Newsgroups: comp.parallel.mpi
From: rahul@fnfaus.fnal.gov (Rahul Singhal)
Subject: Re: Problem with ANL/MSU implementation of MPI on HP-UX
Organization: Fermi National Accelerator Laboratory, Batavia IL
Date: 10 Aug 1994 17:05:16 GMT
Message-ID: <32b1cc$rfg@fnnews.fnal.gov>

I'm not sure if a previous post of mine made it through or not,
because my terminal crashed.  Anyway here is the problem I'm
having.

I've used the following configure lines:

configure -arch=IRIX -device=ch_p4
configure -arch=IRIX -device=ch_p4 -comm=ch_p4 -bopt=g
configure -arch=IRIX -device=ch_p4 -bopt=g

and have recieved the following error every time:


-------------------------------------------------
making p4 in directory usc/usctest

making p4 in directory include

making p4 in directory bin

making p4 in directory servers
        echo not making server on SGI
not making server on SGI
        (cd lib;  make  MPIARCH=IRIX p4inmpi; cd ..)
        ar ruv ../../../../lib/IRIX/ch_p4/libmpi.a p4_globals.o p4_MD.o p4_error
.o p4_args.o p4_alloc.o p4_debug.o p4_procgroup.o p4_tsr.o p4_mon.o p4_broadcast
.o p4_utils.o p4_bm.o p4_rm.o p4_sock_util.o p4_sock_cr.o p4_sock_conn.o p4_sock
_sr.o p4_sock_list.o p4_secure.o p4_shmem_sr.o
a - p4_globals.o
a - p4_MD.o
a - p4_error.o
a - p4_args.o
a - p4_alloc.o
a - p4_debug.o
a - p4_procgroup.o
a - p4_tsr.o
a - p4_mon.o
a - p4_broadcast.o
a - p4_utils.o
a - p4_bm.o
*** Error code 1 (bu21)
*** Error code 1 (bu21)
*** Error code 1 (bu21)
*** Error code 1 (bu21)
*** Error code 1 (bu21)
----------------------------------------------
ar: Error: on write -: No space left on device

make: fatal error.

make: fatal error.

make: fatal error.

make: fatal error.

make: fatal error.
----------------------------------------------
Is there a problem with my configure parameters.  I'm compiling
under version 5.2 for SGIs, if that makes any difference at all.
I have successfully compiled under 4.0.5, but I am now even having
the same error with those machines.  I'm using mpich-Jul22.tar.Z
as the source.  Has there been a change to this source which is
preventing compilation for SGIs?  or am I making an obvious mistake
somewhere along the way, any help would be appreciated.  Thank you
for your time.

Rahul Singhal





