>> 3) The ${AMBERHOME}/test/ambpdb_first_rna test-case gives me a segfault.
>> Since this segfault also occurs with an ambpdb-version compiled in 2004,
>> the input file(s) -either prmtop or inpcrd might be corrupt (?).
>
> This error seems to depend on architecture and compiler, but I don't have all
> the details.
...Right, it is compiler-dependent: gfortran does not zero-initialize
local arrays (and can't be told to unless $version >= 4.3).
"Solution":
Add the lines
C --- zero-initialize neighnum() in case the compiler does not ---
do i=1,MAXATOM
neighnum(i) = 0
enddo
somewhere add the beginning of ${AMBERHOME}/src/etc/forFIRSTteth.f.
This sems to work and doesn't hurt anyway.
Andreas
Received on Fri Apr 18 2008 - 21:12:36 PDT
