[mpich-discuss] Solaris 10 mpich2 dup2 error

[Pavan Balaji]

Sorry, the patch had a typo in it. Please use both r9011 and r9012.

https://trac.mcs.anl.gov/projects/mpich2/changeset/9012

  -- Pavan

On 09/14/2011 07:55 PM, Pavan Balaji wrote:
> Hi Christopher,
>
> I think this should be fixed in r9011. Can you try this patch to see if
> it works for you?
>
> https://trac.mcs.anl.gov/projects/mpich2/changeset/9011
>
>    -- Pavan
>
> On 09/13/2011 01:55 PM, Christopher Jackson wrote:
>> I will be running mph programs on multiple machines. I was simply using the aforementioned command to illustrate the problem I was having. I get the same error when running on multiple machines. But thank you for the response. Solaris is not my first choice for an OS but unfortunately it is what is installed in the CS lab on our campus. I hope to hear more suggestions on possible fixes.
>>
>> Regards,
>> Christopher Jackson
>>
>>
>> On Sep 13, 2011, at 2:49 PM, Rajeev Thakur wrote:
>>
>>> We don't test regularly on Solaris. If you are running on a single machine, try the gforker process manager. (Configure with --pm=gforker). Also, you don't need a hosts file if you are running on a single machine.
>>>
>>> Rajeev
>>>
>>>
>>> On Sep 12, 2011, at 9:26 PM, Christopher Jackson wrote:
>>>
>>>> Why would I get the following error messages when testing out a freshly installed copy of mpich2-1.4.1p1 on Solaris10. Mpich2 seemed to install fine. I have included some information that I thought would be helpful in determining the issue. Thanks in advance for your help.
>>>>
>>>> Error received:
>>>>
>>>>
>>>> clj at cantor% mpiexec -n 1 -f hosts date
>>>> [proxy:0:0 at cantor] HYDU_create_process (./utils/launch/launch.c:36): dup2 error (Bad file number)
>>>> [proxy:0:0 at cantor] launch_procs (./pm/pmiserv/pmip_cb.c:751): create process returned error
>>>> [proxy:0:0 at cantor] HYD_pmcd_pmip_control_cmd_cb (./pm/pmiserv/pmip_cb.c:935): launch_procs returned error
>>>> [proxy:0:0 at cantor] HYDT_dmxu_poll_wait_for_event (./tools/demux/demux_poll.c:77): callback returned error status
>>>> [proxy:0:0 at cantor] main (./pm/pmiserv/pmip.c:226): demux engine error waiting for event
>>>>
>>>> clj at cantor% date
>>>> Mon Sep 12 02:32:41 EDT 2011
>>>>
>>>> clj5096 at cantor% cat hosts
>>>> cantor
>>>>
>>>> clj at cantor% uname -v
>>>> Generic_144489-17
>>>>
>>>> clj at cantor% uname -a
>>>> SunOS cantor 5.10 Generic_144489-17 i86pc i386 i86pc Solaris
>>>>
>>>> clj at cantor% uname -n
>>>> cantor
>>>>
>>>> The configure line I used was as follows:
>>>> ./configure CFLAGS=-Wa,--divide --enable-fast=O3 --enable-cxx --disable-f77 --disable-fc --prefix=/home/clj5096/cluster/mpich2-install |&   tee c.txt
>>>>
>>>> <c.txt><install.log><make.log>_______________________________________________
>>>> mpich-discuss mailing list     mpich-discuss at mcs.anl.gov
>>>> To manage subscription options or unsubscribe:
>>>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>>>
>>> _______________________________________________
>>> mpich-discuss mailing list     mpich-discuss at mcs.anl.gov
>>> To manage subscription options or unsubscribe:
>>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>>
>> _______________________________________________
>> mpich-discuss mailing list     mpich-discuss at mcs.anl.gov
>> To manage subscription options or unsubscribe:
>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>

-- 
Pavan Balaji
http://www.mcs.anl.gov/~balaji