- Pavan
>      >>
>      >> On 7/27/15, 1:36 PM, "Aaron Knister" <aaron.s.knister at nasa.gov
>     <mailto:aaron.s.knister at nasa.gov>> wrote:
>      >>
>      >>> Hi Pavan,
>      >>>
>      >>> I see your reply in the archives but it didn't make it to my
>     inbox so
>      >>> I'm replying to my post. I don't disagree without you about the
>     error
>      >>> being in the SLURM code, but I'm not sure how one would prevent
>     this
>      >>> reliably. SLURM has no expectation that an external library
>     will open
>      >>> something at file descriptor 0 before it reaches the point in
>     the code
>      >>> where it's ready to poll for stdin. Do you have any suggestions?
>      >>>
>      >>> It's been a long while since I've done a git e-mail patch so it
>     might
>      >>> take me a bit to figure out. Should I send the patch to the
>     list or to
>      >>> you directly?
>      >>>
>      >>> Thanks!
>      >>>
>      >>> -Aaron
>      >>>
>      >>> On 7/25/15 10:26 PM, Aaron Knister wrote:
>      >>>> I sent this off to the mvapich list yesterday and it was
>     suggested I
>      >>>> raise it here since this is the upstream project:
>      >>>>
>      >>>> This is a bit of a cross post from a thread I started on the
>     slurm dev
>      >>>> list:
>     http://article.gmane.org/gmane.comp.distributed.slurm.devel/8176
>      >>>>
>      >>>> I'd like to get feedback on the idea that "--input none" be
>     passed to
>      >>>> srun when using the SLURM hydra bootstrap mechanism. I figured it
>      >>>> would be inserted somewhere around here
>      >>>>
>     http://trac.mpich.org/projects/mpich/browser/src/pm/hydra/tools/bootstrap/external/slurm_launch.c#L98.
>      >>>>
>      >>>>
>      >>>> Without this argument I'm getting spurious job aborts and
>     confusing
>      >>>> errors. The gist of it is mpiexec.hydra closes stdin before it
>     exec's
>      >>>> srun. srun then (possibly via the munge libraries) calls some
>     function
>      >>>> that does a look up via nss. We use sssd for AAA so
>     libnss_sssd will
>      >>>> handle this request. Part of the caching mechanism sssd uses will
>      >>>> cause the library to open() the cache file. The lowest fd
>     available is
>      >>>> 0 so the cache file is opened on fd 0. srun then believes it's got
>      >>>> stdin attached and it causes the issues outlined in the slurm dev
>      >>>> post. I think passing "--input none" is the right thing to do here
>      >>>> since hydra has in fact closed stdin to srun. I tested this
>     via the
>      >>>> HYDRA_LAUNCHER_EXTRA_ARGS environment variable and it does
>     resolve the
>      >>>> errors I described.
>      >>>>
>      >>>> Thanks!
>      >>>> -Aaron
>      >>>>
>      >>> --
>      >>> Aaron Knister
>      >>> NASA Center for Climate Simulation (Code 606.2)
>      >>> Goddard Space Flight Center
>      >>> (301) 286-2776 <tel:%28301%29%20286-2776>
>      >>>
>      >>>
>      >> _______________________________________________
>      >> discuss mailing list discuss at mpich.org <mailto:discuss at mpich.org>
>      >> To manage subscription options or unsubscribe:
>      >> https://lists.mpich.org/mailman/listinfo/discuss
>      > _______________________________________________
>      > discuss mailing list discuss at mpich.org <mailto:discuss at mpich.org>
>      > To manage subscription options or unsubscribe:
>      > https://lists.mpich.org/mailman/listinfo/discuss
>
>     --
>     Aaron Knister
>     NASA Center for Climate Simulation (Code 606.2)
>     Goddard Space Flight Center
>     (301) 286-2776 <tel:%28301%29%20286-2776>
>
>
>     -------------- next part --------------
>     A non-text attachment was scrubbed...
>     Name: signature.asc
>     Type: application/pgp-signature
>     Size: 842 bytes
>     Desc: OpenPGP digital signature
>     URL:
>     <http://lists.mpich.org/pipermail/discuss/attachments/20150729/0280399d/attachment-0001.pgp>
>
>     ------------------------------
>
>     Message: 4
>     Date: Thu, 30 Jul 2015 17:40:35 +0300
>     From: Viswanath Krishnamurthy <writetoviswa at gmail.com
>     <mailto:writetoviswa at gmail.com>>
>     To: discuss at mpich.org <mailto:discuss at mpich.org>
>     Subject: Re: [mpich-discuss] Nemesis engine
>     Message-ID:
>
>     <CADhQ-jDZix3e2TmPAPjX3O7GO+Z7vOzSphPgc4Py+B=eRGBypA at mail.gmail.com
>     <mailto:eRGBypA at mail.gmail.com>>
>     Content-Type: text/plain; charset="utf-8"
>
>     Hi All,
>
>     I am currently working on MPICH-version 3.1.4 on Ubuntu..
>     where I get an error stating that
>
>     Assertion failed in
>     file src/mpid/ch3/channels/nemesis/src/ch3_progress.c at Line 252
>
>     The actual problem which I face is that currently even if MPI_Sends have
>     already been dispatched, certain nodes keep waiting for MPI_Recvs which
>     never arrive at all(Using multithreading). When I referred the
>     internet, my
>     understanding is that nemesis is written to handle only one thread
>     receive.
>     Please let me know about the latest patch for nemesis engine or the
>     mpich
>     version which has the changes.
>     *src/mpid/ch3/channels/nemesis/src/ch3_progress.c *
>
>     Thanks,
>     Viswanath
>     -------------- next part --------------
>     An HTML attachment was scrubbed...
>     URL:
>     <http://lists.mpich.org/pipermail/discuss/attachments/20150730/c9d4e337/attachment-0001.html>
>
>     ------------------------------
>
>     Message: 5
>     Date: Thu, 30 Jul 2015 09:51:40 -0500
>     From: Halim Amer <aamer at anl.gov <mailto:aamer at anl.gov>>
>     To: <discuss at mpich.org <mailto:discuss at mpich.org>>
>     Subject: Re: [mpich-discuss] Nemesis engine
>     Message-ID: <55BA39FC.20406 at anl.gov <mailto:55BA39FC.20406 at anl.gov>>
>     Content-Type: text/plain; charset="windows-1252"; format=flowed
>
>     Hi Viswanath,
>
>     Nemesis supports multithreading. Have you initialized the MPI
>     environment with MPI_THREAD_MULTIPLE threading support?
>
>     If you still see the problem after the above initialization, please send
>     us a minimal example code that reproduces it.
>
>     Thank you,
>     --Halim
>
>     Abdelhalim Amer (Halim)
>     Postdoctoral Appointee
>     MCS Division
>     Argonne National Laboratory
>
>     On 7/30/15 9:40 AM, Viswanath Krishnamurthy wrote:
>      > Hi All,
>      >
>      > I am currently working on MPICH-version 3.1.4 on Ubuntu..
>      > where I get an error stating that
>      >
>      > Assertion failed in
>      > file src/mpid/ch3/channels/nemesis/src/ch3_progress.c at Line 252
>      > *
>      > *
>      > The actual problem which I face is that currently even if
>     MPI_Sends have
>      > already been dispatched, certain nodes keep waiting for MPI_Recvs
>     which
>      > never arrive at all(Using multithreading). When I referred the
>     internet,
>      > my understanding is that nemesis is written to handle only one thread
>      > receive. Please let me know about the latest patch for nemesis
>     engine or
>      > the mpich version which has the changes.
>      > *src/mpid/ch3/channels/nemesis/src/ch3_progress.c *
>      >
>      > Thanks,
>      > Viswanath
>      >
>      >
>      >
>      > _______________________________________________
>      > discuss mailing list discuss at mpich.org <mailto:discuss at mpich.org>
>      > To manage subscription options or unsubscribe:
>      > https://lists.mpich.org/mailman/listinfo/discuss
>      >
>
>
>     ------------------------------
>
>     Message: 6
>     Date: Thu, 30 Jul 2015 21:09:04 +0000
>     From: "Dorier, Matthieu" <mdorier at anl.gov <mailto:mdorier at anl.gov>>
>     To: "discuss at mpich.org <mailto:discuss at mpich.org>"
>     <discuss at mpich.org <mailto:discuss at mpich.org>>
>     Subject: [mpich-discuss] Active loop in MPI_Waitany?
>     Message-ID: <37142D5FC373A846ACE4F75AA11EA84D21BA0122 at DITKA.anl.gov
>     <mailto:37142D5FC373A846ACE4F75AA11EA84D21BA0122 at DITKA.anl.gov>>
>     Content-Type: text/plain; charset="iso-8859-1"
>
>     Hi,
>
>     I have a code that looks like this:
>
>     while(true) {
>         do some I/O (HDF5 POSIX output to a remote, parallel file system)
>         wait for communication (MPI_Waitany) from other processes (in
>     the same node and outside the node)
>     }
>
>     I'm measuring the energy consumption of the node that runs this
>     process for the same duration, as a function of the amount of data
>     written in each I/O operation.
>     Surprisingly, the larger the I/O in proposition to the
>     communication, the lower the energy consumption. In other words, the
>     longer I wait in MPI_Waitany, the more I consume.
>
>     Does anyone have a good explanation for that? Is there an active
>     loop in MPI_Waitany? Another reason?
>
>     Thanks!
>
>     Matthieu
>     -------------- next part --------------
>     An HTML attachment was scrubbed...
>     URL:
>     <http://lists.mpich.org/pipermail/discuss/attachments/20150730/6777b87d/attachment-0001.html>
>
>     ------------------------------
>
>     Message: 7
>     Date: Thu, 30 Jul 2015 19:41:21 -0400
>     From: Jeff Hammond <jeff.science at gmail.com
>     <mailto:jeff.science at gmail.com>>
>     To: "discuss at mpich.org <mailto:discuss at mpich.org>"
>     <discuss at mpich.org <mailto:discuss at mpich.org>>
>     Subject: Re: [mpich-discuss] Active loop in MPI_Waitany?
>     Message-ID:
>
>     <CAGKz=uJr5NmO+csEBDOtk67zz+HDEaax_JoLNzHWswZipcPCyA at mail.gmail.com
>     <mailto:uJr5NmO%2BcsEBDOtk67zz%2BHDEaax_JoLNzHWswZipcPCyA at mail.gmail.com>>
>     Content-Type: text/plain; charset="utf-8"
>
>     Seems obvious that Waitany spins on the array of requests until one
>     completes. Is that an active loop by your definition?
>
>     Jeff
>
>     On Thursday, July 30, 2015, Dorier, Matthieu <mdorier at anl.gov
>     <mailto:mdorier at anl.gov>> wrote:
>
>      > Hi,
>      >
>      > I have a code that looks like this:
>      >
>      > while(true) {
>      >    do some I/O (HDF5 POSIX output to a remote, parallel file system)
>      >    wait for communication (MPI_Waitany) from other processes (in
>     the same
>      > node and outside the node)
>      > }
>      >
>      > I'm measuring the energy consumption of the node that runs this
>     process
>      > for the same duration, as a function of the amount of data
>     written in each
>      > I/O operation.
>      > Surprisingly, the larger the I/O in proposition to the
>     communication, the
>      > lower the energy consumption. In other words, the longer I wait in
>      > MPI_Waitany, the more I consume.
>      >
>      > Does anyone have a good explanation for that? Is there an active
>     loop in
>      > MPI_Waitany? Another reason?
>      >
>      > Thanks!
>      >
>      > Matthieu
>      >
>
>
>     --
>     Jeff Hammond
>     jeff.science at gmail.com <mailto:jeff.science at gmail.com>
>     http://jeffhammond.github.io/
>     -------------- next part --------------
>     An HTML attachment was scrubbed...
>     URL:
>     <http://lists.mpich.org/pipermail/discuss/attachments/20150730/8390a38a/attachment.html>
>
>     ------------------------------
>
>     _______________________________________________
>     discuss mailing list
>     discuss at mpich.org <mailto:discuss at mpich.org>
>     https://lists.mpich.org/mailman/listinfo/discuss
>
>     End of discuss Digest, Vol 33, Issue 10
>     ***************************************
>
>
>
>
> _______________________________________________
> discuss mailing list     discuss at mpich.org
> To manage subscription options or unsubscribe:
> https://lists.mpich.org/mailman/listinfo/discuss
>
_______________________________________________
discuss mailing list     discuss at mpich.org
To manage subscription options or unsubscribe:
https://lists.mpich.org/mailman/listinfo/discuss


More information about the discuss mailing list