mwill at penguincomputing.com
Fri Jun 4 09:24:35 PDT 2004
Maybe there is a bug in your code that is triggered in a 3-cpu configuration.
SIGSEGV means segfault, typically an attempt to access memory that does not
belong to you because a pointer has not been initialized.
I don't think it is an MPI problem, bur rather a bug in your c code...
assuming that is what you use ;-)
Did you write the bandpass filter yourself or is it a standard app ?
On Tuesday 01 June 2004 04:25 am, kalyani goli wrote:
> Iam executing parallel program on LINUX , rehat
> version 8, with MPICH-1.2.5 , ch_p4 device.
> While executing the program, i got the following
> could you please tell me what is happening.
> The program is band pass filtering. When iam executing
> with 2 processors, it is getting executed. But with 3
> processors, this error is coming.
> p1_4680: p4_error: net_recv read: probable EOF on
> socket: 1
> bm_list_7292: (0.711428) wakeup_slave: unable to
> interrupt slave 0 pid 7291
> p2_1495: p4_error: interrupt SIGSEGV: 11
> Thanks in advance
> Yahoo! India Matrimony: Find your partner online.
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit
Michael Will, Linux Sales Engineer
NEWS: We have moved to a larger iceberg :-)
NEWS: 300 California St., San Francisco, CA.
Tel: 415-954-2822 Toll Free: 888-PENGUIN
More information about the Beowulf