This barrier is apparently useless since it comes
just before a Bcast.

Anyway, it looks like it is solving an issue that appeared on
TRAVIS (at least on my laptop) so I add it here.

Related to #183

This is useful for faster debugging. E.g. to check a Vector just use
std::cerr<<v<<"\n";

[makedoc]

The kernels were normalised incorrectly and their derivative when the
covariance matrix was non-diagonal was computed wrongly.  This check
affects a number of regtests.  analysis/rt0 changes beecause the kernels
were normalised incorrectly.  basic/rt44, basic/rt48, basic/rt49, basic/rt52
change because the derivatives output in the fes file have changed sign

This makes readdir option consistent with others.

THIS REQUIRES ./configure AGAIN!!

This makes mpi option consistent with others.

THIS REQUIRES ./configure AGAIN!!

Lattice reduction is used in plumed to convert box to its
"least skewed" form. I changed a bit the threashold since it
was giving problems with high optimization flags.
Additionally, I added some warning when too many iterations
are performed (which indicates a numerical issue)

Properly instantiate templates when using high optimization flags
with gcc

I did it only when using the advanced syntax (SWITCH={...}).
I changed the input in some regtests to comply with
this change.

(cherry picked from commit b6ea36c1)

This only affects transpose of non-square Tensors, that as
far as I know are never used in the code.
(cherry picked from commit ec5e599c)

This only affects transpose of non-square Tensors, that as
far as I know are never used in the code.

For reasons lost in the mists of time to set histogram beads you
had to pass an empty string.  This is no longer necessary so I
removed it in order to make this class easier to use

(cherry picked from commit 2d9ac571)

[makedoc]

[makedoc]

Replica suffix is now appended *before* file extension. This is done
irrespectively of the specific extension, so also for plumed.dat and
traj.xyz. As such, I had to change a lot of regtests