Memory error when running Nwchem


Just Got Here
When I am running a simple 2 molecule (1 solvent and 1 solute) setup for classical force field calculation I get an error after successful processing of the prepare task. Any idea what I could do?

Below the program outcome
======================== echo of input deck ========================
start Ac124Ace

echo
print low

title "@U=============== PREPARE FOR CLASSICAL ==============="
prepare
 source Ac124Ace.pdb
solvent name Ace model Ace
new_top new_seq
new_rst
periodic
update lists
ignore
write Ac124Ace.rst
end
task prepare

title "@U=============== CALCULATE ENERGY ==============="
md
  system Ac124Ace
noshake solute
profile
record coord 1
end
task md energy

end
====================================================================
<< header cut >>
          Job information
---------------

   hostname        = MMKnecht0.back.home
program = nwchem
date = Thu May 1 09:26:50 2014

   compiled        = Sat_Feb_08_11:51:42_2014
source = /home/uli/Nwchem/nwchem-src-2014-01-28
nwchem branch = Development
nwchem revision = 25178
ga revision = 10467
input = nwchem.nw
prefix = Ac124Ace.
data base = ./Ac124Ace.db
status = startup
nproc = 1
time left = -1s



          Memory information
------------------

   heap     =   13107201 doubles =    100.0 Mbytes
stack = 13107201 doubles = 100.0 Mbytes
global = 26214400 doubles = 200.0 Mbytes (distinct from heap & stack)
total = 52428802 doubles = 400.0 Mbytes
verify = yes
hardfail = no


          Directory information
---------------------

 0 permanent = .
0 scratch = .




                               NWChem Input Module
-------------------


            @U===============  PREPARE FOR CLASSICAL  ===============
            ---------------------------------------------------------
Prepare Module
--------------



                               NWChem Input Module
-------------------


              @U===============  CALCULATE ENERGY  ===============
              ----------------------------------------------------
MA_verify_allocator_stuff: starting scan ...
stack block 'xsm', handle 71, address 0xb5fe10c4:
current checksum 3053326696 != stored checksum 472081102
stack block 'xsm', handle 71, address 0xb5fe10c4:
current left signature 2864320785 != proper left signature 2863311530
stack block 'wsm', handle 24, address 0xb5fe383c:
current checksum 3053336782 != stored checksum 142524498
stack block 'wsm', handle 24, address 0xb5fe383c:
current left signature 2864971776 != proper left signature 2863311530
MA_verify_allocator_stuff: scan completed

**********
* 0: Unable to allocate temp array in cf_fix 0
**********

Forum Regular
Hi,

The problem is not in the memory allocation itself. The memory allocator internally checks that the memory blocks are sane. Doing that it has found that memory blocks named "xsm" and "wsm" are corrupted by some memory access that was out of bounds. After detecting this issue the code essentially gives up. I'll if I can track down where this happens as this looks like a real bug.

Best wishes, Huub

Forum Regular
Hi,

Could you send the PDB file to Hubertus dot vanDam at PNNL dot gov please?

Huub

Forum Regular
Hi,

The problem seems to be related to your solvent line "solvent name Ace model Ace". What solvent model are you trying to use (the "ACE" model I found is by Schaefer and Karplus, J. Phys. Chem. (1996) vol. 100, pp. 1578-1599, DOI: 10.1021/jp9521621, the default model in the MD code is "SPCE")?

Huub

Just Got Here
Hi Huub
I have sent you the complete tar file as the set of files I am using. Within those you will see that I am using 124 Acetone molecules as solvent around the one molecule - which happens to be Acetone, too. I have created respective fragment and segment files and this approach was working fine with using Methanol as the solvent.

Thanks for your help!

Ulrich


Forum >> NWChem's corner >> Running NWChem