OK, just for clarity,
The 'make test' command just runs a predefined set of small
calculations, and compares the outputs with outputs saved previously
from successful calculations. If you look into the test directories,
those are the files with a ".save" in the end.
So, when you issue 'make test', new output files will be created and,
if any difference with the ".save" files is found, a ".dif" file is
also created. The contents of this file are also placed in the
"TEST_FAILURES.diff" file placed in the main test directory. When you
issue 'make clean' there, only the new outputs and the "*.dif' files
are removed, basically as a preparation to run the tests again.
Gustavo.
On Nov 8, 2007 12:30 PM, Francesco Pietra <chiendarret.yahoo.com> wrote:
>
>
> --- "David A. Case" <case.scripps.edu> wrote:
>
> > On Thu, Nov 08, 2007, Francesco Pietra wrote:
> >
> > > In connection with problems with DOCK6.1 I wanted to check the Linux
> > parallel
> > > installation of Amber9
> > >
> > > export DO_PARALLEL='mpirun -np 4'
> > > cd $AMBERHOME/test
> > > make test.parallel
> > >
> > > was OK
> > >
> > > Then, I had to repeat the test. This time distractedly I commanded "make
> > clean"
> > > before "make test.parallel", which clean resulted in:
> > >
> > > (find antechamber -name '*.out' -o -name '*.log' -o - name '*.lib' -o\
> > -name
> > > "*.dif" |\
> > > while read dif ;\
> > > do \
> > > rm -f $dif ;\
> > > done ;\
> > > cd leap; rm -f *.out *.log *.dif *.lib
> > > while read dif ;\
> > > do \
> > > rm -f $dif ;\
> > > done ;\
> > > rm -f TEST_FAILURES.diff
> > > ----------
> > > All tests PASSED according to screen observation, and top -i showed sander
> > > working on all four nodes, though now there are no more files to check. Is
> > it
> > > possible to restore the original situation?
> >
> > It's not clear to me what the problem is. Can you define what you mean by
> > "the original situation"?
>
> Yes, I was unclear. I wanted to ask whether the "make" may have damaged the
> program as to the capability of carrying out regular tests. If it simply
> removed previous results from tests, I don't care. After that "make", the new
> "make test.parallel" showed PASSED for all programs, as far as I could see from
> the scree; TEST_FAILURES.diff (removed by that "make") was not produced
> (perhaps it would be produced by test.serial, did not try)
> Thanks
> francesco
>
> Typing "make clean" will indeed remove many dif
> > files from previous tests, but you say that you subsequently did "make
> > test.parallel" and everything PASSED.
> >
> > ...dac
> >
> > -----------------------------------------------------------------------
> > The AMBER Mail Reflector
> > To post, send mail to amber.scripps.edu
> > To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
> >
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
> -----------------------------------------------------------------------
>
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Nov 11 2007 - 06:07:45 PST
