Open MPI logo

Open MPI User's Mailing List Archives

  |   Home   |   Support   |   FAQ   |  

This web mail archive is frozen.

This page is part of a frozen web archive of this mailing list.

You can still navigate around this archive, but know that no new mails have been added to it since July of 2016.

Click here to be taken to the new web archives of this list; it includes all the mails that are in this frozen archive plus all new mails that have been sent to the list since it was migrated to the new archives.

Subject: [OMPI users] error when sending a single 'long double' from an amd64 architecture to a ppc and a sparc machine
From: Patrick Heckeler (pheckeler_at_[hidden])
Date: 2008-04-02 19:55:04


Hi everybody,

I tried to submit a single 'long double' from an amd64 architecture to
a ppc, from an amd64 to a sparc machine and from ppc to sparc. The
testprogram and the results are shown below. All other basic datatypes
I have tested are working fine with heterogenous communication. Does
anybody have an idea? I am using openMPI 1.2.4.

Testprogramm:
#include <mpi.h>
#include <iostream>

int main(int argc, char *argv[])
{
  int me, numprocs;
  long double senddata = 3.14;
  long double recvdata;
  MPI_Status status;
  MPI_Init(&argc, &argv);

  MPI_Comm_size(MPI_COMM_WORLD, &numprocs);
  MPI_Comm_rank(MPI_COMM_WORLD, &me);

  if(me==0)
  {
    std::cout << "Long double to send: " << senddata << std::endl;
    MPI_Send(&senddata, 1, MPI_LONG_DOUBLE, 1, 123, MPI_COMM_WORLD);
  }
  else
  {
    MPI_Recv(&recvdata, 1, MPI_LONG_DOUBLE, 0, 123, MPI_COMM_WORLD, &status);
    std::cout << "Long double received: " << recvdata << std::endl;
  }
  MPI_Finalize();
}

--
Output from amd64 to ppc:
unpack there is still room in the input buffer 8 bytes
Long double received: 8.19027e+3364
Long double to send: 3.14
Output from amd64 to sparc:
Long double to send: 3.14
[hannibal:17063] *** An error occurred in MPI_Recv
[hannibal:17063] *** on communicator MPI_COMM_WORLD
[hannibal:17063] *** MPI_ERR_TRUNCATE: message truncated
[hannibal:17063] *** MPI_ERRORS_ARE_FATAL (goodbye)
mpirun noticed that job rank 0 with PID 21916 on node murdock exited
on signal 15 (Terminated).
Output from ppc to sparc:
Long double to send: 3.14
Long double received: 1146.88
*********************************************************
Information of the amd64 machine:
configure:
./configure --prefix=/usr/local/ CFLAGS=-m32 CXXFLAGS=-m32 FFLAGS=-m32
FCFLAGS=-m32 --with-wrapper-cflags=-m32 --with-wrapper-cxxflags=-m32
--with-wrapper-fflags=-m32 --with-wrapper-fcflags=-m32
uname -a:
Linux hannibal 2.6.22-14-server #1 SMP Tue Dec 18 05:52:24 UTC 2007
x86_64 GNU/Linux
less /proc/cpuinfo:
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 33
model name      : Dual Core AMD Opteron(tm) Processor 275
stepping        : 2
cpu MHz         : 2193.774
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
fxsr_opt lm 3dnowext 3dnow pni lahf_lm cmp_legacy
bogomips        : 4390.91
TLB size        : 1024 4K pages
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp
processor       : 1
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 33
model name      : Dual Core AMD Opteron(tm) Processor 275
stepping        : 2
cpu MHz         : 2193.774
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 1
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext
fxsr_opt lm 3dnowext 3dnow pni lahf_lm cmp_legacy
bogomips        : 4387.61
TLB size        : 1024 4K pages
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp
processor       : 2
************************************************************
information of the ppc:
configure:
./configure --prefix=/usr/local --enable-static
--enable-ltdl-convenience --no-create --no-recursion
uname -a:
Linux baracus 2.6.18-5-powerpc #1 Sat Dec 22 21:05:48 UTC 2007 ppc GNU/Linux
less /proc/cpuinfo:
processor       : 0
cpu             : 7410, altivec supported
temperature     : 23-25 C (uncalibrated)
clock           : 533.333332MHz
revision        : 0.3 (pvr 800c 1103)
bogomips        : 66.30
timebase        : 33290001
platform        : PowerMac
machine         : PowerMac3,4
motherboard     : PowerMac3,4 MacRISC2 MacRISC Power Macintosh
detected as     : 69 (PowerMac G4 Silver)
pmac flags      : 00000010
L2 cache        : 1024K unified
pmac-generation : NewWorld
************************************************************
information of the sparc:
configure:
./configure CFLAGS=-mv8plus CXXFLAGS=-mv8plus FFLAGS=-mv8plus
FCFLAGS=-mv8plus --prefix=/usr/local --enable-static
--enable-ltdl-convenience --no-create --no-recursion
uname -a:
Linux murdock 2.6.22-14-sparc64-smp #1 SMP Tue Dec 18 05:40:10 UTC
2007 sparc64 GNU/Linux
less /proc/cpuinfo:
cpu             : TI UltraSparc III (Cheetah)
fpu             : UltraSparc III integrated FPU
prom            : OBP 4.16.4 2004/12/18 05:18
type            : sun4u
ncpus probed    : 2
ncpus active    : 2
D$ parity tl1   : 0
I$ parity tl1   : 0
Cpu0ClkTck      : 000000002cb41780
Cpu1ClkTck      : 000000002cb41780
MMU Type        : Cheetah
State:
CPU0:           online
CPU1:           online
Thanks in advance for your help
-- 
Viele Grüße / Kind regards
Patrick Heckeler
Teckstr. 6
72074 Tübingen
Germany
Mobil: +49172 / 719 23 24
email: pheckeler [at] gmail [dot] com
skype: p_heckeler
ICQ: 200910972