Hi All.
Well, first of all, I would like to thank a lot David for all the help
he gave me up to now. I don't know how to thank for so much help
neither how to apologize for bothering you this much.
Up to now, I could finally compile it with Portland, in both single and
double precision version. But, on the other hand, while the single
precision works properly, the double precision fails in some tests,
like the "mixed" test on the "tutor" directory. It yelds me the
following error message:
_
starting mdrun 'Pure Methanol - Yummie! (and some water)'
1 steps, 20.0 ps.
Back Off! I just backed up traj.trr to ./#traj.trr.10#
Back Off! I just backed up traj.xtc to ./#traj.xtc.10#
step 0
Step 1 Warning: pressure scaling more than 1%, mu: 1.01018 1.02193 1.08117
Step 2 Warning: pressure scaling more than 1%, mu: 1.01124 1.00787 1.08362
Step 3 Warning: pressure scaling more than 1%, mu: 1.03039 1.0523 1.08046
Step 4 Warning: pressure scaling more than 1%, mu: 1.04141 1.08638 1.09351
Step 5 Warning: pressure scaling more than 1%, mu: 1.06415 1.0995 1.06772
Step 6 Warning: pressure scaling more than 1%, mu: 1.04802 1.07721 1.04162
Step 7 Warning: pressure scaling more than 1%, mu: 1.02981 1.03652 1.01836
Step 8 Warning: pressure scaling more than 1%, mu: 1.01626 1.01848 1.00881
Step 9 Warning: pressure scaling more than 1%, mu: 1.01081 1.01178 1.00552
step 130, remaining runtime: 0 s
Step 140 Warning: pressure scaling more than 1%, mu: 1.00411 1.00022 1.0102
step 140, remaining runtime: 0 s
Step 141 Warning: pressure scaling more than 1%, mu: 1.00439 1.00019 1.01173
Step 142 Warning: pressure scaling more than 1%, mu: 1.00395 1.00019 1.01172
Step 143 Warning: pressure scaling more than 1%, mu: 1.00304 1.00017 1.01009
step 240, remaining runtime: 40
s
---
Program mdrun_d, VERSION 3.3
Source code file: stat.c, line: 283
Fatal error:
XTC error
---
"May the Force Be With You" (Star Wars)
___
By the way: this test runs properly and without any error in both
(single and double precision) compilations with GCC, and in the single
precision compilation with Portland. :(
Does anybody have any idea of this kind of error? Is it a known issue, someone has came across this kind of behavior before?
Thanks a lot everybody for all and any help in advance! ;)
Jones
On 4/27/06, David van der Spoel <[EMAIL PROTECTED]> wrote:
Jones de Andrade wrote:> Hi.>> Sorry if it looks like a stupidy question (and it is), but I'm quite bad> on C (fortran addicted).>> I found 4 entries in the file like the one you told me to change. May I
> change all them, or which ones?>only the one in the function that the compiler complains about. gotoline 313 and then upwards from there until you find the offensivedeclaration.> Thanks a lot,
>> Jones>> On 4/26/06, *David van der Spoel* <[EMAIL PROTECTED]> [EMAIL PROTECTED]
>> wrote:>> Jones de Andrade wrote:> > Hi all.> >> > Well, some time ago I decided to bench gromacs on AMD64 machines> > compiled with different compilers (both single and double precision,
> > both internal and external blas and lapack, just FFTw 3 for fourier)> > in order to see which one performs better.> >> > To time, I have available GCC, PGI and Intel compilers. Pathscale
> is a> > future option.> >> > Unfortunatelly, I was able to properly compile it just with GCC> (all 4> > flavours, despite the fact that the external libraries of ACML didn't
> > seem to improve so much the performance as I expected), but failed on> > both PGI and Intel compilers at different points. So, I'm asking for> > help on those compilation issues.
> >> > For Intel, the compilation ends up with a segmentation fault at the> > following point:> >> >> ***
>> > ./mknb -software_invsqrt> > >>> Gromacs nonbonded kernel generator (-h for help)> > >>> Generating single precision functions in C.> > >>> Using Gromacs software version of 1/sqrt(x).
> > make[5]: *** [kernel-stamp] Falha de segmentação> > make[5]: Leaving directory> >> `/home/johannes/src/gromacs/gromacs-3.3/src/gmxlib/nonbonded/nb_kernel'> > make[4]: ** [all-recursive] Erro 1
> >> ***> >> > And, for PGI, it dies at the following point (after using the> > --without-x --without-motif-libraries -- without-motif-includes
> flags> > for configure):> >> >> **
