Merge #2781 #2835

2781: Add support for gdb in ipypresso r=jngrad a=hirschsn

Ipypresso's flag "--gdb" does not work on my system because ipypresso is just a wrapper script executing python with a bunch of parameters. Nevertheless, support of --gdb can be added to ipypresso by a slightly more complicated gdb setup.

This PR adds this support using gdb's exec-wrapper functionality. The implementation is quite hacky by inserting a check for ipython into pypresso.cmakein. Let me know if the functionality itself is something that you want to have upstream. If so, we also could figure out a less hacky solution.

Description of changes:
 - Use ipython as gdb's exec-wrapper, not directly as executable


PR Checklist
------------
 - [ ] Tests?
   - [ ] Interface
   - [ ] Core 
 - [ ] Docs?


2835: Fix some typos in constraints.py r=jngrad a=hirschsn

Fixes some typos in constraints.py.

PR Checklist
------------
 - [ ] Tests?
   - [ ] Interface
   - [ ] Core 
 - [ ] Docs?


Co-authored-by: Steffen Hirschmann <steffen.hirschmann@ipvs.uni-stuttgart.de>
Co-authored-by: Jean-Noel Grad <jgrad@icp.uni-stuttgart.de>
Co-authored-by: Jean-Noël Grad <jgrad@icp.uni-stuttgart.de>

NEWS

@@ -2,6 +2,66 @@
 = ESPRESSO NEWS =
 =================
 
+ESPResSo 4.0.2
+==============
+
+This release provides a number of corrections for the Espresso 4.0 line.
+We recommend that this release be used for all production simulations.
+Please note that a sign error in tabulated interactions was fixed.
+Simulation scripts which worked around this problem might have to be changed.
+Below, please find the list of changes. The numbers in brackets refer to
+ticket numbers on http://github.com/espressomd/espresso
+
+
+Corrections for bugs that may harm simulation results:
+  * A sign error in tabulated interactions was corrected such that
+    the force equals the negative gradient of the potential. (#2519,2520)
+
+  * The flow field of the CPU lattice-Boltzmann implementation was deleted
+    when aspects of the molecular dynamics cell grid were changed; E.g., when
+    interactions, the skin or the parallelization setup were changed.
+    ESPResSo now terminates with an error, when this happens.
+    To avoid this, please setup the CPU lattice-Boltzmann after all
+    other aspects of the system. The GPU LB is not affected in the 4.0
+    release, but was affected in the current development branch. (#2728, #2736)
+
+  * Corrected the force acting on LB Boundaries for the case of
+    agrid and density not equal to 1 (#2624).
+
+  * Corrected the cutoff calculation for the soft sphere interaction. In the
+    previous implementation, the offset parameter was ignored. (#2505)
+
+  * The "three point coupling" of particles to the lattice-Boltzmann method
+    has been removed. While it works in most environments, for some compilers
+    the calculation gives wrong values. This is likely caused by undefined
+    behavior. A corrected implementation is available in
+    ESPResSo's development branch. It cannot be safely backported to 4.0.2,
+    because the code has diverged too far. (#2516, #2517)
+    Users who did not explicitly activate this coupling via couple="3pt" are
+    not affected.
+
+  * The velocity of existing particles was changed when setting or changing
+    the simulation time step (#2480)
+
+
+Further changes:
+  * Fixed the electrokinetic Python interface (#2486)
+
+  * Correction to the installation instructions for mac (#2510)
+
+  * Corrected file permissions (#2470)
+
+  * Minor corrections and extensions to the test suite (#2477, #2552)
+
+  * Fixed a dead-lock in the dipolar Barnes Hutt method on the GPU for
+    recent NVIDIA cards such as RTX 2080 (#2719).
+
+  * Restored Mayavi visualizer's API-compatibility with OpenGL visualizer
+    (#2751)
+
+
+
+
 ESPResSo 4.0.1
 ==============
 

doc/sphinx/analysis.rst

@@ -449,7 +449,7 @@ The short ranged part is given by:
 
 .. math :: p^\text{Coulomb, P3M, dir}_{(k,l)}= \frac{1}{4\pi \epsilon_0 \epsilon_r} \frac{1}{2V} \sum_{\vec{n}}^* \sum_{i,j=1}^N q_i q_j \left( \frac{ \mathrm{erfc}(\beta |\vec{r}_j-\vec{r}_i+\vec{n}|)}{|\vec{r}_j-\vec{r}_i+\vec{n}|^3} +\frac{2\beta \pi^{-1/2} \exp(-(\beta |\vec{r}_j-\vec{r}_i+\vec{n}|)^2)}{|\vec{r}_j-\vec{r}_i+\vec{n}|^2} \right) (\vec{r}_j-\vec{r}_i+\vec{n})_k (\vec{r}_j-\vec{r}_i+\vec{n})_l,
 
-where :math:`\beta` is the P3M splitting parameter, :math:`\vec{n}` identifies the periodic images, the asterix denotes that terms with :math:`\vec{n}=\vec{0}` and i=j are omitted.
+where :math:`\beta` is the P3M splitting parameter, :math:`\vec{n}` identifies the periodic images, the asterisk denotes that terms with :math:`\vec{n}=\vec{0}` and i=j are omitted.
 The long ranged (k-space) part is given by:
 
 .. math :: p^\text{Coulomb, P3M, rec}_{(k,l)}= \frac{1}{4\pi \epsilon_0 \epsilon_r} \frac{1}{2 \pi V^2} \sum_{\vec{k} \neq \vec{0}} \frac{\exp(-\pi^2 \vec{k}^2/\beta^2)}{\vec{k}^2} |S(\vec{k})|^2 \cdot (\delta_{k,l}-2\frac{1+\pi^2\vec{k}^2/\beta^2}{\vec{k}^2} \vec{k}_k \vec{k}_l),

doc/sphinx/electrostatics.rst

@@ -474,10 +474,10 @@ To use the, e.g.,  ``ewald`` solver from SCAFACOS as electrostatics solver for y
 cutoff to :math:`1.5` and tune the other parameters for an accuracy of
 :math:`10^{-3}`, use::
 
-  from espressomd.electrostatics import Scafacos
-  scafacos = Scafacos(prefactor=1, method_name="ewald",
-                      method_params={"ewald_r_cut": 1.5, "tolerance_field": 1e-3})
-  system.actors.add(scafacos)
+   from espressomd.electrostatics import Scafacos
+   scafacos = Scafacos(prefactor=1, method_name="ewald",
+                       method_params={"ewald_r_cut": 1.5, "tolerance_field": 1e-3})
+   system.actors.add(scafacos)
 
 
 For details of the various methods and their parameters please refer to

src/core/electrostatics_magnetostatics/icc.hpp

@@ -18,37 +18,35 @@
   You should have received a copy of the GNU General Public License
   along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */
-//
-
 /** \file
-
-    ICCP3M is a method that allows to take into account the influence
-    of arbitrarily shaped dielectric interfaces.  The dielectric
-    properties of a dielectric medium in the bulk of the simulation
-    box are taken into account by reproducing the jump in the electric
-    field at the inface with charge surface segments. The charge
-    density of the surface segments have to be determined
-    self-consistently using an iterative scheme.  It can at presently
-    - despite its name - be used with P3M, ELCP3M, MMM2D and MMM1D.
-    For details see:<br> S. Tyagi, M. Suzen, M. Sega, C. Holm,
-    M. Barbosa: A linear-scaling method for computing induced charges
-    on arbitrary dielectric boundaries in large system simulations
-    (Preprint)
-
-    To set up ICCP3M first the dielectric boundary has to be modelled
-    by espresso particles 0..n where n has to be passed as a parameter
-    to ICCP3M. This is still a bit inconvenient, as it forces the user
-    to reserve the first n particle ids to wall charges, but as the
-    other parts of espresso do not suffer from a limitation like this,
-    it can be tolerated.
-
-    For the determination of the induced charges only the forces
-    acting on the induced charges has to be determined. As P3M and the
-    other Coulomb solvers calculate all mutual forces, the force
-    calculation was modified to avoid the calculation of the short
-    range part of the source-source force calculation.  For different
-    particle data organisation schemes this is performed differently.
-    */
+ *
+ *  ICCP3M is a method that allows to take into account the influence
+ *  of arbitrarily shaped dielectric interfaces.  The dielectric
+ *  properties of a dielectric medium in the bulk of the simulation
+ *  box are taken into account by reproducing the jump in the electric
+ *  field at the inface with charge surface segments. The charge
+ *  density of the surface segments have to be determined
+ *  self-consistently using an iterative scheme.  It can at presently
+ *  - despite its name - be used with P3M, ELCP3M, MMM2D and MMM1D. For
+ *  details see: S. Tyagi, M. Suzen, M. Sega, M. Barbosa, S. S. Kantorovich,
+ *  C. Holm: An iterative, fast, linear-scaling method for computing induced
+ *  charges on arbitrary dielectric boundaries, J. Chem. Phys. 2010, 132,
+ *  p. 154112, doi:10.1063/1.3376011
+ *
+ *  To set up ICCP3M first the dielectric boundary has to be modelled
+ *  by espresso particles 0..n where n has to be passed as a parameter
+ *  to ICCP3M. This is still a bit inconvenient, as it forces the user
+ *  to reserve the first n particle ids to wall charges, but as the
+ *  other parts of espresso do not suffer from a limitation like this,
+ *  it can be tolerated.
+ *
+ *  For the determination of the induced charges only the forces
+ *  acting on the induced charges has to be determined. As P3M and the
+ *  other Coulomb solvers calculate all mutual forces, the force
+ *  calculation was modified to avoid the calculation of the short
+ *  range part of the source-source force calculation.  For different
+ *  particle data organisation schemes this is performed differently.
+ */
 
 #ifndef CORE_ICCP3M_HPP
 #define CORE_ICCP3M_HPP

src/core/electrostatics_magnetostatics/p3m-common.hpp

@@ -124,7 +124,7 @@ typedef struct {
   int s_ur[6][3];
   /** sizes for send buffers. */
   int s_size[6];
-  /** dimensionof sub meshes to recv. */
+  /** dimension of sub meshes to recv. */
   int r_dim[6][3];
   /** left down corners of sub meshes to recv. */
   int r_ld[6][3];

src/core/electrostatics_magnetostatics/p3m_gpu_cuda.cu

@@ -658,7 +658,7 @@ void assign_forces(const CUDA_particle_data *const pdata, const P3MGpuData p,
   _cuda_check_errors(block, grid, "assign_forces", __FILE__, __LINE__);
 }
 
-/* Init the internal datastructures of the P3M GPU.
+/* Init the internal data structures of the P3M GPU.
  * Mainly allocation on the device and influence function calculation.
  * Be advised: this needs mesh^3*5*sizeof(REAL_TYPE) of device memory.
  * We use real to complex FFTs, so the size of the reciprocal mesh

src/core/object-in-fluid/affinity.hpp

@@ -117,7 +117,7 @@ inline void add_affinity_pair_force(Particle *p1, Particle *p2,
                    ia_params
                        ->affinity_r0) { // Bond does not exist, we are inside
                                         // of possible bond creation area,
-                                        // lets talk about creating a bond
+                                        // let's talk about creating a bond
           // This implementation creates bond always
           for (j = 0; j < 3; j++)
             p1->p.bond_site[j] = unfolded_pos[j] - d[j];
@@ -231,7 +231,7 @@ inline void add_affinity_pair_force(Particle *p1, Particle *p2,
                    ia_params
                        ->affinity_r0) { // Bond does not exist, we are inside
                                         // of possible bond creation area,
-                                        // lets talk about creating a bond
+                                        // let's talk about creating a bond
           double Pon = 1.0 - exp(-ia_params->affinity_Kon * time_step);
           // The probability is given by function Pon(x)= 1 - e^(-x) where x is
           // Kon*dt.
@@ -326,7 +326,7 @@ inline void add_affinity_pair_force(Particle *p1, Particle *p2,
                    ia_params
                        ->affinity_r0) { // Bond does not exist, we are inside
                                         // of possible bond creation area,
-                                        // lets talk about creating a bond
+                                        // let's talk about creating a bond
           double Pon = 1.0 - exp(-ia_params->affinity_Kon * time_step);
           // The probability is given by function Pon(x)= 1 - e^(-x) where x is
           // Kon*dt.
@@ -440,7 +440,7 @@ inline void add_affinity_pair_force(Particle *p1, Particle *p2,
                    ia_params
                        ->affinity_r0) { // Bond does not exist, we are inside
                                         // of possible bond creation area,
-                                        // lets talk about creating a bond
+                                        // let's talk about creating a bond
           double Pon = 1.0 - exp(-ia_params->affinity_Kon * time_step);
           // The probability is given by function Pon(x)= 1 - e^(-x) where x is
           // Kon*dt.
@@ -564,7 +564,7 @@ inline void add_affinity_pair_force(Particle *p1, Particle *p2,
                    ia_params
                        ->affinity_r0) { // Bond does not exist, we are inside
                                         // of possible bond creation area,
-                                        // lets talk about creating a bond
+                                        // let's talk about creating a bond
           double Pon = 1.0 - exp(-ia_params->affinity_Kon * time_step);
           // The probability is given by function Pon(x)= 1 - e^(-x) where x is
           // Kon*dt.
@@ -688,7 +688,7 @@ inline void add_affinity_pair_force(Particle *p1, Particle *p2,
                    ia_params
                        ->affinity_r0) { // Bond does not exist, we are inside
                                         // of possible bond creation area,
-                                        // lets talk about creating a bond
+                                        // let's talk about creating a bond
           double Pon = 1.0 - exp(-ia_params->affinity_Kon * time_step);
           // The probability is given by function Pon(x)= 1 - e^(-x) where x is
           // Kon*dt.

src/python/espressomd/constraints.py

@@ -222,7 +222,7 @@ class _Interpolated(Constraint):
     The data has to have one point of halo in each direction,
     and is shifted by half a grid spacing in the +xyz direction,
     so that the element (0,0,0) has coordinates -0.5 * grid_spacing.
-    The numer of points has to be such that the data spanc the whole
+    The number of points has to be such that the data spans the whole
     box, e.g. the most up right back point has to be at least at
     box + 0.5 * grid_spacing. There are convenience functions on this
     class that can calculate the required grid dimensions and the coordinates.
@@ -508,7 +508,7 @@ class FlowField(_Interpolated):
 
       F = -gamma * (u(r) - v)
 
-    wher v is the velocity of the particle.
+    where v is the velocity of the particle.
 
     """
 
@@ -528,7 +528,7 @@ class HomogeneousFlowField(Constraint):
 
       F = -gamma * (u - v)
 
-    wher v is the velocity of the particle.
+    where v is the velocity of the particle.
 
     Attributes
     ----------

src/python/espressomd/reaction_ensemble.pyx

@@ -243,7 +243,7 @@ cdef class ReactionAlgorithm(object):
         if(self._params["check_for_electroneutrality"]):
             charges = np.array(list(self._params["default_charges"].values()))
             if(np.count_nonzero(charges) == 0):
-                # all partices have zero charge
+                # all particles have zero charge
                 # no need to check electroneutrality
                 return
             total_charge_change = 0.0

src/python/pypresso.cmakein

@@ -54,6 +54,7 @@ fi
 case $1 in
     --gdb)
         shift
+        [ "@PYTHON_FRONTEND@" = "@IPYTHON_EXECUTABLE@" ] && exec gdb -ex "set print thread-events off" -ex "set exec-wrapper sh -c 'exec \"@IPYTHON_EXECUTABLE@\" \"\$@\"'" --args "@PYTHON_EXECUTABLE@" $@
         exec gdb --args @PYTHON_FRONTEND@ $@
         ;;
     --lldb)
@@ -75,6 +76,7 @@ case $1 in
     --gdb=*)
         options="${1#*=}"
         shift
+        [ "@PYTHON_FRONTEND@" = "@IPYTHON_EXECUTABLE@" ] && exec gdb -ex "set print thread-events off" -ex "set exec-wrapper sh -c 'exec \"@IPYTHON_EXECUTABLE@\" \"\$@\"'" ${options} --args "@PYTHON_EXECUTABLE@" $@
         exec gdb ${options} --args @PYTHON_FRONTEND@ $@
         ;;
     --lldb=*)