[gmx-users] rb1 test failed after recompiling gromacs 3.3.2

[Nicolas Sapay]

Justin A. Lemkul wrote:
>
>
> Nicolas Sapay wrote:
>> Nicolas Sapay wrote:
>>> Hello,
>>>
>>> I've run the gmxtest on Gromacs-3.3.2. One simple test failed:
>>>
>>>    Testing rb1 . . . *** glibc detected *** realloc(): invalid next
>>>    size: 0x082c0180 ***
>>>    sh: line 1: 27398 Aborted                 mdrun >mdrun.out 2>&1
>>>    FAILED. Check files in rb1
>>>
>> I have more closely looked the compilation logs and the files used 
>> for the test. There is nothing particular to say about the 
>> compilation (no error, no warning). note that I'm using glibc 2.3.4 
>> compiled on a Linux 2.4.20 system. The error seems to be due to the 
>> -DMOL1 define in grompp.mdp of the rb1 test. When I replace it by 
>> -DBONDS (as in the rb125 test), the test is successful. What the 
>> -DMOL1 statement means? I though you can only use -DFLEXIBLE or 
>> -DPOSRES as defines in a mdp file. I had a look in the 3.3.2 manual, 
>> but found nothing about it.
>>
>
> Anything can be defined in the topology.  Look for the #ifdef 
> statements in topol.top.  I checked the test suite (for mine it's 
> 3.3.3), and found several instances where there is a -DMOL1 flag, but 
> no corresponding definition in the topology:
>
> angles1
> dih1
> rb1
>
> In other tests, where it is defined, it appears to mean that the 
> system should contain one molecule of interest.  No idea why it's 
> missing from these others, and why it doesn't seem to cause problems, 
> at least in angles1 and dih1.
>
> -Justin
>
Thanks for the answer. I can indeed use ANGLES an BONDS (defined in 
topol.top) in place of MOL1, but not RYCKAERT alone... Anyway, if the 
rb125 test is passed, I guess my compilation is safe enough, even for 
the R-B potential.

Nicolas
>>>
>>> And two complex tests failed as well:
>>>
>>>    Testing acetonitrilRF . . . FAILED. Check files in acetonitrilRF
>>>    Testing dec+water . . . FAILED. Check files in dec+water
>>>
>>>
>>> In the case of the rb1 test, I checked all the .out .log files but 
>>> found anything wrong. For acetonitrilRF, I obtain this:
>>>
>>>    -------------------------------------------------------
>>>    Program mdrun, VERSION 3.3.2
>>>    Source code file: network.c, line: 437
>>>
>>>    Routine should not have been called:
>>>    gmx_sumi
>>>    ------------------------------------------------------
>>>
>>>
>>> According to the mailing list, there is something to fix in 
>>> src/mdlib/rf_util.c. For dec+water, I've got the same error than on 
>>> the wiki, so nothing to worry about apparently.Following the advices 
>>> found on the wiki or the mailing list, I have re-compiled FFTW and 
>>> Gromacs (with fixed rf_util.c) from the sources. For FFTW, I've used 
>>> the following command:
>>>
>>>    ./configure --prefix /usr/export/fftw-3.1.2/ --enable-float
>>>    --enable-threads CC=/usr/bin/gcc
>>>
>>>
>>> And for Gromacs:
>>>
>>>    ./configure --prefix /usr/export/gromacs-3.3.2 --without-x
>>>    CC=/usr/bin/gcc CPPFLAGS=-I/usr/export/fftw-3.1.2/include/
>>>    LDFLAGS=-L/usr/export/fftw-3.1.2/lib/
>>>
>>>
>>> I have rerun the test but I still obtain the same error with the rb1 
>>> test (and the dec+water test). Can someone explain to me what the it 
>>> means exactly and how to fix it? (I guess it's related to the 
>>> Ryckaert-Bellemans potential?).
>>>
>>>    Thanks,
>>>    Nicolas
>>>
>>>
>>> ps: I'm using RHE 4.7 and gcc 3.4.6
>>>
>>> _______________________________________________
>>> gmx-users mailing list    gmx-users at gromacs.org
>>> http://www.gromacs.org/mailman/listinfo/gmx-users
>>> Please search the archive at http://www.gromacs.org/search before 
>>> posting!
>>> Please don't post (un)subscribe requests to the list. Use the www 
>>> interface or send it to gmx-users-request at gromacs.org.
>>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>
>> _______________________________________________
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://www.gromacs.org/mailman/listinfo/gmx-users
>> Please search the archive at http://www.gromacs.org/search before 
>> posting!
>> Please don't post (un)subscribe requests to the list. Use the www 
>> interface or send it to gmx-users-request at gromacs.org.
>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: nsapay.vcf
Type: text/x-vcard
Size: 282 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20080821/9c686da6/attachment.vcf>