[gmx-developers] Zero cell size
Roland Schulz
schulzr at ornl.gov
Tue Jul 22 03:20:08 CEST 2008
Hi,
I narrowed down the problem. With gcc (4.2.0 and 4.2.1) and pgi (7.2.2) it
works without a problem. Only with pgi (7.1.6) the problem happens. Thus
either there is code which doesn't conform to the standard or there is a bug
in the this pgi version.
With pgi (7.1.6) I get the problem from the first mail with dlb:
Step 0: The X-size (0.000000) times the triclinic skew factor (1.000000) is
smaller than the smallest allowed cell size (0.600000) for domain
decomposition grid cell 0 0 0
Without dlb the binary compiled with pgi (7.1.6) works fine. There is only a
problem in the dlb part.
The problem from the last mail I can't reproduce. This must have been a bad
built. Sorry.
But this time I did built each binary several times and double checked
everything (including different systems). So I'm sure that the binary
compiled with pgi (7.1.6) reproducable doesn't work with dlb and the ones
with the other compilers do.
BTW: Which compilers and versions are regularly tested?
BTW2: Did you decide something regarding the idea of the auto build system?
If not how about buildbot?
http://buildbot.net/
http://build.webkit.org/waterfall
http://twoalpha.blogspot.com/2007/12/buildbot-continuous-integration-system.html
Roland
On Mon, Jul 21, 2008 at 5:23 PM, Roland Schulz <schulzr at ornl.gov> wrote:
> Hi,
>
> now I get:
> The maximum allowed number of cells is: X 0 Y 0 Z 5
>
> Both with or without dlb. The system dimensions are the same in all three
> dimensions.
>
> With the old binary it still works without dlb.
> I was using an old version of grompp but also with a current grompp I get
> this error.
>
> Could you send me your working tpr file? I would like to check whether it
> is the input system our the different compilation/platform.
>
> Because this is not only with this input
>
> When I run the standard water tutor (without specifying -dd). I get
> Optimizing the DD grid for 4 cells with a minimum initial size of 0.000 nm
> Domain decomposition grid 4 x 0 x 0, separate PME nodes 0
>
> with -dd 2 2 1
> Making 2D domain decomposition 2 x 2 x 1
> starting mdrun 'Pure Water'
> 10000 steps, 20.0 ps.
> _pmii_daemon(SIGCHLD): PE 2 exit signal Segmentation fault
>
> Roland
>
> On Mon, Jul 21, 2008 at 11:50 AM, Berk Hess <hessb at mpip-mainz.mpg.de>
> wrote:
>
>> Hi,
>>
>> There has been a bug like this about two weeks ago.
>> It has been fixed for some time.
>> But today I did a partial commit, which I only noticed now.
>>
>> With the current CVS I can run a system similar to yours
>> with the same options.
>>
>> Could you try CVS update and recompile and run again?
>>
>> Berk.
>>
>>
>> Roland Schulz wrote:
>>
>>> Hi,
>>>
>>> I get the error:
>>> Fatal error:
>>> Step 0: The X-size (0.000000) times the triclinic skew factor (1.000000)
>>> is smaller than the smallest allowed cell size (1.200000) for domain
>>> decomposition grid cell 0 0 0
>>>
>>> I don't quite understand the initialization of cell_x1, cell_x0, thus
>>> have problems to find out the cause.
>>>
>>> This is with RF and mdrun options -nosum -dlb
>>>
>>> I tested both revision 1.145 and 1.148 of domdec.c. (all other files
>>> today version) Thus it is not the recent fatal-fix or 2d PME preparation.
>>>
>>> The simulation runs find without dlb.
>>>
>>> Where should I start looking why the size is 0?
>>>
>>> The below md.log file is for 1M atoms and 256 cores but it crashes the
>>> same with a subsystem of 4200 atoms and 16 cores.
>>>
>>> Thanks
>>> Roland
>>>
>>> the md.log:
>>>
>>>
>>> parameters of the run:
>>> integrator = md
>>> nsteps = 1000
>>> init_step = 0
>>> ns_type = Grid
>>> nstlist = 10
>>> ndelta = 2
>>> nstcomm = 0
>>> comm_mode = Linear
>>> nstcheckpoint = 1000
>>> nstlog = 0
>>> nstxout = 0
>>> nstvout = 0
>>> nstfout = 0
>>> nstenergy = 0
>>> nstxtcout = 0
>>> init_t = 0
>>> delta_t = 0.002
>>> xtcprec = 1000
>>> nkx = 72
>>> nky = 72
>>> nkz = 72
>>> pme_order = 4
>>> ewald_rtol = 1e-05
>>> ewald_geometry = 0
>>> epsilon_surface = 0
>>> optimize_fft = TRUE
>>> ePBC = xyz
>>> bPeriodicMols = FALSE
>>> bContinuation = FALSE
>>> bShakeSOR = FALSE
>>> etc = No
>>> epc = No
>>> epctype = Isotropic
>>> tau_p = 1
>>> ref_p (3x3):
>>> ref_p[ 0]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> ref_p[ 1]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> ref_p[ 2]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> compress (3x3):
>>> compress[ 0]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> compress[ 1]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> compress[ 2]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> refcoord_scaling = No
>>> posres_com (3):
>>> posres_com[0]= 0.00000e+00
>>> posres_com[1]= 0.00000e+00
>>> posres_com[2]= 0.00000e+00
>>> posres_comB (3):
>>> posres_comB[0]= 0.00000e+00
>>> posres_comB[1]= 0.00000e+00
>>> posres_comB[2]= 0.00000e+00
>>> andersen_seed = 815131
>>> rlist = 1.2
>>> rtpi = 0.05
>>> coulombtype = Reaction-Field
>>> rcoulomb_switch = 0
>>> rcoulomb = 1.2
>>> vdwtype = Cut-off
>>> rvdw_switch = 0
>>> rvdw = 1.2
>>> epsilon_r = 1
>>> epsilon_rf = 78
>>> tabext = 1
>>> implicit_solvent = No
>>> gb_algorithm = Still
>>> gb_epsilon_solvent = 80
>>> nstgbradii = 1
>>> rgbradii = 2
>>> gb_saltconc = 0
>>> gb_obc_alpha = 1
>>> gb_obc_beta = 0.8
>>> gb_obc_gamma = 4.85
>>> sa_surface_tension = 2.092
>>> DispCorr = No
>>> free_energy = no
>>> init_lambda = 0
>>> sc_alpha = 0
>>> sc_power = 0
>>> sc_sigma = 0.3
>>> delta_lambda = 0
>>> nwall = 0
>>> wall_type = 9-3
>>> wall_atomtype[0] = -1
>>> wall_atomtype[1] = -1
>>> wall_density[0] = 0
>>> wall_density[1] = 0
>>> wall_ewald_zfac = 3
>>> pull = no
>>> disre_weighting = Conservative
>>> disre_mixed = FALSE
>>> dr_fc = 1000
>>> dr_tau = 0
>>> nstdisreout = 0
>>> orires_fc = 0
>>> orires_tau = 0
>>> nstorireout = 100
>>> dihre-fc = 1000
>>> em_stepsize = 0.01
>>> em_tol = 100
>>> niter = 20
>>> fc_stepsize = 0
>>> nstcgsteep = 1000
>>> nbfgscorr = 10
>>> ConstAlg = Lincs
>>> shake_tol = 1e-04
>>> lincs_order = 8
>>> lincs_warnangle = 30
>>> lincs_iter = 1
>>> bd_fric = 0
>>> ld_seed = 1993
>>> cos_accel = 0
>>> deform (3x3):
>>> deform[ 0]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> deform[ 1]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> deform[ 2]={ 0.00000e+00, 0.00000e+00, 0.00000e+00}
>>> userint1 = 0
>>> userint2 = 0
>>> userint3 = 0
>>> userint4 = 0
>>> userreal1 = 0
>>> userreal2 = 0
>>> userreal3 = 0
>>> userreal4 = 0
>>> grpopts:
>>> nrdf: 2.16e+06
>>> ref_t: 0
>>> tau_t: 0
>>> anneal: No
>>> ann_npoints: 0
>>> acc: 0 0 0
>>> nfreeze: N N N
>>> energygrp_flags[ 0]: 0 0
>>> energygrp_flags[ 1]: 0 0
>>> efield-x:
>>> n = 0
>>> efield-xt:
>>> n = 0
>>> efield-y:
>>> n = 0
>>> efield-yt:
>>> n = 0
>>> efield-z:
>>> n = 0
>>> efield-zt:
>>> n = 0
>>> bQMMM = FALSE
>>> QMconstraints = 0
>>> QMMMscheme = 0
>>> scalefactor = 1
>>> qm_opts:
>>> ngQM = 0
>>>
>>> Initializing Domain Decomposition on 256 nodes
>>> Will use dynamic load balancing
>>> Will sort the charge groups at every domain (re)decomposition
>>> Scaling the initial minimum size with 1/0.8 (option -dds) = 1.25
>>> Optimizing the DD grid for 256 cells with a minimum initial size of 0.750
>>> nm
>>> The maximum allowed number of cells is: X 24 Y 24 Z 37
>>> Domain decomposition grid 8 x 4 x 8, separate PME nodes 0
>>> Domain decomposition nodeid 0, coordinates 0 0 0
>>>
>>> Using two step summing over 64 groups of on average 4.0 processes
>>>
>>> There are 0 dihedral restraints
>>> Table routines are used for coulomb: FALSE
>>> Table routines are used for vdw: FALSE
>>> Cut-off's: NS: 1.2 Coulomb: 1.2 LJ: 1.2
>>> Reaction-Field:
>>> epsRF = 78, rc = 1.2, krf = 0.283823, crf = 1.24204, epsfac = 138.935
>>> The electrostatics potential has its minimum at r = 1.20774
>>> System total charge: -0.000
>>> Generated table with 1100 data points for 1-4 COUL.
>>> Tabscale = 500 points/nm
>>> Generated table with 1100 data points for 1-4 LJ6.
>>> Tabscale = 500 points/nm
>>> Generated table with 1100 data points for 1-4 LJ12.
>>> Tabscale = 500 points/nm
>>>
>>> Enabling SPC water optimization for 240000 molecules.
>>>
>>> Configuring nonbonded kernels...
>>> Testing x86_64 SSE support... present.
>>>
>>>
>>> Removing pbc first time
>>> Done rmpbc
>>>
>>> Initializing LINear Constraint Solver
>>>
>>> ++++ PLEASE READ AND CITE THE FOLLOWING REFERENCE ++++
>>> B. Hess and H. Bekker and H. J. C. Berendsen and J. G. E. M. Fraaije
>>> LINCS: A Linear Constraint Solver for molecular simulations
>>> J. Comp. Chem. 18 (1997) pp. 1463-1472
>>> -------- -------- --- Thank You --- -------- --------
>>>
>>> The number of constraints is 144000
>>>
>>> ++++ PLEASE READ AND CITE THE FOLLOWING REFERENCE ++++
>>> S. Miyamoto and P. A. Kollman
>>> SETTLE: An Analytical Version of the SHAKE and RATTLE Algorithms for
>>> Rigid
>>> Water Models
>>> J. Comp. Chem. 13 (1992) pp. 952-962
>>> -------- -------- --- Thank You --- -------- --------
>>>
>>>
>>> Linking all bonded interactions to atoms
>>> There are 1092000 inter charge-group exclusions,
>>> will use an extra communication step for exclusion forces for
>>> Reaction-Field
>>>
>>> The maximum number of communication pulses is: X 1 Y 1 Z 1
>>> The minimum size for domain decomposition cells is 1.200 nm
>>> The requested allowed shrink of DD cells (option -dds) is: 0.80
>>> The allowed shrink of domain decomposition cells is: X 0.52 Y 0.26 Z 0.34
>>>
>>> The maximum allowed distance for atoms involved in interactions is:
>>> two-body bonded interactions (-rdd) 1.200 nm
>>> multi-body bonded interactions (-rdd) 1.200 nm
>>>
>>>
>>> Making 3D domain decomposition grid 8 x 4 x 8, home cell index 0 0 0
>>>
>>>
>>> Will not sum the energies at every step,
>>> therefore the energy file does not contain exact averages and
>>> fluctuations.
>>>
>>> WARNING:
>>> Because of the no energy summing option setting nstcomm (was 0) to
>>> nstlist (10)
>>>
>>> There are: 1008000 Atoms
>>> Charge group distribution at step 0: 1262 1274 1251 1282 1277 1242 1289
>>> 1260 1264 1271 1277 1274 1249 1285 1263 1263 1261 1274 1250 1281 1277 1240
>>> 1289 1259 1265 1271 1278 1275 1249 1287 1263 1264 1292 1267 1299 1275 1277
>>> 1301 1265 1284 1251 1283 1269 1272 1280 1270 1265 1289 1292 1267 1299 1275
>>> 1277 1301 1265 1284 1251 1283 1269 1272 1280 1270 1265 1289 1239 1263 1266
>>> 1234 1250 1265 1240 1257 1282 1252 1277 1264 1266 1281 1260 1255 1239 1263
>>> 1266 1234 1250 1265 1240 1257 1282 1252 1277 1264 1266 1281 1260 1255 1249
>>> 1279 1263 1254 1275 1255 1272 1272 1254 1252 1246 1245 1263 1236 1252 1260
>>> 1249 1279 1262 1254 1275 1254 1272 1271 1254 1252 1247 1245 1263 1237 1252
>>> 1261 1262 1274 1251 1282 1277 1242 1289 1260 1264 1271 1277 1274 1249 1285
>>> 1263 1263 1261 1274 1250 1281 1277 1240 1289 1259 1265 1271 1278 1275 1249
>>> 1287 1263 1264 1291 1267 1299 1274 1277 1301 1264 1284 1251 1283 1269 1272
>>> 1280 1270 1265 1289 1291 1267 1299 1274 1277 1301 1264 1284 1251 1283 1269
>>> 1272 1280 1270 1265 1289 1240 1263 1266 1235 1250 1265 1241 1257 1282 1252
>>> 1277 1264 1266 1281 1260 1255 1240 1263 1266 1235 1250 1265 1241 1257 1282
>>> 1252 1277 1264 1266 1281 1260 1255 1249 1279 1263 1254 1275 1255 1272 1272
>>> 1254 1252 1246 1245 1263 1236 1252 1260 1249 1279 1262 1254 1275 1254 1272
>>> 1271 1254 1252 1247 1245 1263 1237 1252 1261
>>>
>>> ------------------------------
>>> -------------------------
>>> Program mdrun, VERSION 3.3.99_development_20080718
>>> Source code file: domdec.c, line: 2220
>>>
>>> Fatal error:
>>> Step 0: The X-size (0.000000) times the triclinic skew factor (1.000000)
>>> is smaller than the smallest allowed cell size (1.200000) for domain
>>> decomposition grid cell 0 0 0
>>> -------------------------------------------------------
>>>
>>> Thanx for Using GROMACS - Have a Nice Day
>>> This email was Anti Virus checked by Astaro Security Gateway.
>>> http://www.astaro.com
>>> ------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> gmx-developers mailing list
>>> gmx-developers at gromacs.org
>>> http://www.gromacs.org/mailman/listinfo/gmx-developers
>>> Please don't post (un)subscribe requests to the list. Use the www
>>> interface or send it to gmx-developers-request at gromacs.org.
>>>
>>
>> _______________________________________________
>> gmx-developers mailing list
>> gmx-developers at gromacs.org
>> http://www.gromacs.org/mailman/listinfo/gmx-developers
>> Please don't post (un)subscribe requests to the list. Use the www
>> interface or send it to gmx-developers-request at gromacs.org.
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-developers/attachments/20080721/40d28ff8/attachment.html>
More information about the gromacs.org_gmx-developers
mailing list