Any possible remedy to run the job

ajay — Thu, 12 Feb 2009 05:57:11 GMT

Hi,

Here I am usingSGI Altix 450 server with following configuration:

System Model : ALTIX 450 Server with Itanium2 - 32 cores @1.6 GHz

& 64GB memory

OS : SLES 10 with PPK 5

Installed compliers: 1) Intel C++ 10 Compiler Professional

2) Intel Fortron 10 Compiler Professional

I have been trying to execute the below code on SGI Altix 450 server and getting the error Segmentation fault. Where the code is exactly crashing and the snapshot of the error is appended below. The problem occurs at the segment of the code where the first parallel statement is invoked.

Code:

ajay@aquamarine:~/temp> cat runNGRI

#!/bin/sh

export OMP_NUM_THREADS=8

export F_UFMTENDIAN=big

/home/ajay/temp/magLem37p192n08-v01b-Altx < in-test

=============================

Snapshot of the error:

ajay@aquamarine:~/temp> ./runNGRI

!--- PROGRAM MagLem VER_0c 12.08.2007 ---!

! Started at 31:36

! Reading control parameters!

! Reading physical parameters!

! Reading start information!

! Reading output information!

! Reading mantle information!

! Reading inner core information!

! The log-file exists already !

! I add _BIS to the tag and create new files!

INPUT PARAMETERS:

&control

mode = 12 n_time_steps = 1 tag

= b2c03a_BIS

n_tScale = 0 n_lScale = 0 alpha

= 0.60

enscale = 1.00 l_update_v = T l_update_b = T

l_update_s = T l_update_LEM = T dtstart

= 0.0000E+00

dtmax = 0.0000E+00 courfac = 2.00 alffac

= 1.00

n_cour_step = 4 difnu = 0.00 difeta

= 0.00

difkap = 0.00 difLEM = 0.00 ldif = 1

ldifexp = -1 l_correct_AMe = F l_correct_AMz = F

l_non_rot = F l_grav = T l_runTimeLimit= F

runHours = 0 runMinutes = 0 runSeconds

= 0

tEND = 0.0000E+00 nThreadsRun = 8

&phys_param

ra = 0.1200E+06 pr = 0.3000E+00 epsc0 =

0.0000E+00

ra_LEM = 0.0000E+00 pr_LEM = 0.3000E+01 epsc0_LEM =

0.0000E+00

prmag = 0.0000E+00 ek = 0.1000E-02 radratio =

0.3500E+00

Lewis Nr. = 0.1000E+02

ktops = 1 kbots = 1

ktopv = 2 kbotv = 2

ktopb = 1 kbotb = 1

ktopLEM = 1 kbotLEM = 1

Bottom boundary l,m,S:

Top boundary l,m,S:

&start_field

l_start_file= F l_reset_t = F start_file =

no_start_file

inform = -1 scale_s = 1.00 scale_LEM = 1.00

scale_b = 1.00 scale_v = 1.00

init_s1 = 0 amp_s1 = 1.000

init_s2 = 0 amp_s2 = 0.000

init_LEM1 = 0 amp_LEM1 = 1.000

init_LEM2 = 0 amp_LEM2 = 0.000

init_v1 = 0 amp_v1 = 0.000

init_b1 = 0 amp_b1 = 1.000

imagcon = 0 tipdipole = 0.0000E+00

&output_control

n_graph_step = 0 n_graphs = 1

t_graph_start = 0.0000E+00 t_graph_stop = 0.0000E+00 dt_graph

= 0.0000E+00

n_rst_step = 0 n_rsts = 1

t_rst_start = 0.0000E+00 t_rst_stop = 0.0000E+00 dt_rst

= 0.0000E+00

n_stores = 1 n_log_step = 100 n_logs

= 0

t_log_start = 0.0000E+00 t_log_stop = 0.0000E+00 dt_log

= 0.0000E+00

n_p_step = 0 n_ps = 0

t_p_start = 0.0000E+00 t_p_stop = 0.0000E+00 dt_p

= 0.0000E+00

n_spec_step = 0 n_specs = 0

t_spec_start = 0.0000E+00 t_spec_stop = 0.0000E+00 dt_spec

= 0.0000E+00

n_cmb_step = 0 n_cmbs = 0

t_cmb_start = 0.0000E+00 t_cmb_stop = 0.0000E+00 dt_cmb

= 0.0000E+00

l_movie = F n_movie_step = 0

n_movie_frames= 0

t_movie_start = 0.0000E+00 t_movie_stop = 0.0000E+00 dt_movie

= 0.0000E+00

l_ravrg = F l_lFlowScl = F

ngform = 0 ns_r_gra = 1 ns_theta_gra

= 1

ns_phi_gra = 1 ns_r_ic_gra = 1 ns_r_mov

= 1

ns_theta_mov = 1 ns_phi_mov = 1 ns_r_ic_mov

= 1

ns_theta_sur = 2 ns_phi_sur = 2 l_average

= 0

l_cmb_field = 0 l_dt_cmb_field= 0 l_save_out

= T

l_true_time = F lVerbose = T l_rMagSpec

= F

l_DTrMagSpec = F l_max_cmb = F l_max_v_r

= F

l_hel = F l_AM = F l_power

= F

l_drift = F l_TO = F l_TOmovie

= F

l_storeBpot = F l_storeVpot = F l_RMS

= F

&mantle

nRotMa = 0 conductance_ma= 0.0000E+00 rho_ratio_ma

= 0.1000E+01

omega_ma1 = 0.0000E+00 omegaOsz_ma1 = 0.0000E+00 tShift_ma1

= 0.0000E+00

omega_ma2 = 0.0000E+00 omegaOsz_ma2 = 0.0000E+00 tShift_ma2

= 0.0000E+00

&inner_core

nRotIc = 0 sigma_ratio = 0.0000E+00 rho_ratio_ic

= 0.1000E+01

omega_ic1 = 0.0000E+00 omegaOsz_ic1 = 0.0000E+00 tShift_ic1

= 0.0000E+00

omega_ic2 = 0.0000E+00 omegaOsz_ic2 = 0.0000E+00 tShift_ic2

= 0.0000E+00

BIC = 0.0000E+00

! Uneven load balancing in LM blocks!

! Load percentage of last block: 2.60223048327138

1 1 269 269

2 270 538 269

3 539 807 269

4 808 1076 269

5 1077 1345 269

6 1346 1614 269

7 1615 1883 269

8 1884 2145 269

!-- Blocking information:

! Number of LM-blocks: 8

! Size of LM-blocks: 269

! nChunk : 512

! nThreads: 8

! nThreadsRB: 8

! nThreadsMLB: 8

! Number of theta blocks: 6

! size of theta blocks: 16

! ideal size (nfs): 16

! Max thread number available : 32

! Max thread number demanded : 8

! Max thread number compiled : 8

! Number of threads I will use: 8

! Fourier factors= 2 3 4 4

! NO STARTFILE READ, SETTING Z10!

! Using dtMax time step: 0.100000D-04

! Thermo-comp. convection without mag field [conv+heat+comp]

! Normalized OC moment of inertia: 0.143646D+02 ! Normalized IC moment of inertia: 0.758441D-01 ! Normalized MA moment of inertia: 0.284846D+03 ! Normalized IC volume: 0.653962D+00 ! Normalized OC volume: 0.145988D+02

! Grid parameters:

n_r_max = 37 = number of radial grid points

n_cheb_max = 35

max cheb deg.= 34

n_phi_max = 192 = no of longitude grid points

n_theta_max = 96 = no of latitude grid points

n_r_ic_max = 17 = number of radial grid points in IC

n_cheb_ic_max= 14

max cheb deg = 28

l_max = 64 = max degree of Plm

m_max = 64 = max oder of Plm

lm_max = 2145 = no of l/m combinations

minc = 1 = longitude symmetry wave no

! STARTING TIME INTEGRATION AT:

start_time = 0.0000000000D+00

step no = 0

start dt = 1.0000D-05

! STARTING STEP_TIME !

! Starting time integration!

! Starting time step 1

! Starting radial loop!

Entering wallTime

Exiting wallTime

Entering RadialLoop

./runNGRI: line 4: 28576 Segmentation fault (error)

/home/ajay/temp/magLem37p192n08-v01b-Altx

Please suggest possible remedy.

With kind regards,

Ajay

topic Any possible remedy to run the job in Intel® Moderncode for Parallel Architectures

Any possible remedy to run the job