Merge pull request #746 from edoapra/release-7-2-0

changes from master branch

.github/workflows/docker_actions.yml

@@ -36,23 +36,18 @@ jobs:
           key: ${{ matrix.folder}}-${{ matrix.fc}}/${{ matrix.archs}}-nwchem-dockeractions-v002
       - name: Qemu
         id: qemu
-        uses: docker/setup-qemu-action@v1
+        uses: docker/setup-qemu-action@v2
         with:
-          image: crazymax/binfmt:v6.0.0
+          image: tonistiigi/binfmt:qemu-v6.0.0-10
       - name: Available platforms
         run: |
           echo ${{ steps.qemu.outputs.platforms }}
           docker images
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1
+        uses: docker/setup-buildx-action@v2
         with:
           driver-opts: |
-              image=moby/buildkit:master
-      - name: Get Arch
-        id: get-arch
-        run: |
-          echo "::set-output name=arch::$(echo ${{matrix.archs }} | sed 's/linux//'|sed  's/\///g' )"
-        shell: bash
+              image=moby/buildkit:v0.10.4
       - name: fetch cache
         if: (steps.setup-cache.outputs.cache-hit == 'true') && ( matrix.folder  != 'helloworld' )
         run: |
@@ -62,14 +57,13 @@ jobs:
           rsync -av ~/cache/libext* cache/. ; \
           echo "libext cache fetched" ; \
           fi
-      - uses: haya14busa/action-cond@v1
-        id: cache-hit-reporter
-        with:
-          cond: ${{ steps.setup-cache.outputs.cache-hit != '' }}
-          if_true: "Y"
-          if_false: "N"
+          echo "cache_hit=Y" >> $GITHUB_ENV
+      - name: negative-cache-hit-reporter
+        if: (steps.setup-cache.outputs.cache-hit != 'true')
+        run: |
+          echo "cache_hit=N" >> $GITHUB_ENV
       - name: build_schedule
-        uses: docker/build-push-action@v2
+        uses: docker/build-push-action@v3
         with:
           push: false
           context: ${{ matrix.folder }}
@@ -78,7 +72,9 @@ jobs:
           tags: nwchem_image
           build-args: |
             FC=${{ matrix.fc }}
-            CACHE_HIT=${{ steps.cache-hit-reporter.outputs.value }}
+            CACHE_HIT=${{ env.cache_hit }}
+            NWCHEM_BRANCH=${{ github.ref_name }}
+            GITHUB_REPOSITORY_OWNER=${{ github.repository_owner }}
       - name: store cache
         run: |
             mkdir -p ~/cache/

.github/workflows/github_actions.yml

@@ -129,6 +129,16 @@ jobs:
             nwchem_modules: "tce"
             fc: gfortran-11
             cc: gcc-11
+          - os: ubuntu-20.04
+            experimental: true
+            mpi_impl: mpich
+            armci_network: MPI-TS
+            nwchem_modules: "nwdft solvation driver"
+            fc: gfortran
+            cc: gcc
+            use_libxc: -1
+            blas: "internal"
+            blas_size: 8
           - os: ubuntu-22.04
             experimental: true
             mpi_impl: mpich
@@ -136,6 +146,7 @@ jobs:
             nwchem_modules: "qmandpw qmd"
             fc: gfortran-11
             cc: gcc-11
+            use_libxc: -1
           - os: ubuntu-20.04
             experimental: true
             mpi_impl: mpich
@@ -220,7 +231,7 @@ jobs:
             armci_network: MPI-TS
             nwchem_modules: "tinyqmpw"
             fc: ifort
-            cc: icc
+            cc: icx
             use_libxc: 1
           - os: macos-11
             experimental: true 
@@ -252,7 +263,7 @@ jobs:
             armci_network: MPI-TS
             nwchem_modules: "nwdft solvation driver"
             fc: ifort
-            cc: icc
+            cc: icx
             use_simint: 1
           - os: ubuntu-20.04
             experimental: true 
@@ -357,14 +368,7 @@ jobs:
         with:
           path: |
             ~/cache
-            ~/apps
           key: ${{ matrix.os }}-${{ matrix.mpi_impl}}-${{ matrix.fc}}-xcode${{ matrix.xcode_version}}-simint${{ matrix.use_simint}}-blas${{ matrix.blas}}-blas_size${{ matrix.blas_size}}-libxc${{matrix.use_libxc}}-tblite${{matrix.use_tblite}}-simd${{ env.simd }}-nwchemcache-v003
-      - name: fetch apps cache
-        if: steps.setup-cache.outputs.cache-hit == 'true'
-        run: |
-          if [[ -f ~/apps/oneapi/setvars.sh ]]; then \
-          echo "apps oneapi cache fetched" ; \
-          export GOT_APPSCACHE="Y" ; fi
       - name: build environment
         run: |
           pwd

.github/workflows/lint_python.yml

@@ -14,12 +14,13 @@ jobs:
       - uses: actions/checkout@v3
       - uses: actions/setup-python@v4
         with:
-          python-version: '3.10'
+          python-version: '3.11'
           check-latest: true
       - run: pip install --upgrade pip wheel
       - run: pip install bandit black codespell flake8 flake8-2020 flake8-bugbear
                          flake8-comprehensions isort mypy pytest pyupgrade safety
       - run: python -m pip  install --upgrade requests
+      - run: python -m pip install 'setuptools>=65.5.1'
       - run: bandit --recursive --skip B101,B110,B306,B307,B311,B605,B607 .
       - run: black --check . || true
       - run: codespell --count

QA/doqmtests.mpi

@@ -519,8 +519,6 @@ let "myexit+=$?"
 let "myexit+=$?"
 ./runtests.mpi.unix procs $np carbon_fon
 let "myexit+=$?"
-./runtests.mpi.unix procs $np ch5n_nbo
-let "myexit+=$?"
 ./runtests.mpi.unix procs $np tpss tpssh
 let "myexit+=$?"
 #
@@ -532,7 +530,7 @@ let "myexit+=$?"
 ./runtests.mpi.unix procs $np h2o2_fde
 let "myexit+=$?"
 # tddft gradients
-./runtests.mpi.unix procs $np tddftgrad_co_cis tddftgrad_n2 tddftgrad_co tddftgrad_n2_uks tddftgrad_ch2o
+./runtests.mpi.unix procs $np tddftgrad_co_cis tddftgrad_n2 tddftgrad_co tddftgrad_n2_uks tddftgrad_ch2o tddftgrad_h2o_cis_lda
 let "myexit+=$?"
 # new xc functionals for 7.0.0
 ./runtests.mpi.unix procs $np dft_rscan
@@ -543,6 +541,12 @@ let "myexit+=$?"
 let "myexit+=$?"
 ./runtests.mpi.unix procs $np rt_tddft_mocap
 let "myexit+=$?"
+./runtests.mpi.unix procs $np rt_tddft_cytosine_cam_cdfit
+let "myexit+=$?"
+./runtests.mpi.unix procs $np rt_tddft_tcne_spinkick
+let "myexit+=$?"
+./runtests.mpi.unix procs $np rt_tddft_water_abs_spec
+let "myexit+=$?"
 ./runtests.mpi.unix procs $np dft_ne_n12
 let "myexit+=$?"
 ./runtests.mpi.unix procs $np dft_mfm_ch3
@@ -578,6 +582,10 @@ if [[ ! -z "${USE_TBLITE}" ]]; then
 let "myexit+=$?"
 ./runtests.mpi.unix procs $np xtb_siosi7
 let "myexit+=$?"
+./runtests.mpi.unix procs $np h2o_ccca o2_ccca
+let "myexit+=$?"
+./runtests.mpi.unix procs $np dftd3_c6cn
+let "myexit+=$?"
 fi
 #--- if we only want quick tests get out now!
 if [[ "$what" == "fast" ]]; then
@@ -605,6 +613,10 @@ let "myexit+=$?"
 let "myexit+=$?"
 ./runtests.mpi.unix procs $np dft_r2scan0
 let "myexit+=$?"
+./runtests.mpi.unix procs $np scf_dftguess
+let "myexit+=$?"
+./runtests.mpi.unix procs $np back_to_atguess
+let "myexit+=$?"
 #
 ./runtests.mpi.unix procs $np ccsdt_ompt_w3pvdz
 let "myexit+=$?"
@@ -685,7 +697,8 @@ echo 'Please check the final values for the last optimized structure.'
 let "myexit+=$?"
 ./runtests.mpi.unix procs $np dielsalder
 let "myexit+=$?"
-./runtests.mpi.unix procs $np talc
+np_small=$(get_limit $np 16)
+./runtests.mpi.unix procs $np_small talc
 let "myexit+=$?"
 echo 'The intermediate energy values for the numerical gradient '
 echo 'test may cause some differences.'

QA/nwparse.pl

@@ -18,7 +18,7 @@
 #           Richland, WA 99352-0999
 #
 $quiet = 0;
-$debug = 1;
+$debug = 0;
 $num_argv = @ARGV;
 
 if ($num_argv == 0) {
@@ -402,7 +402,7 @@ foreach $filename (@FILES_TO_PARSE) {
           }
         }
 
-	if (/Excitation energy/ || /Rotatory /) {
+	if (/Excitation energy/ || /Rotatory / || /IBO loc: largest element in /) {
 	    if ($debug) {print "\ndebug: $_";}
 	    @line_tokens = split(' ');
 	    $num_line_tokens = @line_tokens;

QA/runtests.mpi.unix

@@ -32,7 +32,9 @@
 export HYDRA_DEBUG=0
 # find memory leaks using  this glibc feature that
 # initialized memory blocks to non-zero values
-export MALLOC_PERTURB_=$(($RANDOM % 255 + 1))
+if [[ -z "${USE_ASAN}" ]]; then
+    export MALLOC_PERTURB_=$(($RANDOM % 255 + 1))
+fi    
 
 if [[ -z "${NWCHEM_TARGET}" ]]; then
     UNAME_S=$(uname -s)

QA/tests/back_to_atguess/back_to_atguess.nw

@@ -0,0 +1,41 @@
+#
+# $Id$
+#
+
+title "Water geometry optimization in cc-pvdz basis set"
+
+start h2o
+
+geometry autosym
+  O    0.0    1.0    -0.02
+  H   -0.74   1.0    -0.76
+  H    0.74   1.0    -0.76
+  O    0.0    0.0    -0.02
+  H   -0.74   0.0    -0.76
+  H    0.74   0.0    -0.76
+end
+
+basis
+  H library cc-pvdz
+  O library cc-pvdz
+end
+
+driver
+  clear
+end
+
+scf; print low; end
+
+#task scf optimize
+dft
+vectors input hcore
+maxiter 1
+end
+task dft ignore
+
+set back_to_atguess 2
+dft
+maxiter 99
+end
+task dft
+

QA/tests/bas_details/bas_details.nw

@@ -0,0 +1,64 @@
+start
+
+title "test basis details keyword"
+
+echo
+
+geometry
+C     0.000000     0.000000     0.000000
+H     0.000000     0.000000     1.089000
+H     1.026719     0.000000    -0.363000
+H    -0.513360    -0.889165    -0.363000
+H    -0.513360     0.889165    -0.363000
+end
+
+basis b1 
+*  details library aug-cc-pvdz
+end
+
+basis b2 bse 
+*  library aug-cc-pvdz
+end
+
+basis b3 spherical
+*   library aug-cc-pvdz
+end
+
+basis b4 
+*   library aug-cc-pvdz
+end
+
+set "ao basis" b1
+
+dft
+vectors  output b1.mos
+print low
+end
+
+task dft
+
+set "ao basis" b2
+
+dft
+vectors input project b1 b1.mos output b2.mos
+end
+
+task dft
+
+set "ao basis" b3
+
+dft
+vectors input project b2 b2.mos output b3.mos
+end
+
+task dft
+
+
+set "ao basis" b4
+
+dft
+vectors input project b3 b3.mos output b4.mos
+end
+
+task dft
+

QA/tests/bas_details/bas_details.out

@@ -0,0 +1,836 @@
+ argument  1 = /Users/edo/nwchem/nwchem-edoapra-master/QA/tests/bas_details/bas_details.nw
+  NWChem w/ OpenMP: maximum threads =    1
+ 
+
+
+============================== echo of input deck ==============================
+start
+
+title "test basis details keyword"
+
+echo
+
+geometry
+C     0.000000     0.000000     0.000000
+H     0.000000     0.000000     1.089000
+H     1.026719     0.000000    -0.363000
+H    -0.513360    -0.889165    -0.363000
+H    -0.513360     0.889165    -0.363000
+end
+
+basis b1
+*  details library aug-cc-pvdz
+end
+
+basis b2 bse
+*  library aug-cc-pvdz
+end
+
+basis b3 spherical
+*   library aug-cc-pvdz
+end
+
+basis b4
+*   library aug-cc-pvdz
+end
+
+set "ao basis" b1
+
+dft
+vectors  output b1.mos
+print low
+end
+
+task dft
+
+set "ao basis" b2
+
+dft
+vectors input project b1 b1.mos output b2.mos
+end
+
+task dft
+
+set "ao basis" b3
+
+dft
+vectors input project b2 b2.mos output b3.mos
+end
+
+task dft
+
+
+set "ao basis" b4
+
+dft
+vectors input project b3 b3.mos output b4.mos
+end
+
+task dft
+
+================================================================================
+
+
+                                         
+                                         
+ 
+ 
+             Northwest Computational Chemistry Package (NWChem) 7.2.0
+             --------------------------------------------------------
+ 
+ 
+                    Environmental Molecular Sciences Laboratory
+                       Pacific Northwest National Laboratory
+                                Richland, WA 99352
+ 
+                              Copyright (c) 1994-2022
+                       Pacific Northwest National Laboratory
+                            Battelle Memorial Institute
+ 
+             NWChem is an open-source computational chemistry package
+                        distributed under the terms of the
+                      Educational Community License (ECL) 2.0
+             A copy of the license is included with this distribution
+                              in the LICENSE.TXT file
+ 
+                                  ACKNOWLEDGMENT
+                                  --------------
+
+            This software and its documentation were developed at the
+            EMSL at Pacific Northwest National Laboratory, a multiprogram
+            national laboratory, operated for the U.S. Department of Energy
+            by Battelle under Contract Number DE-AC05-76RL01830. Support
+            for this work was provided by the Department of Energy Office
+            of Biological and Environmental Research, Office of Basic
+            Energy Sciences, and the Office of Advanced Scientific Computing.
+
+
+           Job information
+           ---------------
+
+    hostname        = WE40672
+    program         = /Users/edo/nwchem/nwchem-edoapra-master/bin/MACX64/nwchem
+    date            = Wed Dec 14 18:04:30 2022
+
+    compiled        = Wed_Dec_14_18:03:49_2022
+    source          = /Users/edo/nwchem/nwchem-edoapra-master
+    nwchem branch   = 7.2.0
+    nwchem revision = nwchem_on_git-4232-g9621f8a386
+    ga revision     = 5.8.1
+    use scalapack   = T
+    input           = /Users/edo/nwchem/nwchem-edoapra-master/QA/tests/bas_details/bas_details.nw
+    prefix          = bas_details.
+    data base       = ./bas_details.db
+    status          = startup
+    nproc           =        1
+    time left       =     -1s
+
+
+
+           Memory information
+           ------------------
+
+    heap     =   26214396 doubles =    200.0 Mbytes
+    stack    =   26214401 doubles =    200.0 Mbytes
+    global   =   52428800 doubles =    400.0 Mbytes (distinct from heap & stack)
+    total    =  104857597 doubles =    800.0 Mbytes
+    verify   = yes
+    hardfail = no 
+
+
+           Directory information
+           ---------------------
+ 
+  0 permanent = .
+  0 scratch   = .
+ 
+ 
+ 
+ 
+                                NWChem Input Module
+                                -------------------
+ 
+ 
+                            test basis details keyword
+                            --------------------------
+
+ Scaling coordinates for geometry "geometry" by  1.889725989
+ (inverse scale =  0.529177249)
+
+ TD  symmetry detected
+
+          ------
+          auto-z
+          ------
+  Looking for out-of-plane bends
+  no constraints, skipping   0.000000000000000E+000
+  no constraints, skipping   0.000000000000000E+000
+ 
+ 
+                             Geometry "geometry" -> ""
+                             -------------------------
+ 
+ Output coordinates in angstroms (scale by  1.889725989 to convert to a.u.)
+ 
+  No.       Tag          Charge          X              Y              Z
+ ---- ---------------- ---------- -------------- -------------- --------------
+    1 C                    6.0000     0.00000000     0.00000000     0.00000000
+    2 H                    1.0000    -0.62873455     0.62873455     0.62873455
+    3 H                    1.0000     0.62873455    -0.62873455     0.62873455
+    4 H                    1.0000    -0.62873455    -0.62873455    -0.62873455
+    5 H                    1.0000     0.62873455     0.62873455    -0.62873455
+ 
+      Atomic Mass 
+      ----------- 
+ 
+      C                 12.000000
+      H                  1.007825
+ 
+
+ Effective nuclear repulsion energy (a.u.)      13.4477252140
+
+            Nuclear Dipole moment (a.u.) 
+            ----------------------------
+        X                 Y               Z
+ ---------------- ---------------- ----------------
+     0.0000000000     0.0000000000     0.0000000000
+ 
+      Symmetry information
+      --------------------
+ 
+ Group name              Td       
+ Group number             42
+ Group order              24
+ No. of unique centers     2
+ 
+      Symmetry unique atoms
+ 
+     1    2
+ 
+
+
+                                Z-matrix (autoz)
+                                -------- 
+
+ Units are Angstrom for bonds and degrees for angles
+ 
+      Type          Name      I     J     K     L     M      Value
+      ----------- --------  ----- ----- ----- ----- ----- ----------
+    1 Stretch                  1     2                       1.08900
+    2 Stretch                  1     3                       1.08900
+    3 Stretch                  1     4                       1.08900
+    4 Stretch                  1     5                       1.08900
+    5 Bend                     2     1     3               109.47122
+    6 Bend                     2     1     4               109.47122
+    7 Bend                     2     1     5               109.47122
+    8 Bend                     3     1     4               109.47122
+    9 Bend                     3     1     5               109.47122
+   10 Bend                     4     1     5               109.47122
+ 
+ 
+            XYZ format geometry
+            -------------------
+     5
+ geometry
+ C                     0.00000000     0.00000000     0.00000000
+ H                    -0.62873455     0.62873455     0.62873455
+ H                     0.62873455    -0.62873455     0.62873455
+ H                    -0.62873455    -0.62873455    -0.62873455
+ H                     0.62873455     0.62873455    -0.62873455
+ 
+ ==============================================================================
+                                internuclear distances
+ ------------------------------------------------------------------------------
+       center one      |      center two      | atomic units |  angstroms
+ ------------------------------------------------------------------------------
+    2 H                |   1 C                |     2.05791  |     1.08900
+    3 H                |   1 C                |     2.05791  |     1.08900
+    4 H                |   1 C                |     2.05791  |     1.08900
+    5 H                |   1 C                |     2.05791  |     1.08900
+ ------------------------------------------------------------------------------
+                         number of included internuclear distances:          4
+ ==============================================================================
+
+
+
+ ==============================================================================
+                                 internuclear angles
+ ------------------------------------------------------------------------------
+        center 1       |       center 2       |       center 3       |  degrees
+ ------------------------------------------------------------------------------
+    2 H                |   1 C                |   3 H                |   109.47
+    2 H                |   1 C                |   4 H                |   109.47
+    2 H                |   1 C                |   5 H                |   109.47
+    3 H                |   1 C                |   4 H                |   109.47
+    3 H                |   1 C                |   5 H                |   109.47
+    4 H                |   1 C                |   5 H                |   109.47
+ ------------------------------------------------------------------------------
+                            number of included internuclear angles:          6
+ ==============================================================================
+
+
+
+
+
+ Summary of "b1" -> "" (cartesian)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ *                        aug-cc-pvdz                 on all atoms 
+
+
+
+
+ Summary of "b2" -> "" (cartesian)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ *                        aug-cc-pvdz                 on all atoms 
+
+
+
+
+ Summary of "b3" -> "" (spherical)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ *                        aug-cc-pvdz                 on all atoms 
+
+
+
+
+ Summary of "b4" -> "" (cartesian)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ *                        aug-cc-pvdz                 on all atoms 
+
+
+ 
+                                 NWChem DFT Module
+                                 -----------------
+ 
+ 
+                            test basis details keyword
+ 
+ 
+                      Basis "ao basis" -> "b1" (spherical)
+                      -----
+  C (Carbon)
+  ----------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  6.66500000E+03  0.000692
+  1 S  1.00000000E+03  0.005329
+  1 S  2.28000000E+02  0.027077
+  1 S  6.47100000E+01  0.101718
+  1 S  2.10600000E+01  0.274740
+  1 S  7.49500000E+00  0.448564
+  1 S  2.79700000E+00  0.285074
+  1 S  5.21500000E-01  0.015204
+ 
+  2 S  6.66500000E+03 -0.000146
+  2 S  1.00000000E+03 -0.001154
+  2 S  2.28000000E+02 -0.005725
+  2 S  6.47100000E+01 -0.023312
+  2 S  2.10600000E+01 -0.063955
+  2 S  7.49500000E+00 -0.149981
+  2 S  2.79700000E+00 -0.127262
+  2 S  5.21500000E-01  0.544529
+ 
+  3 S  1.59600000E-01  1.000000
+ 
+  4 S  4.69000000E-02  1.000000
+ 
+  5 P  9.43900000E+00  0.038109
+  5 P  2.00200000E+00  0.209480
+  5 P  5.45600000E-01  0.508557
+ 
+  6 P  1.51700000E-01  1.000000
+ 
+  7 P  4.04100000E-02  1.000000
+ 
+  8 D  5.50000000E-01  1.000000
+ 
+  9 D  1.51000000E-01  1.000000
+ 
+  H (Hydrogen)
+  ------------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  1.30100000E+01  0.019685
+  1 S  1.96200000E+00  0.137977
+  1 S  4.44600000E-01  0.478148
+ 
+  2 S  1.22000000E-01  1.000000
+ 
+  3 S  2.97400000E-02  1.000000
+ 
+  4 P  7.27000000E-01  1.000000
+ 
+  5 P  1.41000000E-01  1.000000
+ 
+
+
+ Summary of "ao basis" -> "b1" (spherical)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ C                        aug-cc-pvdz                9       23   4s3p2d
+ H                        aug-cc-pvdz                5        9   3s2p
+
+
+ 
+  Caching 1-el integrals 
+  WARNING: movecs_in_org=atomic not equal to movecs_in=./b1.mos
+   Time after variat. SCF:      0.4
+   Time prior to 1st pass:      0.4
+
+
+         Total DFT energy =      -40.098721470454
+      One electron energy =      -79.817762448480
+           Coulomb energy =       32.731732491192
+    Exchange-Corr. energy =       -6.460416727190
+ Nuclear repulsion energy =       13.447725214025
+
+ Numeric. integr. density =       10.000005610713
+
+     Total iterative time =      0.3s
+
+
+ 
+                  Occupations of the irreducible representations
+                  ----------------------------------------------
+ 
+                     irrep           alpha         beta
+                     --------     --------     --------
+                     a1                2.0          2.0
+                     a2                0.0          0.0
+                     e                 0.0          0.0
+                     t1                0.0          0.0
+                     t2                3.0          3.0
+ 
+
+ Task  times  cpu:        0.4s     wall:        0.4s
+ 
+ 
+                                NWChem Input Module
+                                -------------------
+ 
+ 
+ 
+                                 NWChem DFT Module
+                                 -----------------
+ 
+ 
+                            test basis details keyword
+ 
+ 
+                      Basis "ao basis" -> "b2" (spherical)
+                      -----
+  C (Carbon)
+  ----------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  6.66500000E+03  0.000692
+  1 S  1.00000000E+03  0.005329
+  1 S  2.28000000E+02  0.027077
+  1 S  6.47100000E+01  0.101718
+  1 S  2.10600000E+01  0.274740
+  1 S  7.49500000E+00  0.448564
+  1 S  2.79700000E+00  0.285074
+  1 S  5.21500000E-01  0.015204
+ 
+  2 S  6.66500000E+03 -0.000146
+  2 S  1.00000000E+03 -0.001154
+  2 S  2.28000000E+02 -0.005725
+  2 S  6.47100000E+01 -0.023312
+  2 S  2.10600000E+01 -0.063955
+  2 S  7.49500000E+00 -0.149981
+  2 S  2.79700000E+00 -0.127262
+  2 S  5.21500000E-01  0.544529
+ 
+  3 S  1.59600000E-01  1.000000
+ 
+  4 S  4.69000000E-02  1.000000
+ 
+  5 P  9.43900000E+00  0.038109
+  5 P  2.00200000E+00  0.209480
+  5 P  5.45600000E-01  0.508557
+ 
+  6 P  1.51700000E-01  1.000000
+ 
+  7 P  4.04100000E-02  1.000000
+ 
+  8 D  5.50000000E-01  1.000000
+ 
+  9 D  1.51000000E-01  1.000000
+ 
+  H (Hydrogen)
+  ------------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  1.30100000E+01  0.019685
+  1 S  1.96200000E+00  0.137977
+  1 S  4.44600000E-01  0.478148
+ 
+  2 S  1.22000000E-01  1.000000
+ 
+  3 S  2.97400000E-02  1.000000
+ 
+  4 P  7.27000000E-01  1.000000
+ 
+  5 P  1.41000000E-01  1.000000
+ 
+
+
+ Summary of "ao basis" -> "b2" (spherical)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ C                        aug-cc-pvdz                9       23   4s3p2d
+ H                        aug-cc-pvdz                5        9   3s2p
+
+
+ 
+  Caching 1-el integrals 
+  WARNING: movecs_in_org=project not equal to movecs_in=./b2.mos
+   Time after variat. SCF:      0.7
+   Time prior to 1st pass:      0.7
+
+
+         Total DFT energy =      -40.098721474766
+      One electron energy =      -79.817417679667
+           Coulomb energy =       32.731340846306
+    Exchange-Corr. energy =       -6.460369855430
+ Nuclear repulsion energy =       13.447725214025
+
+ Numeric. integr. density =       10.000005610781
+
+     Total iterative time =      0.2s
+
+
+ 
+                  Occupations of the irreducible representations
+                  ----------------------------------------------
+ 
+                     irrep           alpha         beta
+                     --------     --------     --------
+                     a1                2.0          2.0
+                     a2                0.0          0.0
+                     e                 0.0          0.0
+                     t1                0.0          0.0
+                     t2                3.0          3.0
+ 
+
+ Task  times  cpu:        0.3s     wall:        0.3s
+ 
+ 
+                                NWChem Input Module
+                                -------------------
+ 
+ 
+ 
+                                 NWChem DFT Module
+                                 -----------------
+ 
+ 
+                            test basis details keyword
+ 
+ 
+                      Basis "ao basis" -> "b3" (spherical)
+                      -----
+  C (Carbon)
+  ----------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  6.66500000E+03  0.000692
+  1 S  1.00000000E+03  0.005329
+  1 S  2.28000000E+02  0.027077
+  1 S  6.47100000E+01  0.101718
+  1 S  2.10600000E+01  0.274740
+  1 S  7.49500000E+00  0.448564
+  1 S  2.79700000E+00  0.285074
+  1 S  5.21500000E-01  0.015204
+ 
+  2 S  6.66500000E+03 -0.000146
+  2 S  1.00000000E+03 -0.001154
+  2 S  2.28000000E+02 -0.005725
+  2 S  6.47100000E+01 -0.023312
+  2 S  2.10600000E+01 -0.063955
+  2 S  7.49500000E+00 -0.149981
+  2 S  2.79700000E+00 -0.127262
+  2 S  5.21500000E-01  0.544529
+ 
+  3 S  1.59600000E-01  1.000000
+ 
+  4 S  4.69000000E-02  1.000000
+ 
+  5 P  9.43900000E+00  0.038109
+  5 P  2.00200000E+00  0.209480
+  5 P  5.45600000E-01  0.508557
+ 
+  6 P  1.51700000E-01  1.000000
+ 
+  7 P  4.04100000E-02  1.000000
+ 
+  8 D  5.50000000E-01  1.000000
+ 
+  9 D  1.51000000E-01  1.000000
+ 
+  H (Hydrogen)
+  ------------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  1.30100000E+01  0.019685
+  1 S  1.96200000E+00  0.137977
+  1 S  4.44600000E-01  0.478148
+ 
+  2 S  1.22000000E-01  1.000000
+ 
+  3 S  2.97400000E-02  1.000000
+ 
+  4 P  7.27000000E-01  1.000000
+ 
+  5 P  1.41000000E-01  1.000000
+ 
+
+
+ Summary of "ao basis" -> "b3" (spherical)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ C                        aug-cc-pvdz                9       23   4s3p2d
+ H                        aug-cc-pvdz                5        9   3s2p
+
+
+ 
+  Caching 1-el integrals 
+  WARNING: movecs_in_org=project not equal to movecs_in=./b3.mos
+   Time after variat. SCF:      1.0
+   Time prior to 1st pass:      1.0
+
+
+         Total DFT energy =      -40.098721474784
+      One electron energy =      -79.817393093582
+           Coulomb energy =       32.731313156390
+    Exchange-Corr. energy =       -6.460366751616
+ Nuclear repulsion energy =       13.447725214025
+
+ Numeric. integr. density =       10.000005610778
+
+     Total iterative time =      0.2s
+
+
+ 
+                  Occupations of the irreducible representations
+                  ----------------------------------------------
+ 
+                     irrep           alpha         beta
+                     --------     --------     --------
+                     a1                2.0          2.0
+                     a2                0.0          0.0
+                     e                 0.0          0.0
+                     t1                0.0          0.0
+                     t2                3.0          3.0
+ 
+
+ Task  times  cpu:        0.3s     wall:        0.3s
+ 
+ 
+                                NWChem Input Module
+                                -------------------
+ 
+ 
+ 
+                                 NWChem DFT Module
+                                 -----------------
+ 
+ 
+                            test basis details keyword
+ 
+ 
+                      Basis "ao basis" -> "b4" (cartesian)
+                      -----
+  C (Carbon)
+  ----------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  6.66500000E+03  0.000692
+  1 S  1.00000000E+03  0.005329
+  1 S  2.28000000E+02  0.027077
+  1 S  6.47100000E+01  0.101718
+  1 S  2.10600000E+01  0.274740
+  1 S  7.49500000E+00  0.448564
+  1 S  2.79700000E+00  0.285074
+  1 S  5.21500000E-01  0.015204
+ 
+  2 S  6.66500000E+03 -0.000146
+  2 S  1.00000000E+03 -0.001154
+  2 S  2.28000000E+02 -0.005725
+  2 S  6.47100000E+01 -0.023312
+  2 S  2.10600000E+01 -0.063955
+  2 S  7.49500000E+00 -0.149981
+  2 S  2.79700000E+00 -0.127262
+  2 S  5.21500000E-01  0.544529
+ 
+  3 S  1.59600000E-01  1.000000
+ 
+  4 S  4.69000000E-02  1.000000
+ 
+  5 P  9.43900000E+00  0.038109
+  5 P  2.00200000E+00  0.209480
+  5 P  5.45600000E-01  0.508557
+ 
+  6 P  1.51700000E-01  1.000000
+ 
+  7 P  4.04100000E-02  1.000000
+ 
+  8 D  5.50000000E-01  1.000000
+ 
+  9 D  1.51000000E-01  1.000000
+ 
+  H (Hydrogen)
+  ------------
+            Exponent  Coefficients 
+       -------------- ---------------------------------------------------------
+  1 S  1.30100000E+01  0.019685
+  1 S  1.96200000E+00  0.137977
+  1 S  4.44600000E-01  0.478148
+ 
+  2 S  1.22000000E-01  1.000000
+ 
+  3 S  2.97400000E-02  1.000000
+ 
+  4 P  7.27000000E-01  1.000000
+ 
+  5 P  1.41000000E-01  1.000000
+ 
+
+
+ Summary of "ao basis" -> "b4" (cartesian)
+ ------------------------------------------------------------------------------
+       Tag                 Description            Shells   Functions and Types
+ ---------------- ------------------------------  ------  ---------------------
+ C                        aug-cc-pvdz                9       25   4s3p2d
+ H                        aug-cc-pvdz                5        9   3s2p
+
+
+ 
+  Caching 1-el integrals 
+  WARNING: movecs_in_org=project not equal to movecs_in=./b4.mos
+   Time after variat. SCF:      1.3
+   Time prior to 1st pass:      1.3
+
+
+         Total DFT energy =      -40.100836427564
+      One electron energy =      -79.804187567806
+           Coulomb energy =       32.702475992294
+    Exchange-Corr. energy =       -6.446850066076
+ Nuclear repulsion energy =       13.447725214025
+
+ Numeric. integr. density =       10.000005618093
+
+     Total iterative time =      0.3s
+
+
+ 
+                  Occupations of the irreducible representations
+                  ----------------------------------------------
+ 
+                     irrep           alpha         beta
+                     --------     --------     --------
+                     a1                2.0          2.0
+                     a2                0.0          0.0
+                     e                 0.0          0.0
+                     t1                0.0          0.0
+                     t2                3.0          3.0
+ 
+
+ Task  times  cpu:        0.4s     wall:        0.4s
+ 
+ 
+                                NWChem Input Module
+                                -------------------
+ 
+ 
+ Summary of allocated global arrays
+-----------------------------------
+  No active global arrays
+
+
+MA_summarize_allocated_blocks: starting scan ...
+heap block 'gridpts', handle 74, address 0x7fedf6985698:
+	type of elements:		double precision
+	number of elements:		33554432
+	address of client space:	0x7fedf6985700
+	index for client space:		17581915250195
+	total number of bytes:		268435568
+MA_summarize_allocated_blocks: scan completed: 1 heap block, 0 stack blocks
+MA usage statistics:
+
+	allocation statistics:
+					      heap	     stack
+					      ----	     -----
+	current number of blocks	         1	         0
+	maximum number of blocks	        25	        55
+	current total bytes		 268435568	         0
+	maximum total bytes		 273176464	  22512552
+	maximum total K-bytes		    273177	     22513
+	maximum total M-bytes		       274	        23
+ 
+ 
+                                     CITATION
+                                     --------
+                Please cite the following reference when publishing
+                           results obtained with NWChem:
+ 
+          E. Apra, E. J. Bylaska, W. A. de Jong, N. Govind, K. Kowalski,
+       T. P. Straatsma, M. Valiev, H. J. J. van Dam, Y. Alexeev, J. Anchell,
+       V. Anisimov, F. W. Aquino, R. Atta-Fynn, J. Autschbach, N. P. Bauman,
+     J. C. Becca, D. E. Bernholdt, K. Bhaskaran-Nair, S. Bogatko, P. Borowski,
+         J. Boschen, J. Brabec, A. Bruner, E. Cauet, Y. Chen, G. N. Chuev,
+      C. J. Cramer, J. Daily, M. J. O. Deegan, T. H. Dunning Jr., M. Dupuis,
+   K. G. Dyall, G. I. Fann, S. A. Fischer, A. Fonari, H. Fruchtl, L. Gagliardi,
+      J. Garza, N. Gawande, S. Ghosh, K. Glaesemann, A. W. Gotz, J. Hammond,
+       V. Helms, E. D. Hermes, K. Hirao, S. Hirata, M. Jacquelin, L. Jensen,
+   B. G. Johnson, H. Jonsson, R. A. Kendall, M. Klemm, R. Kobayashi, V. Konkov,
+      S. Krishnamoorthy, M. Krishnan, Z. Lin, R. D. Lins, R. J. Littlefield,
+      A. J. Logsdail, K. Lopata, W. Ma, A. V. Marenich, J. Martin del Campo,
+   D. Mejia-Rodriguez, J. E. Moore, J. M. Mullin, T. Nakajima, D. R. Nascimento,
+    J. A. Nichols, P. J. Nichols, J. Nieplocha, A. Otero-de-la-Roza, B. Palmer,
+    A. Panyala, T. Pirojsirikul, B. Peng, R. Peverati, J. Pittner, L. Pollack,
+   R. M. Richard, P. Sadayappan, G. C. Schatz, W. A. Shelton, D. W. Silverstein,
+   D. M. A. Smith, T. A. Soares, D. Song, M. Swart, H. L. Taylor, G. S. Thomas,
+            V. Tipparaju, D. G. Truhlar, K. Tsemekhman, T. Van Voorhis,
+      A. Vazquez-Mayagoitia, P. Verma, O. Villa, A. Vishnu, K. D. Vogiatzis,
+        D. Wang, J. H. Weare, M. J. Williamson, T. L. Windus, K. Wolinski,
+        A. T. Wong, Q. Wu, C. Yang, Q. Yu, M. Zacharias, Z. Zhang, Y. Zhao,
+                                and R. J. Harrison
+                        "NWChem: Past, present, and future
+                         J. Chem. Phys. 152, 184102 (2020)
+                               doi:10.1063/5.0004997
+ 
+                                      AUTHORS
+                                      -------
+  E. Apra, E. J. Bylaska, N. Govind, K. Kowalski, M. Valiev, D. Mejia-Rodriguez,
+       A. Kunitsa, N. P. Bauman, A. Panyala, W. A. de Jong, T. P. Straatsma,
+   H. J. J. van Dam, D. Wang, T. L. Windus, J. Hammond, J. Autschbach, A. Woods,
+    K. Bhaskaran-Nair, J. Brabec, K. Lopata, S. A. Fischer, S. Krishnamoorthy,
+     M. Jacquelin, W. Ma, M. Klemm, O. Villa, Y. Chen, V. Anisimov, F. Aquino,
+     S. Hirata, M. T. Hackler, E. Hermes, L. Jensen, J. E. Moore, J. C. Becca,
+      V. Konjkov, T. Risthaus, M. Malagoli, A. Marenich, A. Otero-de-la-Roza,
+        J. Mullin, P. Nichols, R. Peverati, J. Pittner, Y. Zhao, P.-D. Fan,
+        A. Fonari, M. J. Williamson, R. J. Harrison, J. R. Rehr, M. Dupuis,
+     D. Silverstein, D. M. A. Smith, J. Nieplocha, V. Tipparaju, M. Krishnan,
+     B. E. Van Kuiken, A. Vazquez-Mayagoitia, M. Swart, Q. Wu, T. Van Voorhis,
+     A. A. Auer, M. Nooijen, L. D. Crosby, E. Brown, G. Cisneros, G. I. Fann,
+   H. Fruchtl, J. Garza, K. Hirao, R. A. Kendall, J. A. Nichols, K. Tsemekhman,
+    K. Wolinski, J. Anchell, D. E. Bernholdt, P. Borowski, T. Clark, D. Clerc,
+   H. Dachsel, M. J. O. Deegan, K. Dyall, D. Elwood, E. Glendening, M. Gutowski,
+   A. C. Hess, J. Jaffe, B. G. Johnson, J. Ju, R. Kobayashi, R. Kutteh, Z. Lin,
+   R. Littlefield, X. Long, B. Meng, T. Nakajima, S. Niu, L. Pollack, M. Rosing,
+   K. Glaesemann, G. Sandrone, M. Stave, H. Taylor, G. Thomas, J. H. van Lenthe,
+                               A. T. Wong, Z. Zhang.
+
+ Total times  cpu:        1.6s     wall:        1.6s

QA/tests/ch5n_nbo/ch5n_nbo.nw

@@ -19,7 +19,12 @@ geometry
  H            .566733    1.075185    -.889981
  H           -.423217   -1.094815     .824662
  H           -.423217   -1.094815    -.824662
- symmetry c1
+# symmetry c1
+end
+
+scf
+ sym off
+ adapt off
 end
 
 task SCF energy

QA/tests/dftd3_c6cn/dftd3_c6cn.nw

@@ -0,0 +1,62 @@
+echo
+charge 0
+
+geometry noautoz
+O          0.92504        0.99938        2.07185
+O          1.68765        1.57908       -0.29568
+H          3.55204        2.68516        0.28628
+C          0.29661        0.89268        3.31517
+C          2.11209        1.80315        2.02170
+C          2.43675        2.01653        0.57987
+H          0.67784        1.56486        4.44460
+O         -1.08891       -0.62239        2.15648
+O         -1.72707       -1.54708       -0.16299
+H         -3.68169       -2.46110        0.48375
+C         -0.80690        0.02845        3.35278
+C         -2.33997       -1.33013        2.13140
+C         -2.56045       -1.79603        0.72196
+H         -1.50976       -0.14582        4.51669
+Ca         0.00000        0.00000        0.00000
+O         -0.92505        0.99938       -2.07185
+O         -1.68765        1.57908        0.29568
+H         -3.55205        2.68516       -0.28628
+C         -0.29662        0.89268       -3.31516
+C         -2.11210        1.80315       -2.02170
+C         -2.43676        2.01653       -0.57987
+H         -0.67784        1.56486       -4.44460
+O          1.08890       -0.62239       -2.15647
+O          1.72706       -1.54708        0.16299
+H          3.68169       -2.46110       -0.48374
+C          0.80690        0.02845       -3.35278
+C          2.33996       -1.33013       -2.13140
+C          2.56044       -1.79603       -0.72196
+H          1.50975       -0.14582       -4.51669
+H          2.97119        1.25775        2.53560
+H          1.91900        2.79894        2.54200
+H          1.85225        2.99521        0.59220
+H         -3.18534       -0.63249        2.44484
+H         -2.28804       -2.21902        2.84328
+H         -2.02490       -2.77456        0.95706
+H         -2.97119        1.25775       -2.53560
+H         -1.91901        2.79894       -2.54200
+H         -1.85225        2.99521       -0.59220
+H          3.18533       -0.63249       -2.44484
+H          2.28803       -2.21902       -2.84328
+H          2.02489       -2.77456       -0.95706
+
+end 
+
+basis "ao basis" spherical
+* library def2-svp
+end
+
+
+
+dft
+direct
+xc cpbe96 xpbe96
+adft
+disp vdw 4
+convergence fast
+end
+task dft

QA/tests/h2o2_fde/h2o2_fde.nw

@@ -1,7 +1,6 @@
 start dimer
   title "Water dimer DFT using fragment initial guess"
-  geometry print nocenter noautoz dimer
-  symmetry c1 tol 0.0
+  geometry print nocenter noautoz noautosym dimer
     O   -0.595   1.165  -0.048 
     H    0.110   1.812  -0.170 
     H   -1.452   1.598  -0.154 
@@ -9,14 +8,12 @@ start dimer
     H    0.175  -2.013   0.348
     H    0.177  -0.480   0.010
   end
-  geometry print nocenter noautoz h2o1 
-  symmetry c1 tol 0.0
+  geometry print nocenter noautoz noautosym h2o1 
     O   -0.595   1.165  -0.048
     H    0.110   1.812  -0.170
     H   -1.452   1.598  -0.154
   end
-  geometry print nocenter noautoz h2o2 
-  symmetry c1 tol 0.0
+  geometry print nocenter noautoz noautosym h2o2 
     O    0.724  -1.284   0.034
     H    0.175  -2.013   0.348
     H    0.177  -0.480   0.010
@@ -27,19 +24,26 @@ start dimer
   end
 
   set dft:no_prune true
+  dft
+   sym off
+   adapt off
+   grid xfine
+   tolerances tight
+   convergence energy 1d-9
+  end
 
   set geometry h2o1
-  dft; vectors input atomic output h2o1.movecs; grid xfine; tolerances tight; end
+  dft; vectors input atomic output h2o1.movecs;  end
   task dft
 
   set geometry h2o2
-  dft; vectors input atomic output h2o2.movecs; grid xfine; tolerances tight; end
+  dft; vectors input atomic output h2o2.movecs;  end
   task dft
 
   set geometry dimer
   dft
   vectors input fragment h2o1.movecs h2o2.movecs \
-          output dimer.movecs; tolerances tight; grid xfine;
+          output dimer.movecs; 
   end
   task dft
 
@@ -53,23 +57,23 @@ start dimer
   # Spin-paired Thomas Fermi
   set geometry h2o1
   dft; frozemb geometry h2o2 vectors h2o2.movecs; frozemb_ts tf; \
-  vectors input h2o1.movecs output h2o1.fde.movecs; odft; grid xfine; tolerances tight; end
+  vectors input h2o1.movecs output h2o1.fde.movecs; odft;  end
   task dft
 
   # Spin-unpaired 
   dft; frozemb geometry h2o2 vectors h2o2.movecs; frozemb_ts tf; \
-  vectors input h2o1.movecs output h2o1.fde.movecs; odft; grid xfine; tolerances tight; end
+  vectors input h2o1.movecs output h2o1.fde.movecs; odft;  end
   task dft
 
   # Spin-paired von Weisacker
   set geometry h2o1
   dft; frozemb geometry h2o2 vectors h2o2.movecs; frozemb_ts vw; \
-  vectors input h2o1.movecs output h2o1.fde.movecs; odft; grid xfine; tolerances tight; end
+  vectors input h2o1.movecs output h2o1.fde.movecs; odft; end
   task dft
 
   # Spin-unpaired 
   dft; frozemb geometry h2o2 vectors h2o2.movecs; frozemb_ts vw; \
-  vectors input h2o1.movecs output h2o1.fde.movecs; odft; grid xfine; tolerances tight; end
+  vectors input h2o1.movecs output h2o1.fde.movecs; odft; end
   task dft
 
   unset dft:frozemb

QA/tests/hi_zora_sf/hi_zora_sf.nw

@@ -11,12 +11,11 @@ basis spherical
 end
 
 set dft:no_prune T
-set dft:job_grid_acc 1d-20
 set dft:tol_rho 1d-30
 
 dft
- grid xfine
- grid lebedev H 350 18 I 350 18 
+ grid huge
+ tolerances acccoul 15
  xc b3lyp
 end
 

QA/tests/localize-ibo-aa/localize-ibo-aa.nw

@@ -71,7 +71,7 @@ convergence energy 1e-10
 end
 
 property
- localization ibo
+ localization ibo 2
 end
 
 task scf property

QA/tests/localize-ibo-aa/localize-ibo-aa.out

@@ -1,4 +1,4 @@
-running on 4 processors
+running on 6 processors
  argument  1 = acrylic-acid.nw
  
 
@@ -77,12 +77,14 @@ convergence energy 1e-10
 end
 
 property
- localization ibo
+ localization ibo 2
 end
 
+#task scf energy
 task scf property
 
 
+
 ================================================================================
 
 
@@ -90,7 +92,7 @@ task scf property
                                          
  
  
-             Northwest Computational Chemistry Package (NWChem) 7.0.1
+             Northwest Computational Chemistry Package (NWChem) 7.2.0
              --------------------------------------------------------
  
  
@@ -98,7 +100,7 @@ task scf property
                        Pacific Northwest National Laboratory
                                 Richland, WA 99352
  
-                              Copyright (c) 1994-2020
+                              Copyright (c) 1994-2022
                        Pacific Northwest National Laboratory
                             Battelle Memorial Institute
  
@@ -123,21 +125,21 @@ task scf property
            Job information
            ---------------
 
-    hostname        = ja31
-    program         = /home/workspace/jochena/nwchem/github-fork-ibo-devel/bin/LINUX64/nwchem
-    date            = Thu May  5 15:37:05 2022
+    hostname        = ja04
+    program         = /home/workspace/jochena/nwchem/github-fork-ibovir-devel/bin/LINUX64/nwchem
+    date            = Tue Feb 14 12:19:49 2023
 
-    compiled        = Thu_May_05_15:36:59_2022
-    source          = /home/workspace/jochena/nwchem/github-fork-ibo-devel
-    nwchem branch   = 7.0.0
-    nwchem revision = nwchem_on_git-3953-g107e8544ce
-    ga revision     = 5.8.1
-    use scalapack   = T
+    compiled        = Tue_Feb_14_12:19:44_2023
+    source          = /home/workspace/jochena/nwchem/github-fork-ibovir-devel
+    nwchem branch   = 7.2.0
+    nwchem revision = nwchem_on_git-4622-g8f291f7cfd
+    ga revision     = 5.8.0
+    use scalapack   = F
     input           = acrylic-acid.nw
     prefix          = testjob.
     data base       = ./testjob.db
     status          = startup
-    nproc           =        3
+    nproc           =        5
     time left       =     -1s
 
 
@@ -145,10 +147,10 @@ task scf property
            Memory information
            ------------------
 
-    heap     =    6553596 doubles =     50.0 Mbytes
-    stack    =    6553601 doubles =     50.0 Mbytes
+    heap     =    6553598 doubles =     50.0 Mbytes
+    stack    =    6553595 doubles =     50.0 Mbytes
     global   =   13107200 doubles =    100.0 Mbytes (distinct from heap & stack)
-    total    =   26214397 doubles =    200.0 Mbytes
+    total    =   26214393 doubles =    200.0 Mbytes
     verify   = yes
     hardfail = no 
 
@@ -265,7 +267,7 @@ task scf property
 
   library name resolved from: environment
   library file name is: <
- /home/workspace/jochena/nwchem/github-fork-ibo-devel/src/basis/libraries/>
+ /home/workspace/jochena/nwchem/github-fork-ibovir-devel/src/basis/libraries/>
   
 
 
@@ -278,7 +280,7 @@ task scf property
 
   library name resolved from: environment
   library file name is: <
- /home/workspace/jochena/nwchem/github-fork-ibo-devel/src/basis/libraries/>
+ /home/workspace/jochena/nwchem/github-fork-ibovir-devel/src/basis/libraries/>
   
                       Basis "iao basis" -> "" (spherical)
                       -----
@@ -502,22 +504,22 @@ task scf property
 
               iter       energy          gnorm     gmax       time
              ----- ------------------- --------- --------- --------
-                 1     -265.3310324443  1.42D+00  2.92D-01      0.2
-                 2     -265.4360577045  4.13D-01  8.78D-02      0.4
-                 3     -265.4514376314  3.64D-02  9.22D-03      0.8
-                 4     -265.4516512505  1.03D-03  2.34D-04      1.3
-                 5     -265.4516513710  7.80D-06  1.92D-06      1.9
+                 1     -265.3310324443  1.42D+00  2.92D-01      0.1
+                 2     -265.4360577045  4.13D-01  8.78D-02      0.2
+                 3     -265.4514376314  3.64D-02  9.22D-03      0.4
+                 4     -265.4516512505  1.03D-03  2.34D-04      0.6
+                 5     -265.4516513711  7.80D-06  1.92D-06      0.9
 
 
        Final RHF  results 
        ------------------ 
 
-         Total SCF energy =   -265.451651371048
-      One-electron energy =   -682.442839763839
+         Total SCF energy =   -265.451651371050
+      One-electron energy =   -682.442839763841
       Two-electron energy =    255.473969802116
  Nuclear repulsion energy =    161.517218590675
 
-        Time for solution =      1.8s
+        Time for solution =      0.9s
 
 
              Final eigenvalues
@@ -558,7 +560,7 @@ task scf property
                        -------------------------------------
  
  Vector    6  Occ=2.000000D+00  E=-1.441125D+00
-              MO Center= -1.1D+00,  2.6D-02, -3.4D-17, r^2= 1.2D+00
+              MO Center= -1.1D+00,  2.6D-02, -2.1D-17, r^2= 1.2D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     58      0.413566  5 O  s                 30      0.293845  3 C  s          
@@ -566,7 +568,7 @@ task scf property
     57      0.208468  5 O  s                 45      0.150994  4 O  s          
  
  Vector    7  Occ=2.000000D+00  E=-1.338769D+00
-              MO Center= -9.1D-01, -4.1D-01, -2.6D-17, r^2= 1.4D+00
+              MO Center= -9.1D-01, -4.1D-01,  6.6D-17, r^2= 1.4D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     44      0.424994  4 O  s                 58     -0.352816  5 O  s          
@@ -575,7 +577,7 @@ task scf property
     57     -0.176450  5 O  s          
  
  Vector    8  Occ=2.000000D+00  E=-1.070923D+00
-              MO Center=  1.1D+00,  2.8D-01, -4.7D-17, r^2= 1.5D+00
+              MO Center=  1.1D+00,  2.8D-01,  3.4D-17, r^2= 1.5D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     16      0.400200  2 C  s                  2      0.337172  1 C  s          
@@ -583,7 +585,7 @@ task scf property
     17      0.160365  2 C  s          
  
  Vector    9  Occ=2.000000D+00  E=-8.974613D-01
-              MO Center=  2.3D-01,  2.6D-01, -9.8D-17, r^2= 3.3D+00
+              MO Center=  2.3D-01,  2.6D-01, -1.0D-16, r^2= 3.3D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     30     -0.305983  3 C  s                  2      0.303431  1 C  s          
@@ -592,7 +594,7 @@ task scf property
      1     -0.152955  1 C  s          
  
  Vector   10  Occ=2.000000D+00  E=-7.642057D-01
-              MO Center=  1.6D-01,  3.0D-01,  1.4D-16, r^2= 4.1D+00
+              MO Center=  1.6D-01,  3.0D-01,  5.3D-16, r^2= 4.1D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     60      0.316925  5 O  px                16     -0.241192  2 C  s          
@@ -602,7 +604,7 @@ task scf property
      4      0.156757  1 C  px                71      0.155748  6 H  s          
  
  Vector   11  Occ=2.000000D+00  E=-7.114488D-01
-              MO Center=  2.7D-01,  9.8D-03, -1.4D-16, r^2= 3.1D+00
+              MO Center=  2.7D-01,  9.8D-03, -3.3D-16, r^2= 3.1D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     19      0.248395  2 C  py                81      0.212079  8 H  s          
@@ -612,7 +614,7 @@ task scf property
     44      0.168207  4 O  s                 46     -0.165171  4 O  px         
  
  Vector   12  Occ=2.000000D+00  E=-6.942998D-01
-              MO Center= -5.7D-01, -1.8D-01,  7.4D-17, r^2= 3.0D+00
+              MO Center= -5.7D-01, -1.8D-01, -2.1D-16, r^2= 3.0D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     61      0.332130  5 O  py                47     -0.285481  4 O  py         
@@ -622,7 +624,7 @@ task scf property
     33      0.150516  3 C  py         
  
  Vector   13  Occ=2.000000D+00  E=-6.290129D-01
-              MO Center= -9.9D-01,  1.2D-01,  2.6D-16, r^2= 1.6D+00
+              MO Center= -9.9D-01,  1.2D-01, -3.2D-16, r^2= 1.6D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     62      0.438625  5 O  pz                65      0.308876  5 O  pz         
@@ -630,7 +632,7 @@ task scf property
     51      0.167682  4 O  pz                37      0.158757  3 C  pz         
  
  Vector   14  Occ=2.000000D+00  E=-6.234118D-01
-              MO Center=  8.0D-01,  1.3D-02, -1.6D-16, r^2= 3.8D+00
+              MO Center=  8.0D-01,  1.3D-02,  4.5D-16, r^2= 3.8D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
      4      0.334286  1 C  px                76      0.262842  7 H  s          
@@ -639,7 +641,7 @@ task scf property
     18     -0.161739  2 C  px         
  
  Vector   15  Occ=2.000000D+00  E=-5.982438D-01
-              MO Center= -2.3D-02,  2.4D-01, -5.0D-17, r^2= 3.5D+00
+              MO Center= -2.3D-02,  2.4D-01, -3.8D-17, r^2= 3.5D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     61      0.305744  5 O  py                18      0.273635  2 C  px         
@@ -649,7 +651,7 @@ task scf property
     33     -0.180575  3 C  py                71     -0.170295  6 H  s          
  
  Vector   16  Occ=2.000000D+00  E=-5.496397D-01
-              MO Center=  1.1D+00,  6.0D-02,  2.5D-17, r^2= 3.3D+00
+              MO Center=  1.1D+00,  6.0D-02,  3.0D-16, r^2= 3.3D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
      5      0.298431  1 C  py                71     -0.271981  6 H  s          
@@ -658,7 +660,7 @@ task scf property
     33      0.163556  3 C  py                61     -0.150120  5 O  py         
  
  Vector   17  Occ=2.000000D+00  E=-4.889073D-01
-              MO Center= -9.0D-01, -2.5D-01, -6.7D-17, r^2= 2.0D+00
+              MO Center= -9.0D-01, -2.5D-01, -5.3D-16, r^2= 2.0D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     62     -0.430815  5 O  pz                48      0.407766  4 O  pz         
@@ -666,7 +668,7 @@ task scf property
     34      0.166236  3 C  pz         
  
  Vector   18  Occ=2.000000D+00  E=-4.582158D-01
-              MO Center= -5.2D-01, -7.3D-01, -9.6D-17, r^2= 1.9D+00
+              MO Center= -5.2D-01, -7.3D-01, -1.6D-16, r^2= 1.9D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     46      0.549426  4 O  px                49      0.417524  4 O  px         
@@ -674,7 +676,7 @@ task scf property
     64     -0.174821  5 O  py                18      0.151160  2 C  px         
  
  Vector   19  Occ=2.000000D+00  E=-4.009704D-01
-              MO Center=  1.1D+00,  1.5D-01, -2.0D-16, r^2= 2.0D+00
+              MO Center=  1.1D+00,  1.5D-01, -4.1D-16, r^2= 2.0D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     20      0.384445  2 C  pz                 6      0.352495  1 C  pz         
@@ -682,7 +684,7 @@ task scf property
     48     -0.214439  4 O  pz                51     -0.173233  4 O  pz         
  
  Vector   20  Occ=0.000000D+00  E= 9.067747D-02
-              MO Center=  8.9D-01,  1.5D-02,  2.7D-17, r^2= 3.0D+00
+              MO Center=  8.9D-01,  1.5D-02,  9.9D-17, r^2= 3.0D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
      9      0.621145  1 C  pz                23     -0.459696  2 C  pz         
@@ -692,7 +694,7 @@ task scf property
     65      0.155474  5 O  pz         
  
  Vector   21  Occ=0.000000D+00  E= 1.933962D-01
-              MO Center= -2.6D+00,  1.3D-01, -3.0D-16, r^2= 2.6D+00
+              MO Center= -2.6D+00,  1.3D-01,  2.0D-16, r^2= 2.6D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     87      1.532433  9 H  s                 17     -0.762684  2 C  s          
@@ -702,7 +704,7 @@ task scf property
     60      0.181785  5 O  px                 3      0.180050  1 C  s          
  
  Vector   22  Occ=0.000000D+00  E= 1.990748D-01
-              MO Center=  2.4D+00,  1.2D+00, -1.1D-16, r^2= 4.4D+00
+              MO Center=  2.4D+00,  1.2D+00,  5.2D-17, r^2= 4.4D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
      3      1.857859  1 C  s                 77     -1.791874  7 H  s          
@@ -712,7 +714,7 @@ task scf property
      8      0.218519  1 C  py                35      0.169789  3 C  px         
  
  Vector   23  Occ=0.000000D+00  E= 2.314259D-01
-              MO Center=  1.7D+00, -2.0D-01, -9.3D-17, r^2= 6.0D+00
+              MO Center=  1.7D+00, -2.0D-01,  9.0D-16, r^2= 6.0D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
      3      1.936960  1 C  s                 72     -1.805329  6 H  s          
@@ -721,7 +723,7 @@ task scf property
      8     -0.489914  1 C  py                 5     -0.154259  1 C  py         
  
  Vector   24  Occ=0.000000D+00  E= 2.626515D-01
-              MO Center=  1.3D-01, -2.5D-02, -4.7D-16, r^2= 2.9D+00
+              MO Center=  1.3D-01, -2.5D-02,  7.6D-16, r^2= 2.9D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     37     -0.804941  3 C  pz                23      0.756689  2 C  pz         
@@ -731,7 +733,7 @@ task scf property
     62      0.158924  5 O  pz         
  
  Vector   25  Occ=0.000000D+00  E= 2.688800D-01
-              MO Center=  1.9D+00,  3.8D-01,  5.4D-16, r^2= 6.9D+00
+              MO Center=  1.9D+00,  3.8D-01, -1.8D-15, r^2= 6.9D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     77     -1.706163  7 H  s                 72      1.689090  6 H  s          
@@ -741,7 +743,7 @@ task scf property
     87     -0.319009  9 H  s                  3     -0.288918  1 C  s          
  
  Vector   26  Occ=0.000000D+00  E= 3.855495D-01
-              MO Center= -4.6D-01, -2.3D-01,  1.1D-16, r^2= 3.7D+00
+              MO Center= -4.6D-01, -2.3D-01, -1.7D-16, r^2= 3.7D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     31      4.026240  3 C  s                 45     -1.439431  4 O  s          
@@ -751,7 +753,7 @@ task scf property
     50     -0.560923  4 O  py                82     -0.528079  8 H  s          
  
  Vector   27  Occ=0.000000D+00  E= 4.475195D-01
-              MO Center=  5.8D-01,  2.5D-01, -9.4D-17, r^2= 3.9D+00
+              MO Center=  5.8D-01,  2.5D-01,  3.4D-16, r^2= 3.9D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     21      3.533345  2 C  px                 3     -2.731556  1 C  s          
@@ -761,7 +763,7 @@ task scf property
      7      0.878572  1 C  px                30      0.524028  3 C  s          
  
  Vector   28  Occ=0.000000D+00  E= 4.523376D-01
-              MO Center=  1.5D+00, -1.4D-01,  9.3D-17, r^2= 4.3D+00
+              MO Center=  1.5D+00, -1.4D-01, -3.7D-16, r^2= 4.3D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     17      6.048068  2 C  s                  3     -4.672145  1 C  s          
@@ -771,7 +773,7 @@ task scf property
     59     -0.983223  5 O  s                 36      0.975052  3 C  py         
  
  Vector   29  Occ=0.000000D+00  E= 4.789847D-01
-              MO Center=  2.6D-01,  3.3D-01, -1.0D-16, r^2= 3.8D+00
+              MO Center=  2.6D-01,  3.3D-01,  2.1D-18, r^2= 3.8D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     17      2.858248  2 C  s                  3     -2.090889  1 C  s          
@@ -781,7 +783,7 @@ task scf property
     35      0.814942  3 C  px                63      0.614144  5 O  px         
  
  Vector   30  Occ=0.000000D+00  E= 5.702926D-01
-              MO Center=  1.4D+00,  5.8D-01,  2.3D-15, r^2= 2.8D+00
+              MO Center=  1.4D+00,  5.8D-01, -1.9D-15, r^2= 2.8D+00
    Bfn.  Coefficient  Atom+Function         Bfn.  Coefficient  Atom+Function  
   ----- ------------  ---------------      ----- ------------  ---------------
     17      2.332617  2 C  s                 31     -1.303425  3 C  s          
@@ -831,7 +833,7 @@ task scf property
      2   1 1 0      0.077897      0.000000      0.000002
      2   1 0 1      0.000000      0.000000      0.000000
      2   0 2 0    -24.880694      0.000000     90.697964
-     2   0 1 1      0.000000      0.000000      0.000000
+     2   0 1 1     -0.000000      0.000000      0.000000
      2   0 0 2    -22.450650      0.000000      0.000000
  
 
@@ -842,19 +844,20 @@ task scf property
           *** ORBITAL LOCALIZATION ***
           ****************************
 
-           IAO-IBO localization
-           --------------------
+                IAO-IBO localization
+                --------------------
+               (occ and virt orbitals)
 
- IBO loc: largest element in C(iao, trans) S C(iao) -1:      0.000000000000
+ IBO loc: largest element in C(iao,T) S C(iao) -1:          0.00000000
  Significant deviations from zero may indicate
  elevated numerical noise in the IAO generation
 
 
- IBO loc: largest element of C(MO, trans) C(MO) -1:      0.000000000000
+ IBO loc: largest element in C(MO,T) C(MO) -1:          0.00000000
  should be zero, for CMOs in the IAO basis
 
            iter   Max. delocal   Mean delocal    Converge
-           ----   ------------   ------------   ---------
+           ----   ------------   ------------    --------
               1   6.4744549430   3.2718681346    0.00D+00
               2   2.9502374379   1.7407413313    7.84D-01
               3   2.1535052593   1.5156908043    6.80D-01
@@ -863,10 +866,10 @@ task scf property
               6   2.1555070209   1.5083589529    8.63D-05
               7   2.1555070209   1.5083589538    6.36D-06
               8   2.1555070209   1.5083589538    9.22D-08
-              9   2.1555070209   1.5083589538    0.00D+00
+              9   2.1555070209   1.5083589538    3.73D-09
 
- IBO loc: largest element of C(MO, trans) C(MO) -1:      0.000000000000
- should be zero, for LMOs in the IAO basis
+ IBO loc: largest element in C(MO,T) C(MO) -1:          0.00000000
+ should be zero, for IBOs in the IAO basis
 
  IAO-IBO localized orbitals
 
@@ -890,9 +893,59 @@ task scf property
    18      -0.507592 2.000      4( 0.93)     3( 0.04)     5( 0.01)     2( 0.01)
    19      -0.424380 2.000      2( 0.53)     1( 0.43)     3( 0.04)
  
+
+ IBO localization (occ): IBOs will be stored
+ in file locorb.movecs, number 
+          1  to         19
+
+
+ IBO loc: largest element in C(iao,T) S C(iao) -1:          0.00000000
+ Significant deviations from zero may indicate
+ elevated numerical noise in the IAO generation
+
+ non-zero singular values:         10
+
+ IBO loc: largest element in C(MO,T) C(MO) -1:          0.00000000
+ should be zero, for CMOs in the IAO basis
+
+           iter   Max. delocal   Mean delocal    Converge
+           ----   ------------   ------------    --------
+              1   5.7963239473   4.2732750386    0.00D+00
+              2   2.4941110119   2.1820228090    7.23D-01
+              3   2.1364906389   1.9920797446    4.32D-01
+              4   2.1348072188   1.9904017968    1.39D-01
+              5   2.1348013497   1.9903805756    9.02D-03
+              6   2.1348017491   1.9903804736    7.84D-04
+              7   2.1348017624   1.9903804751    5.13D-05
+              8   2.1348017624   1.9903804751    3.41D-06
+              9   2.1348017624   1.9903804751    2.26D-07
+             10   2.1348017624   1.9903804751    1.34D-08
+             11   2.1348017624   1.9903804751    5.27D-09
+
+ IBO loc: largest element in C(MO,T) C(MO) -1:          0.00000000
+ should be zero, for IBOs in the IAO basis
+
+ IAO-IBO localized orbitals
+
+    1       0.209314 0.000      1( 0.55)     2( 0.44)
+    2       0.251712 0.000      3( 0.64)     4( 0.25)     5( 0.07)     2( 0.03)
+    3       0.441767 0.000      9( 0.69)     5( 0.30)
+    4       0.571110 0.000      7( 0.57)     1( 0.42)
+    5       0.576430 0.000      6( 0.58)     1( 0.41)
+    6       0.588168 0.000      8( 0.57)     2( 0.42)
+    7       0.612417 0.000      3( 0.61)     5( 0.34)     4( 0.04)
+    8       0.678788 0.000      2( 0.50)     3( 0.47)     4( 0.02)
+    9       0.845663 0.000      3( 0.59)     4( 0.40)
+   10       0.862369 0.000      1( 0.50)     2( 0.50)
+ 
+
+ IBO localization (vir): IBOs will be stored
+ in file locorb.movecs, number 
+         20  to         29
+
  Exiting Localization driver routine
 
- Task  times  cpu:        2.0s     wall:        2.1s
+ Task  times  cpu:        1.0s     wall:        1.2s
  
  
                                 NWChem Input Module
@@ -914,7 +967,7 @@ MA usage statistics:
 	current number of blocks	         0	         0
 	maximum number of blocks	        22	        14
 	current total bytes		         0	         0
-	maximum total bytes		    139392	  22511336
+	maximum total bytes		    139376	  22511288
 	maximum total K-bytes		       140	     22512
 	maximum total M-bytes		         1	        23
  
@@ -954,10 +1007,10 @@ MA usage statistics:
                                       -------
   E. Apra, E. J. Bylaska, N. Govind, K. Kowalski, M. Valiev, D. Mejia-Rodriguez,
        A. Kunitsa, N. P. Bauman, A. Panyala, W. A. de Jong, T. P. Straatsma,
-        H. J. J. van Dam, D. Wang, T. L. Windus, J. Hammond, J. Autschbach,
+   H. J. J. van Dam, D. Wang, T. L. Windus, J. Hammond, J. Autschbach, A. Woods,
     K. Bhaskaran-Nair, J. Brabec, K. Lopata, S. A. Fischer, S. Krishnamoorthy,
      M. Jacquelin, W. Ma, M. Klemm, O. Villa, Y. Chen, V. Anisimov, F. Aquino,
-    S. Hirata, M. T. Hackler, Eric Hermes, L. Jensen, J. E. Moore, J. C. Becca,
+     S. Hirata, M. T. Hackler, E. Hermes, L. Jensen, J. E. Moore, J. C. Becca,
       V. Konjkov, T. Risthaus, M. Malagoli, A. Marenich, A. Otero-de-la-Roza,
         J. Mullin, P. Nichols, R. Peverati, J. Pittner, Y. Zhao, P.-D. Fan,
         A. Fonari, M. J. Williamson, R. J. Harrison, J. R. Rehr, M. Dupuis,
@@ -972,4 +1025,4 @@ MA usage statistics:
    K. Glaesemann, G. Sandrone, M. Stave, H. Taylor, G. Thomas, J. H. van Lenthe,
                                A. T. Wong, Z. Zhang.
 
- Total times  cpu:        2.0s     wall:        2.1s
+ Total times  cpu:        1.0s     wall:        1.2s

QA/tests/o2_ccca/o2_ccca.nw

@@ -2,7 +2,7 @@ start o2_ccca
 
 title "O2, ccCA test"
 
-memory stack 1000 heap 100 global 4000 mb
+memory total 5000 mb global 4000 mb
 
 geometry units au
 O       0.0000000000   0.0000000000  -2.0000

QA/tests/pspw_stress/pspw_stress.nw

@@ -42,7 +42,7 @@ set nwpw:minimizer 2
 task pspw energy
 
 nwpw
-  tolerances 1.0d-9 1.0d-9
+  tolerances 1d-12 1d-12
 end
 driver
    clear

QA/tests/rt_tddft_cytosine_cam_cdfit/rt_tddft_cytosine_cam_cdfit.nw

@@ -0,0 +1,252 @@
+##
+## Real-time TDDFT QA
+## rt_tddft_cytosine_cam_cdfit
+## 
+## This will test range-separated XC and charge density fitting by
+## propagating from a converged ground state in the absence of an
+## external field.  Since we are at a variational minimum, the dipole
+## moment and energy should be time-independent and consistent with
+## the DFT SCF values.
+##
+## Suggested checks:
+## nw_rtparse -xdipole -px -c myrun.out cytosine.out
+## nw_rtparse -xdipole -py -c myrun.out cytosine.out
+## nw_rtparse -xdipole -pz -c myrun.out cytosine.out
+## nw_rtparse -xenergy -c myrun.out cytosine.out
+##
+
+title "Cytosine CD fitting + CAM-B3LYP"
+
+memory 1800 mb
+
+start cytosine
+echo
+
+geometry "system" units angstroms noautosym nocenter noautoz
+  C     -0.21392   1.48233  -2.56453
+  C      1.70616   2.50220  -1.71935
+  C      2.22400   1.26962  -1.26854
+  C      1.47919   0.17758  -1.49356
+  H     -0.16957  -0.54127  -2.42952
+  H      2.00878   4.39665  -1.75063
+  H      3.23491   3.58949  -1.32727
+  H      3.09143   1.21549  -0.77806
+  H      1.72181  -0.74069  -1.17471
+  N      0.28958   0.27064  -2.13203
+  N      0.52437   2.60665  -2.33531
+  N      2.40270   3.63507  -1.53819
+  O     -1.30179   1.51461  -3.16181
+end
+
+set geometry "system"
+
+
+##
+## H, C, N, O: cc-pvdz
+##
+basis spherical
+H    S
+     13.0100000              0.0196850        
+      1.9620000              0.1379770        
+      0.4446000              0.4781480        
+H    S
+      0.1220000              1.0000000        
+H    P
+      0.7270000              1.0000000        
+C    S
+   6665.0000000              0.0006920             -0.0001460        
+   1000.0000000              0.0053290             -0.0011540        
+    228.0000000              0.0270770             -0.0057250        
+     64.7100000              0.1017180             -0.0233120        
+     21.0600000              0.2747400             -0.0639550        
+      7.4950000              0.4485640             -0.1499810        
+      2.7970000              0.2850740             -0.1272620        
+      0.5215000              0.0152040              0.5445290        
+C    S
+      0.1596000              1.0000000        
+C    P
+      9.4390000              0.0381090        
+      2.0020000              0.2094800        
+      0.5456000              0.5085570        
+C    P
+      0.1517000              1.0000000        
+C    D
+      0.5500000              1.0000000        
+N    S
+   9046.0000000              0.0007000             -0.0001530        
+   1357.0000000              0.0053890             -0.0012080        
+    309.3000000              0.0274060             -0.0059920        
+     87.7300000              0.1032070             -0.0245440        
+     28.5600000              0.2787230             -0.0674590        
+     10.2100000              0.4485400             -0.1580780        
+      3.8380000              0.2782380             -0.1218310        
+      0.7466000              0.0154400              0.5490030        
+N    S
+      0.2248000              1.0000000        
+N    P
+     13.5500000              0.0399190        
+      2.9170000              0.2171690        
+      0.7973000              0.5103190        
+N    P
+      0.2185000              1.0000000        
+N    D
+      0.8170000              1.0000000        
+O    S
+  11720.0000000              0.0007100             -0.0001600        
+   1759.0000000              0.0054700             -0.0012630        
+    400.8000000              0.0278370             -0.0062670        
+    113.7000000              0.1048000             -0.0257160        
+     37.0300000              0.2830620             -0.0709240        
+     13.2700000              0.4487190             -0.1654110        
+      5.0250000              0.2709520             -0.1169550        
+      1.0130000              0.0154580              0.5573680        
+O    S
+      0.3023000              1.0000000        
+O    P
+     17.7000000              0.0430180        
+      3.8540000              0.2289130        
+      1.0460000              0.5087280        
+O    P
+      0.2753000              1.0000000        
+O    D
+      1.1850000              1.0000000        
+end
+
+
+# ##
+# ## H, C, N, O: ahlrichs coulomb fitting
+# ##
+basis "cd basis"
+H    S
+      9.30813000E+00         0.03446618       
+      2.30671800E+00         0.12253380       
+H    S
+      7.52012000E-01         1.0000000        
+H    S
+      2.73978000E-01         1.0000000        
+H    P
+      2.03270400E+00         1.0000000        
+H    P
+      7.90252000E-01         1.0000000        
+H    D
+      2.01954800E+00         1.0000000        
+C    S
+      5.91553927E+02         0.31582020       
+      1.72117940E+02         0.87503863       
+      5.47992590E+01         2.30760524       
+C    S
+      1.89590940E+01         1.0000000        
+C    S
+      7.05993000E+00         1.0000000        
+C    S
+      2.79484900E+00         1.0000000        
+C    S
+      1.15863400E+00         1.0000000        
+C    S
+      4.94324000E-01         1.0000000        
+C    S
+      2.12969000E-01         1.0000000        
+C    P
+      3.27847358E-01         1.0000000        
+C    P
+      7.86833659E-01         1.0000000        
+C    P
+      1.97101832E+00         1.0000000        
+C    D
+      4.01330100E+00         1.0000000        
+C    D
+      1.24750500E+00         1.0000000        
+C    D
+      4.08148000E-01         1.0000000        
+C    F
+      9.00000000E-01         1.0000000        
+N    S
+      7.91076935E+02         0.41567506       
+      2.29450184E+02         1.14750694       
+      7.28869600E+01         3.01935767       
+N    S
+      2.51815960E+01         1.0000000        
+N    S
+      9.37169700E+00         1.0000000        
+N    S
+      3.71065500E+00         1.0000000        
+N    S
+      1.53946300E+00         1.0000000        
+N    S
+      6.57553000E-01         1.0000000        
+N    S
+      2.83654000E-01         1.0000000        
+N    P
+      4.70739194E-01         1.0000000        
+N    P
+      1.12977407E+00         1.0000000        
+N    P
+      2.83008403E+00         1.0000000        
+N    D
+      5.83298650E+00         1.0000000        
+N    D
+      1.73268650E+00         1.0000000        
+N    D
+      5.45242500E-01         1.0000000        
+N    F
+      1.82648000E+00         1.0000000        
+O    S
+      9.57843253E+02         0.56249624       
+      2.81967425E+02         1.49108985       
+      9.01998320E+01         3.86547733       
+O    S
+      3.11382990E+01         1.0000000        
+O    S
+      1.14937320E+01         1.0000000        
+O    S
+      4.48404900E+00         1.0000000        
+O    S
+      1.82350400E+00         1.0000000        
+O    S
+      7.60903000E-01         1.0000000        
+O    S
+      3.20292000E-01         1.0000000        
+O    P
+      6.14708863E-01         1.0000000        
+O    P
+      1.47530127E+00         1.0000000        
+O    P
+      3.69562968E+00         1.0000000        
+O    D
+      7.65267200E+00         1.0000000        
+O    D
+      2.21786800E+00         1.0000000        
+O    D
+      6.82337000E-01         1.0000000        
+O    F
+      2.19178082E+00         1.0000000        
+end
+
+
+##
+## CAM-B3LYP with tight convergence
+##
+## (no need for "direct" since coulomb part done with charge density fitting)
+##
+dft
+  xc xcamb88 1.00 lyp 0.81 vwn_5 0.19 hfexch 1.00
+  cam 0.33 cam_alpha 0.19 cam_beta 0.46
+  convergence density 1d-9
+  tolerances tight acccoul 10
+end
+task dft energy
+
+
+##
+## No excitation, just propagate a few steps with full checking and profiling
+##
+rt_tddft
+  tmax 1.0
+  dt 0.2
+
+  nchecks *
+  checklvl 3
+  nrestarts 0
+  prof
+end
+task dft rt_tddft

QA/tests/rt_tddft_tcne_spinkick/rt_tddft_tcne_spinkick.nw

@@ -0,0 +1,115 @@
+##
+## Real-time TDDFT QA
+## rt_tddft_tcne_spinkick
+## 
+## This tests open shell propagation and spin-dependent excitation
+## rules.  Here we excite the alpha spin with a positive kick, and the
+## beta spin with a negative kick.
+##
+## Suggested checks:
+## nw_rtparse -xdipole -salpha -pz -c myrun.out tcne.out
+## nw_rtparse -xdipole -sbeta -pz -c myrun.out tcne.out
+## nw_rtparse -xenergy -c myrun.out tcne.out
+## nw_rtparse -xS2 -c myrun.out tcne.out
+##
+
+title "Tetracyanoethylene anion"
+
+echo
+
+start tcne
+echo
+
+
+##
+## optimized with cc-pvdz/B3LYP
+##
+geometry "system" units angstroms noautosym nocenter noautoz
+ C    -1.77576486     0.66496556     0.00004199
+ N    -2.94676621     0.71379797     0.00004388
+ C    -0.36046718     0.62491168     0.00003506
+ C     0.36049301    -0.62492429    -0.00004895
+ C     1.77579907    -0.66504145    -0.00006082
+ N     2.94680364    -0.71382258    -0.00006592
+ C    -0.31262746    -1.87038951    -0.00011201
+ N    -0.85519492    -2.90926164    -0.00016331
+ C     0.31276207     1.87031662     0.00010870
+ N     0.85498782     2.90938919     0.00016857
+end
+
+set geometry "system"
+
+
+##
+## C, N: 6-31G
+##
+basis
+C    S
+   3047.5249000              0.0018347        
+    457.3695100              0.0140373        
+    103.9486900              0.0688426        
+     29.2101550              0.2321844        
+      9.2866630              0.4679413        
+      3.1639270              0.3623120        
+C    SP
+      7.8682724             -0.1193324              0.0689991        
+      1.8812885             -0.1608542              0.3164240        
+      0.5442493              1.1434564              0.7443083        
+C    SP
+      0.1687144              1.0000000              1.0000000        
+N    S
+   4173.5110000              0.0018348        
+    627.4579000              0.0139950        
+    142.9021000              0.0685870        
+     40.2343300              0.2322410        
+     12.8202100              0.4690700        
+      4.3904370              0.3604550        
+N    SP
+     11.6263580             -0.1149610              0.0675800        
+      2.7162800             -0.1691180              0.3239070        
+      0.7722180              1.1458520              0.7408950        
+N    SP
+      0.2120313              1.0000000              1.0000000        
+end
+
+charge -1
+
+dft
+  mult 2
+  odft
+  xc b3lyp
+end
+task dft energy
+
+
+##
+## Excite alpha and beta spins with different kick excitations.  Also
+## do full checking.
+##
+rt_tddft
+  tmax 5.0
+  dt 0.2
+
+  nrestarts 0
+  nchecks *
+  checklvl 3
+
+  field "ka"
+    type delta
+    polarization z
+    max 0.001
+    spin alpha
+  end
+
+  field "kb"
+    type delta
+    polarization z
+    max -0.001
+    spin beta
+  end
+
+  excite "system" with "ka"
+  excite "system" with "kb"
+end
+task dft rt_tddft
+

QA/tests/scf_dftguess/scf_dftguess.nw

@@ -0,0 +1,56 @@
+#
+# $Id$
+#
+
+title "3 Water molecules scf with DFT hfexch"
+
+start h2o
+
+geometry autosym
+       O          -0.167787    1.645761    0.108747
+       H           0.613411    1.102620    0.113724
+       H          -0.093821    2.209720   -0.643619
+       O           1.517569   -0.667424   -0.080674
+       H           1.989645   -1.098799    0.612047
+       H           0.668397   -1.091798   -0.139744
+       O          -1.350388   -0.964879   -0.092208
+       H          -1.908991   -1.211298    0.626207
+       H          -1.263787   -0.018107   -0.055536 
+end
+
+
+basis
+  H library aug-cc-pvdz
+  O library aug-cc-pvtz
+end
+set dftguess t
+scf
+#uhf
+#triplet
+end
+task scf
+eof
+driver
+  clear
+end
+set movecs:tanalyze 0d0
+task dft
+eof
+dft
+xc hfexch
+end
+scf
+print 'final vector analysis' "final vector analysis"
+end
+#task dft
+#eof
+set forces:fakezero t
+task scf optimize
+eof
+task dft gradient
+eof
+
+scf; print low; end
+task scf optimize
+
+

QA/tests/tddftgrad_h2o_cis_lda/tddftgrad_h2o_cis_lda.nw

@@ -0,0 +1,30 @@
+echo
+title h2o
+start h2o
+geometry units bohr 
+symmetry c1
+    H -1.57866004 0.0  0.40931277
+    O  0.00000000 0.0 -0.81862553
+    H  1.57866004 0.0  0.40931277
+end  
+basis "ao basis"   
+  * library 6-31G   
+end
+dft 
+  xc slater 1.0 pw91lda 1.0 
+#  xc slater 1.0 pw91lda 1.0 hfexch 1d-8
+end 
+tddft
+ cis
+ nroots 10
+ algorithm 1
+ notriplet
+ civecs
+ grad
+   root 1
+ end
+end
+#set tddft:tdaloc .false.
+task tddft gradient
+task tddft gradient numerical
+

README.md

@@ -1,5 +1,5 @@
 <!---[![Build Status](https://img.shields.io/endpoint.svg?url=https%3A%2F%2Factions-badge.atrox.dev%2Fnwchemgit%2Fnwchem%2Fbadge%3Fref%3Dmaster&style=flat)](https://actions-badge.atrox.dev/nwchemgit/nwchem/goto?ref=master)--->
-[![NWChem_CI](https://img.shields.io/github/workflow/status/nwchemgit/nwchem/NWChem_CI)](https://github.com/nwchemgit/nwchem/actions)   
+[![NWChem_CI](https://img.shields.io/github/actions/workflow/status/nwchemgit/nwchem/github_actions.yml?style=plastic)](https://github.com/nwchemgit/nwchem/actions)   
 <!---[![Build Travis-CI Status](https://img.shields.io/travis/nwchemgit/nwchem.svg)](https://travis-ci.org/nwchemgit/nwchem/builds)--->
 [![License](https://img.shields.io/badge/license-ECL2-blue.svg)](https://raw.githubusercontent.com/nwchemgit/nwchem/master/LICENSE.md)  
 [![Github Downloads All Releases](https://img.shields.io/github/downloads/nwchemgit/nwchem/total.svg)](https://github.com/nwchemgit/nwchem/releases)   

contrib/git.nwchem/github_actions_selfhosted.txt

@@ -141,8 +141,11 @@ jobs:
           export TRAVIS_BUILD_DIR=`pwd`
           unset USE_SIMINT
           mytee=tee
-          if  [[ `uname -s` == "Darwin" ]]; then mytee=gtee; fi
-          source travis/nwchem.bashrc; cd QA && ./doqmtests.mpi 3 | $mytee >(ts "%d%b%Y %H:%M:%S" > ../../../../doqmtests.mpi.log.` date +%b%d%Y_%H:%M`)
+          if  [[ `uname -s` == "Darwin" ]]; then mytee=gtee; dist=Darwin; fi
+          if  test -f "/usr/lib/os-release"; then dist=$(grep ID= /etc/os-release |grep -v VERSION|head -1 |cut -c4-| sed 's/\"//g') ; fi
+          if [[ $dist == "fedora" ]]; then nprocs=2; else nprocs=3 ;fi
+          echo nprocs is $nprocs
+          source travis/nwchem.bashrc; cd QA && ./doqmtests.mpi $nprocs | $mytee >(ts "%d%b%Y %H:%M:%S" > ../../../../doqmtests.mpi.log.` date +%b%d%Y_%H:%M`)
         shell: bash
       - name: Check if QA testing has failed
         if: ${{ failure() }} &&  steps.qa_test.outcome == 'failure'

release.notes.7.2.0

@@ -0,0 +1,51 @@
+NWChem Version 7.2.0 Release Notes
+
+NWChem is now available on Github at
+https://github.com/nwchemgit/nwchem
+
+Documentation available from the NWChem website at
+https://nwchemgit.github.io
+
+NWChem 7.2.0 is released as open-source under the ECL 2.0 license.
+
+NWChem 7.2.0 will be released with the latest Global Arrays Toolkit (v5.8.2).
+
+NEW FUNCTIONALITY 
+
+Modules:  
+ Molecular GW with Gaussian basis 
+
+Interfaces:  
+ * Plumed
+ * Libxc  
+ * tblite   
+
+Compilation:
+Added the enviroment variable USE_HWOPT: when USE_HWOPT=n all hardware (non-dynamic) optimizations are not used (e.g. gcc -march=native)
+
+Solvation module:
+* COSMO updates   
+
+Gaussian DFT module:
+* New functionality: exact two-component relativistic Hamiltonian (X2C)
+* New functionality: Auxiliary Density Functional Theory (ADFT) XC method  
+* New functionality: Resolution of Identity Time-Dependft DFT (RITDFT)
+* New DFT XC functionals
+ - r2SCAN0
+ - r2scan-D3 and rscan-D3
+ - n12-sx
+ - mn15 and mn15-l
+ - mn12-sx and mm12-l
+ - revm11
+ - wb97, wb97-d3 and wb97x
+ - r2SCAN, r2SCAN-L
+
+Basis set 
+* added autoaux fitting basis sets to libraries.bse
+* new "bse" input option
+
+
+   
+BUG FIXES/ENHANCEMENTS
+
+TBD

src/NWints/simint/libsimint_source/build_simint.sh

@@ -238,6 +238,7 @@ cd ../..
 echo ln -sf  simint.l${SIMINT_MAXAM}_p${PERMUTE_SLOW}_d${DERIV}.install simint_install
 ln -sf  simint.l${SIMINT_MAXAM}_p${PERMUTE_SLOW}_d${DERIV}.install simint_install
 cd simint_install/lib
+strip --strip-debug libsimint.a
 ln -sf libsimint.a libnwc_simint.a
 export SIMINT_HOME=${SRC_HOME}/simint.l${SIMINT_MAXAM}_p${PERMUTE_SLOW}_d${DERIV}.install
 echo 'SIMINT library built with maximum angular momentum='${SIMINT_MAXAM}

src/basis/bas_input.F

@@ -1135,11 +1135,13 @@ c     Now read the basis for this magical atom type in using
 c     the given tag
 c
  10   if (.not. inp_a(field)) field=' '
+#ifdef IGNORE_HEH
 c
 c     ignore spherical vs cartesian from bse for He/H
 c     since they do assign spherical when the other elements are cartesian
 c
       if (.not.inp_compare(.false., 'h', string(1:1))) then
+#endif
       if(read_sphere) then
          if (inp_compare(.false., 'spherical', field)) then
             if (.not. bas_set_spherical(basis, .true.))
@@ -1147,7 +1149,9 @@ c
      &           (' bas_set_spherical failed ',911, BASIS_ERR)
          endif
       endif
+#ifdef IGNORE_HEH
       endif
+#endif
       if (.not. inp_read()) call errquit
      $     ('bas_tag_lib: premature EOF', 0, INPUT_ERR)
 c     

src/config/makefile.h

@@ -19,7 +19,7 @@
 # For development tree 
 #RELEASE := 
 # For current release tree
-RELEASE := 7.0.0
+RELEASE := 7.2.0
 
 #
 ifndef NWCHEM_TOP
@@ -243,10 +243,10 @@ endif
 # specified below.  Use of MPI requires substituting the tcgmsg-mpi
 # wrapper for the normal tcgmsg library.
 # the 2 following environmental variables are need for linking
-# LIBMPI - represents the name of mpi library (with -l)
-# MPI_LIB - represents the path to the mpi library
-#LIBMPI =  -lmpich
-#MPI_LIB= /usr/local/lib
+# NWLIBMPI - represents the name of mpi library (with -l)
+# NWMPI_LIB - represents the path to the mpi library
+#NWLIBMPI =  -lmpich
+#NWMPI_LIB= /usr/local/lib
 
 #JN: under the new structure, tools should be listed first as
 # their header files are needed for dependency analysis of
@@ -292,9 +292,6 @@ ifdef BUILD_SCALAPACK
         $(error )
     endif
 
-    ifndef SCALAPACK_SIZE
-        SCALAPACK_SIZE=8
-    endif
     SCALAPACK=-L$(NWCHEM_TOP)/src/libext/lib -lnwc_scalapack
 endif
 
@@ -317,15 +314,6 @@ ifdef BUILD_ELPA
 endif
 
 
-ifdef BUILD_MPICH
-    NW_CORE_SUBDIRS += libext
-    PATH := $(NWCHEM_TOP)/src/libext/bin:$(PATH)
-    MPI_INCLUDE = $(shell PATH=$(NWCHEM_TOP)/src/libext/bin:$(PATH) $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_include)
-    MPI_LIB     = $(shell PATH=$(NWCHEM_TOP)/src/libext/bin:$(PATH)  $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_lib)
-    LIBMPI      = $(shell PATH=$(NWCHEM_TOP)/src/libext/bin:$(PATH) $(NWCHEM_TOP)/src/tools/guess-mpidefs --libmpi)
-    LIBMPI	+= $(shell /usr/local/bin/pkg-config --libs-only-L hwloc 2> /dev/null)
-endif
-
 
 ifndef EXTERNAL_GA_PATH
     NW_CORE_SUBDIRS += tools
@@ -1105,7 +1093,7 @@ ifeq ($(TARGET),MACX)
     ifeq ($(FC),g77)
         #g77, only decent one form Fink http://fink.sf.net
         #gcc version 3.4 20031015 (experimental)
-        _G77V33= $(shell g77 -v  2>&1|egrep spec|head -n 1|awk ' /3.3/  {print "Y"}')
+        _G77V33= $(shell g77 -v  2>&1|grep spec|head -n 1|awk ' /3.3/  {print "Y"}')
         FDEBUG= -O1 -g
         FOPTIONS   = -fno-second-underscore -fno-globals -Wno-globals
         FOPTIMIZE  = -O3 -fno-inline-functions -funroll-loops
@@ -1142,7 +1130,7 @@ ifeq ($(TARGET),MACX)
         GNUMAJOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __GNUC__ |cut -c18-)
 
         ifdef GNUMAJOR
-            GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | egrep __GNUC_MINOR | cut -c24)
+            GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __GNUC_MINOR | cut -c24)
             GNU_GE_4_6 = $(shell [ $(GNUMAJOR) -gt 4 ] || [ $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 6 ] && echo true)
             GNU_GE_4_8 = $(shell [ $(GNUMAJOR) -gt 4 ] || [ $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 8 ] && echo true)
             GNU_GE_6 = $(shell [ $(GNUMAJOR) -ge 6  ] && echo true)
@@ -1260,8 +1248,8 @@ ifeq ($(TARGET),MACX)
     ifeq ($(FC),xlf) 
         LDOPTIONS = -Wl,-multiply_defined -Wl,warning
     else
-#       _GCC4= $(shell gcc -v  2>&1|egrep spec|head -n 1|awk ' / 3./  {print "N";exit}; / 2./ {print "N";exit};{print "Y"}')
-        _GCC4= $(shell $(CC) -dM -E - < /dev/null | egrep __VERS | cut -c22|awk ' /3/  {print "N";exit}; /2/ {print "N";exit};{print "Y"}')
+#       _GCC4= $(shell gcc -v  2>&1|grep spec|head -n 1|awk ' / 3./  {print "N";exit}; / 2./ {print "N";exit};{print "Y"}')
+        _GCC4= $(shell $(CC) -dM -E - < /dev/null | grep __VERS | cut -c22|awk ' /3/  {print "N";exit}; /2/ {print "N";exit};{print "Y"}')
         ifeq ($(_GCC4),Y) 
 #           EXTRA_LIBS += 
         else
@@ -1353,7 +1341,7 @@ ifeq ($(TARGET),MACX64)
         GNUMAJOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __GNUC__ |cut -c18-)
 
         ifneq ($(strip $(GNUMAJOR)),)
-            GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | egrep __GNUC_MINOR | cut -c24)
+            GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __GNUC_MINOR | cut -c24)
             GNU_GE_4_6 = $(shell [ $(GNUMAJOR) -gt 4 -o \( $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 6 \) ] && echo true)
             GNU_GE_4_8 = $(shell [ $(GNUMAJOR) -gt 4 -o \( $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 8 \) ] && echo true)
             GNU_GE_6 = $(shell [ $(GNUMAJOR) -ge 6  ] && echo true)
@@ -1427,11 +1415,11 @@ ifeq ($(TARGET),MACX64)
 
 
     ifeq ($(FC),ifort)
-        _IFCV11= $(shell ifort -logo  2>&1|egrep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 11) {print "Y";exit}}')
-        _IFCV12= $(shell ifort -logo  2>&1|egrep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 12) {print "Y";exit}}')
-        _IFCV14= $(shell ifort -logo  2>&1|egrep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 14) {print "Y";exit}}')
-        _IFCV15ORNEWER=$(shell ifort -logo  2>&1|egrep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 15) {print "Y";exit}}')
-        _IFCV17=$(shell ifort -logo  2>&1|egrep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 17) {print "Y";exit}}')
+        _IFCV11= $(shell ifort -logo  2>&1|grep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 11) {print "Y";exit}}')
+        _IFCV12= $(shell ifort -logo  2>&1|grep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 12) {print "Y";exit}}')
+        _IFCV14= $(shell ifort -logo  2>&1|grep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 14) {print "Y";exit}}')
+        _IFCV15ORNEWER=$(shell ifort -logo  2>&1|grep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 15) {print "Y";exit}}')
+        _IFCV17=$(shell ifort -logo  2>&1|grep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 17) {print "Y";exit}}')
         DEFINES  += -DIFCV8 -DIFCLINUX
 
         ifdef USE_I4FLAGS
@@ -1594,7 +1582,7 @@ ifeq ($(TARGET),$(findstring $(TARGET),LINUX CYGNUS CYGWIN))
         GNUMAJOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __GNUC__ |cut -c18-)
 
         ifdef GNUMAJOR
-            GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | egrep __VERS | cut -c24)
+            GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __VERS | cut -c24)
             GNU_GE_4_6 = $(shell [ $(GNUMAJOR) -gt 4 ] || [ $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 6 ] && echo true)
             GNU_GE_4_8 = $(shell [ $(GNUMAJOR) -gt 4 ] || [ $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 8 ] && echo true)
             GNU_GE_6 = $(shell [ $(GNUMAJOR) -ge 6  ] && echo true)
@@ -1643,7 +1631,7 @@ ifeq ($(TARGET),$(findstring $(TARGET),LINUX CYGNUS CYGWIN))
         _CPU = $(shell uname -m  )
 
         ifeq ($(FC),g77)
-            _G77V33= $(shell g77 -v  2>&1|egrep spec|head -n 1|awk ' /3.3/  {print "Y"}')
+            _G77V33= $(shell g77 -v  2>&1|grep spec|head -n 1|awk ' /3.3/  {print "Y"}')
             FOPTIONS   += -fno-second-underscore   
             FOPTIONS   += -fno-f90  -ffixed-line-length-72 -ffixed-form
             FOPTIMIZE  +=  -O2  -malign-double -finline-functions 
@@ -1669,7 +1657,7 @@ ifeq ($(TARGET),$(findstring $(TARGET),LINUX CYGNUS CYGWIN))
             _CPU=i786
         else
             ifeq ($(_CPU),i686)
-                _GOTSSE2= $(shell cat /proc/cpuinfo | egrep sse2 | tail -n 1 | awk ' /sse2/  {print "Y"}')
+                _GOTSSE2= $(shell cat /proc/cpuinfo | grep sse2 | tail -n 1 | awk ' /sse2/  {print "Y"}')
                 ifeq ($(_GOTSSE2),Y) 
                     _CPU=i786
                 endif
@@ -1750,7 +1738,7 @@ ifeq ($(TARGET),$(findstring $(TARGET),LINUX CYGNUS CYGWIN))
             ifdef  USE_DEBUG
                 FOPTIONS += -g
             endif
-            _IFCV7= $(shell ifort -v  2>&1|egrep "Version "|head -n 1|awk '/7./ {print "Y"; exit}')
+            _IFCV7= $(shell ifort -v  2>&1|grep "Version "|head -n 1|awk '/7./ {print "Y"; exit}')
             ifneq ($(_IFCV7),Y)
                 DEFINES+= -DIFCV8
                 ifeq ($(FC),ifc)
@@ -2212,7 +2200,7 @@ ifneq ($(TARGET),LINUX)
             else
                 GNUMAJOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __GNUC__ |cut -c18-)
                 ifdef GNUMAJOR
-                    GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | egrep __GNUC_MINOR | cut -c24)
+                    GNUMINOR=$(shell $(FC) -dM -E - < /dev/null 2> /dev/null | grep __GNUC_MINOR | cut -c24)
                     GNU_GE_4_6 = $(shell [ $(GNUMAJOR) -gt 4 ] || [ $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 6 ] && echo true)
                     GNU_GE_4_8 = $(shell [ $(GNUMAJOR) -gt 4 ] || [ $(GNUMAJOR) -eq 4 -a $(GNUMINOR) -ge 8 ] && echo true)
                 endif
@@ -2304,9 +2292,9 @@ ifneq ($(TARGET),LINUX)
             endif
 
             ifeq ($(FC),ifort)
-                _IFCV9= $(shell ifort -v  2>&1|egrep "Version "|head -n 1|awk '/9./ {print "Y"}; /10./ {print "Y"; exit}')
-                _IFCV81= $(shell ifort -v  2>&1|egrep "Version "|head -n 1|awk ' /8.1/  {print "Y";exit}; /9./ {print "Y"; exit}; /10./ {print "Y"; exit}')
-                _IFCV8= $(shell ifort -v  2>&1|egrep "Version "|head -n 1|awk ' /8./  {print "Y";exit}; /9./ {print "Y"; exit}; /10./ {print "Y"; exit}')
+                _IFCV9= $(shell ifort -v  2>&1|grep "Version "|head -n 1|awk '/9./ {print "Y"}; /10./ {print "Y"; exit}')
+                _IFCV81= $(shell ifort -v  2>&1|grep "Version "|head -n 1|awk ' /8.1/  {print "Y";exit}; /9./ {print "Y"; exit}; /10./ {print "Y"; exit}')
+                _IFCV8= $(shell ifort -v  2>&1|grep "Version "|head -n 1|awk ' /8./  {print "Y";exit}; /9./ {print "Y"; exit}; /10./ {print "Y"; exit}')
 
                 ifeq ($(_IFCV8),Y)
                     DEFINES+= -DIFCV8
@@ -2315,7 +2303,7 @@ ifneq ($(TARGET),LINUX)
                 ifeq ($(_IFCV81),Y)
                     DEFINES+= -DIFCV81
                 endif
-                ITANIUMNO = $(shell   cat /proc/cpuinfo | egrep family | head -n 1  2>&1 | awk ' /Itanium 2/ { print "-tpp2"; exit };/Itanium/ { print "-tpp1"}')
+                ITANIUMNO = $(shell   cat /proc/cpuinfo | grep family | head -n 1  2>&1 | awk ' /Itanium 2/ { print "-tpp2"; exit };/Itanium/ { print "-tpp1"}')
                 FOPTIONS   += -auto -w -ftz $(ITANIUMNO)
                 ifdef  USE_GPROF
                     FOPTIONS += -qp
@@ -2410,20 +2398,20 @@ ifneq ($(TARGET),LINUX)
             # support for traditional Intel(R) Fortran compiler
             ifeq ($(_FC),ifort)
                 ifeq ($(shell $(CNFDIR)/check_env.sh $(USE_HWOPT)),1)
-                    _GOTSSE3= $(shell cat /proc/cpuinfo | egrep sse3 | tail -n 1 | awk ' /sse3/  {print "Y"}')
-                    _GOTSSE42= $(shell cat /proc/cpuinfo | egrep sse4_2 | tail -n 1 | awk ' /sse4_2/  {print "Y"}')
-                    _GOTAVX= $(shell cat /proc/cpuinfo | egrep avx | tail -n 1 | awk ' /avx/  {print "Y"}')
-                    _GOTAVX2= $(shell cat /proc/cpuinfo | egrep fma | tail -n 1 | awk ' /fma/  {print "Y"}')
-                    _GOTAVX512F= $(shell cat /proc/cpuinfo | egrep avx512f | tail -n 1 | awk ' /avx512f/  {print "Y"}')
+                    _GOTSSE3= $(shell cat /proc/cpuinfo | grep sse3 | tail -n 1 | awk ' /sse3/  {print "Y"}')
+                    _GOTSSE42= $(shell cat /proc/cpuinfo | grep sse4_2 | tail -n 1 | awk ' /sse4_2/  {print "Y"}')
+                    _GOTAVX= $(shell cat /proc/cpuinfo | grep avx | tail -n 1 | awk ' /avx/  {print "Y"}')
+                    _GOTAVX2= $(shell cat /proc/cpuinfo | grep fma | tail -n 1 | awk ' /fma/  {print "Y"}')
+                    _GOTAVX512F= $(shell cat /proc/cpuinfo | grep avx512f | tail -n 1 | awk ' /avx512f/  {print "Y"}')
                 endif
                 _IFCE = $(shell ifort -V  2>&1 |head -1 |awk ' /64/ {print "Y";exit};')
-                _IFCV7= $(shell ifort -v  2>&1|egrep "Version "|head -n 1|awk ' /7./  {print "Y";exit}')
-                _IFCV11= $(shell ifort -logo  2>&1|egrep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 11) {print "Y";exit}}')
-                _IFCV12= $(shell ifort -logo  2>&1|egrep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 12) {print "Y";exit}}')
-                _IFCV14= $(shell ifort -logo  2>&1|egrep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 14) {print "Y";exit}}')
-                _IFCV15ORNEWER=$(shell ifort -logo  2>&1|egrep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 15) {print "Y";exit}}')
-                _IFCV17=$(shell ifort -logo  2>&1|egrep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 17) {print "Y";exit}}')
-                _IFCV18=$(shell ifort -logo  2>&1|egrep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 18) {print "Y";exit}}')
+                _IFCV7= $(shell ifort -v  2>&1|grep "Version "|head -n 1|awk ' /7./  {print "Y";exit}')
+                _IFCV11= $(shell ifort -logo  2>&1|grep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 11) {print "Y";exit}}')
+                _IFCV12= $(shell ifort -logo  2>&1|grep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 12) {print "Y";exit}}')
+                _IFCV14= $(shell ifort -logo  2>&1|grep "Version "|head -n 1|sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 14) {print "Y";exit}}')
+                _IFCV15ORNEWER=$(shell ifort -logo  2>&1|grep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 15) {print "Y";exit}}')
+                _IFCV17=$(shell ifort -logo  2>&1|grep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 17) {print "Y";exit}}')
+                _IFCV18=$(shell ifort -logo  2>&1|grep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 18) {print "Y";exit}}')
 
 #               Intel EM64T is required
                 ifneq ($(_IFCE),Y)
@@ -2456,12 +2444,12 @@ ifneq ($(TARGET),LINUX)
 #
                 ifeq ($(_IFCV15ORNEWER), Y)
 		  IFORTVER=$(shell ifort -v 2>&1|cut -d " " -f 3)
-                ifeq ($(IFORTVER),2021.7.0)
-                   $(info     )
-                   $(info     ifort 2021.7.0 not validated)
-                   $(info     )
-                   $(error )
-                endif
+#                ifeq ($(IFORTVER),2021.7.0)
+#                   $(info     )
+#                   $(info     ifort 2021.7.0 not validated)
+#                   $(info     )
+#                   $(error )
+#                endif
 #                   fpp seems to get lost with ifort 15 in the offload bit
 #                   only use EXPLICITF for offload because otherwise we want debugging to be easy
 #                   FOPTIONS +=  -Qoption,fpp,-P -Qoption,fpp,-c_com=no  -allow nofpp_comments 
@@ -2619,7 +2607,7 @@ ifneq ($(TARGET),LINUX)
             endif
 
             ifeq ($(_CC),icc)
-                ICCV15ORNEWER=$(shell icc -V  2>&1|egrep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 15) {print "Y";exit}}')
+                ICCV15ORNEWER=$(shell icc -V  2>&1|grep "Version "|head -n 1 | sed 's/.*Version \([0-9][0-9]\).*/\1/' | awk '{if ($$1 >= 15) {print "Y";exit}}')
                 ifdef USE_KNL
                     COPTIONS   +=   -xMIC-AVX512 -ftz
                     DEFINES+= -DINTEL_64ALIGN
@@ -3317,7 +3305,7 @@ ifeq ($(BUILDING_PYTHON),python)
 
     ifndef GOT_PYTHONCONFIG	  
         PYMAJOR:=$(word 1, $(subst ., ,$(PYTHONVERSION)))
-        errorpythonconfig:$
+        errorpythonconfig:
 	    $(info )
 	    $(info python-config not found in your PATH)
 	    $(info Please install the packages)
@@ -3455,14 +3443,6 @@ else
     CORE_LIBS += $(BLASOPT)
 endif
 
-ifdef NWCHEM_LINK_CUDA
-    ifeq ($(_FC),pgf90)
-       CORE_LIBS += -acc -cuda -cudalib=cublas
-    endif
-    ifeq ($(_FC),gfortran)
-       CORE_LIBS +=  -fopenacc -lcublas
-    endif
-endif
 
 
 ifdef BLASOPT
@@ -3532,39 +3512,97 @@ endif
 ifdef USE_MPI 
     #ifeq ($(FC),$(findstring $(FC),mpifrt mpfort mpif77 mpxlf mpif90 ftn scorep-ftn))
     ifeq ($(FC),$(findstring $(FC), ftn scorep-ftn))
-        LIBMPI =
-        MPI_INCLUDE =
-        MPI_LIB =
-    else
+        NWLIBMPI =
+        NWMPI_INCLUDE =
+        NWMPI_LIB =
+    else ifdef BUILD_MPICH
+        NW_CORE_SUBDIRS += libext
+        PATH := $(NWCHEM_TOP)/src/libext/bin:$(PATH)
+        NWMPI_INCLUDE = $(shell PATH=$(NWCHEM_TOP)/src/libext/bin:$(PATH) $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_include)
+        NWMPI_LIB     = $(shell PATH=$(NWCHEM_TOP)/src/libext/bin:$(PATH)  $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_lib)
+        NWLIBMPI      = $(shell PATH=$(NWCHEM_TOP)/src/libext/bin:$(PATH) $(NWCHEM_TOP)/src/tools/guess-mpidefs --libmpi)
+        NWLIBMPI	+= $(shell pkg-config --libs-only-L hwloc 2> /dev/null)
+        ifeq ($(NWCHEM_TARGET),MACX64)
+           GOT_BREW = $(shell command -v brew 2> /dev/null)
+           ifdef GOT_BREW
+              NWLIBMPI	+= -L$(shell brew --prefix)/lib
+           endif
+	endif
+    else ifdef FORCE_MPI_ENV
         ifndef MPI_INCLUDE
-            # check if mpif90 is present
-            MPIF90YN = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_include)
-            ifeq ($(MPIF90YN),mpif90notfound)
-                errormpif90:
-	            $(info )
-	            $(info mpif90 not found. Please add its location to PATH)
-	            $(info e.g. export PATH=/usr/local/bin:/usr/lib64/openmpi/bin:...)
-	            $(info )
-            endif
-            MPI_INCLUDE = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_include)
+            errormpi1:
+                $(info )
+                $(info FORCE_MPI_ENV set but MPI_INCLUDE not set)
+                $(info )
+                $(error )
+        else
+            NWMPI_INCLUDE = $(MPI_INCLUDE)
         endif
         ifndef MPI_LIB
-            MPI_LIB     = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_lib)
-        endif
+            errormpi2:
+                $(info )
+                $(info FORCE_MPI_ENV set but MPI_LIB not set)
+                $(info )
+                $(error )
+        else
+           NWMPI_LIB = $(MPI_LIB)
+	endif
         ifndef LIBMPI
-            LIBMPI      = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --libmpi)
+            errormpi3:
+                $(info )
+                $(info FORCE_MPI_ENV set but LIBMPI not set)
+                $(info )
+                $(error )
+        else
+           NWLIBMPI = $(LIBMPI)
+	endif
+    else
+        ifeq ($(shell pwd), $(NWCHEM_TOP)/src)
+	    ifndef FORCE_MPI_ENV
+                ifdef LIBMPI
+                    $(info ***warning LIBMPI ignored since FORCE_MPI_ENV not set***)
+                endif
+                ifdef MPI_LIB
+                    $(info ***warning MPI_LIB ignored since FORCE_MPI_ENV not set***)
+ 	        endif
+                ifdef MPI_INCLUDE
+                    $(info ***warning MPI_INCLUDE ignored since FORCE_MPI_ENV not set***)
+ 	        endif
+	    endif
+	endif
+        # check if mpif90 is present
+        MPIF90YN = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_include)
+        ifeq ($(MPIF90YN),mpif90notfound)
+            errormpif90:
+                $(info )
+                $(info mpif90 not found. Please add its location to PATH)
+                $(info e.g. export PATH=/usr/local/bin:/usr/lib64/openmpi/bin:...)
+                $(info )
+                $(error )
         endif
+        NWMPI_INCLUDE = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_include)
+        NWMPI_LIB     = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --mpi_lib)
+        NWLIBMPI      = $(shell $(NWCHEM_TOP)/src/tools/guess-mpidefs --libmpi)
     endif
 
-    ifdef MPI_LIB 
-        CORE_LIBS += $(patsubst -L-L%,-L%,-L$(MPI_LIB))
+    ifdef NWMPI_INCLUDE
+      LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(NWMPI_INCLUDE))
+    endif
+    ifdef NWMPI_LIB 
+        CORE_LIBS += $(patsubst -L-L%,-L%,-L$(NWMPI_LIB))
     endif 
     ifdef OLD_GA
-        CORE_LIBS += -ltcgmsg-mpi $(LIBMPI) 
+        CORE_LIBS += -ltcgmsg-mpi $(NWLIBMPI) 
     else
-        CORE_LIBS += $(LIBMPI) 
+        CORE_LIBS += $(NWLIBMPI) 
     endif
 else 
+    errornousempi:
+        $(info )
+        $(info please set the env. variable USE_MPI)
+        $(info and provide a working MPI installation)
+        $(info )
+        $(error )
     ifdef OLD_GA
       CORE_LIBS += -ltcgmsg 
     else
@@ -3608,13 +3646,13 @@ endif
 ifdef USE_LIBXC
     DEFINES += -DUSE_LIBXC
     EXTRA_LIBS += -L$(NWCHEM_TOP)/src/libext/libxc/install/lib
-    EXTRA_LIBS += -lxcf03 -lxc
+    EXTRA_LIBS += -lnwc_xcf03 -lnwc_xc
 endif
 
 # we use an external libxc library out of LIBXC_DIR
-ifdef LIBXC_DIR
+ifdef LIBXC_LIB
     DEFINES += -DUSE_LIBXC
-    EXTRA_LIBS += -L$(LIBXC_DIR)/lib
+    EXTRA_LIBS += -L$(LIBXC_LIB)
     EXTRA_LIBS += -lxcf03 -lxc
 endif
 
@@ -3636,7 +3674,7 @@ ifdef BUILD_PLUMED
     LD_LIBRARY_PATH := $(NWCHEM_TOP)/src/libext/lib:$(LD_LIBRARY_PATH)
     DEFINES += -DUSE_PLUMED
     PLUMED_HOME=$(NWCHEM_TOP)/src/libext
-    PLUMED_DYNAMIC_LIBS=$(shell test -x $(NWCHEM_TOP)/src/libext/bin/plumed && $(NWCHEM_TOP)/src/libext/bin/plumed info --configuration|egrep DYNAMIC_LIBS| cut -c 14-)
+    PLUMED_DYNAMIC_LIBS=$(shell test -x $(NWCHEM_TOP)/src/libext/bin/plumed && $(NWCHEM_TOP)/src/libext/bin/plumed info --configuration|grep DYNAMIC_LIBS| cut -c 14-)
     PLUMED_HASMPI = $(shell test -x $(NWCHEM_TOP)/src/libext/bin/plumed && $(NWCHEM_TOP)/src/libext/bin/plumed info --configuration|grep program_can_run_mpi|cut -c 21-21)
 endif
 ifdef USE_PLUMED
@@ -3650,8 +3688,8 @@ ifdef USE_PLUMED
 	    $(info  Please add to your PATH the directory where the plumed command is found )
 	    $(info )
     endif
-    PLUMED_HOME = $(shell plumed info --configuration|egrep prefix=|head -1|cut -c 8-)
-    PLUMED_DYNAMIC_LIBS = $(shell plumed info --configuration|egrep DYNAMIC_LIBS| cut -c 14-)
+    PLUMED_HOME = $(shell plumed info --configuration|grep prefix=|head -1|cut -c 8-)
+    PLUMED_DYNAMIC_LIBS = $(shell plumed info --configuration|grep DYNAMIC_LIBS| cut -c 14-)
     PLUMED_HASMPI = $(plumed info --configuration|grep program_can_run_mpi|cut -c 21-21)
     #PLUMED_LOAD= /home/edo/tahoma/apps/plumed262.intel20u2/lib/libplumed.a -ldl  -lstdc++ -lfftw3 -lz -ldl -llapack -lblas   -rdynamic -Wl,-Bsymbolic -fopenmp 
     ifndef PLUMED_DYNAMIC_LIBS
@@ -3737,6 +3775,7 @@ ifdef TCE_OPENACC
         FOPTIONS += -acc
         LDOPTIONS += -acc
     endif
+    NWCHEM_LINK_CUDA=1
 endif
 
 ifndef HIP
@@ -3753,6 +3792,15 @@ ifdef USE_F90_ALLOCATABLE
     DEFINES += -DUSE_F90_ALLOCATABLE
 endif
 
+ifdef NWCHEM_LINK_CUDA
+    ifeq ($(_FC),pgf90)
+       CORE_LIBS += -acc -cuda -cudalib=cublas
+    endif
+    ifeq ($(_FC),gfortran)
+       CORE_LIBS +=  -fopenacc -lcublas
+    endif
+endif
+
 ifdef GWCMPLX
   ifdef GWEN
     errorgw:
@@ -3884,8 +3932,24 @@ ifeq ($(shell echo $(BLASOPT) |awk '/lblas/ {print "Y"; exit}'),Y)
     DEFINES += -DBLAS_NOTHREADS
 endif
 
-ifndef BLAS_SIZE
-    LIB_DEFINES += -DUSE_INTEGER8
+ifneq ($(or $(SCALAPACK),$(SCALAPACK_LIB)),)
+   ifndef SCALAPACK_SIZE
+      $(info     )
+      $(info You must set)
+      $(info SCALAPACK_SIZE)
+      $(info )
+      $(error )
+   endif
+endif
+ifneq ($(or $(BLASOPT),$(BLAS_LIB)),)
+   ifndef BLAS_SIZE
+      $(info     )
+      $(info You must set)
+      $(info BLAS_SIZE)
+      $(info see https://nwchemgit.github.io/Compiling-NWChem.html#how-to-deal-with-integer-size-of-linear-algebra-libraries)
+      $(info )
+      $(error )
+   endif
 endif
 ifeq ($(BLAS_SIZE),8)
     LIB_DEFINES += -DUSE_INTEGER8
@@ -3969,7 +4033,7 @@ ifndef FLINT
 
     ifdef TCE_CUDA
         ifdef USE_TTLG
-            CUDA_VERS_GE8=$(shell nvcc --version|egrep rel|  awk '/release 9/ {print "Y";exit}; /release 8/ {print "Y";exit};{print "N"}')
+            CUDA_VERS_GE8=$(shell nvcc --version|grep rel|  awk '/release 9/ {print "Y";exit}; /release 8/ {print "Y";exit};{print "N"}')
             ifeq ($(CUDA_VERS_GE8),N)
                 CUDA_FLAGS = -O3 -Xcompiler -std=c++11 -DNOHTIME -Xptxas --warn-on-spills $(CUDA_ARCH) 
             else

src/ddscf/int_dip_ga.F

@@ -18,6 +18,11 @@
       integer ijshell, ilo, ihi, jlo, jhi
       integer l_buf, l_scr
       integer k_buf, k_scr
+      logical do_mirr
+      logical util_mirrmat
+      external util_mirrmat
+      integer g_x_mirr,g_y_mirr,g_z_mirr
+      integer g_x_org,g_y_org,g_z_org
 !
       double precision center(3)
 !
@@ -27,6 +32,34 @@
       call ga_zero(g_x)
       call ga_zero(g_y)
       call ga_zero(g_z)
+      g_x_mirr=0
+      g_y_mirr=0
+      g_z_mirr=0
+      do_mirr=util_mirrmat(1,g_x,g_x_mirr,
+     D     .false., .true.)
+      do_mirr=do_mirr.and.util_mirrmat(1,g_y,g_y_mirr,
+     D     .false., .true.)
+      do_mirr=do_mirr.and.util_mirrmat(1,g_z,g_z_mirr,
+     D     .false., .true.)
+      if(do_mirr) then
+         g_x_org=g_x
+         g_x=g_x_mirr
+         g_y_org=g_y
+         g_y=g_y_mirr
+         g_z_org=g_z
+         g_z=g_z_mirr
+            
+      else
+         if(g_x_mirr.ne.0) then
+            if(.not.ga_destroy(g_x_mirr))call errquit(
+     $           'could not destroy mirrx handle',1, GA_ERR)
+         endif
+         if(g_y_mirr.ne.0) then
+            if(.not.ga_destroy(g_y_mirr))call errquit(
+     $           'could not destroy mirry handle',1, GA_ERR)
+         endif
+      endif
+
       if(oscfps) call pstat_on(ps_int_dip)
 !
 !     grab basis set info type stuff
@@ -65,7 +98,8 @@
 !
 !     get the integrals we want
 !
-               call int_mpole(ibas, ishell, jbas, jshell,
+!               call int_mpole(ibas, ishell, jbas, jshell,
+               call int_mpole(jbas, jshell, ibas, ishell,
      $              1,
      $              center,
      $              mscratch, dbl_mb(k_scr), max1e, dbl_mb(k_buf))
@@ -84,6 +118,20 @@
       if(.not.MA_pop_stack(l_buf))
      $   call errquit('int_dip_ga:pop failed',0, MA_ERR)
 !
+      if (do_mirr) then
+         g_x=g_x_org
+         call util_mirrmerge(g_x_mirr,g_x)
+         if(.not.ga_destroy(g_x_mirr))call errquit(
+     $        'could not destroy g_x_mirr handle',1, GA_ERR)
+         g_y=g_y_org
+         call util_mirrmerge(g_y_mirr,g_y)
+         if(.not.ga_destroy(g_y_mirr))call errquit(
+     $        'could not destroy g_y_mirr handle',1, GA_ERR)
+         g_z=g_z_org
+         call util_mirrmerge(g_z_mirr,g_z)
+         if(.not.ga_destroy(g_z_mirr))call errquit(
+     $        'could not destroy g_x_mirr handle',1, GA_ERR)
+      endif
       call ga_sync()
 !
       if (util_print('multipole', print_debug)) then
@@ -101,15 +149,16 @@
       implicit none
 #include "global.fh"
       integer g_x, g_y, g_z, ilo, ihi, jlo, jhi
-      double precision buf(jlo:jhi,3,ilo:ihi)
+!      double precision buf(jlo:jhi,3,ilo:ihi)
+      double precision buf(ilo:ihi,3,jlo:jhi)
       integer i, j
-      do i = ilo,ihi
+c      do i = ilo,ihi
          do j = jlo, jhi
-            call ga_put(g_x,i,i,j,j,buf(j,1,i),1)
-            call ga_put(g_y,i,i,j,j,buf(j,2,i),1)
-            call ga_put(g_z,i,i,j,j,buf(j,3,i),1)
+            call ga_put(g_x,ilo,ihi,j,j,buf(ilo,1,j),1)
+            call ga_put(g_y,ilo,ihi,j,j,buf(ilo,2,j),1)
+            call ga_put(g_z,ilo,ihi,j,j,buf(ilo,3,j),1)
          end do
-      end do
+c      end do
       end
 
       subroutine int_qdr_ga(ibas, jbas, g_xx, g_xy, g_xz,

src/ddscf/localize.F

@@ -276,9 +276,9 @@ c
       gamma_tol = 1d-10
 c
       if (ga_nodeid() .eq. 0) then
-         write(6,2)
- 2       format(/10x,' iter   Max. delocal   Mean delocal    Converge'/
-     $        10x,' ----   ------------   ------------   ---------')
+        write(6,9001)
+ 9001   format(/10x,' iter   Max. delocal   Mean delocal    Converge'/
+     &          10x,' ----   ------------   ------------    --------')
          call util_flush(6)
       end if
 c
@@ -323,9 +323,9 @@ c
          call ga_dgop(2, d , 1, '+')
 c
          if (ga_nodeid() .eq. 0) then
-            write(6,1) iter, dmax, d/dble(nloc), gamma_max
- 1          format(10x, i5, 2f15.10, 1p,d12.2,d12.2)
-            call util_flush(6)
+           write(6,'(10x, i5, 2f15.10, 1p,d12.2,d12.2)')
+     &       iter, dmax, d/dble(nloc), gamma_max
+           call util_flush(6)
          end if
          call ga_sync
 c
@@ -439,8 +439,8 @@ c
                   end if
                end do
             end do
-            write(6,77) s, (list(a), pop(a), a=1,nlist)
- 77         format(i5, 100(2x,i4,'(',f5.2,')'))
+            write(6,9002) s, (list(a), pop(a), a=1,nlist)
+ 9002         format(i5, 100(2x,i4,'(',f5.2,')'))
          end do
          call util_flush(6)
       end if
@@ -498,10 +498,10 @@ c
       u21 = util_random(12345)
 c
       if (ga_nodeid() .eq. 0) then
-         write(6,2)
- 2       format(/10x,' iter   Max. dipole2   Mean dipole2    Converge'/
-     $        10x,' ----   ------------   ------------   ---------')
-         call util_flush(6)
+        write(6,9001)
+ 9001   format(/10x,' iter   Max. dipole2   Mean dipole2    Converge'/
+     &          10x,' ----   ------------   ------------    --------')
+        call util_flush(6)
       end if
 c
       dprev = 0.0d0
@@ -544,9 +544,9 @@ c
          call ga_dgop(2, d , 1, '+')
 c
          if (ga_nodeid() .eq. 0) then
-            write(6,1) iter, dmax, d/dble(nloc), gamma_max
- 1          format(10x, i5, 2f17.8, 1p,2d12.2)
-            call util_flush(6)
+           write(6,'(10x, i5, 2f17.8, 1p,2d12.2)')
+     &       iter, dmax, d/dble(nloc), gamma_max
+           call util_flush(6)
          end if
          call ga_sync
 c
@@ -663,8 +663,8 @@ c
                   end if
                end do
             end do
-            write(6,77) s, (list(a), pop(a), a=1,nlist)
- 77         format(i5, 100(2x,i4,'(',f5.2,')'))
+            write(6,9002) s, (list(a), pop(a), a=1,nlist)
+ 9002         format(i5, 100(2x,i4,'(',f5.2,')'))
          end do
          call util_flush(6)
       end if
@@ -674,16 +674,19 @@ c
       end
 c
 c     =================================================================
-c      
-      subroutine localizeIBO(basis, minbas,
-     &  c, nloc, iloc, nbf, nmo,
-     &  g_c, g_smat, g_iao, mnbf, eval, occ)
 c
-c     IAO-IBO localization
-c     
-      
+      subroutine localizeIBO(minbas, c, g_c, nloc, nbf, mnbf,
+     &  natoms)
+c
+c     IBO localization.  Similar to Pipek-Mezey, but in an orthonormal
+c     IAO basis. Analysis of the resulting IBOs is done in the calling
+c     routine.
+c
+c     written by Jochen Autschbach, jochena@buffalo.edu,
+c     adapted from routine localizePM
+
       implicit none
-      
+
 #include "errquit.fh"
 #include "nwc_const.fh"
 #include "mafdecls.fh"
@@ -692,136 +695,92 @@ c
 #include "bas.fh"
 #include "util.fh"
 #include "stdio.fh"
-c
-c     Localize the nloc orbitals in iloc(*) by mixing with each other
-c
-      integer basis, minbas, nloc, iloc(*), nbf, nmo, mnbf
-      double precision c(mnbf, 2), eval(nbf), occ(nbf)
-      integer g_c, g_smat, g_iao
-      integer maxat, nlist
-*............................   these should be dynamically allocated ?
-      parameter (maxat = nw_max_atom)
-      integer list(maxat)
-      double precision pop(maxat)
 
-      integer iter, ss, s, tt, t, a, u, bflo, bfhi, natoms, geom
+c     subroutine arguments:
+
+      integer minbas  ! minimal basis for IAOs (in)
+      integer nloc    ! number of orbitals to localize (in)
+      integer nbf     ! size of AO basis (in)
+      integer mnbf    ! size of minimal basis (in)
+      integer natoms  ! number of atoms (in)
+      double precision c(nbf, 2) ! memory used for MO transformations
+      integer g_c     ! orbital coefficients in IAO basis (inout)
+
+c     local variables:
+
+      integer nlist
+      integer iter, s, t, a, u, bflo, bfhi, geom
       double precision ast, bst, qast, qat, qas, gamma, cosg, sing, d,
      &  qs, dprev, tol, dmax, gamma_tol, gamma_max, tmp
       double precision minval, swap
       integer nrot, set, pair, neven
-      integer ntmp1, ntmp2
       logical master, debug
-      integer imo, jmo, kmo
-
-      integer g_tmp1, g_tmp2, g_cibocc
-
-      character*(11) pname
-
-      character*(256) lmotrans
-      logical file_write_ga
-      external file_write_ga      
+      integer g_tmp
+      character*(12) pname
 
 c     -----------------------------------------------------------------
 
-      pname = 'iao-ibo loc'
+      pname = 'iao-ibo iter'
       master = (ga_nodeid() .eq. 0)
       debug = .false.  ! true, during development
 
       if (debug.and.master) write(luout,*) 'entering '//pname
-c
-      if (.not. bas_geom(basis, geom)) call errquit
-     $     (pname//': basis ', 0, BASIS_ERR)
+
+c     sanity checks, so we don't crash later:
+
+      if (.not. bas_geom(minbas, geom)) call errquit
+     &     (pname//': minbas ', 0, BASIS_ERR)
       if (.not. geom_ncent(geom, natoms)) call errquit
-     $     (pname//': geom',0, GEOM_ERR)
-c
-      if (natoms.gt.maxat) call errquit
-     &  (pname//': maxat too small ', 911, UNKNOWN_ERR)
+     &  (pname//': geom',0, GEOM_ERR)
+      if (mnbf>nbf) call errquit
+     &  (pname//': mnbf>nbf', 0, UNKNOWN_ERR)
+      if (nloc>nbf) call errquit
+     &  (pname//': nloc>nbf', 0, UNKNOWN_ERR)
+      if (nloc>mnbf) call errquit
+     &  (pname//': nloc>mnbf', 0, UNKNOWN_ERR)
 
-c     note sure if we need this for sure, but later we assume this
-      if (nloc.gt.mnbf) call errquit
-     &  (pname//': nloc > mnbf', 66, UNKNOWN_ERR)
-      if (nloc.gt.nbf) call errquit
-     &  (pname//': nloc > nbf', 66, UNKNOWN_ERR)      
-
-
-c     DEBUG some things related to the basis function centers
+c     DEBUG some things related to the minimal basis centers
       if (debug) then
         do a = 1, natoms
-          if (.not. bas_ce2bfr(basis, a, bflo, bfhi))
-     $      call errquit('localized: basis ', 0, BASIS_ERR)
-          if (master) then
-            write(luout,'(1x,a,3i4)') 'ao basis, atom = ', a, bflo, bfhi
-          end if
           if (.not. bas_ce2bfr(minbas, a, bflo, bfhi))
-     &      call errquit('localized: basis ', 0, BASIS_ERR)
+     &      call errquit(pname//': minbas', 0, BASIS_ERR)
           if (master) then
-            write(luout,'(1x,a,3i4)') 'minbas  , atom = ', a, bflo, bfhi
+            write(luout,'(1x,a,3i4)') 'minbas, atom = ', a, bflo, bfhi
           end if
         end do
       end if ! debug
 
-c     transform the set of occupied MOs to be localized to the IAO basis
-c     via C(iao, trans) S C(MO)
-
-      if (.not. ga_create(MT_DBL, mnbf, nloc , 'loc:cibocc',
-     &  mnbf, 0, g_cibocc)) call errquit(pname//': cibocc',0, GA_ERR)
-
-      if (.not. ga_create(MT_DBL, nbf, nloc , 'loc:tmp1',
-     &  nbf, 0, g_tmp1)) call errquit(pname//': tmp1',0, GA_ERR)
-
-      call ga_dgemm('n', 'n', nbf, nloc, nbf, 
-     &  1.0d0, g_smat, g_c, 0.0d0, g_tmp1)
-
-      call ga_dgemm('t', 'n', mnbf, nloc, nbf, 
-     &  1.0d0, g_iao, g_tmp1, 0.0d0, g_cibocc)
-
-      if (.not. ga_destroy(g_tmp1))
-     &  call errquit(pname//': ga_destroy failed g_tmp1',0, GA_ERR)
-
-c      if (debug) call ga_print(g_cibocc)
-
-      if (debug .and. master) write(luout,*) 'cibocc done'
-
-c     save a copy of the starting MOs for later
-
-      if (.not.ga_duplicate(g_cibocc, g_tmp2, 'g_tmp2')) call
-     &  errquit(pname//': ga_dup cibocc', 0, GA_ERR)
-
-      call ga_copy(g_cibocc, g_tmp2)
-
 c     the MOs in the orthonormal IAO basis must form an orthogonal
 c     matrix. test that before localization, and again later
 
-      if (.not. ga_create(MT_DBL, nloc, nloc , 'loc:tmp1',
-     &  mnbf, 0, g_tmp1)) call errquit(pname//': tmp1',0, GA_ERR)
+      if (.not. ga_create(MT_DBL, nloc, nloc , 'ibo:tmp',
+     &  0, 0, g_tmp)) call errquit(pname//': tmp',0, GA_ERR)
 
-      call ga_dgemm('t', 'n', nloc, nloc, mnbf, 
-     &  1.0d0, g_cibocc, g_cibocc, 0.0d0, g_tmp1)
+      call ga_dgemm('t', 'n', nloc, nloc, mnbf,
+     &  1.0d0, g_c, g_c, 0.0d0, g_tmp)
 
-      call ga_add_todiag(g_tmp1, -1.0d0)
+      call ga_add_todiag(g_tmp, -1.0d0)
 
-      call ga_maxelt(g_tmp1, tmp)
+      call ga_maxelt(g_tmp, tmp)
       if (master) then
-        write(luout,'(/1x,a,1x,f18.12)')
-     &    'IBO loc: largest element of C(MO, trans) C(MO) -1: ',
+        write(luout,'(/1x,a,1x,f18.8)')
+     &    'IBO loc: largest element in C(MO,T) C(MO) -1: ',
      &    tmp
         write(luout,'(1x,a)')
      &    'should be zero, for CMOs in the IAO basis'
       end if
 
-      if (.not. ga_destroy(g_tmp1))
-     &  call errquit(pname//': ga_destroy failed g_tmp1',0, GA_ERR)
+      if (.not. ga_destroy(g_tmp))
+     &  call errquit(pname//': ga_destroy failed g_tmp',0, GA_ERR)
 
-      
-c
       tol = 1d-8
       gamma_tol = 1d-10
 c
       if (master) then
-         write(luout,2)
- 2       format(/10x,' iter   Max. delocal   Mean delocal    Converge'/
-     $        10x,' ----   ------------   ------------   ---------')
-         call util_flush(luout)
+        write(luout,9001)
+ 9001   format(/10x,' iter   Max. delocal   Mean delocal    Converge'/
+     &          10x,' ----   ------------   ------------    --------')
+        call util_flush(luout)
       end if
 c
       dprev = 0.0d0
@@ -832,25 +791,24 @@ c
 c
 c     Analyze convergence by forming P
 c
-         d = 0.0d0
-         dmax = 0.0d0
-         do ss = 1+ga_nodeid(), nloc, ga_nnodes()
-            s = iloc(ss)
-            call ga_get(g_cibocc, 1, mnbf, s, s, c(1,1), 1)
-            qs = 0.0d0
-            do a = 1, natoms
-               if (.not. bas_ce2bfr(minbas, a, bflo, bfhi))
-     &              call errquit(pname//': basis ', 1, BASIS_ERR)
-               qas  = 0.0d0
-               do u = bflo, bfhi
-                  qas  = qas  + c(u,1)*c(u,1)
-               end do
-               qs = qs + qas**2
+        d = 0.0d0
+        dmax = 0.0d0
+        do s = 1+ga_nodeid(), nloc, ga_nnodes()
+          call ga_get(g_c, 1, mnbf, s, s, c(1,1), 1)
+          qs = 0.0d0
+          do a = 1, natoms
+            if (.not. bas_ce2bfr(minbas, a, bflo, bfhi))
+     &        call errquit(pname//': basis ', 1, BASIS_ERR)
+            qas  = 0.0d0
+            do u = bflo, bfhi
+              qas  = qas  + c(u,1)*c(u,1)
             end do
-*            write(luout,*) ' ds ', s, 1.0d0/qs
-            dmax = max(dmax, 1.0d0/qs)
-            d = d + 1.0d0/qs
-         end do
+            qs = qs + qas**2
+          end do
+*         write(luout,*) ' ds ', s, 1.0d0/qs
+          dmax = max(dmax, 1.0d0/qs)
+          d = d + 1.0d0/qs
+        end do
 c
 #ifdef NWCHEM_USE_GOP_ABSMAX
          call ga_dgop(1, gamma_max, 1, 'absmax')
@@ -861,84 +819,77 @@ c
          call ga_dgop(1, gamma_max, 1, 'max')
          call ga_dgop(1, dmax, 1, 'max')
 #endif
-         call ga_dgop(2, d , 1, '+')
+        call ga_dgop(2, d , 1, '+')
 c
-         if (ga_nodeid() .eq. 0) then
-            write(luout,1) iter, dmax, d/dble(nloc), gamma_max
- 1          format(10x, i5, 2f15.10, 1p,d12.2,d12.2)
-            call util_flush(luout)
-         end if
-         call ga_sync
+        if (ga_nodeid() .eq. 0) then
+          write(luout,'(10x, i5, 2f15.10, 1p,d12.2,d12.2)')
+     &      iter, dmax, d/dble(nloc), gamma_max
+          call util_flush(luout)
+        end if
+        call ga_sync
+
+        if (iter.gt.1 .and. gamma_max.lt.tol) goto 1000
+        gamma_max = 0.0d0
+
+c       Loop over pairs with as much parallelism as possible
+
+        neven = nloc + mod(nloc,2)
+        do set = 1, neven-1
+          do pair = 1+ga_nodeid(), neven/2, ga_nnodes()
+            call localize_pairs(neven, set, pair, s, t)
+            if (t .le. nloc) then
+
+              call ga_get(g_c, 1, mnbf, s, s, c(1,1), 1)
+              call ga_get(g_c, 1, mnbf, t, t, c(1,2), 1)
 c
-**         if (abs(d-dprev)/dble(nloc) .lt. tol) goto 1000
-**         dprev = d
-         if (iter.gt.1 .and. gamma_max.lt.tol) goto 1000
-         gamma_max = 0.0d0
+c             Form rotation information
 c
-c     Loop over pairs with as much parallelism as possible
+              ast = 0.0d0
+              bst = 0.0d0
+              do a = 1, natoms
+                if (.not. bas_ce2bfr(minbas, a, bflo, bfhi))
+     &            call errquit(pname//': basis ', 2,
+     &            BASIS_ERR)
 c
-         neven = nloc + mod(nloc,2)
-         do set = 1, neven-1
-            do pair = 1+ga_nodeid(), neven/2, ga_nnodes()
-               call localize_pairs(neven, set, pair, ss, tt)
-               if (tt .le. nloc) then
-                  s = iloc(ss)
-                  t = iloc(tt)
-*                  write(luout,*) nloc, neven, set, pair, ss, tt, s, t
-                  call ga_get(g_cibocc, 1, mnbf, s, s, c(1,1), 1)
-                  call ga_get(g_cibocc, 1, mnbf, t, t, c(1,2), 1)
-c     
-c     Form rotation information
-c     
-                  ast = 0.0d0
-                  bst = 0.0d0
-                  do a = 1, natoms
-                     if (.not. bas_ce2bfr(minbas, a, bflo, bfhi))
-     &                    call errquit(pname//': basis ', 2,
-     &                BASIS_ERR)
-c     
-                     qast = 0.0d0
-                     qas  = 0.0d0
-                     qat  = 0.0d0
-                     do u = bflo, bfhi
-                        qast = qast + c(u,2)*c(u,1) + c(u,1)*c(u,2)
-                        qas  = qas  + c(u,1)*c(u,1)
-                        qat  = qat  + c(u,2)*c(u,2)
-                     end do
-                     qast = qast * 0.5d0
-c     
-c                    ast = ast + qast**2 - 0.25d0*(qas - qat)**2
-c                    bst = bst + qast*(qas - qat)
-                     tmp = -1.0d0*(qas**4 + qat**4)
-                     tmp = tmp + 6.0d0*((qas**2 + qat**2)* qast**2)
-                     tmp = tmp + qas**3 * qat + qas * qat**3
-                     ast = ast + 0.25d0*tmp
-                     tmp = 4.0d0 * qast*(qas**3 - qat**3)
-                     bst = bst + 0.25d0*tmp
-                  end do
-c     
-                  gamma = 0.25d0*acos(-ast/sqrt(ast**2+bst**2))
-                  gamma = sign(gamma,bst)
-                  gamma_max = max(gamma_max, abs(gamma))
-*                  if (iter .eq. 1 .and. abs(gamma).lt.0.01d0) then
-*                     gamma = (util_random(0)-0.5d0)*3.14d0
-*                  endif
-c     
-                  if (abs(gamma) .gt. gamma_tol) then
-                     nrot = nrot + 1
-                     cosg = cos(gamma)
-                     sing = sin(gamma)
-c     
-c     Do the rotation of C 
-c     
-                     call drot(mnbf, c(1,1), 1, c(1,2), 1, cosg, sing)
-                     call ga_put(g_cibocc, 1, mnbf, s, s, c(1,1), 1)
-                     call ga_put(g_cibocc, 1, mnbf, t, t, c(1,2), 1)
-                  end if
-               end if
-            end do
-            call ga_sync
-         end do
+                qast = 0.0d0
+                qas  = 0.0d0
+                qat  = 0.0d0
+                do u = bflo, bfhi
+                  qast = qast + c(u,2)*c(u,1) + c(u,1)*c(u,2)
+                  qas  = qas  + c(u,1)*c(u,1)
+                  qat  = qat  + c(u,2)*c(u,2)
+                end do
+                qast = qast * 0.5d0
+c
+c               ast = ast + qast**2 - 0.25d0*(qas - qat)**2
+c               bst = bst + qast*(qas - qat)
+                tmp = -1.0d0*(qas**4 + qat**4)
+                tmp = tmp + 6.0d0*((qas**2 + qat**2)* qast**2)
+                tmp = tmp + qas**3 * qat + qas * qat**3
+                ast = ast + 0.25d0*tmp
+                tmp = 4.0d0 * qast*(qas**3 - qat**3)
+                bst = bst + 0.25d0*tmp
+              end do
+c
+              gamma = 0.25d0*acos(-ast/sqrt(ast**2+bst**2))
+              gamma = sign(gamma,bst)
+              gamma_max = max(gamma_max, abs(gamma))
+
+              if (abs(gamma) .gt. gamma_tol) then
+                nrot = nrot + 1
+                cosg = cos(gamma)
+                sing = sin(gamma)
+c
+c               Do the rotation of C
+c
+                call drot(mnbf, c(1,1), 1, c(1,2), 1, cosg, sing)
+                call ga_put(g_c, 1, mnbf, s, s, c(1,1), 1)
+                call ga_put(g_c, 1, mnbf, t, t, c(1,2), 1)
+              end if
+            end if
+          end do
+          call ga_sync
+        end do
       end do
 c
  1000 continue
@@ -947,178 +898,29 @@ c
 c     the MOs in the orthonormal IAO basis must form an orthogonal
 c     matrix. test that after localization
 
-      if (.not. ga_create(MT_DBL, nloc, nloc , 'loc:tmp1',
-     &  mnbf, 0, g_tmp1)) call errquit(pname//': tmp1',0, GA_ERR)
+      if (.not. ga_create(MT_DBL, nloc, nloc , 'ibo:tmp',
+     &  mnbf, 0, g_tmp)) call errquit(pname//': tmp',0, GA_ERR)
 
-      call ga_dgemm('t', 'n', nloc, nloc, mnbf, 
-     &  1.0d0, g_cibocc, g_cibocc, 0.0d0, g_tmp1)
+      call ga_dgemm('t', 'n', nloc, nloc, mnbf,
+     &  1.0d0, g_c, g_c, 0.0d0, g_tmp)
 
-      call ga_add_todiag(g_tmp1, -1.0d0)
+      call ga_add_todiag(g_tmp, -1.0d0)
 
-      call ga_maxelt(g_tmp1, tmp)
+      call ga_maxelt(g_tmp, tmp)
       if (master) then
-        write(luout,'(/1x,a,1x,f18.12)')
-     &    'IBO loc: largest element of C(MO, trans) C(MO) -1: ',
+        write(luout,'(/1x,a,1x,f18.8)')
+     &    'IBO loc: largest element in C(MO,T) C(MO) -1: ',
      &    tmp
         write(luout,'(1x,a)')
-     &    'should be zero, for LMOs in the IAO basis'
+     &    'should be zero, for IBOs in the IAO basis'
       end if
 
-c     Following steps:
-      
-c     1. while we have array g_tmp1 allocated, also calculate the
-c     localization transform. The starting MOs were saved in g_tmp2 in
-c     the IAO basis
+      if (.not. ga_destroy(g_tmp))
+     &  call errquit(pname//': ga_destroy failed g_tmp',0, GA_ERR)
 
-c     2. calculate LMO energies and order the LMO set by energy
+      return
+      end
 
-c     3.  save the localization transform to a file, instead of
-c     generating it again in the calling routine and saving it from
-c     there
-
-c     4. We then transform the CMOs to re-ordered LMOs in the
-c     AO basis and store the transformed MOs back in g_c      
-
-
-c     CMOs(iao, trans) * LMOs(iao) = localization transform -> g_tmp1
-      call ga_dgemm('t', 'n', nloc, nloc, mnbf, 
-     &  1.0d0, g_tmp2, g_cibocc, 0.0d0, g_tmp1)
-
-      if (.not. ga_destroy(g_tmp2))
-     &  call errquit(pname//': ga_destroy failed g_tmp2',0, GA_ERR)
-
-c     transform MO energies. We assume that the first nloc MOs were
-c     localized, and that the Fock matrix was diagonal in the basis of
-c     input MOs. We also assume (this is tested upon entry) that
-c     nloc is smaller than the column sizes of arays eval, c(:,1:2)
-c     so we don't segfault when using c(:,:) as temp memory
-
-      do jmo = 1,nloc
-        call ga_get(g_tmp1, 1, nloc, jmo, jmo, c(1,1), 1)
-        tmp = 0.0d0
-        do imo = 1,nloc
-          tmp = tmp + eval(imo) * c(imo,1)**2
-        end do ! jmo
-        c(jmo,2) = tmp ! temp storage for transformed MO energies
-      end do ! imo
-      do imo = 1,nloc
-        eval(imo) = c(imo,2)
-      end do
-      
-c     sort LMOs by energy. We'll have to do this on the master node, or
-c     otherwise avoid having multiple nodes working on the same data
-
-      if (master) then
-      do imo = 1,nloc -1
-        minval = eval(imo)
-c       find lowest eval(kmo) below eval(imo)
-        kmo = 0
-        do jmo = imo+1, nloc
-          if (eval(jmo).lt.minval) then
-            kmo = jmo
-            minval = eval(jmo)
-          end if
-        end do
-c       if kmo > 0 we swap imo and jmo
-        if (kmo.ne.0) then
-          if (debug.and.master) write(luout,*) 'swap ',imo,kmo
-c         swap eigenvalues
-          swap = eval(imo)
-          eval(imo) = eval(kmo)
-          eval(kmo) = swap
-c         swap columns of LMO transformation
-          call ga_get(g_tmp1, 1, nloc, imo, imo, c(1,1), 1)
-          call ga_get(g_tmp1, 1, nloc, kmo, kmo, c(1,2), 1)
-          call ga_put(g_tmp1, 1, nloc, imo, imo, c(1,2), 1)
-          call ga_put(g_tmp1, 1, nloc, kmo, kmo, c(1,1), 1)
-c         swap LMOs in IAO basis, for the analysis below
-          call ga_get(g_cibocc, 1, mnbf, imo, imo, c(1,1), 1)
-          call ga_get(g_cibocc, 1, mnbf, kmo, kmo, c(1,2), 1)
-          call ga_put(g_cibocc, 1, mnbf, imo, imo, c(1,2), 1)
-          call ga_put(g_cibocc, 1, mnbf, kmo, kmo, c(1,1), 1)          
-        end if
-      end do
-      end if ! master
-
-      call ga_sync
-
-c     write transformation to scratch file
-      
-      call util_file_name('lmotrans',.true.,.true.,lmotrans)
-      if(.not.file_write_ga(lmotrans,g_tmp1)) call errquit
-     &  (pname//': could not write lmotrans',0, DISK_ERR)
-
-      if (debug.and.master) write (luout,*)
-     &  pname//': lmos reordered and lmotrans written'
-
-c     transform CMOs to re-ordered LMOs
-      
-      if (.not. ga_create(MT_DBL, nbf, nloc , 'loc:tmp2',
-     &  nbf, 0, g_tmp2)) call errquit(pname//': tmp2',0, GA_ERR)      
-
-      call ga_dgemm('n', 'n', nbf, nloc, nloc, 
-     &  1.0d0, g_c, g_tmp1, 0.0d0, g_tmp2)
-
-      call ga_copy(g_tmp2, g_c)
-
-      if (debug.and.master) write (luout,*) pname//': g_c now has LMOs'
-
-      if (.not. ga_destroy(g_tmp1))
-     &  call errquit(pname//': ga_destroy failed g_tmp1',0, GA_ERR)
-
-      if (.not. ga_destroy(g_tmp2))
-     &  call errquit(pname//': ga_destroy failed g_tmp2',0, GA_ERR)
-
-c     Analyze localization of each MO:
-c     per LMO, a list of atomic populations is printed
-c     in decreasing magnitude, with the polulations in parentheses
-
-      if (master) then
-         write(luout,'(/1x,a/)') 'IAO-IBO localized orbitals'
-         do ss = 1, nloc
-            s = iloc(ss)
-            call ga_get(g_cibocc,  1, mnbf, s, s, c(1,1), 1)
-            nlist = 0
-            do a = 1, natoms
-               if (.not. bas_ce2bfr(minbas, a, bflo, bfhi))
-     &              call errquit('localized: basis ', 3,
-     &          BASIS_ERR)
-               qas  = 0.0d0
-               do u = bflo, bfhi
-                  qas  = qas  + c(u,1)*c(u,1)
-               end do
-               if (abs(qas) .gt. 0.01d0) then
-                  nlist = nlist + 1
-                  list(nlist) = a
-                  pop(nlist) = qas
-               end if
-            end do
-            do u = 1, nlist
-               do t = 1, u-1
-                  if (abs(pop(t)).lt.abs(pop(u))) then
-                     tmp = pop(u)
-                     pop(u) = pop(t)
-                     pop(t) = tmp
-                     tt = list(u)
-                     list(u) = list(t)
-                     list(t) = tt
-                  end if
-               end do
-            end do
-            write(luout,77) s, eval(s),
-     &        occ(s),(list(a), pop(a), a=1,nlist)
- 77         format(i5, 1x, f14.6,1x, f5.3, 1x,100(2x,i4,'(',f5.2,')'))
-          end do
-          write(luout,*)
-         call util_flush(luout)
-       end if
-
-       if (.not. ga_destroy(g_cibocc))
-     &   call errquit(pname//': ga_destroy failed g_cibocc',0, GA_ERR)
-
-      if (debug.and.master) write(luout,*) 'leaving '//pname
-
-      end      
 c
 c     =================================================================
 c      

src/ddscf/movecs_pr_anal.F

@@ -249,7 +249,8 @@ c
 c     create a global array to store x, y, z and x^2, y^2, z^2 for each AO
 c
       if (.not. ga_create(mt_dbl, 6*nbf, nbf, 'GXLM',
-     $     32,32,g_xlm)) call errquit('mo_r_and_r2: g_xlm',6*nbf*nbf,
+c     $     32,32,g_xlm)) call errquit('mo_r_and_r2: g_xlm',6*nbf*nbf,
+     $     nbf,32,g_xlm)) call errquit('mo_r_and_r2: g_xlm',6*nbf*nbf,
      &       GA_ERR)
 c
 c     compute dipoles and quadrupole components for each AO
@@ -301,6 +302,7 @@ c
 #include "geom.fh"
 #include "stdio.fh"
 #include "util_params.fh"
+#include "util.fh"
 c     
       integer basis             ! [input] basis 
       double precision center(3) ! [input] the expansion center
@@ -318,6 +320,9 @@ c
       integer ilo, ihi, jlo, jhi, idim, jdim, ind, i, j, l, ioff
       integer l_scr, k_scr, l_mp, k_mp
       integer lmax              ! Maximum value of L = 2
+      integer clo, chi, rlo, rhi
+      integer type, dim1, dim2
+      logical doloop, oprint
 c
       if (.not. bas_geom(basis, geom)) call errquit
      $     ('multipole: bad basis', 0, BASIS_ERR)
@@ -349,11 +354,30 @@ c
      &     call errquit('xlm_pole: cannot allocate scratch', maxscr,
      &       MA_ERR)
 c     
+      call ga_distribution(g_xlm,
+     .     ga_nodeid(), rlo, rhi, clo, chi)
+      doloop=.not.(rlo.eq.0.and.rhi.eq.-1.and.clo.eq.0.and.chi.eq.-1)
+      call ga_inquire(g_xlm, type, dim1, dim2)
+      oprint=(ga_nodeid() .eq. 0 .and. 
+     $     util_print('debug_xlm_make', print_debug))
+      if(oprint) then
+         write(luout,*) ' dim1 dim2 ',dim1,dim2
+         write(luout,*) ' rlo rhi ',rlo,rhi
+         write(luout,*) ' clo chi ',clo,chi
+      endif
       call ga_zero(g_xlm)
+       nproc=0
+       if(doloop) nproc=1
+       call ga_igop(65536,nproc,1,'+')
+       if(oprint) then
+          write(luout,*) ' using ',nproc,
+     C      ' nodes out of ',ga_nnodes()
+          call util_flush(6)
+       endif
 c
+      if (doloop) then
       ijshell = -1
       me = ga_nodeid()
-      nproc = ga_nnodes()
       do ishell = 1, nshell
          if (.not. bas_cn2bfr(basis, ishell, ilo, ihi)) call errquit
      &        ('xlm_pole: bas_cn2bfr failed for basis', basis,
@@ -429,6 +453,7 @@ c
          end do
       end do
 c
+      endif
       call ga_sync
 c
 c      write(LuOut,*) ' THE AO MPOLES '

src/esp/esp_plt.F

@@ -8,7 +8,9 @@ c
 #include "esp_common.fh"
 #include "global.fh"
 #include "msgids.fh"
+#include "geom.fh"
 #include "util.fh"
+#include "errquit.fh"
 c
       real*8 x(3,natoms),q(ndim,4)
       real*8 g(*)
@@ -83,6 +85,7 @@ c
 c
       call util_file_name('xyz',.false.,.false.,xyzfil)
       open(unit=lfnxyz,file=xyzfil,form='formatted',status='unknown')
+#if 0      
       write(lfnxyz,2000) natoms
  2000 format('* Coordinates in XYZ format',/,i5)
       do 5 i=1,natoms
@@ -90,6 +93,11 @@ c
      + (10.0d0*cau2nm*x(j,i),j=1,3),q(i,4)
  2001 format(a,t3,4f12.6)
     5 continue
+#else
+      if(.not.geom_print_xyz(igeom, lfnxyz))
+     A     call errquit(' failed geom_print_xyz',
+     E     0, GEOM_ERR)
+#endif      
       close(lfnxyz)
       endif
 c

src/geom/GNUmakefile

@@ -38,7 +38,7 @@ include ../config/makelib.h
             FOPTIMIZE += -O0
     endif
   endif
-  ifeq ($(FC),ifort)
+  ifeq ($(_FC),ifort)
             FOPTIONS += -save
   endif
   ifeq ($(_FC),xlf)

src/libext/elpa/build_elpa.sh

@@ -9,7 +9,6 @@ SHORTVERSION=2021.11.001
 VERSION=new_release_2021.11.001
 #https://gitlab.mpcdf.mpg.de/elpa/elpa/-/archive/new_release_2020.11.001/elpa-new_release_2020.11.001.tar.gz
 echo mpif90 is `which mpif90`
-export ARFLAGS=rU
 if [ -f  elpa-${VERSION}.tar.gz ]; then
     echo "using existing"  elpa-${VERSION}.tar.gz
 else
@@ -21,8 +20,14 @@ tar xzf elpa-${VERSION}.tar.gz
 ln -sf elpa-${VERSION} elpa
 cd elpa
 UNAME_S=$(uname -s)
+if [[ ${UNAME_S} == Linux ]]; then
+    export ARFLAGS=rU
+fi
 if [[ ${UNAME_S} == Darwin ]]; then
-    export FORTRAN_CPP=$(find  /usr/local/Cellar/gcc/`brew list --versions gcc|cut -c 5-`/bin -name cpp*)
+if [[  -z "$HOMEBREW_PREFIX" ]]; then
+    HOMEBREW_PREFIX=/usr/local
+fi
+    export FORTRAN_CPP=$(find  "$HOMEBREW_PREFIX"/Cellar/gcc/`brew list --versions gcc|cut -c 5-`/bin -name cpp*)
     if ! [ -x "$(command -v $FORTRAN_CPP)" ]; then
 	echo
 	echo cpp from gcc homebrew missing
@@ -153,7 +158,11 @@ fi
 
 if [ ! -f  configure ]; then
     sh ./autogen.sh
-fi    
+fi
+# patch affinity
+rm -f check_thread_affinity.patch
+wget https://raw.githubusercontent.com/conda-forge/elpa-feedstock/main/recipe/check_thread_affinity.patch
+patch -p2 -s -N < check_thread_affinity.patch
 mkdir -p build
 cd build
 if [[ !  -z "${BUILD_SCALAPACK}" ]]; then
@@ -197,11 +206,13 @@ unset SCALAPACK_FCFLAGS
 unset SCALAPACK_LDFLAGS
 echo mpif90 is `which mpif90`
 echo MPIF90 is "$MPIF90"
-    make V=1 -j1 FC=$MPIF90 CC=$MPICC -l0.0001
+    make V=0 -j1 FC=$MPIF90 CC=$MPICC -l0.0001
 if [[ "$?" != "0" ]]; then
     echo " "
     echo "Elpa compilation failed"
     echo " "
+    echo "****** config.log *****"
+    cat config.log
     exit 1
 fi
     make V=0   install

src/libext/libxc/GNUmakefile

@@ -5,8 +5,8 @@
 
 include ../../config/makefile.h
 
-install/lib/libxc.a:
-	./build_libxc.sh 
+install/lib/libnwc_xc.a:
+	./build_libxc.sh $(LIBXC_VERSION)
 
 LIB_TARGETS += libxc
 

src/libext/libxc/build_libxc.sh

@@ -7,14 +7,25 @@ check_tgz() {
     [ -f $1 ] && gunzip -t $1 > /dev/null && myexit=1
     echo $myexit
 }
-
-VERSION=5.2.2
+if [ $# -eq 0 ];  then
+    VERSION=6.1.0
+else
+    VERSION=$1
+fi
+VERSION_MAJOR=$(echo $VERSION | cut -d . -f 1)
+if [[ "$VERSION_MAJOR" -lt 4 ]]; then
+    echo
+    echo "LIBXC unsupported version " "$VERSION"
+    echo "please use 4.0.0 and later versions"
+    echo
+    exit  1
+fi
 TGZ=libxc-${VERSION}.tar.gz
 if [ `check_tgz $TGZ` == 1 ]; then
     echo "using existing $TGZ"
 else
     echo "downloading $TGZ"
-    curl -L https://gitlab.com/libxc/libxc/-/archive/${VERSION}/libxc-${VERSION}.tar.gz -o $TGZ
+    curl -sS -L https://gitlab.com/libxc/libxc/-/archive/${VERSION}/libxc-${VERSION}.tar.gz -o $TGZ
     if [ `check_tgz $TGZ` != 1 ]; then
 	rm -f libxc-${VERSION}.tar.gz
 	curl -L https://github.com/ElectronicStructureLibrary/libxc/archive/refs/tags/${VERSION}.tar.gz -o $TGZ
@@ -78,7 +89,7 @@ fi
 
 cd libxc
 # patch pk09 to avoid compiler  memory problems
-patch -p0 -N < ../pk09.patch
+#patch -p0 -N < ../pk09.patch
 mkdir -p build
 cd build
 if [[ -z "${NWCHEM_TOP}" ]]; then
@@ -100,10 +111,22 @@ else
     fcflags=" "
 fi
 rm -rf libxc/build
+if [[ "${USE_HWOPT}" == "n" ]]; then
+    enable_xhost_flag=OFF
+else
+    enable_xhost_flag=ON
+fi
+
 $CMAKE -E env CFLAGS="$cflags" LDFLAGS="$ldflags" FCFLAGS="$fcflags" FFLAGS="$fcflags" \
 $CMAKE  -DCMAKE_INSTALL_PREFIX=${NWCHEM_TOP}/src/libext/libxc/install -DCMAKE_C_COMPILER=$CC -DENABLE_FORTRAN=ON -DCMAKE_Fortran_COMPILER=$FC -DDISABLE_KXC=OFF \
+-DENABLE_XHOST="$enable_xhost_flag" \
+-DENABLE_FORTRAN03=ON \
 -DCMAKE_INSTALL_LIBDIR="lib" -DCMAKE_BUILD_TYPE=Release ..
 
 make -j4 | tee make.log
 make install
+strip --strip-debug ../../install/lib/libxc.a
+strip --strip-debug ../../install/lib/libxcf03.a
+ln -sf  ../../install/lib/libxc.a ../../install/lib/libnwc_xc.a
+ln -sf  ../../install/lib/libxcf03.a ../../install/lib/libnwc_xcf03.a
 

src/libext/openblas/build_openblas.sh

@@ -8,8 +8,16 @@ if [ -f  OpenBLAS-${VERSION}.tar.gz ]; then
     echo "using existing"  OpenBLAS-${VERSION}.tar.gz
 else
     rm -rf OpenBLAS*
-    curl -L https://github.com/xianyi/OpenBLAS/archive/v${VERSION}.tar.gz -o OpenBLAS-${VERSION}.tar.gz
+    tries=1 ; until [ "$tries" -ge 6 ] ; do
+		  if [ "$tries" -gt 1 ]; then sleep 9; echo attempt no.  $tries ; fi
+		  curl -L https://github.com/xianyi/OpenBLAS/archive/v${VERSION}.tar.gz -o OpenBLAS-${VERSION}.tar.gz ;
+		  # check tar.gz integrity
+		  gzip -t OpenBLAS-${VERSION}.tar.gz >&  /dev/null
+		  if [ $? -eq 0 ]; then break ;  fi
+		  tries=$((tries+1)) ;  done
 fi
+gzip -t OpenBLAS-${VERSION}.tar.gz >&  /dev/null
+if [ $? -ne 0 ]; then echo  "openBLAS tarball not ready"; rm -f OpenBLAS-${VERSION}.tar.gz; exit 1 ; fi
 tar xzf OpenBLAS-${VERSION}.tar.gz
 ln -sf OpenBLAS-${VERSION} OpenBLAS
 cd OpenBLAS
@@ -65,8 +73,10 @@ if [[ "${NWCHEM_TARGET}" == "LINUX" ]]; then
 else
   binary=64
 fi
-if [ -n "${USE_DYNAMIC_ARCH}" ]; then
-    FORCETARGET+="DYNAMIC_ARCH=1 DYNAMIC_OLDER=1"
+if [[ -n "${USE_DYNAMIC_ARCH}" ]] || [[ "${USE_HWOPT}" == "n" ]]; then
+    if [[ "$arch" == "x86_64" ]]; then
+	FORCETARGET+="DYNAMIC_ARCH=1 DYNAMIC_OLDER=1"
+    fi
 fi
 #cray ftn wrapper
 if [[ ${FC} == ftn ]]; then
@@ -136,10 +146,21 @@ elif  [[ -n ${FC} ]] && [[ "${FC}" == "ifort" ]] || [[ "${FC}" == "ifx" ]]; then
     LAPACK_FPFLAGS_VAL=" -fp-model source -O2 -g "
 else
 	#assuming gfortran
+    FORCETARGET+=' F_COMPILER=GFORTRAN '
     LAPACK_FPFLAGS_VAL=" "
         if [[ ${BLAS_SIZE} == 8 ]]; then
-	    LAPACK_FLAGS_VAL+=" -fdefault-integer-8"
-	fi
+           LAPACK_FPFLAGS_VAL+=" -fdefault-integer-8"
+       fi
+fi
+if  [[ -n ${CC} ]] && [[ "${CC}" == "amdclang" ]]; then
+    let VERSIONEQ15=$(expr `${CC} -dM -E - < /dev/null 2> /dev/null|egrep 15|grep __clang_major__ |cut  -d ' ' -f 3 ` \= 15)
+    if [[ ${VERSIONEQ15} == 1 ]]; then
+       echo "amdclang 15 buggy. reduced optimization to O1"
+       FORCETARGET+=' COMMON_OPT=-O1'
+    fi
+fi
+if [[   -z "${FC}" ]]; then
+    FC=gfortran
 fi
 if [[   -z "${CC}" ]]; then
     CC=cc
@@ -179,6 +200,13 @@ else
     THREADOPT="1"
     MYNTS="128"
 fi
+# cross compilation
+GOTMINGW64=$("$CC" -dM -E - </dev/null 2> /dev/null |grep MINGW64|cut -c21)
+if [[ "${GOTMINGW64}" == "1" ]]; then
+    FORCETARGET+=HOSTCC=\"gcc\"
+    THREADOPT="0"
+    MYNTS="1"
+fi
 
 #we want openblas to use pthreads and not openmp.
 #but NWChem and OpenBLAS both use USE_OPENMP
@@ -187,7 +215,8 @@ if [[  ! -z "${USE_OPENMP}" ]]; then
     unset USE_OPENMP
     NWCHEM_USE_OPENMP=1
 fi
-echo make $FORCETARGET LAPACK_FPFLAGS=$LAPACK_FPFLAGS_VAL  INTERFACE64=$sixty4_int BINARY=$binary NUM_THREADS=$MYNTS NO_CBLAS=1 NO_LAPACKE=1 DEBUG=0 USE_THREAD=$THREADOPT  libs netlib -j4
+echo FC is $FC
+echo make FC=$FC $FORCETARGET LAPACK_FPFLAGS=$LAPACK_FPFLAGS_VAL  INTERFACE64=$sixty4_int BINARY=$binary NUM_THREADS=$MYNTS NO_CBLAS=1 NO_LAPACKE=1 DEBUG=0 USE_THREAD=$THREADOPT  libs netlib -j4
 echo
 echo OpenBLAS compilation in progress
 echo output redirected to libext/openblas/OpenBLAS/openblas.log
@@ -195,7 +224,7 @@ echo
 if [[ ${_FC} == xlf ]]; then
  make FC="xlf -qextname" $FORCETARGET  LAPACK_FPFLAGS="$LAPACK_FPFLAGS_VAL"  INTERFACE64="$sixty4_int" BINARY="$binary" NUM_THREADS=$MYNTS NO_CBLAS=1 NO_LAPACKE=1 DEBUG=0 USE_THREAD="$THREADOPT" libs netlib -j4 >& openblas.log
 else
- make $FORCETARGET  LAPACK_FPFLAGS="$LAPACK_FPFLAGS_VAL"  INTERFACE64="$sixty4_int" BINARY="$binary" NUM_THREADS=128 NO_CBLAS=1 NO_LAPACKE=1 DEBUG=0 USE_THREAD="$THREADOPT" libs netlib -j4 >& openblas.log
+ make FC=$FC $FORCETARGET  LAPACK_FPFLAGS="$LAPACK_FPFLAGS_VAL"  INTERFACE64="$sixty4_int" BINARY="$binary" NUM_THREADS=128 NO_CBLAS=1 NO_LAPACKE=1 DEBUG=0 USE_THREAD="$THREADOPT"  libs netlib -j4 >& openblas.log
 fi
 if [[ "$?" != "0" ]]; then
     tail -500 openblas.log
@@ -206,6 +235,7 @@ if [[ "$?" != "0" ]]; then
 fi
 
 mkdir -p ../../lib
+strip --strip-debug libopenblas*-*.a
 cp libopenblas.a ../../lib/libnwc_openblas.a
 #make PREFIX=. install
 if [[  ! -z "${NWCHEM_USE_OPENMP}" ]]; then

src/libext/scalapack/build_scalapa.sh

@@ -30,7 +30,17 @@ if [[  -z "${NWCHEM_TOP}" ]]; then
 fi
 if [[ ! -z "${BUILD_MPICH}" ]]; then
     export PATH=${NWCHEM_TOP}/src/libext/bin:$PATH
-    export LDFLAGS=`pkg-config --libs-only-L hwloc`
+    if [ -x "$(command -v pkg-config1)" ]; then
+        export LDFLAGS=`pkg-config --libs-only-L hwloc`
+    else
+	if [ -x "$(command -v brew)" ]; then
+	    export LDFLAGS=-L`brew --prefix`/lib/
+	else
+	    echo 'WARNING: cannot guess the location of the hwloc library'
+#	    exit 1
+	fi
+    fi
+    echo LDFLAGS for hwloc is $LDFLAGS
 fi
 if [[ "$FC" = "ftn"  ]] || [[ ! -z "$USE_CMAKE_MASTER" ]] ; then
     get_cmake_master
@@ -110,7 +120,13 @@ if [[ -f "scalapack-$COMMIT.tar.gz" ]]; then
 else
     echo "downloading"  "scalapack-$COMMIT.tar.gz"
     rm -f scalapack-$COMMIT.tar.gz
-    curl -L https://github.com/Reference-ScaLAPACK/scalapack/archive/$COMMIT.tar.gz -o scalapack-$COMMIT.tar.gz
+    tries=1 ; until [ "$tries" -ge 6 ] ; do
+		  if [ "$tries" -gt 1 ]; then sleep 9; echo attempt no.  $tries ; fi
+		  curl -L https://github.com/Reference-ScaLAPACK/scalapack/archive/$COMMIT.tar.gz -o scalapack-$COMMIT.tar.gz
+		  # check tar.gz integrity
+		  gzip -t scalapack-$COMMIT.tar.gz >&  /dev/null
+		  if [ $? -eq 0 ]; then break ;  fi
+		  tries=$((tries+1)) ;  done
 fi
 tar xzf scalapack-$COMMIT.tar.gz
 ln -sf scalapack-$COMMIT scalapack
@@ -193,6 +209,9 @@ echo MPICH_CC is "$MPICH_CC"
 if [[  -z "$I_MPI_F90"   ]] ; then
     export I_MPI_F90="$FC"
 fi
+if [[  -z "$I_MPI_CC"   ]] ; then
+    export I_MPI_CC="$CC"
+fi
 echo I_MPI_F90 is "$I_MPI_F90"
 if [[  -z "$PE_ENV"   ]] ; then
     #check if mpif90 and FC are consistent
@@ -282,6 +301,7 @@ if [[ "$?" != "0" ]]; then
     exit 1
 fi
 mkdir -p ../../../lib
+strip --strip-debug lib/libscalapack.a
 cp lib/libscalapack.a ../../../lib/libnwc_scalapack.a
 if [[ "$KNL_SWAP" == "1" ]]; then
     module swap  craype-haswell craype-mic-knl

src/libext/tblite/build_tblite_cmake.sh

@@ -219,5 +219,5 @@ fi
 $CMAKE --install _build
 
 cd ..
-
+strip --strip-debug ../lib/libtblite.a
 ln -sf  ../lib/libtblite.a  ../lib/libnwc_tblite.a

src/mp2_grad/mp2_back_transform.F

@@ -55,7 +55,8 @@ c
       integer l_c_t, k_c_t
       integer l_act, k_act, l_actsh, k_actsh
       integer twopdmunit
-      integer junk, ninseg, ierr, i, j, ish, ishlo, ishhi, shmax, tdim
+      integer junk(2), ninseg, ierr, i, j, ish, ishlo, ishhi, shmax,
+     t     tdim
       integer nsh, natoms, nactive, nblock
 c
       integer k_scr, l_scr, k_lab, l_lab, k_eri, l_eri, leneri, lenscr
@@ -482,10 +483,6 @@ c
          else
             call ga_copy(g_buf_trans,g_buf)
          endif
-cc      call ga_print(g_buf_trans)
-cc      call ga_print(g_buf)
-c               call ga_sync()
-c         call ga_mask_sync(.true.,.false.)
          call ga_sync()
 c     
 c     Now have locally (a,1:nbfpairlocal).  Loop thru local shell pairs
@@ -524,7 +521,6 @@ c
             ptr = ptr + count*(oseg_hi-oseg_lo+1)
          end do
 c     
-cedo         call ga_mask_sync(.true.,.false.)
          call ga_sync
 c     
       end do                    ! End of i
@@ -1158,27 +1154,12 @@ c     check for boundaries of buff()
                write(6,*) ga_nodeid(),' i_1.gt.ihi ',i_1,ihi
                call errquit(' mp2copybck: i boundary passed ',0,0)
             endif
-#if 0
-            call ga_get(g_a_trans,i_0,i_1,
-     A           jlo,jhi,buff(i_0,jlo),i_ld)
-#else
             call ga_get(g_a_trans,i_0,i_1,
      A           jlo,jhi,buff,i_ld)
             call ga_put(g_a, i_0, i_1, jlo, jhi, buff, i_ld)
-#endif
          enddo
       endif
 c     everybody needs to have the buffer filled
       call ga_sync()
-cold      if(dowork) call ga_put(g_a, ilo, ihi, jlo, jhi, buff, ihi-ilo+1)
-      call ga_sync()
-#if 0
-      call ga_sync()
-      tr=ga_ddot(g_a,g_a)
-      tr_tr=ga_ddot(g_a_trans,g_a_trans)
-      if(abs(tr-tr_tr).gt.1d-6)then
-         if(ga_nodeid().eq.0)write(6,*) 'tr ',tr,' tr_tr ',tr_tr
-      endif
-#endif
       return
       end

src/mp2_grad/mp2_energy.F

@@ -717,6 +717,10 @@ c
             endif
          endif
          call pstat_on(ps_laifock)
+         if (doprint) then
+            write(6,111) ' Start laifock', util_wallsec()
+            call util_flush(6)
+         end if
          call mp2_lai_fock_uhf_prepar(
      $        g_p_ab_a,g_p_ab_b,g_p_ij_a,g_p_ij_b,
      $        noa,nob,nva,nvb,g_tmpa, g_tmpb,nmo)
@@ -725,6 +729,10 @@ c
      $        noa,nob,nva,nvb,g_lai_a,g_lai_b,rtdb,tol2e_fock)
          call mp2_lai_fock_uhf_tidy(g_tmpa, g_tmpb)
          call pstat_off(ps_laifock)
+         if (doprint) then
+            write(6,111) ' Done laifock', util_wallsec()
+            call util_flush(6)
+         end if
 c
          if (omatrix) then
             call ga_print(g_lai_a)
@@ -782,11 +790,20 @@ c
          if(.not.ga_create(mt_dbl,nmo,nmo,'tot beta dens',nmo,0,
      $        g_p_tot_b))call errquit(
      $        'mp2_energy: ga_create failed',0, GA_ERR)
-c     
+c
+         if (doprint) then
+            write(6,111) ' Start make_tot_dens', util_wallsec()
+            call util_flush(6)
+         end if
+         
          call mp2_make_tot_dens(g_p_ij_a,g_p_ij_b,g_p_ab_a,g_p_ab_b,
      $        g_rhs,g_p_tot_a,g_p_tot_b,noa,nva,nob,nvb,nbf,nmo,
      $        int_mb(k_map_a), int_mb(k_map_b),
      $        g_vecs_a,g_vecs_b,basis,scftype)
+         if (doprint) then
+            write(6,111) ' Done make_tot_dens', util_wallsec()
+            call util_flush(6)
+         end if
 c
          if (otrace) then
             call mp2_print_trace_norm(g_p_tot_a,'g_p_tot_a',nmo)
@@ -931,6 +948,7 @@ c
       endif
 c     
       call ga_sync
+ 111  format(a,' at ', f10.1,'s')
 c     
       end
       subroutine mp2_print_trace_norm(g, text, n)

src/mp2_grad/mp2_grad_doit.F

@@ -333,7 +333,7 @@ c
 
          end if
          call pstat_on(ps_maket)
-         call mp2_make_tuhf(nbf,noa_lo,noa_hi,nva_lo,nva_hi,
+         call mp2_make_tuhf(rtdb,nbf,noa_lo,noa_hi,nva_lo,nva_hi,
      $        sym_lo_oa,sym_hi_oa,sym_lo_va,sym_hi_va,num_oa,num_va,
      $        nob_lo,nob_hi,nvb_lo,nvb_hi,sym_lo_ob,sym_hi_ob,
      $        sym_lo_vb,sym_hi_vb,num_ob,num_vb,oseg_lo,oseg_hi,

src/mp2_grad/mp2_make_dip_mom.F

@@ -22,6 +22,8 @@ c
       double precision mp2y, scfy
       double precision mp2z, scfz
       double precision ndip(3)
+      logical ga_create_atom_blocked
+      external ga_create_atom_blocked
       integer geom
 c
       if (.not. bas_cando_mpoles(basis)) return
@@ -31,15 +33,9 @@ c
 c
 c     Generate dipole integrals
 c
-*ga:1:0
-      if(.not.ga_create(mt_dbl,nbf,nbf,' ao integs x ',
-     $     0,0,g_x)) call errquit('mp2_dipole: ga',0, GA_ERR)
-*ga:1:0
-      if(.not.ga_create(mt_dbl,nbf,nbf,' ao integs y ',
-     $     0,0,g_y)) call errquit('mp2_dipole: ga',0, GA_ERR)
-*ga:1:0
-      if(.not.ga_create(mt_dbl,nbf,nbf,' ao integs z',
-     $     0,0,g_z)) call errquit('mp2_dipole: ga',0, GA_ERR)
+      g_x = ga_create_atom_blocked(geom, basis,'ao integs x')
+      g_y = ga_create_atom_blocked(geom, basis,'ao integs y')
+      g_z = ga_create_atom_blocked(geom, basis,'ao integs z')
 c
       call int_init(rtdb,1,basis)
       call int_dip_ga(basis, basis, g_x, g_y, g_z)
@@ -48,8 +44,7 @@ c
 c     Generate SCF AO density
 c
 *ga:1:0
-      if(.not.ga_create(mt_dbl, nbf, nbf,' AO density  ',
-     $     0,0,g_dens)) call errquit('mp2_dipole: ga',0, GA_ERR)
+      g_dens = ga_create_atom_blocked(geom, basis,'ao density')
       call ga_dgemm('n', 't', nbf, nbf, noa, 1.0d0, g_vecs_a,
      $     g_vecs_a, 0.0d0, g_dens)
       call ga_dgemm('n', 't', nbf, nbf, nob, 1.0d0, g_vecs_b,
@@ -62,8 +57,7 @@ c
 c     Form the MP2 AO density
 c
 *ga:1:0
-      if(.not.ga_create(mt_dbl,nmo,nbf,'scratch',0,0,g_scratch))
-     $     call errquit('mp2_grad: scratch',1, GA_ERR)
+      g_scratch = ga_create_atom_blocked(geom, basis,'scratch')
       call ga_dgemm('n','t',nmo,nbf,nmo,1.0d0,g_p_tot_a,g_vecs_a,
      $     0.0d0,g_scratch)
       call ga_dgemm('n','n',nbf,nbf,nmo,1.0d0,g_vecs_a,g_scratch,

src/mp2_grad/mp2_make_tot_dens.F

@@ -26,6 +26,8 @@
       character*256 p_a_ao
       character*256 p_b_ao
       integer base
+      integer l_buf,k_buf
+      integer ilo,ihi,jlo,jhi,dowork
 c     
       call util_file_name('p_a_ao',.true.,.true.,p_a_ao)
       call util_file_name('p_b_ao',.true.,.true.,p_b_ao)
@@ -48,22 +50,48 @@ c$$$  call ga_put(g_p_tot_a,i,i,a+noa,a+noa,temp,1)
 c$$$  end do
 c$$$  end do
 c     
+      call ga_distribution(g_p_tot_a, ga_nodeid(), 
+     T     ilo, ihi, jlo, jhi)
+c     check if distrib by columns
+      if ((ilo.eq.0).and.(ihi.eq.-1))  then
+         dowork=0
+      else
+         if(ilo.ne.1.and.ihi.ne.nmo) call
+     R        errquit(' mp2maketotdens bug ',0,0)
+         dowork=1
+      endif
       ioff = 0
-      do i = 1+ga_nodeid(),noa,ga_nnodes() ! Loop thru SCF order
+      if(dowork.eq.1) then
+      do i = 1,noa
+         ii = map_a(i)
+         if(ii.ge.jlo.and.ii.le.jhi) then
          do a = 1,nva
-            ii = map_a(i)
             aa = map_a(a+noa)-noa
             ia = ioff+a+(i-1)*nva
             call ga_get(g_rhs,ia,ia,1,1,xx,1)
             call ga_put(g_p_tot_a,aa+noa,aa+noa,ii,ii,xx,1)
             call ga_put(g_p_tot_a,ii,ii,aa+noa,aa+noa,xx,1)
          end do
+         endif
       end do
-c     
-      call ga_copy_patch('n',g_p_ij_a,1,noa,1,noa,g_p_tot_a,
-     $     1,noa,1,noa)
-      call ga_copy_patch('n',g_p_ab_a,1,nva,1,nva,g_p_tot_a,
-     $     noa+1,nmo,noa+1,nmo)
+      if (.not. ma_push_get(mt_dbl, nmo, 'mp2: maket',
+     $     l_buf, k_buf)) call errquit
+     $     ('mp2: insufficient memory : bug ', nmo, MA_ERR)
+
+      do i=jlo,jhi
+         if(i.le.noa) then
+            call ga_get(g_p_ij_a,1,noa,i,i,dbl_mb(k_buf),noa)
+            call ga_put(g_p_tot_a,1,noa,i,i,dbl_mb(k_buf),noa)
+         else
+            call ga_get(g_p_ab_a,1,nva,i-noa,i-noa,dbl_mb(k_buf),nva)
+            call ga_put(g_p_tot_a,noa+1,nmo,i,i,dbl_mb(k_buf),nva)
+         endif
+      enddo
+      if (.not. ma_pop_stack(l_buf)) call errquit
+     $     ('mp2: pop_stack failed ', nmo, MA_ERR)
+      endif
+      call ga_mask_sync(.true.,.false.)
+      call ga_sync()
 c     
       if (scftype .eq. 'UHF') then
          base = (nva*noa)
@@ -71,15 +99,18 @@ c
          base = 0               ! Alpha is the same as beta
       end if
       ioff = base
-      do i = 1+ga_nodeid(),nob,ga_nnodes() ! Loop thru SCF order
+      if (dowork.eq.1) then
+      do i = 1,nob
+         ii = map_b(i)
+         if(ii.ge.jlo.and.ii.le.jhi) then
          do a = 1,nvb
-            ii = map_b(i)
             aa = map_b(a+nob)-nob
             ia = ioff+a+(i-1)*nvb
             call ga_get(g_rhs,ia,ia,1,1,xx,1)
             call ga_put(g_p_tot_b,aa+nob,aa+nob,ii,ii,xx,1)
             call ga_put(g_p_tot_b,ii,ii,aa+nob,aa+nob,xx,1)
          end do
+         endif
       end do
 c$$$  call ga_copy_patch('n',g_rhs,base+1,base+(nvb*nob),
 c$$$  $     1,1,g_p_tot_b,nob+1,nmo,1,nob)
@@ -90,11 +121,25 @@ c$$$  call ga_get(g_rhs,indx,indx,1,1,temp,1)
 c$$$  call ga_put(g_p_tot_b,i,i,a+nob,a+nob,temp,1)
 c$$$  end do
 c$$$  end do
-c     
-      call ga_copy_patch('n',g_p_ij_b,1,nob,1,nob,g_p_tot_b,
-     $     1,nob,1,nob)
-      call ga_copy_patch('n',g_p_ab_b,1,nvb,1,nvb,g_p_tot_b,
-     $     nob+1,nmo,nob+1,nmo)
+c
+      if (.not. ma_push_get(mt_dbl, nmo, 'mp2: maket',
+     $     l_buf, k_buf)) call errquit
+     $     ('mp2: insufficient memory : bug ', nmo, MA_ERR)
+
+      do i=jlo,jhi
+         if(i.le.nob) then
+            call ga_get(g_p_ij_b,1,nob,i,i,dbl_mb(k_buf),nob)
+            call ga_put(g_p_tot_b,1,nob,i,i,dbl_mb(k_buf),nob)
+         else
+            call ga_get(g_p_ab_b,1,nvb,i-nob,i-nob,dbl_mb(k_buf),nvb)
+            call ga_put(g_p_tot_b,nob+1,nmo,i,i,dbl_mb(k_buf),nvb)
+         endif
+      enddo
+      if (.not. ma_pop_stack(l_buf)) call errquit
+     $     ('mp2: pop_stack failed ', nmo, MA_ERR)
+      
+      endif
+      call ga_sync()
 c     
 *ga:1:0
       if(.not.ga_create(mt_dbl,nmo,nbf,'scratch',0,0,g_scratch))

src/mp2_grad/mp2_make_tot_wdens.F

@@ -46,7 +46,7 @@ c
       call ga_copy_patch('n',g_w_ai_a,1,noa_hi,1,nva,g_w_tot_a,
      $     1, noa_hi, noa_hi+1, nmo)
 c     
-      call ga_transpose(g_w_ai_a,g_w_ai_a_tr)
+      call mp2_transpwai(g_w_ai_a,g_w_ai_a_tr)
       call ga_copy_patch('n',g_w_ai_a_tr,1,nva,1,noa_hi,g_w_tot_a,
      $     noa_hi+1,nmo,1,noa_hi)
       call ga_copy_patch('n',g_w_ij_a,1,noa_hi,1,noa_hi,g_w_tot_a,
@@ -55,7 +55,7 @@ c
      $     noa_hi+1,nmo,noa_hi+1,nmo)
       call ga_copy_patch('n',g_w_ai_b,1,nob_hi,1,nvb,g_w_tot_b,
      $     1, nob_hi, nob_hi+1,nmo)
-      call ga_transpose(g_w_ai_b,g_w_ai_b_tr)
+      call mp2_transpwai(g_w_ai_b,g_w_ai_b_tr)
       call ga_copy_patch('n',g_w_ai_b_tr,1,nvb,1,nob_hi,g_w_tot_b,
      $     nob_hi+1,nmo,1,nob_hi)
       call ga_copy_patch('n',g_w_ij_b,1,nob_hi,1,nob_hi,g_w_tot_b,
@@ -148,3 +148,44 @@ c
 c
       end
 
+      subroutine mp2_transpwai(g,g_tr)
+      implicit none
+#include "mafdecls.fh"
+#include "global.fh"
+#include "errquit.fh"
+      integer g,g_tr
+c
+      integer ilo,ihi,jlo,jhi
+      integer i,j
+      integer k_buf,l_buf
+      integer n_row,n_col,gtype
+      integer dowork
+c
+      call ga_inquire(g_tr, gtype, n_row, n_col)
+      call ga_distribution(g_tr, ga_nodeid(), 
+     T     ilo, ihi, jlo, jhi)
+      if ((ilo.eq.0).and.(ihi.eq.-1))  then
+         dowork=0
+      else
+c     check if distrib by columns
+         if(ilo.ne.1.and.ihi.ne.n_row) call
+     R        errquit(' mp2transpwai bug ',0,0)
+         dowork=1
+      endif
+      if(dowork.eq.1) then
+         if (.not. ma_push_get(mt_dbl, n_row, 'mp2: wdens',
+     $        l_buf, k_buf)) call errquit
+     $        ('mp2: insufficient memory : bug ', n_row, MA_ERR)
+      do j=jlo,jhi
+         call ga_get(g,j,j,1,n_row,dbl_mb(k_buf),1)
+          if (.not.ma_verify_allocator_stuff()) stop ' ma ck 2b'
+          call ga_put(g_tr,1,n_row,j,j,dbl_mb(k_buf),n_row)
+           if (.not.ma_verify_allocator_stuff()) stop ' ma ck 2b'
+      enddo
+      if (.not. ma_pop_stack(l_buf)) call errquit
+     $     ('mp2transp pop_stack failed ', n_row, MA_ERR)
+       
+      endif
+      call ga_sync()
+      return
+      end

src/mp2_grad/mp2_make_tuhf.F

@@ -1,4 +1,4 @@
-      subroutine mp2_make_tuhf(nbf,noa_lo,noa_hi,nva_lo,nva_hi,
+      subroutine mp2_make_tuhf(rtdb,nbf,noa_lo,noa_hi,nva_lo,nva_hi,
      $     sym_lo_oa,sym_hi_oa,sym_lo_va,sym_hi_va,num_oa,num_va,
      $     nob_lo,nob_hi,nvb_lo,nvb_hi, sym_lo_ob,sym_hi_ob,
      $     sym_lo_vb,sym_hi_vb,num_ob,num_vb,   
@@ -16,6 +16,7 @@
 #include "rtdb.fh"
 #include "eaf.fh"
 #include "util.fh"
+      integer rtdb
       character*8 task
       integer nbf,i
       integer noa_lo,noa_hi,nva_lo,nva_hi
@@ -49,6 +50,12 @@ c
       parameter (eps=1d-14)
 c
       logical otdebug
+      logical do_mirra,do_mirrb,do_mirr
+      integer g_ta_mirr,g_tb_mirr,g_ta_org,g_tb_org
+      logical util_mirrmat,tuhfreplicated
+      external util_mirrmat
+      integer util_mirrmem,mem_mirr
+      external util_mirrmem
 c
       integer l_ia_uv, k_ia_uv, l_tmp, k_tmp, l_ia_jb, k_ia_jb
       double precision ddot
@@ -82,6 +89,48 @@ c
 *ga:1:0
       if(.not.ga_create(mt_dbl,nob*nvb,nva,'T_b',(nob*nvb),0,g_t_b))
      $     call errquit('could not alloc ga for T_b',1, GA_ERR)
+      if (.not.rtdb_get(rtdb, 'mp2:tuhfreplicated', mt_log, 1, 
+     T     tuhfreplicated)) tuhfreplicated=.false.
+      do_mirr=.false.
+      do_mirra=.false.
+      do_mirrb=.false.
+      if(tuhfreplicated) then
+         mem_mirr=
+     M        util_mirrmem(1,g_t_a)+
+     M        util_mirrmem(1,g_t_b)
+         if(ga_nodeid().eq.0) then
+            write(6,14) ' needed by mirroring: ', mem_mirr
+            write(6,14) ' ga_memory_avail()  : ', ga_memory_avail()
+         endif
+ 14      format(a,i10,' bytes')
+         if(ga_memory_avail().gt. mem_mirr) then
+            do_mirra=util_mirrmat(1,g_t_a,g_ta_mirr,
+     D           .false., .true.)
+            do_mirrb=util_mirrmat(1,g_t_b,g_tb_mirr,
+     D           .false., .true.)
+            else
+               if(ga_nodeid().eq.0)
+     W              write(6,*) ' insuff. memory  for mirroring '
+            endif
+      endif
+      if(do_mirra.and.do_mirrb) then
+         if(ga_nodeid().eq.0)
+     w        write(6,*) 'mp2_make_tuhf mirroring'
+         do_mirr=.true.
+         g_ta_org=g_t_a
+         g_tb_org=g_t_b
+         g_t_a=g_ta_mirr
+         g_t_b=g_tb_mirr
+      else
+         if(do_mirra) then
+            if(.not.ga_destroy(g_ta_mirr))call errquit(
+     $           'could not destroy mirrt_a handle',1, GA_ERR)
+            write(6,10) ga_nodeid(),'destroyed g_ta_mirr',g_ta_mirr
+ 10         format(i5,a,i5)
+         endif
+         g_ta_org=0
+         g_tb_org=0
+      endif
       tunitptra=1
       tunitptrb=1
       do i=oseg_lo,oseg_hi 
@@ -189,6 +238,12 @@ c
             enddo
          enddo
          call ga_sync()
+         if (do_mirr) then
+            call ga_mask_sync(.true.,.false.)
+            call ga_merge_mirrored(g_t_a)
+            call ga_mask_sync(.true.,.false.)
+            call ga_merge_mirrored(g_t_b)
+         endif
          do a_a=nva_lo_local,nva_hi_local
             counta = 0
             do b=nva_lo,nva_hi
@@ -265,8 +320,21 @@ c
       esing = esing + (-4.0d0*eab + eaa)/4.0d0
       etrip = etrip + (-3.0d0*eaa)/4.0d0
 c
-      if(.not.ga_destroy(g_t_b))call errquit(
-     $     'could not destroy t_b handle',1, GA_ERR)
+      if(do_mirr) then
+         if(.not.ga_destroy(g_t_b))call errquit(
+     $        'could not destroy mirr t_b handle',1, GA_ERR)
+         if(.not.ga_destroy(g_t_a))call errquit(
+     $        'could not destroy mirr t_a handle',1, GA_ERR)
+             g_t_a=g_ta_org
+             g_t_b=g_tb_org
+       endif
+
+       if(.not.ga_destroy(g_t_b))then
+          if(ga_nodeid().eq.0) call ga_summarize(1)
+          call ga_sync()
+        call errquit(
+     $         'could not destroy t_b handle',1, GA_ERR)
+        endif
       if(.not.ga_destroy(g_t_a))call errquit(
      $     'could not destroy t_a handle',1, GA_ERR)
 c

src/nwchem.F

@@ -30,7 +30,7 @@ c
 c  $Id$
 c
 c  ======================================================================================================
-C> \mainpage Northwest Computational Chemistry Package (NWChem) 7.0.1
+C> \mainpage Northwest Computational Chemistry Package (NWChem) 7.2.0
 C>
 C> NWChem is an open-source computational chemistry package distributed under the terms of
 C> the Educational Community License (ECL) 2.0
@@ -461,7 +461,7 @@ c
       write(LuOut,*)
       write(LuOut,*)
       call util_print_centered(LuOut,
-     $     'Northwest Computational Chemistry Package (NWChem) 7.0.1',
+     $     'Northwest Computational Chemistry Package (NWChem) 7.2.0',
      $     40, .true.)
       write(LuOut,*)
       write(LuOut,*)
@@ -557,7 +557,7 @@ C
 C  Print version information to ecce.out file
 C
       compiled(inp_strlen(compiled)+1:inp_strlen(compiled)+25) = 
-     $         ' Version 7.0.1'
+     $         ' Version 7.2.0'
       call ecce_print_version(compiled(1:inp_strlen(compiled)))
 C
       If ( Inp_StrLen(BatchID) .gt. 0) then

src/nwdft/grid/grid_acc_def.F

@@ -25,7 +25,8 @@ c
       double precision grid_acc(num_gridtypes)
       double precision job_grid_acc
       character*16 element
-      character*16 tag
+      character*16 tag, tag_mod
+      integer lenb
       double precision eps
       parameter (eps = 1.d-20)
       data gridtypes /'xcoarse', 'coarse', 'medium', 'fine', 'xfine',
@@ -418,7 +419,10 @@ c
 c hack for nbo
             if(tag(3:4).eq.'gh') iptr=5
             if(tag(1:1).eq.'X'.or.tag(1:1).eq.'x') iptr=2
-            if (.not. geom_tag_to_element(tag(iptr:), symbol, 
+            lenb=inp_strlen(tag)
+            tag_mod=' '
+            tag_mod(1:lenb-iptr+1)=tag(iptr:lenb)
+            if (.not. geom_tag_to_element(tag_mod, symbol, 
      &           element, i_atomic_number)) then
                if (inp_compare(.false.,tag(1:2),'bq')) then 
                   i_atomic_number = 0

src/nwdft/grid/grid_atom_type_info.F

@@ -26,7 +26,8 @@ c
       logical same_atom, same_bq, isbq
 c
       character*16 element
-      character*16 tag
+      character*16 tag,tag_mod
+      integer lenb
       character*2 symbol
 c
       logical lnewtype
@@ -179,7 +180,10 @@ c           Hence they should always have i_atomic_number .eq. 0
 c           no matter what follows the X.
 c
             if(tag(1:1).eq.'X'.or.tag(1:1).eq.'x') iptr=2
-            if (.not. geom_tag_to_element(tag(iptr:), symbol, 
+            lenb=inp_strlen(tag)
+            tag_mod=' '
+            tag_mod(1:lenb-iptr+1)=tag(iptr:lenb)
+            if (.not. geom_tag_to_element(tag_mod, symbol, 
      &           element, i_atomic_number)) then
                if (inp_compare(.false.,tag(1:2),'bq')) then 
                   if(bqdontcare) then

src/nwdft/input_dft/dft_inpana.F

@@ -212,7 +212,7 @@ c
 c     
 c     write noc (consistent with definition in ddscf) to rtdb
 c     
-      if (.not. rtdb_put(rtdb, 'dft:noc', mt_int, 2, noc))
+      if (.not. rtdb_put(rtdb, 'dft:noc', mt_int, 2, noc(1)))
      &   call errquit('inpana: rtdb_put of noc failed', 0, RTDB_ERR)
 cHvD
       use_nwxc = util_module_avail("nwxc")
@@ -1049,3 +1049,115 @@ c
       endif
       return
       end
+      subroutine dft_sethfx(rtdb)
+      implicit none
+#include "geom.fh"
+#include "rtdb.fh"
+#include "mafdecls.fh"
+#include "cdft.fh"
+#include "errquit.fh"
+      integer rtdb
+c
+      character*10 pname
+      integer i,nopen
+      double precision thresh,tol2e
+      pname='dft_sethfx'
+c     multiplicity
+      if(rtdb_get(rtdb, 'scf:nopen', MT_INT, 1, nopen)) then
+         mult=nopen+1
+         ipol=2
+         if (.not. rtdb_put(rtdb, 'dft:ipol', mt_int, 1, ipol))
+     &   call errquit(pname//': rtdb_put failed', 0, RTDB_ERR)
+         if (.not. rtdb_put(rtdb, 'dft:mult', mt_int, 1, mult))
+     &   call errquit(pname//': rtdb_put failed', 1, RTDB_ERR)
+      endif
+      do i = 1, numfunc
+         xfac(i)=0.d0
+         cfac(i)=0.d0
+         lcfac(i)=.false.
+         nlcfac(i)=.false.
+         lxfac(i)=.false.
+         nlxfac(i)=.false.
+         xccomb(i)=.false.
+      enddo
+      lxfac(1) = .true.
+      nlxfac(1) = .true.
+      xfac(1) = 1.0d0
+      if (.not. rtdb_put(rtdb, 'dft:cfac',
+     &     mt_dbl, numfunc, cfac(1)))
+     &     call errquit(pname//': rtdb_put failed', 2, RTDB_ERR)
+      if (.not. rtdb_put(rtdb, 'dft:xfac',
+     &      mt_dbl, numfunc, xfac(1)))
+     &     call errquit(pname//': rtdb_put failed', 3, RTDB_ERR)
+      if (.not. rtdb_put(rtdb, 'dft:lcfac',
+     &     mt_log, numfunc, lcfac(1)))
+     &     call errquit(pname//': rtdb_put failed', 4, RTDB_ERR)
+      if (.not. rtdb_put(rtdb, 'dft:nlcfac',
+     &     mt_log, numfunc, nlcfac(1)))
+     &     call errquit(pname//': rtdb_put failed', 5, RTDB_ERR)
+      if (.not. rtdb_put(rtdb, 'dft:lxfac',
+     &     mt_log, numfunc, lxfac(1)))
+     &     call errquit(pname//': rtdb_put failed', 6, RTDB_ERR)
+      if (.not. rtdb_put(rtdb, 'dft:nlxfac',
+     &    mt_log, numfunc, nlxfac(1)))
+     &     call errquit(pname//': rtdb_put failed', 7, RTDB_ERR)
+      if (.not. rtdb_cput(rtdb, 'dft:theory', 1, 'dft'))
+     $        call errquit('rtdb_cput failed',0, RTDB_ERR)
+      if (.not.rtdb_put(rtdb,'dft:e_conv',mt_dbl,1,1d-8))
+     .     call errquit(pname//': rtdbput econv failed',0, RTDB_ERR)
+      if (.not.rtdb_get(rtdb, 'scf:thresh', MT_DBL, 1, thresh))
+     &     thresh=1d-4
+      d_conv=thresh*thresh
+cdbg      write(6,*) ' d_conv is ',d_conv
+      if(.not. rtdb_put(rtdb, 'dft:d_conv', mt_dbl, 1, d_conv))
+     &      call errquit('dft_rdinput: rtdb_put failed', 106, RTDB_ERR)
+      if (.not. rtdb_get(rtdb, 'scf:tol2e', MT_DBL, 1, tol2e))
+     *     tol2e=1d-12
+      itol2e=-log10(tol2e)
+cdbg      write(6,*) ' itol2e is ',itol2e
+      if (.not. rtdb_put(rtdb, 'dft:itol2e', mt_int, 1, 12))
+     .     call errquit(pname//': rtdbput itol2e failed',0, RTDB_ERR)
+      return
+      end
+      subroutine dft_sethfx_cleanup(rtdb)
+      implicit none
+#include "geom.fh"
+#include "bas.fh"
+#include "rtdb.fh"
+#include "mafdecls.fh"
+#include "cdft.fh"
+#include "errquit.fh"
+      integer level_org
+      common /dftg/ level_org
+      integer rtdb
+      character*18 pname
+      integer i
+      pname='dft_sethfx_cleanup'
+      call util_print_set_level(level_org)
+      return
+      end
+      logical function dft_guesshfx(rtdb)
+      implicit none
+#include "rtdb.fh"
+#include "mafdecls.fh"
+#include "cdft.fh"
+#include "errquit.fh"
+      integer level_org
+      common /dftg/ level_org
+      integer rtdb
+c
+      logical dftguess
+c      
+      dft_guesshfx=.false.
+      if(.not.rtdb_get(rtdb,'dftguess',mt_log,1,dftguess)) 
+     D     dftguess=.false.
+      if(dftguess) then
+         if(.not.rtdb_delete(rtdb,'dftguess'))
+     D        call errquit(' rtdb_delete failed',0,RTDB_ERR)
+         call dft_sethfx(rtdb)
+         dft_guesshfx=.true.
+         call util_print_get_level(level_org)
+         call util_print_set_level(10)
+      endif
+      return
+      end

src/nwdft/libxc/GNUmakefile

@@ -12,12 +12,20 @@ LIBRARY = libnwdft.a
 
 LIB_INCLUDES = -I../include 
 ifdef USE_LIBXC
+LIB_DEFINES += -DXC_MAJOR_VERSION=$(shell grep XC_MAJOR_VERSION ../../libext/libxc/install/include/xc_version.h |cut -d " " -f 3)
+LIB_DEFINES += -DLIBXC_NP_SIZE=$(shell ./libxc_findsizenp.sh ../../libext/libxc/install/include)
 LIB_INCLUDES += -I../../libext/libxc/install/include
 endif
-ifdef LIBXC_DIR
-LIB_INCLUDES += -I$(LIBXC_DIR)/include
+ifdef LIBXC_INCLUDE
+LIB_DEFINES += -DXC_MAJOR_VERSION=$(shell grep XC_MAJOR_VERSION $(LIBXC_INCLUDE)/xc_version.h |cut -d " " -f 3)
+LIB_DEFINES += -DLIBXC_NP_SIZE=$(shell ./libxc_findsizenp.sh $(LIBXC_INCLUDE))
+LIB_INCLUDES += -I$(LIBXC_INCLUDE)
+endif
+ifdef LIBXC_MODDIR
+LIB_INCLUDES += -I$(LIBXC_MODDIR)
 endif
 LIB_INCLUDES += -I.
 
+
 include ../../config/makefile.h
 include ../../config/makelib.h

src/nwdft/libxc/libxc.fh

@@ -1,3 +1,8 @@
+#ifdef USE_LIBXC
+#if XC_MAJOR_VERSION < 4
+#error "need libxc version > 3 "
+#endif
+#endif
       integer, parameter :: maxfunc = 100
 
       integer :: libxc_nfuncs

src/nwdft/libxc/libxc_findsizenp.sh

@@ -0,0 +1,18 @@
+#!/usr/bin/env bash
+npstring=$(egrep np "$1"/xc.h|head -1)
+np_array=($npstring)
+count=-1
+sizenp_string=' '
+for substring in "${np_array[@]}"; do
+    [[ $substring == "np," ]] && sizenp_string=${np_array[$count]} && break
+    ((++count))
+done
+if [[ "$sizenp_string" == 'int' ]]; then
+    size_np=4	       	       
+elif [[ "$sizenp_string" == 'size_t' ]]; then
+    size_np=8
+else
+    echo "unexpected sizenp_string $sizenp_string"
+    exit 1
+fi
+echo $size_np

src/nwdft/libxc/nwchem_libxc_compute.F

@@ -57,14 +57,20 @@
       double precision :: Amat3(nq,NCOL_AMAT3)
       double precision :: Cmat3(nq,NCOL_CMAT3)
 
+#ifdef USE_LIBXC
       double precision :: fac 
 
       double precision, external :: ddot
       logical gga,mgga,dolap
       logical,external :: nwchem_libxc_family
 
+#if LIBXC_NP_SIZE == 4
+      integer*4 nqs
+#elif LIBXC_NP_SIZE == 8
       integer(c_size_t) :: nqs
-#ifdef USE_LIBXC      
+#else
+#error "unexpected value for LIBXC_NP_SIZE"
+#endif
       type(xc_f03_func_t) :: xcfunc      
 
       gga = .false.
@@ -154,22 +160,39 @@
 
         call xc_f03_func_init(xcfunc,libxc_funcs(ifunc),polarized)
         call xc_f03_func_set_dens_threshold(xcfunc, tol_rho)
-        call xc_f03_func_set_sigma_threshold(xcfunc, tol_rho**2)
+#if XC_MAJOR_VERSION > 4
+        call xc_f03_func_set_sigma_threshold(xcfunc, tol_rho)
         call xc_f03_func_set_zeta_threshold(xcfunc, 1d-10)
-
+#endif
         select case(libxc_family(ifunc))
+#if XC_MAJOR_VERSION > 4
         case (XC_FAMILY_LDA, XC_FAMILY_HYB_LDA)
-
+#else
+        case (XC_FAMILY_LDA)
+#endif
           if ((.not.do_2nd) .and. (.not.do_3rd)) then
             call xc_f03_lda_exc_vxc(xcfunc,nqs,
      $            dbl_mb(krho),dbl_mb(kexc),dbl_mb(kvrho))
           elseif (.not.do_3rd) then
+#if XC_MAJOR_VERSION > 4
             call xc_f03_lda_exc_vxc_fxc(xcfunc,nqs,dbl_mb(krho),
      $            dbl_mb(kexc),dbl_mb(kvrho),dbl_mb(kv2rho2))
+#else
+            call xc_f03_lda_exc_vxc(xcfunc,nqs,dbl_mb(krho),
+     $                              dbl_mb(kexc),dbl_mb(kvrho))
+            call xc_f03_lda_fxc(xcfunc,nqs,dbl_mb(krho),dbl_mb(kv2rho2))
+#endif
           else
+#if XC_MAJOR_VERSION > 4
             call xc_f03_lda_exc_vxc_fxc_kxc(xcfunc,nqs,dbl_mb(krho),
      $            dbl_mb(kexc),dbl_mb(kvrho),dbl_mb(kv2rho2),
      $            dbl_mb(kv3rho3))
+#else
+            call xc_f03_lda_exc_vxc(xcfunc,nqs,dbl_mb(krho),
+     $                              dbl_mb(kexc),dbl_mb(kvrho))
+            call xc_f03_lda_fxc(xcfunc,nqs,dbl_mb(krho),dbl_mb(kv2rho2))
+            call xc_f03_lda_kxc(xcfunc,nqs,dbl_mb(krho),dbl_mb(kv3rho3))
+#endif
           endif
 
 
@@ -186,17 +209,43 @@
      $                                dbl_mb(kexc),
      $                                dbl_mb(kvrho),dbl_mb(kvsigma))
             elseif (.not.do_3rd) then
+#if XC_MAJOR_VERSION > 4
               call xc_f03_gga_exc_vxc_fxc(xcfunc,nqs,
      $             dbl_mb(krho),dbl_mb(ksigma),dbl_mb(kexc),
      $             dbl_mb(kvrho),dbl_mb(kvsigma),dbl_mb(kv2rho2),
      $             dbl_mb(kv2rhosig),dbl_mb(kv2sig2))
+#else
+              call xc_f03_gga_exc_vxc(xcfunc,nqs,
+     $                                dbl_mb(krho),dbl_mb(ksigma),
+     $                                dbl_mb(kexc),
+     $                                dbl_mb(kvrho),dbl_mb(kvsigma))
+              call xc_f03_gga_fxc(xcfunc,nqs,
+     $                                dbl_mb(krho),dbl_mb(ksigma),
+     $                                dbl_mb(kv2rho2),dbl_mb(kv2rhosig),
+     $                                dbl_mb(kv2sig2))
+#endif
             else
+#if XC_MAJOR_VERSION > 4
               call xc_f03_gga_exc_vxc_fxc_kxc(xcfunc,nqs,
      $             dbl_mb(krho),dbl_mb(ksigma),dbl_mb(kexc),
      $             dbl_mb(kvrho),dbl_mb(kvsigma),dbl_mb(kv2rho2),
      $             dbl_mb(kv2rhosig),dbl_mb(kv2sig2),
      $             dbl_mb(kv3rho3),dbl_mb(kv3rho2sig),
      $             dbl_mb(kv3rhosig2),dbl_mb(kv3sig3))
+#else
+              call xc_f03_gga_exc_vxc(xcfunc,nqs,
+     $                                dbl_mb(krho),dbl_mb(ksigma),
+     $                                dbl_mb(kexc),
+     $                                dbl_mb(kvrho),dbl_mb(kvsigma))
+              call xc_f03_gga_fxc(xcfunc,nqs,
+     $                            dbl_mb(krho),dbl_mb(ksigma),
+     $                            dbl_mb(kv2rho2),dbl_mb(kv2rhosig),
+     $                            dbl_mb(kv2sig2))
+              call xc_f03_gga_kxc(xcfunc,nqs,
+     $                            dbl_mb(krho),dbl_mb(ksigma),
+     $                            dbl_mb(kv3rho3),dbl_mb(kv3rho2sig),
+     $                            dbl_mb(kv3rhosig2),dbl_mb(kv3sig3))
+#endif
             endif
           else
             if ((.not.do_2nd).and.(.not.do_3rd)) then
@@ -204,17 +253,41 @@
      $                            dbl_mb(krho),dbl_mb(ksigma),
      $                            dbl_mb(kvrho),dbl_mb(kvsigma))
             elseif (.not.do_3rd) then
+#if XC_MAJOR_VERSION > 4
               call xc_f03_gga_vxc_fxc(xcfunc,nqs,
      $             dbl_mb(krho),dbl_mb(ksigma),
      $             dbl_mb(kvrho),dbl_mb(kvsigma),dbl_mb(kv2rho2),
      $             dbl_mb(kv2rhosig),dbl_mb(kv2sig2))
+#else
+              call xc_f03_gga_vxc(xcfunc,nqs,
+     $                            dbl_mb(krho),dbl_mb(ksigma),
+     $                            dbl_mb(kvrho),dbl_mb(kvsigma))
+              call xc_f03_gga_fxc(xcfunc,nqs,
+     $                            dbl_mb(krho),dbl_mb(ksigma),
+     $                            dbl_mb(kv2rho2),dbl_mb(kv2rhosig),
+     $                            dbl_mb(kv2sig2))
+#endif
             else
+#if XC_MAJOR_VERSION > 4
               call xc_f03_gga_vxc_fxc_kxc(xcfunc,nqs,
      $             dbl_mb(krho),dbl_mb(ksigma),
      $             dbl_mb(kvrho),dbl_mb(kvsigma),dbl_mb(kv2rho2),
      $             dbl_mb(kv2rhosig),dbl_mb(kv2sig2),
      $             dbl_mb(kv3rho3),dbl_mb(kv3rho2sig),
      $             dbl_mb(kv3rhosig2),dbl_mb(kv3sig3))
+#else
+              call xc_f03_gga_vxc(xcfunc,nqs,
+     $                            dbl_mb(krho),dbl_mb(ksigma),
+     $                            dbl_mb(kvrho),dbl_mb(kvsigma))
+              call xc_f03_gga_fxc(xcfunc,nqs,
+     $                            dbl_mb(krho),dbl_mb(ksigma),
+     $                            dbl_mb(kv2rho2),dbl_mb(kv2rhosig),
+     $                            dbl_mb(kv2sig2))
+              call xc_f03_gga_kxc(xcfunc,nqs,
+     $                            dbl_mb(krho),dbl_mb(ksigma),
+     $                            dbl_mb(kv3rho3),dbl_mb(kv3rho2sig),
+     $                            dbl_mb(kv3rhosig2),dbl_mb(kv3sig3))
+#endif
             endif
             call dfill(nq,0d0,dbl_mb(kexc),1)
           endif
@@ -222,8 +295,8 @@
         case (XC_FAMILY_MGGA, XC_FAMILY_HYB_MGGA)
           gga = .true.
           mgga = .true.
-          dolap = .true.
-
+          dolap = iand(libxc_flags(ifunc),xc_flags_needs_laplacian).eq.
+     $            xc_flags_needs_laplacian
           
           if (iand(libxc_flags(ifunc),xc_flags_have_exc).eq.
      $                                xc_flags_have_exc) then

src/nwdft/libxc/nwchem_libxc_read.F

@@ -41,7 +41,11 @@
       xcfamily = xc_f03_func_info_get_family(xcinfo)
 
       select case(xcfamily)
+#if XC_MAJOR_VERSION > 4
       case (XC_FAMILY_LDA, XC_FAMILY_HYB_LDA)
+#else
+      case (XC_FAMILY_LDA)
+#endif
       case (XC_FAMILY_GGA, XC_FAMILY_HYB_GGA)
       case (XC_FAMILY_MGGA, XC_FAMILY_HYB_MGGA)
       case default
@@ -50,7 +54,11 @@
       endselect
 
       select case(xcfamily)
+#if XC_MAJOR_VERSION > 4
       case (XC_FAMILY_HYB_LDA, XC_FAMILY_HYB_GGA, XC_FAMILY_HYB_MGGA)
+#else
+      case (XC_FAMILY_HYB_GGA, XC_FAMILY_HYB_MGGA)
+#endif
         if (iand(xcflags,xc_flags_hyb_cam).eq.xc_flags_hyb_cam) then
           call xc_f03_hyb_cam_coef(xcfunc,cam_omega,cam_alpha,cam_beta)
 

src/nwdft/libxc/nwchem_libxc_util.F

@@ -73,9 +73,7 @@
       subroutine nwchem_libxc_print
       use,intrinsic :: iso_c_binding
 #ifdef USE_LIBXC
-      use xc_f03_lib_m, only: xc_f03_version_string,
-     $                        xc_f03_reference,
-     $                        xc_f03_functional_get_name
+      use xc_f03_lib_m, only: xc_f03_functional_get_name
 #endif
       implicit none
 
@@ -109,9 +107,7 @@
       subroutine nwchem_libxc_print_header
       use,intrinsic :: iso_c_binding
 #ifdef USE_LIBXC
-      use xc_f03_lib_m, only: xc_f03_version_string,
-     $                        xc_f03_reference,
-     $                        xc_f03_functional_get_name
+      use xc_f03_lib_m, only: xc_f03_version_string
 #endif
       implicit none
 

src/nwdft/lr_tddft/tddft_iguess.F

@@ -371,7 +371,7 @@ c
       else if (lewin) then
        if (nroots.gt.newin) nroots = newin
       else if (nroots.gt.nov(1)+nov(2)) then
-        if (nodezero.and.util_print('initial guess',print_default))
+        if (nodezero.and.util_print('initial guess',print_low))
      1    write(LuOut,9040) nroots,nov(1)+nov(2)
         nroots=nov(1)+nov(2)
         if (.not.rtdb_put(rtdb,'tddft:nroots',mt_int,1,nroots))

src/nwdft/lr_tddft/tddft_prod_v.F

@@ -58,7 +58,6 @@ c
       character*4 targetsym    ! Symmetry of target excited state
       double precision cwall   ! Wall timings
       double precision ccpu    ! CPU timings
-      double precision dcpu    ! CPU timings
       integer algorithm        ! =1 multiple fock build; =2: conventional
       integer mult             ! Ground state spin multiplicity
       integer l_irs(2),k_irs(2)! Irreps
@@ -116,6 +115,8 @@ c      double precision, allocatable, dimension(:,:) :: eye
       integer iwhich
 c
       nodezero=(ga_nodeid().eq.0)
+      ccpu=0d0
+      cwall=0d0
 cjb
 c      call ga_print(g_movecs(1))
 c

src/nwdft/lr_tddft_grad/tddft_grad_compute_g.F

@@ -616,7 +616,7 @@ c
         call tddft_grad_compute_dao(ipol,nao,nocc,g_mo,g_d)
 c DEBUG
 c        call ga_print(g_d)
-        if (tddft_grad_util_print('tddft grad g',print_debug)) then
+        if (util_print('tddft grad g',print_debug)) then
           if (oroot) write(LuOut,*)'DEBUG: '//pname//'D'
           call tddft_grad_print_array(ipol,1,g_d,dble(ipol))
         endif
@@ -640,7 +640,7 @@ c
         enddo
 c DEBUG
 c        call ga_print(g_p_ao)
-        if (tddft_grad_util_print('tddft grad g',print_debug)) then
+        if (util_print('tddft grad g',print_debug)) then
           if (oroot) write(LuOut,*)'DEBUG: '//pname//'P'
 c          call tddft_grad_print_array(ipol,1,g_p_ao,dble(ipol))
           call tddft_grad_print_array(ipol,1,g_d(3),dble(ipol))
@@ -798,7 +798,7 @@ c
 c Daniel (11-30-12): Print X+Y and X-Y in the AO basis
 c DEBUG
 c        call ga_print(g_d(5))
-        if (tddft_grad_util_print('tddft grad g',print_debug)) then
+        if (util_print('tddft grad g',print_debug)) then
           if (oroot) write(LuOut,*)'DEBUG: '//pname//'(X+Y)'
           call tddft_grad_print_array(ipol,1,g_d(5),dble(ipol))
         endif
@@ -818,7 +818,7 @@ c Daniel (11-30-12): Print W (note, this contains the ground state
 c density matrix contribution as well).
 c DEBUG
 c        call ga_print(g_wp_ao)
-        if (tddft_grad_util_print('tddft grad g',print_debug)) then
+        if (util_print('tddft grad g',print_debug)) then
           if (oroot) write(LuOut,*)'DEBUG: '//pname//'W'
           call tddft_grad_print_array(1,1,g_wp_ao,dble(ipol))
         endif
@@ -1140,7 +1140,7 @@ c Daniel (4-8-13): Need conditional statement here for CAM functionals
 c
 c       Print results if so required
 c
-        if (tddft_grad_util_print('tddft grad terms',print_high)) then
+        if (util_print('tddft grad terms',print_high)) then
           if (oroot) then
             write(LuOut,5)ir
     5       format(' The Excited State Energy Gradient by Terms',

src/nwdft/lr_tddft_grad/tddft_grad_compute_p.F

@@ -102,7 +102,7 @@ c
         call nga_scale_patch(g_tp(ip),blo,bhi,0.5d0)
       enddo
 c
-      if (tddft_grad_util_print(pname,print_debug)) then
+      if (util_print(pname,print_debug)) then
         oroot = ga_nodeid().eq.0
         if (oroot) write(LuOut,*)'DEBUG: '//pname//'P'
         call tddft_grad_print_array(ipol,nroots,g_tp,dble(ipol))

src/nwdft/lr_tddft_grad/tddft_grad_compute_r.F

@@ -120,17 +120,22 @@ c
 c
 c Daniel (2-8-13): This is a local TDA variable, so that we don't need
 c to change the global tda variable.
-      logical tdaloc
+      logical tdaloc, tddft_grad_tdaloc
+      external tddft_grad_tdaloc
+      logical oprint
+      integer g_r_x(2),g_r_org(2)
 c
       double precision Exc(2)    ! Exchange-correlation energy
 c
       pname="tddft_grad_compute_r: "
       iwhich = 0  ! call to tddft_nga_cont()
+      write(6,*) ' 0 tda ',tda
 c
 c     1. Do the T contributions
 c
 c     1.1. Create global array to hold Tuv
 c
+      oprint=util_print('tddft_grad_r',print_debug)
       idim(1) = nroot*ipol
       idim(2) = nao
       idim(3) = nao
@@ -180,6 +185,7 @@ c fock_xc:triplet here for restricted triplet calculations to work.
 c
       call tddft_nga_cont(rtdb,ihdl_geom,ihdl_bfao,g_tuv,g_apbt,g_ambt,
      +nao,ipol,tol2e,tda,oskel,kfac,lhashf,.false.,nroot,iwhich,lhascd)
+      write(6,*) ' 2 tda ',tda
 c Daniel (2-26-13): Reset fock_xc:triplet here for restricted triplet 
 c calculations to work.
       if (otriplet) then
@@ -217,11 +223,11 @@ c
 c
 c     1. Done T contributions
 cDEBUG
-c     if (tddft_grad_util_print('tddft grad r',print_debug)) then
-c       oroot = ga_nodeid().eq.0
-c       if (oroot) write(LuOut,*)'DEBUG A: tddft_grad_comp_r: R'
-c       call tddft_grad_print_array(ipol,nroot,g_r,dble(ipol))
-c     endif
+      if (oprint) then
+       oroot = ga_nodeid().eq.0
+       if (oroot) write(LuOut,*)'DEBUG A: tddft_grad_comp_r: R_T'
+       call tddft_grad_print_array(ipol,nroot,g_r,dble(ipol))
+      endif
 cDEBUG
 c
 c     2. Do the (X+Y) and (X-Y) contributions
@@ -278,32 +284,10 @@ c
      +    call errquit(pname//'failed to create g_apby',0,GA_ERR)
       if (.not.nga_create(mt_dbl,3,idim,'vectors (A-B)Y',ichnk,g_amby))
      +    call errquit(pname//'failed to create g_amby',0,GA_ERR)
-c
+      tdaloc=tddft_grad_tdaloc(rtdb,tda,lhashf)
+c     
 c     2.5 Compute (A+B)X, (A-B)X, (A+B)Y and (A-B)Y
 c
-c Daniel (1-5-13): We manipulate the code here because the
-c R vector has the same number of terms for RPA and CIS.  This is a
-c consequence of (X-Y) = X.  It might be a good idea to avoid doing this
-c part for the Y vector since Y = 0.  Note that the coupling matrix
-c expressions H^+[V] and H^-[V] can both be nonzero for CIS, so it isn't
-c okay to skip the anti-symmetric part in the tddft_nga_cont routine.  
-c What CIS does is makes the Y vector contribution zero in the following
-c routines.
-c Daniel (2-8-13): Set the local TDA variable so that we don't change
-c the global one.
-      if (lhashf) then
-        if (tda) then
-          tdaloc = .false. ! For CIS calculations
-        else
-          tdaloc = .false. ! For RPA calculations
-        endif
-      else 
-        if (tda) then
-          tdaloc = .true.  ! For TDDFT/TDA calculations
-        else
-          tdaloc = .false. ! For TDDFT calculations
-        endif
-      endif
       call tddft_nga_cont(rtdb,ihdl_geom,ihdl_bfao,g_x,g_apbx,g_ambx,
      +nao,ipol,tol2e,tdaloc,oskel,kfac,lhashf,otriplet,nroot,iwhich,
      +lhascd)
@@ -349,6 +333,21 @@ c     2.10 Transform (A+B)(X+Y) to MO basis occupied-occupied block only
 c
       call tddft_grad_trans_ao2mo(ipol,nao,nfc,naoc,nocc,nav,nfv,
      +     nroot,1.0d0,0.0d0,"ij",g_mo,g_apbx,g_hij,"ij")
+      if(oprint)then
+c     extra ga to keep track of X
+         if (.not. ga_duplicate(g_r(1), g_r_x(1), 'gtmp2'))
+     c        call errquit('tdgcr: gadupl failed',0,0)
+         call ga_zero(g_r_x(1))
+         g_r_org(1)=g_r(1)
+         g_r(1)=g_r_x(1)
+         if(ipol.eq.2) then
+         if (.not. ga_duplicate(g_r(2), g_r_x(2), 'gtmp2'))
+     c           call errquit('tdgcr: gadupl failed',0,0)
+         call ga_zero(g_r_x(2))
+         g_r_org(2)=g_r(2)
+         g_r(2)=g_r_x(2)
+         endif
+      endif
 c
 c     2.11 Add -sum_j (X+Y)ja [(A+B)(X+Y)ji] to Ria
 c
@@ -546,11 +545,21 @@ c
 c     2. Done (X+Y) and (X-Y) contributions
 c
 cDEBUG
-c     if (tddft_grad_util_print('tddft grad r',print_debug)) then
-c       oroot = ga_nodeid().eq.0
-c       if (oroot) write(LuOut,*)'DEBUG B: tddft_grad_comp_r: R'
-c       call tddft_grad_print_array(ipol,nroot,g_r,dble(ipol))
-c     endif
+      if (oprint) then
+       oroot = ga_nodeid().eq.0
+       if (oroot) write(LuOut,*)'DEBUG B: tddft_grad_comp_r: R_x'
+       call tddft_grad_print_array(ipol,nroot,g_r_x,dble(ipol))
+       call ga_add(1d0,g_r_x(1),1d0,g_r_org(1),g_r_org(1))
+       g_r(1)=g_r_org(1)
+       if (.not.ga_destroy(g_r_x(1))) 
+     +     call errquit(pname//'failed to destroy grx',1,GA_ERR)
+       if(ipol.eq.2) then
+       call ga_add(1d0,g_r_x(2),1d0,g_r_org(2),g_r_org(2))
+       g_r(2)=g_r_org(2)
+       if (.not.ga_destroy(g_r_x(2))) 
+     +     call errquit(pname//'failed to destroy grx',2,GA_ERR)
+       endif
+      endif
 cDEBUG
 c
 c     3. Do the Gxc contributions
@@ -664,14 +673,14 @@ c
         enddo
         call ga_sync()
 cDEBUG
-c     if (tddft_grad_util_print('tddft grad r',print_debug)) then
-c       oroot = ga_nodeid().eq.0
-c       if (oroot) write(LuOut,*)'DEBUG C: tddft_grad_comp_r: den'
-c       call tddft_grad_print_array(ipol,nroot,int_mb(k_den2),
-c    +                              dble(ipol))
-c       call tddft_grad_print_array(ipol,nroot,int_mb(k_den2+1),
-c    +                              dble(ipol))
-c     endif
+        if (oprint) then
+           oroot = ga_nodeid().eq.0
+           if (oroot) write(LuOut,*)'DEBUG C: tddft_grad_comp_r: den'
+       call tddft_grad_print_array(ipol,nroot,int_mb(k_den2),
+     +          dble(ipol))
+       call tddft_grad_print_array(ipol,nroot,int_mb(k_den2+1),
+     +      dble(ipol))
+      endif
 cDEBUG
 c Daniel (1-10-13): For a single restricted root ngxc = 1 and there are 2
 c density matrices fed to fock_xc, the GS density matrix D and the
@@ -689,7 +698,7 @@ c fock_xc.
      +               int_mb(k_gxc), Exc, ipol, .false.)
         call ga_sync()
 c DEBUG
-      if (tddft_grad_util_print('tddft grad r',print_debug)) then
+      if (oprint) then
         oroot = ga_nodeid().eq.0
         if (oroot) write(LuOut,*)'DEBUG: '//pname//'gxc'
         call tddft_grad_print_array(ipol,nroot,int_mb(k_gxc),
@@ -734,9 +743,9 @@ c
       endif  ! xc_gotxc()
 c
 c     Debug print
-      if (tddft_grad_util_print('tddft grad r',print_debug)) then
+      if (oprint) then
         oroot = ga_nodeid().eq.0
-        if (oroot) write(LuOut,*)'DEBUG: '//pname//'R'
+        if (oroot) write(LuOut,*)'DEBUG: '//pname//'R_all'
         call tddft_grad_print_array(ipol,nroot,g_r,dble(ipol))
       endif
 c

src/nwdft/lr_tddft_grad/tddft_grad_compute_t.F

@@ -147,7 +147,7 @@ c (X-Y)ias(X-Y)ibs
 c
       enddo ! ip = 1, ipol
 c
-      if (tddft_grad_util_print('tddft grad t',print_debug)) then
+      if (util_print('tddft grad t',print_debug)) then
         oroot = ga_nodeid().eq.0
         if (oroot) write(LuOut,*)'DEBUG: '//pname//'T'
         call tddft_grad_print_array(ipol,nroot,g_tp,dble(ipol))

src/nwdft/lr_tddft_grad/tddft_grad_compute_w.F

@@ -141,7 +141,8 @@ c
       parameter (oskel=.false.)
       double precision Exc(2)    ! Exchange-correlation energy
 c
-      logical tdaloc
+      logical tdaloc, tddft_grad_tdaloc
+      external tddft_grad_tdaloc
       logical doitw,doitz
       logical doitxpy1,doitxpy2
       logical doitxmy1,doitxmy2
@@ -599,7 +600,7 @@ c Daniel (2-12-13): We need to set triplet here for fock_xc.
      +               int_mb(k_gxc), Exc, ipol, .false.)
         call ga_sync()
 c DEBUG
-      if (tddft_grad_util_print('tddft grad w',print_debug)) then
+      if (util_print('tddft grad w',print_debug)) then
         oroot = ga_nodeid().eq.0
         if (oroot) write(LuOut,*)'DEBUG: '//pname//'gxc'
         call tddft_grad_print_array(ipol,nroots,int_mb(k_gxc),
@@ -746,29 +747,7 @@ c
 c
 c     - Compute (A+B)X, (A-B)X, (A+B)Y and (A-B)Y
 c
-c Daniel (1-7-13): We manipulate the code here because the
-c R vector has the same number of terms for RPA and CIS.  This is a
-c consequence of (X-Y) = X.  It might be a good idea to avoid doing this
-c part for the Y vector since Y = 0.  Note that the coupling matrix
-c expressions H^+[V] and H^-[V] can both be nonzero for CIS, so it isn't
-c okay to skip the anti-symmetric part in the tddft_nga_cont routine.  
-c What CIS does is makes the Y vector contribution zero in the following
-c routines.
-c Daniel (2-8-13): Set the local TDA variable so that we don't change
-c the global one.
-      if (lhashf) then
-        if (tda) then
-          tdaloc = .false. ! For CIS and TDDFT/TDA (hybrid) calculations
-        else
-          tdaloc = .false. ! For RPA and TDDFT (hybrid) calculations
-        endif
-      else
-        if (tda) then
-          tdaloc = .true.  ! For TDDFT/TDA (pure) calculations
-        else
-          tdaloc = .false. ! For TDDFT (pure) calculations
-        endif
-      endif
+      tdaloc=tddft_grad_tdaloc(rtdb,tda,lhashf)
       call tddft_nga_cont(rtdb,ihdl_geom,ihdl_bfao,g_x,g_apbx,g_ambx,
      +   nao,ipol,tol2e,tdaloc,oskel,kfac,lhashf,otriplet,nroots,iwhich,
      +   lhascd)
@@ -1190,7 +1169,7 @@ c
       enddo
       call ga_sync()
 c
-      if (tddft_grad_util_print('tddft grad w',print_debug)) then
+      if (util_print('tddft grad w',print_debug)) then
         oroot = ga_nodeid().eq.0
         if (oroot) write(LuOut,*)'DEBUG: '//pname//'W'
         call tddft_grad_print_array(ipol,nroots,g_w,dble(ipol))

src/nwdft/lr_tddft_grad/tddft_grad_solve.F

@@ -60,7 +60,6 @@ c
 c
       logical odiff,oprint
       parameter (odiff=.true.)
-      parameter (oprint=.true.)
 c
 c     Input:
 c
@@ -124,35 +123,20 @@ c
       logical oroot
 c
       character*32 pname
-      logical tdaloc
+      logical tdaloc, tddft_grad_tdaloc
+      external tddft_grad_tdaloc
+      logical oprint_debug
 c
       pname= "tddft_grad_solve: "
+      oprint=util_print('tddft_grad_solve',print_medium)
+      oprint_debug=util_print('tddft_grad_solvecphf',print_debug)
 c
 c     Initialize the common block for tddft_grad_solve_product and
 c     tddft_grad_solve_precond
 c
       oroot = ga_nodeid().eq.0
 c
-c Daniel (1-5-13): For a CIS calculation, we still need to solve the
-c full Z vector equation (i.e. we need to have the equation solved when
-c B =/= 0).  This is confirmed by running an RPA calculation with B = 0,
-c since there were no further modifications beyond making B = 0 for
-c that formulation of the linear response equations and you obtain the 
-c correct answer for the CIS gradients that way.
-c
-      if (lhashf) then
-        if (tda) then
-          tdaloc = .false. ! For CIS calculations
-        else
-          tdaloc = .false. ! For RPA calculations
-        endif
-      else 
-        if (tda) then
-          tdaloc = .true.  ! For TDDFT/TDA calculations
-        else
-          tdaloc = .false. ! For TDDFT calculations
-        endif
-      endif
+      tdaloc=tddft_grad_tdaloc(rtdb,tda,lhashf)
 c End test
       tdg_tol2e   = tol2e
       tdg_kfac    = kfac
@@ -177,7 +161,6 @@ c End test
       tdg_g_mo(1) = g_mo(1)
       tdg_g_mo(2) = g_mo(2)
       tdg_e_mo    = iptr_mo_e
-c      tdg_tda     = tda
       tdg_tda     = tdaloc
       tdg_oskel   = oskel
       tdg_lhashf  = lhashf
@@ -224,6 +207,14 @@ c
         call ga_copy(g_b,g_x)
         call tddft_grad_solve_precond(g_x,0.0d0)
       endif
+      if(oprint_debug) then
+         if(oroot) write(luout,*) ' LHS'
+         call util_flush(luout)
+         call ga_print(g_x)
+         if(oroot) write(luout,*) ' RHS'
+         call util_flush(luout)
+         call ga_print(g_b)
+      endif
 c
       call ga_lkain(rtdb,g_x,g_b,tddft_grad_solve_product,
      +              tddft_grad_solve_precond,solve_thresh,
@@ -254,7 +245,7 @@ c
       if (.not.ga_destroy(g_x)) 
      +   call errquit(pname//'failed to destroy g_x',0, GA_ERR)
 c
-      if (tddft_grad_util_print('tddft grad z',print_debug)) then
+      if (util_print('tddft grad z',print_debug)) then
         if (oroot) write(LuOut,*)'DEBUG: '//pname//'Z'
         call tddft_grad_print_array(ipol,nroots,g_z,dble(ipol))
       endif
@@ -266,4 +257,41 @@ c     Reset fock_xc:triplet here for restricted triplet calculations
       endif
 c
       end
-c $Id$
+c  We manipulate the code here because the
+c R vector has the same number of terms for RPA and CIS.  This is a
+c consequence of (X-Y) = X.  It might be a good idea to avoid doing this
+c part for the Y vector since Y = 0.  Note that the coupling matrix
+c expressions H^+[V] and H^-[V] can both be nonzero for CIS, so it isn't
+c okay to skip the anti-symmetric part in the tddft_nga_cont routine.  
+c What CIS does is makes the Y vector contribution zero in the following
+c routines.
+c Set the local TDA variable so that we don't change
+c the global one.
+      logical function tddft_grad_tdaloc(rtdb,tda,lhashf)
+      implicit none
+#include "rtdb.fh"
+#include "mafdecls.fh"
+      integer rtdb
+      logical tda
+      logical lhashf
+c
+      logical tdaloc_rtdb
+c      
+      if (lhashf) then
+        if (tda) then
+          tddft_grad_tdaloc = .false. ! For CIS calculations
+        else
+          tddft_grad_tdaloc = .false. ! For RPA calculations
+        endif
+      else 
+        if (tda) then
+cold          tddft_grad_tdaloc = .true.  ! For TDDFT/TDA calculations
+          tddft_grad_tdaloc = .false.  ! For TDDFT/TDA calculations
+        else
+          tddft_grad_tdaloc = .false. ! For TDDFT calculations
+        endif
+      endif
+      if (rtdb_get(rtdb,'tddft:tdaloc',mt_log,1,tdaloc_rtdb))
+     +     tddft_grad_tdaloc=tdaloc_rtdb
+      return
+      end

src/nwdft/lr_tddft_grad/tddft_gradient.F

@@ -168,6 +168,9 @@ c     Get the start times
 c
 c     Init pstat
       call dft_pstat_init(rtdb)
+c      
+      call util_print_push
+      call util_print_rtdb_load(rtdb,'tddft')
 c
 c     Print header
       if (oroot) call tddft_grad_print_header(luout)
@@ -225,7 +228,7 @@ c     Print calculation information
 c
 c     Print start times
       if (oroot) then
-        if (tddft_grad_util_print('tddft grad time',print_default)) then
+        if (util_print('tddft grad time',print_default)) then
           write(luout,*)
           write(luout,"(' Start at time  cpu:',f11.1,'s  wall:',f11.1,
      +                  's')")-tddft_grad_cpu,-tddft_grad_wall
@@ -435,7 +438,7 @@ c     Print end times and wrap up
       tddft_grad_cpu  = tddft_grad_cpu  + util_cpusec()
       tddft_grad_wall = tddft_grad_wall + util_wallsec()
       if (oroot) then
-        if (tddft_grad_util_print('tddft grad time',print_default)) then
+        if (util_print('tddft grad time',print_default)) then
           write(LuOut,*)
           write(LuOut,"(' TDDFT Gradient time  cpu:',f11.1,'s  wall:',
      +                  f11.1,'s')")tddft_grad_cpu,tddft_grad_wall
@@ -496,6 +499,7 @@ c
 c <-- MN solvation models 
 c
       call dft_pstat_print
+      call util_print_pop
       return
       end
 c $Id$

src/nwdft/scf_dft/dft_scf.F

@@ -248,6 +248,7 @@ C     RTDB flag for printing out all matrices
       logical lprint_mats  !print matrices to stdout?
       logical lsave_mats   !save matrices to file?
       logical molden, forceatguess, save_evals, densmat
+      integer back_to_atguess
       character*256 file_mat1,movecs_in_org
       integer ma_type, no_sflip,l_at_flip,k_at_flip
       character*26 date
@@ -301,7 +302,7 @@ c
       oprint_info = util_print('common', print_debug)
       oprint_parm = util_print('parameters', print_default)
       oprint_3c2e = util_print('3c 2e integrals', print_default)
-      oprint_conv = util_print('convergence', print_default)
+      oprint_conv = util_print('convergence', print_low)
       oprint_conv_details = util_print('convergence details', 
      &                                 print_high)
       oprint_vecs = util_print('intermediate vectors', print_high)
@@ -862,9 +863,13 @@ c       optionally, g_x2c_c2u is also allocated and kept.
       end if                    ! do_x2c
 c
 c     == determine guess ==
+      if(.not.rtdb_get(rtdb,'back_to_atguess',mt_int,1,
+     N     back_to_atguess)) 
+     .     back_to_atguess=0
       if(.not.rtdb_get(rtdb,'forceatguess',mt_log,1,forceatguess)) 
      .     forceatguess=.false.
-      if(forceatguess) movecs_in='atomic'
+ 1993 if(forceatguess) movecs_in='atomic'
+      if(movecs_in.eq.'atomic') back_to_atguess=0
       call dft_guessin(movecs_in,ldmix,ncanorg,fon,
      ,     vecs_or_dens,
      ,     ipol,nbf_ao,g_movecs,g_gmovecs,
@@ -1820,8 +1825,21 @@ c
       call dft_scfcvg(rms, derr, Etold, Etnew,
      &                e_conv, d_conv, g_conv, ipol, 
      &                iter, iterations, idone, rtdb,
-     &                converged, diising)
+     &     converged, diising,
+     B     back_to_atguess)
 c
+      if(movecs_in.ne.'atomic'.and.back_to_atguess.lt.0) then
+         movecs_in='atomic'
+         if(ga_nodeid().eq.0)
+     W        write(luout,*) ' restart from atomic guess'
+         if (.not. ga_destroy(g_denso(1))) call errquit
+     &        ('dft_scf: could not destroy g_denso', 0, GA_ERR)
+         if(ipol.eq.2) then
+            if (.not. ga_destroy(g_denso(2))) call errquit
+     &           ('dft_scf: could not destroy g_dens2', 0, GA_ERR)
+         endif
+         goto 1993
+      endif
       if (oprint_time)
      &     call dft_tstamp(' called scfcvg ')
 c

src/nwdft/scf_dft/dft_scfcvg.F

@@ -1,6 +1,7 @@
       subroutine dft_scfcvg(rms, derr, etold, etnew, e_conv, 
      &                      d_conv, g_conv, ipol, iter, iterations, 
-     &                      idone, rtdb, converged, diising)
+     &     idone, rtdb, converged, diising,
+     B     back_to_atguess)
 c
 c     $Id$
 c
@@ -21,6 +22,7 @@ c
       integer rtdb              ! [input]
       logical converged         ! [output]
       logical diising           ! [input]
+      integer back_to_atguess   ! [in/out]
 c
 #include "mafdecls.fh"
 #include "rtdb.fh"
@@ -89,6 +91,13 @@ c     Check over-all convergence.
 c
       converged = e_conv_logical.and.d_conv_logical.and.g_conv_logical
       if (converged)idone = 1
+c
+c     try at guess if d_e > 1d-1 after back_to_atguess iterations
+c
+      if(.not.converged.and.back_to_atguess.gt.0) then
+         if(abde.gt.1d-1.and.iter.gt.back_to_atguess)
+     B        back_to_atguess=-1
+      endif
 c     
 c     Check iteration value.
 c     

src/nwdft/so_dft/dft_scf_so.F

@@ -1945,7 +1945,7 @@ c
       call dft_scfcvg(rms, derr, Etold, Etnew, 
      &     e_conv, d_conv, g_conv, ipol, 
      &     iter, iterations, idone, rtdb,
-     &     converged, diising)
+     &     converged, diising, 0)
 c
 #ifdef LOOSE_START
       if (delta.lt.1.d-3)then

src/nwdft/xc/xc_vdw_util.F

@@ -42,6 +42,7 @@ C>
       double precision cnj !< [Input] \f$\mathrm{CN}^B(R)\f$
       double precision top,bottom,dist,c6_ref
       double precision cna,cnb
+      double precision c6_mem,dist_save
       integer i,j
       double precision eps
       parameter (eps=1d-90)
@@ -49,6 +50,8 @@ c
       c6cn=0.0d0
       top=0.0d0
       bottom=0.0d0
+      c6_mem=-1.d+99
+      dist_save=1.0d99
       do i=1,maxcn(iat)
          do j=1,maxcn(jat)
             c6_ref=c6ab(iat,jat,i,j,1)
@@ -56,6 +59,10 @@ c
                cna=c6ab(iat,jat,i,j,2)
                cnb=c6ab(iat,jat,i,j,3)
                dist=(cna-cni)**2+(cnb-cnj)**2
+               if (dist.lt.dist_save) then
+                  dist_save=dist
+                  c6_mem=c6_ref
+               endif
                top=top+dexp(k3*dist)*c6_ref
                bottom=bottom+dexp(k3*dist)
             endif
@@ -65,8 +72,7 @@ c
       if (bottom.gt.eps) then
          c6cn=top/bottom
       else
-         c6cn=0.0d0
-         call errquit(' c6cn returning zero ',0,0)
+         c6cn=c6_mem
       endif
       return
       end

src/nwpw/GNUmakefile

@@ -11,7 +11,14 @@
 
 
       LIBRARY = libnwpw.a
+$(info USE_OPENMP $(origin USE_OPENMP))
 
+ifdef USE_OPENMP
+# QA issues
+override undefine USE_OPENMP
+endif
+
+$(info USE_OPENMP $(origin USE_OPENMP))
 
   LIB_DEFINES =
 

src/nwpw/nwpwlib/C1dB/GNUmakefile

@@ -42,9 +42,6 @@ endif
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
 else
   LIB_DEFINES += -DTCGMSG
 endif

src/nwpw/nwpwlib/C3dB/GNUmakefile

@@ -47,9 +47,6 @@ endif
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
 else
   LIB_DEFINES += -DTCGMSG
 endif

src/nwpw/nwpwlib/CMatrix/GNUmakefile

@@ -43,9 +43,6 @@ endif
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
 else
   LIB_DEFINES += -DTCGMSG
 endif

src/nwpw/nwpwlib/D1dB/GNUmakefile

@@ -43,9 +43,6 @@ endif
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
 else
   LIB_DEFINES += -DTCGMSG
 endif

src/nwpw/nwpwlib/D3dB/GNUmakefile

@@ -41,9 +41,6 @@ endif
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
 else
   USES_BLAS += D3dB-tcgmsg.F
   LIB_DEFINES += -DTCGMSG

src/nwpw/nwpwlib/DMatrix/GNUmakefile

@@ -43,9 +43,6 @@ endif
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
 else
   LIB_DEFINES += -DTCGMSG
 endif

src/nwpw/nwpwlib/K1dB/GNUmakefile

@@ -41,9 +41,6 @@ endif
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
 else
   LIB_DEFINES += -DTCGMSG
 endif

src/nwpw/nwpwlib/Parallel/GNUmakefile

@@ -40,9 +40,6 @@ include ../../../config/makelib.h
 
 ifeq ($(USE_MPIF),y)
   LIB_DEFINES += -DMPI
-  ifneq ($(MPI_INCLUDE),)
-    LIB_INCLUDES += $(patsubst -I-I%,-I%,-I$(MPI_INCLUDE))
-  endif
   ifeq ($(DEV_GA),y)
     LIB_DEFINES += -DGA_DEV
   endif

src/nwpw/nwpwlib/nwpwxc/GNUmakefile

@@ -13,18 +13,26 @@
                    nwpwxc_c_vs98.o nwpwxc_c_lsda.o nwpwxc_c_vwn1.o \
                    nwpwxc_c_vwn1_rpa.o nwpwxc_c_vwn2.o nwpwxc_c_vwn3.o nwpwxc_c_vwn4.o \
                    nwpwxc_c_vwn5.o nwpwxc_c_op.o \
-                   nwpwxc_cs00.o nwpwxc_vdw3a.o nwpwxc_vdw3b.o nwpwxc_vdw3c.o \
+                   nwpwxc_cs00.o nwpwxc_vdw3b.o nwpwxc_vdw3c.o \
                    nwpwxcP_xc_b97.o nwpwxc_x_b97.o nwpwxc_c_b97.o nwpwxc_xc_kt1.o \
                    nwpwxc_c_spbe.o nwpwxc_x_ssbD_1.o nwpwxc_c_perdew81.o \
                    nwpwxc_c_perdew86.o nwpwxc_x_pw91.o nwpwxc_x_pw6.o nwpwxc_x_b86b.o \
                    nwpwxc_x_scan.o nwpwxc_c_scan.o
+ifdef NO_NWPWXC_VDW3A
+             OBJ_OPTIMIZE += nwpwxc_vdw3a_stub.o
+else
+             OBJ_OPTIMIZE += nwpwxc_vdw3a.o
+endif
 
              OBJ = nwpwxc_block.o nwpwxc_add.o nwpwxc_query.o nwpwxc_nwchem.o \
                    nwpwxc_vdw3d.o
 
       LIBRARY = libnwpwlib.a
 
-    USES_BLAS = nwpwxc_vdw3a.F nwpwxc_vdw3c.F nwpwxc_eval.F nwpwxc_c_vwn_a.fh
+    USES_BLAS = nwpwxc_vdw3c.F nwpwxc_eval.F nwpwxc_c_vwn_a.fh
+ifndef NO_NWPWXC_VDW3A
+    USES_BLAS += nwpwxc_vdw3a.F 
+endif
 
       HEADERS = nwpwxc.fh nwpwxc_param.fh
 
@@ -32,7 +40,9 @@
 
 include ../../../config/makefile.h
 include ../../../config/makelib.h
+ifndef NO_NWPWXC_VDW3A
 nwpwxc_vdw3a.F:
-	./build_dftd3a.sh 
+	./build_dftd3a.sh
+endif
 
 

src/nwpw/nwpwlib/nwpwxc/build_dftd3a.sh

@@ -5,14 +5,24 @@ do_exit(){
 	echo ' '
 	exit 1
     }
-rm -f dftd3.f nwpwxc_vdw3a.F
-export PATH=`pwd`:$PATH
+check_patch(){
 if [[ ! -x "$(command -v patch)" ]]; then
     #try to download busybox for x86_64 linux
-    if [[ $(uname -s) == "Linux" ]]  && [[ $(uname -m) == "x86_64" ]] ; then
+    if [[ $(uname -s) == "Linux" ]] ; then
 	echo "patch command missing"
 	echo "downloading busybox to use patch command"
-	wget https://www.busybox.net/downloads/binaries/1.31.0-defconfig-multiarch-musl/busybox-x86_64 -O patch
+	if [[ $(uname -m) == "x86_64" ]] ; then
+	    barch=x86_64
+	elif [[ $(uname -m) == "aarch64" ]] ; then
+	    barch=armv8l
+	elif [[ $(uname -m) == "ppc64" ]] ; then
+	    barch=powerpc64
+	elif [[ $(echo armv6l| awk ' /arm*/ { print "arm"}') == "arm" ]] ; then
+	    barch=armv5l
+	else
+	    do_exit
+	fi
+	wget https://www.busybox.net/downloads/binaries/1.31.0-defconfig-multiarch-musl/busybox-$barch -O patch
 	if [ "$?" != 0 ]; then
 	    do_exit
 	else
@@ -22,7 +32,18 @@ if [[ ! -x "$(command -v patch)" ]]; then
 	do_exit
     fi
 fi
-URL1="https://www.chemie.uni-bonn.de/pctc/mulliken-center/software/dft-d3/"
+}
+
+if [[ ! -z "${NO_NWPWXC_VDW3A}" ]]; then
+    rm -f nwpwxc_vdw3a.F
+    echo ' removed nwpwxc_vdw3a.F source'
+    exit 0
+fi
+check_patch
+rm -f dftd3.f nwpwxc_vdw3a.F
+export PATH=`pwd`:$PATH
+#URL1="https://www.chemie.uni-bonn.de/pctc/mulliken-center/software/dft-d3/"
+URL1="https://www.chemiebn.uni-bonn.de/pctc/mulliken-center/software/dft-d3/"
 URL2="https://web.archive.org/web/20210527062154if_/https://www.chemie.uni-bonn.de/pctc/mulliken-center/software/dft-d3/"
 declare -a urls=("$URL1" "$URL1" "$URL1" "$URL2" "$URL2")
 TGZ=dftd3.tgz
@@ -47,10 +68,12 @@ else
     echo chchc ${COM1} ${urls[$tries]}/"$TGZ" ${COM2}
     tries=1 ; until [ "$tries" -ge 5 ] ; do
     ${COM1} ${urls[$tries]}/"$TGZ" ${COM2} && break
-    tries=$((tries+1)) ; echo attempt no.  $tries    ; sleep 5 ;  done
+    tries=$((tries+1)) ; echo attempt no.  $tries    ; sleep 9 ;  done
 fi
 if [[ ! -f "$TGZ" ]]; then
     echo "download failed"
+    echo "if internet connectivity is missing"
+    echo "set NO_NWPWXC_VDW3A=1"
     exit 1
 fi
 tar xzf dftd3.tgz dftd3.f

src/nwpw/nwpwlib/nwpwxc/nwpwxc_vdw3a_stub.F

@@ -0,0 +1,28 @@
+      subroutine  nwpwxc_vdw3_dftd3()
+      call errquit(' dft-d3 functionality missing',0,0)
+      end subroutine
+#if 0      
+      subroutine nwpwxc_printoptions
+      end subroutine nwpwxc_printoptions
+      subroutine nwpwxc_setfuncpar(func,version,TZ,s6,rs6,s18,rs18,alp)
+      end subroutine nwpwxc_setfuncpar
+      subroutine nwpwxc_rdpar(dtmp,version,s6,s18,rs6,rs18,alp)
+      end subroutine nwpwxc_rdpar
+      subroutine nwpwxc_edisp(max_elem,maxc,n,xyz,iz,c6ab,mxc,
+      end subroutine nwpwxc_edisp
+      subroutine nwpwxc_adisp(max_elem,maxc,n,xyz,iz,c6ab,mxc,
+      end subroutine nwpwxc_adisp
+c      subroutine nwpwxc_gdisp(max_elem,maxc,n,xyz,iz,c6ab,mxc,r2r4,r0ab,rcov,
+      subroutine nwpwxc_gdisp(max_elem,maxc,n,xyz,iz,c6ab,mxc,
+      end subroutine nwpwxc_gdisp
+      subroutine nwpwxc_get_dC6_dCNij(maxc,max_elem,c6ab,mxci,mxcj,
+      end subroutine nwpwxc_get_dC6_dCNij
+      subroutine nwpwxc_getc6(maxc,max_elem,c6ab,mxc,iat,jat,nci,ncj,c6)
+      end subroutine nwpwxc_getc6
+      subroutine nwpwxc_ncoord(natoms,rcov,iz,xyz,cn,cn_thr)
+      end subroutine nwpwxc_ncoord
+      subroutine nwpwxc_loadc6(fname,maxc,max_elem,c6ab,maxci)
+      end subroutine nwpwxc_loadc6
+      subroutine nwpwxc_limit(iat,jat,iadr,jadr)
+      end subroutine nwpwxc_limit
+#endif

src/peigs/doc/GNUmakefile

@@ -1,12 +1,8 @@
-#
-# $Id$
-#
 
 
-all:	; texi2dvi peigs.texi
-	  dvips peigs.dvi
+all:	; texi2pdf peigs.texi
 
-#.EXPORT_ALL_VARIABLES:
+.EXPORT_ALL_VARIABLES:
 
 .PHONY :	clean
 

src/property/GNUmakefile

@@ -63,6 +63,7 @@
          CalcPerturbedTDPmat1.o \
          CalcPerturbedTDPmat1_opt.o \
          localization_driver.o \
+         ibo_localization.o \
          hnd_vec_write.o \
 	 giao_b1_movecs_tools.o \
          aor_r1_beta_anl_tools.o\

src/property/ibo_localization.F

@@ -0,0 +1,967 @@
+
+      subroutine ibo_localization(rtdb, geom, ltyp, basis, g_movecs,
+     &  nocc, nvir, nmo, nbf, natoms, eval, occ, c, pop, list)
+
+c     =================================================================
+c     IAO construction and generation of occupied or virtual IBOs.
+c     Written by J. Autschbach, jochena@buffalo.edu
+c
+c     The formalism is described in G. Knizia, J. Chem. Theory
+c     Comput. 2013, 9, 4834 (with additional useful information provided
+c     in Senjean et al., J. Chem. Theory Comput. 2021, 17, 1337 and the
+c     SI of that article). What is implemented here corresponds to
+c     Knizia's original formulation, and the equations for p = 4 in
+c     Appendix D as given in the 'bugfixed' version of the article on
+c     the author's web page [https://sites.psu.edu/knizia/software/
+c     (accessed 02/2023)].
+
+c     g_movecs and arrays occ and eval are for the MOs of a given
+c     spin (alpha or beta)
+
+c     The localization routine called from here is in
+c     src/ddscf/localize.F
+
+c     =================================================================
+
+      implicit none
+
+#include "errquit.fh"
+#include "mafdecls.fh"
+#include "global.fh"
+#include "msgids.fh"
+#include "geom.fh"
+#include "rtdb.fh"
+#include "bas.fh"
+#include "util.fh"
+#include "stdio.fh"
+#include "apiP.fh"
+#include "prop.fh"
+#include "bgj.fh"
+
+
+c     subroutine arguments:
+
+      integer rtdb, geom, basis
+      character*(3) ltyp
+      integer g_movecs
+      integer nocc, nvir, nmo, nbf, natoms
+      double precision eval(nbf), occ(nbf), c(nbf,2)
+      double precision pop(natoms)
+      integer list(natoms)
+
+c     local GA handles:
+
+      integer g_s2, g_s12, g_p12, g_p21
+      integer g_ctilde, g_iao, g_mo
+      integer g_temp, g_tmp1, g_tmp2, g_cib, g_u, g_vt
+      integer g_smat
+
+c     local variables:
+
+      logical master, debug
+      integer dbg, info
+
+      integer ncore, imo, jmo, kmo
+
+      integer n1, n2, nsing, noff
+
+      integer s, nlist, a, i, t, tt, u, bflo, bfhi
+
+      double precision rtemp, qas
+
+      double precision small
+      parameter (small=1d-8)
+
+      integer
+     &  mnbf, iao_mxprim, iao_mxang, iao_mxcont,
+     &  iao_mxnbf_cn, iao_mxnbf_ce, iao_nshells,
+     &  minbas
+
+      double precision minval, swap
+      integer l_val, k_val
+
+      logical int_normalize
+      external int_normalize
+
+      integer ga_create_atom_blocked
+      external ga_create_atom_blocked
+
+
+      character*(256) lmotrans
+      logical file_write_ga
+      external file_write_ga
+
+      character*(16) pname
+
+c     =================================================================
+
+      pname = 'ibo_localization'
+
+      dbg = 0
+      master =  ga_nodeid().eq.0
+      debug = (dbg>0) .and. master ! .true. during development
+
+      if (ltyp.ne.'occ' .and. ltyp.ne.'vir')  call errquit
+     &     (pname//': loc. type unknown', 0, BASIS_ERR)
+
+      if(debug) then
+        if (ltyp.eq.'occ') write (luout,*)
+     &       'entering occupied IBO localization'
+        if (ltyp.eq.'vir') write (luout,*)
+     &    'entering virtual IBO localization'
+      end if
+
+      if (.not. geom_num_core(rtdb, geom, 'ddscf', ncore)) ncore = 0
+
+      if (debug) write (luout,*) 'ncore = ',ncore
+
+
+
+c     ------------------------------
+c     construct IAOs in the AO basis
+c     ------------------------------
+
+c     Note:
+c     Basis 1 is the AO basis used in the SCF calculation
+c     Basis 2 is the minimal auxiliary basis
+
+c     AO Overlap Matrix S1 -> g_smat:
+
+      g_smat  = ga_create_atom_blocked(geom, basis, 'loc:smat')
+      call ga_zero(g_smat)
+      call int_1e_ga(basis, basis, g_smat, 'overlap', .false.)
+
+c     auxiliary basis overlap S2 -> g_s2 basis needs to defined in
+c     the input as "iao basis". we will now create the basis here and
+c     then calculate the overlap and the mixed ao-iao basis overlap
+c     S12
+
+      if (.not. bas_create(minbas, 'iao basis'))
+     &  call errquit(pname//': cannot create iao bas', 86, BASIS_ERR)
+
+      if (.not. bas_rtdb_load(rtdb, geom, minbas, 'iao basis'))
+     &  call errquit(pname//': iao basis not present', 86, BASIS_ERR)
+
+      if (.not. int_normalize(rtdb,minbas))
+     &  call errquit(pname//': int_normalize iao', 86, INT_ERR)
+
+      if (.not. bas_numbf(minbas, mnbf))
+     &  call errquit(pname//': bas err iao mnbf', 86, BASIS_ERR)
+
+      if (.not. bas_nprim_cn_max(minbas,iao_mxprim))
+     &  call errquit(pname//': bas err iao mxprim', 86, BASIS_ERR)
+
+      if (.not. bas_high_angular(minbas,iao_mxang))
+     &  call errquit(pname//': bas err iao mxang', 86, BASIS_ERR)
+
+      if (.not. bas_ncontr_cn_max(minbas,iao_mxcont))
+     &  call errquit(pname//': bas err iao mxcont', 86, BASIS_ERR)
+
+      if (.not. bas_nbf_cn_max(minbas,iao_mxnbf_cn))
+     &  call errquit(pname//': bas err iao mcnbf_cn', 86, BASIS_ERR)
+
+      if (.not. bas_nbf_ce_max(minbas,iao_mxnbf_ce))
+     &  call errquit(pname//': bas err iao mxnbf_ce', 86, BASIS_ERR)
+
+      if (.not. bas_numcont(minbas,iao_nshells))
+     &  call errquit(pname//': bas err iao nshells', 86, BASIS_ERR)
+
+c     not sure if it makes much sense in case the minimal basis for
+c     the IAOs is larger than our regular basis, so let's check that
+c     here and exit with an error if nbf < mnbf
+
+      if (nbf.lt.mnbf) call errquit
+     &  (pname//': nbf < mnbf. cannot handle!', 66, UNKNOWN_ERR)
+
+c     create overlap for minbas, and the mixed basis-minbas
+c     overlap S12.
+c     Then we calculate the projectors P12 and P21
+c     via solving S1 P12 = S12 for P12 etc.
+
+      g_s2  = ga_create_atom_blocked(geom, minbas, 'loc:s2')
+      call ga_zero(g_s2)
+      call int_1e_ga(minbas, minbas, g_s2, 'overlap', .false.)
+
+      if (.not. ga_create(MT_DBL, nbf, mnbf, 'loc:s12',
+     &  nbf, 0, g_s12)) call errquit(pname//': S12',61, GA_ERR)
+
+      call ga_zero(g_s12)
+      call int_1e_ga(basis, minbas, g_s12, 'overlap', .false.)
+
+      if (.not.ga_duplicate(g_s12, g_p12, 'g_p12')) call
+     &  errquit(pname//': ga_dup P12',62,GA_ERR)
+
+      call ga_copy(g_s12, g_p12)
+
+#ifdef SCALAPACK
+      info = ga_solve(g_smat, g_p12)
+      if (debug) write(luout,*)
+     &  'info from ga_solve P12:', info
+#else
+      if (debug) write(luout,*) 'call 1 ma_solve'
+      call ma_solve(g_smat, g_p12)
+#endif
+
+      if (.not. ga_create(MT_DBL, mnbf, nbf, 'loc:p21',
+     &  mnbf, 0, g_p21)) call errquit(pname//': P21',63, GA_ERR)
+
+      call ga_transpose(g_s12, g_p21)
+      if (debug) write(luout,*) 'P12 assembled'
+
+c     S12 no longer needed
+      if (.not. ga_destroy(g_s12))
+     &  call errquit(pname//': ga_destroy failed g_s12',61, GA_ERR)
+
+c : scalapack throws a 'ga_to_SL: rows/cols error' so we're using
+c       ma_solve for the time being
+c#ifdef SCALAPACK
+c        info = ga_solve(g_s2, g_p21)
+c        if (debug) write(luout,*)
+c     &    'info from ga_solve P21:', info
+c#else
+      if (debug) write(luout,*) 'call 2 ma_solve'
+      call ma_solve(g_s2, g_p21)
+c#endif
+
+c     g_s2 no longer needed
+      if (.not. ga_destroy(g_s2)) call errquit(
+     &  pname//': error destroying g_s2',0, GA_ERR)
+
+      if (dbg>1) then
+        call ga_print(g_p12)
+        call ga_print(g_p21)
+      end if
+
+c     store P12 * P21 in array g_temp; we'll need it soon
+
+      if (.not.ga_duplicate(g_smat, g_temp, 'g_temp')) call
+     &  errquit(pname//': ga_dup temp', 71, GA_ERR)
+
+      call ga_dgemm('n', 'n', nbf, nbf, mnbf,
+     &  1.0d0, g_p12, g_p21, 0.0d0, g_temp)
+
+      if (debug) write (luout,*) 'p12 p21 -> temp'
+
+c     P21 is no longer needed
+
+      if (.not. ga_destroy(g_p21))
+     &  call errquit(pname//': ga_destroy failed g_p21',61, GA_ERR)
+
+c     construct IAOS from occ. MOs:
+
+c     copy the relevant CMOs to g_mo
+
+      if (.not. ga_create(MT_DBL, nbf, nocc, 'loc:g_mo',
+     &  nbf, 0, g_mo)) call errquit(pname//': g_mo',0, GA_ERR)
+
+      call ga_copy_patch('n',
+     &  g_movecs, 1, nbf, 1, nocc,
+     &  g_mo,     1, nbf, 1, nocc)
+
+      if (debug) write (luout,*) 'movecs(occ) -> mo'
+
+c     create C-tilde from Appendix C of Knizia's IBO paper.  g_temp
+c     holds P12 * P21; we won't need it after the next matrix
+c     multiplication
+
+      if (.not. ga_create(MT_DBL, nbf, nocc, 'loc:ctilde',
+     &  nbf, 0, g_ctilde)) call errquit('loc_driver: sc',0, GA_ERR)
+
+      call ga_dgemm('n', 'n', nbf, nocc, nbf,
+     &  1.0d0, g_temp, g_mo, 0.0d0, g_ctilde)
+
+      if (.not. ga_destroy(g_temp))
+     &  call errquit(pname//': ga_destroy failed g_temp',71, GA_ERR)
+
+      if (debug) write (luout,*) 'pre sym_orth 1'
+      call sym_orth(g_ctilde, g_smat, nbf, nocc)
+
+      if (dbg>1) call ga_print(g_ctilde)
+
+c     now assemble the matrix products to form matrix A from Appendix
+c     C of the IBO paper. The result will be accumulated in array
+c     g_iao which are the IAOs in basis 1. The shape of the array is
+c     the same as P12
+
+      if (.not.ga_duplicate(g_smat, g_tmp1, 'g_tmp1')) call
+     &  errquit(pname//': ga_dup tmp1', 72, GA_ERR)
+
+      if (.not.ga_duplicate(g_smat, g_tmp2, 'g_tmp2')) call
+     &  errquit(pname//': ga_dup tmp2', 72, GA_ERR)
+
+      if (.not.ga_duplicate(g_smat, g_temp, 'g_temp')) call
+     &  errquit(pname//': ga_dup temp', 72, GA_ERR)
+
+      if (.not.ga_duplicate(g_p12, g_iao, 'g_iao')) call
+     &  errquit(pname//': ga_dup iao', 72, GA_ERR)
+
+
+c     C * C(T)
+      call ga_dgemm('n', 't', nbf, nbf, nocc,
+     &  1.0d0, g_mo, g_mo, 0.0d0, g_temp)
+
+c     C * C(T) * S1 -> tmp1
+      call ga_dgemm('n', 'n', nbf, nbf, nbf,
+     &  1.0d0, g_temp, g_smat, 0.0d0, g_tmp1)
+
+c     tilde-C * tilde-C(T)
+      call ga_dgemm('n', 't', nbf, nbf, nocc,
+     &  1.0d0, g_ctilde, g_ctilde, 0.0d0, g_temp)
+
+c     tilde-C * tilde-C(T) * S1 -> tmp2
+      call ga_dgemm('n', 'n', nbf, nbf, nbf,
+     &  1.0d0, g_temp, g_smat, 0.0d0, g_tmp2)
+
+c     C * C(T) * S1  * tilde-C * tilde-C(T) * S1
+
+      call ga_dgemm('n', 'n', nbf, nbf, nbf,
+     &  1.0d0, g_tmp1, g_tmp2, 0.0d0, g_temp)
+
+c     multiply prev. result by P12 and store in g_iao
+
+      call ga_dgemm('n', 'n', nbf, mnbf, nbf,
+     &  1.0d0, g_temp, g_p12, 0.0d0, g_iao)
+
+c     form C * C(T) -1
+      call ga_add_todiag(g_tmp1, -1.0d0)
+
+c     form tilde-C * tilde-C(T) -1
+      call ga_add_todiag(g_tmp2, -1.0d0)
+
+c     multiply the two prev. results with each other, then multiply
+c     the result with P12 and add to g_iao
+
+      call ga_dgemm('n', 'n', nbf, nbf, nbf,
+     &  1.0d0, g_tmp1, g_tmp2, 0.0d0, g_temp)
+
+      call ga_dgemm('n', 'n', nbf, mnbf, nbf,
+     &  1.0d0, g_temp, g_p12, 1.0d0, g_iao)
+
+c     g_ctilde no longer needed
+
+      if (.not. ga_destroy(g_ctilde))
+     &  call errquit(pname//': ga_destroy failed ctilde',62, GA_ERR)
+
+c     delete temp arrays before calling sym_orth, which allocates its
+c     own temp memory
+
+      if (.not. ga_destroy(g_tmp1))
+     &  call errquit(pname//': ga_destroy failed g_tmp1',71, GA_ERR)
+
+      if (.not. ga_destroy(g_tmp2))
+     &  call errquit(pname//': ga_destroy failed g_tmp2',71, GA_ERR)
+
+      if (.not. ga_destroy(g_temp))
+     &  call errquit(pname//': ga_destroy failed g_temp',71, GA_ERR)
+c     last step: orthonormalize g_iao
+
+      if (debug) write (luout,*) 'pre sym_orth 2'
+      call sym_orth(g_iao, g_smat, nbf, mnbf)
+
+      call ga_sync
+
+      if (dbg>1) call ga_print(g_iao)
+
+c     check orthogonality of IAOs w.r.t. metric S1
+c     g_p12 is used for scratch
+
+      if (.not. ga_create(MT_DBL, mnbf, mnbf, 'g_temp',
+     &  mnbf, 0, g_temp)) call errquit(pname//': temp',73, GA_ERR)
+
+      call ga_dgemm('n', 'n', nbf, mnbf, nbf,
+     &  1.0d0, g_smat, g_iao, 0.0d0, g_p12)
+
+      call ga_dgemm('t', 'n', mnbf, mnbf, nbf,
+     &  1.0d0, g_iao, g_p12, 0.0d0, g_temp)
+
+      if (dbg>1) call ga_print(g_temp)
+
+      call ga_add_todiag(g_temp, -1.0d0)
+
+      call ga_maxelt(g_temp, rtemp)
+      if (master) then
+        write(luout,'(/1x,a,1x,f18.8)')
+     &    'IBO loc: largest element in C(iao,T) S C(iao) -1: ',
+     &    rtemp
+        write(luout,'(1x,a/1x,a/)')
+     &    'Significant deviations from zero may indicate',
+     &    'elevated numerical noise in the IAO generation'
+      end if
+
+      if (.not. ga_destroy(g_temp))
+     &  call errquit(pname//': ga_destroy failed g_temp',73, GA_ERR)
+
+c     g_p12 no longer needed
+
+      if (.not. ga_destroy(g_p12))
+     &  call errquit(pname//': ga_destroy failed g_p12',61, GA_ERR)
+
+c     ---------------------------------------------------------------
+c     IAOs are now in array g_iao. Next, generate occupied or virtual
+c     IBOs, depending on the input settings (ltyp)
+c     ---------------------------------------------------------------
+
+      if (ltyp.eq.'occ') then
+
+c       -----------------------
+c       generate occupied IBOs:
+c       -----------------------
+
+c       note: g_mo already holds the occupied MOs
+
+        if (debug) then
+          write(luout,*) 'movecs(occ) -> mo'
+        end if
+
+c       few more sanity check, just in case
+        if (nocc.gt.mnbf) call errquit
+     &    (pname//': nocc > mnbf', 66, UNKNOWN_ERR)
+        if (nocc.gt.nbf) call errquit
+     &    (pname//': nocc > nbf', 66, UNKNOWN_ERR)
+
+c       transform the set of occupied MOs to be localized to the IAO
+c       basis via C(iao,T) S C(MO)
+
+        if (.not. ga_create(MT_DBL, mnbf, nocc , 'loc:cib',
+     &    mnbf, 0, g_cib))
+     &    call errquit(pname//': cib',0, GA_ERR)
+
+        if (.not. ga_create(MT_DBL, nbf, nocc , 'loc:tmp1',
+     &    0, 0, g_tmp1)) call errquit(pname//': tmp1',0, GA_ERR)
+
+        call ga_dgemm('n', 'n', nbf, nocc, nbf,
+     &    1.0d0, g_smat, g_mo, 0.0d0, g_tmp1)
+
+        call ga_dgemm('t', 'n', mnbf, nocc, nbf,
+     &    1.0d0, g_iao, g_tmp1, 0.0d0, g_cib)
+
+        if (.not. ga_destroy(g_tmp1))
+     &    call errquit(pname//': ga_destroy failed g_tmp1',0, GA_ERR)
+
+        if (dbg>1) call ga_print(g_cib)
+
+        if (debug) write(luout,*) 'cib done'
+
+c       save a copy of the starting MOs for later
+
+        if (.not.ga_duplicate(g_cib, g_tmp2, 'g_tmp2')) call
+     &    errquit(pname//': ga_dup cib', 0, GA_ERR)
+
+        call ga_copy(g_cib, g_tmp2)
+
+c       perform localization of the MOs in IAO basis:
+
+        call localizeIBO(minbas, c, g_cib, nocc, nbf, mnbf,
+     &    natoms)
+
+
+      else if (ltyp.eq.'vir') then
+
+c       ----------------------
+c       generate virtual IBOs:
+c       ----------------------
+
+c       store the virtual MOs in g_mo. The array is allocated, so we
+c       destroy it frst
+
+        if (.not. ga_destroy(g_mo)) call errquit(
+     &    pname//': error destroying g_mo',1, GA_ERR)
+
+        if (.not. ga_create(MT_DBL, nbf, nvir, 'loc:g_mo',
+     &    nbf, 0, g_mo)) call errquit(pname//': g_mo',1, GA_ERR)
+
+        call ga_copy_patch('n',
+     &    g_movecs, 1, nbf, nocc+1, nmo,
+     &    g_mo,     1, nbf, 1,      nvir)
+
+c       sanity check: for virtual localization, assume no. of
+c       virtuals >= (mnbf-nocc)
+
+        if (nvir.lt.(mnbf-nocc)) call errquit
+     &    (pname//': nvir < mnbf-nocc. Cannot handle', 66,
+     &    UNKNOWN_ERR)
+
+        if (debug) then
+          write(luout,*) 'movecs(vir) -> mo'
+          write(luout,*) 'nocc, nvir, nmo, nbf, mnbf',
+     &      nocc, nvir, nmo, nbf, mnbf
+        end if
+
+c       transform the set of virtual MOs to the IAO basis
+c       via C(iao,T) S C(MO), store in g_tmp2 (and keep the array)
+
+        if (.not. ga_create(MT_DBL, mnbf, nvir , 'loc:tmp2',
+     &    0, 0, g_tmp2)) call errquit(pname//': tmp2',0, GA_ERR)
+
+        if (.not. ga_create(MT_DBL, nbf, nvir , 'loc:tmp1',
+     &    0, 0, g_tmp1)) call errquit(pname//': tmp1',0, GA_ERR)
+
+        call ga_dgemm('n', 'n', nbf, nvir, nbf,
+     &    1.0d0, g_smat, g_mo, 0.0d0, g_tmp1)
+
+        call ga_dgemm('t', 'n', mnbf, nvir, nbf,
+     &    1.0d0, g_iao, g_tmp1, 0.0d0, g_tmp2)
+
+        if (.not. ga_destroy(g_tmp1))
+     &    call errquit(pname//': ga_destroy failed g_tmp1',0, GA_ERR)
+
+        if (debug .and. master) write(luout,*) 'cib before svd'
+
+c       now perform SVD on the matrix C(vir) that we just calculated and
+c       stored in array g_tmp2.  The "eigenvector" matrix U gives the
+c       true virtual MOs in the IAO basis that we want to localize
+
+c       #(sing. values) = min(mnbf,nvir) = mnbf , assuming virtuals.
+c       The non-zero singular values must number mnbf - nocc, which
+c       means we will localize a number of virtuals that's smaller than
+c       nvir and also smaller than min(mnbf,nvir).
+
+        nsing=min(mnbf,nvir)
+
+        if (.not. ma_push_get(MT_DBL, nsing, 'svd-vals', l_val, k_val))
+     &    call errquit(pname//': svd-vals', nsing, MA_ERR)
+
+        if (.not. ga_create(MT_DBL, mnbf, nsing,'g_u', 0, 0, g_u))
+     &    call errquit(pname//': svd-U', mnbf*nsing, GA_ERR)
+
+        if (.not. ga_create(MT_DBL, nsing, nvir, 'g_vt',0, 0, g_vt))
+     &    call errquit(pname//': svd-vT', nsing*nvir, GA_ERR)
+
+        if (debug) write(luout,*) 'svd: m,n,nsing=',mnbf,nvir,nsing
+
+        call wrap_dgesvd(g_tmp2, g_u, g_vt, dbl_mb(k_val))
+
+c       wrap_dgesvd leaves the content of the array in arg 1 intact, no
+c       matter what the options in the call to dgesvd are inside the
+c       routine, so g_tmp2 still contains the initial (redundant)
+c       virtual MO set expressed in the IAO basis. We'll use it later.
+
+        if (debug) then
+          write(luout,*) pname//' singular values:'
+          do imo = 1,mnbf
+            write(luout,*) imo, dbl_mb(k_val+imo-1)
+          end do
+        end if
+
+c       count the non-zero singular values: re-define nsing
+
+        nsing = 0
+        do imo = 1,mnbf
+          if (dbl_mb(k_val+imo-1) .gt.small) nsing = nsing+1
+        end do
+        if (master) write(luout,'(1x,a,i10)')
+     &    'non-zero singular values: ',nsing
+
+c       we should have as many non-zero singlular values as we have
+c       additional IAOs beyond the number of occupied MOs. Check that,
+c       and print a warning if the values differ.
+
+        if (nsing .ne. (mnbf-nocc)) then
+          if (master)
+     &      write(luout,'(/1x,a/1x,a/1x,a/1x,a,i6/1x,a,i6/1x,a)')
+     &      'WARNING: number of non-zero singular values differs from',
+     &      ' the number of IAOs minus the number of occupied MOs.',
+     &      ' Will continue suspiciously!',
+     &      ' nsing       : ', nsing,
+     &      ' mnbf - nocc : ', mnbf - nocc,
+     &      ' The numbers in the two previous lines should be the same.'
+        end if
+
+c       array g_cib will hold the left-singular vectors corresponding to
+c       non-zero singlular values. Since dgesvd returns them in
+c       descending order, and they are by definition positive, we simply
+c       take the first nsing vectors
+
+        if (.not. ga_create(MT_DBL, mnbf, nsing,'loc:cib', 0, 0, g_cib))
+     &    call errquit(pname//': cib', mnbf*nsing, GA_ERR)
+
+        call ga_copy_patch('n',
+     &    g_u,   1, mnbf, 1, nsing,
+     &    g_cib, 1, mnbf, 1, nsing )
+
+c       dellocate memory used for SVD
+
+        if (.not. ga_destroy(g_vt))
+     &    call errquit(pname//': ga_destroy failed g_vt',0, GA_ERR)
+
+        if (.not. ga_destroy(g_u))
+     &    call errquit(pname//': ga_destroy failed g_u',0, GA_ERR)
+
+        if (.not. ma_pop_stack(l_val)) call errquit(
+     &    pname//': error MA pop svd-vals',0, MA_ERR)
+
+        if (dbg>1) call ga_print(g_cib)
+
+        if (debug) write(luout,*) 'cib done'
+
+c       perform localization of the MOs in IAO basis:
+
+        call localizeIBO(minbas, c, g_cib, nsing, nbf, mnbf,
+     &    natoms)
+
+      end if ! ltyp occ or vir
+
+c     localization interations done
+
+c     calculate the localization transform. The starting MOs were
+c     saved in g_tmp2 in the IAO basis
+c     CMOs(iao,T) * LMOs(iao) = localization transform -> g_tmp1
+
+      if (ltyp.eq.'occ') then
+
+        if (.not. ga_create(MT_DBL, nocc, nocc, 'loc:tmp1',
+     &    0, 0, g_tmp1)) call errquit(pname//': tmp1',0, GA_ERR)
+
+        n1 = nocc ! dimensioning used later
+        n2 = nocc
+        noff = 0
+
+        call ga_dgemm('t', 'n', nocc, nocc, mnbf,
+     &    1.0d0, g_tmp2, g_cib, 0.0d0, g_tmp1)
+
+      if (debug) write(luout,*) 'ibo occ: transformation in g_tmp1'
+
+      elseif (ltyp.eq.'vir') then
+
+        n1 = nvir ! dimensioning used later
+        n2 = nsing
+        noff = nocc ! offset for virtual orbitals
+
+        if (.not. ga_create(MT_DBL, nvir, nsing, 'loc:tmp1',
+     &    0, 0, g_tmp1)) call errquit(pname//': tmp1',0, GA_ERR)
+
+        call ga_dgemm('t', 'n', nvir, nsing, mnbf,
+     &    1.0d0, g_tmp2, g_cib, 0.0d0, g_tmp1)
+
+      end if ! ltyp
+
+      if(noff+n1 > nbf) call errquit (pname//
+     &  ': noff + n1 > nbf', noff, UNKNOWN_ERR)
+
+      if (.not. ga_destroy(g_tmp2))
+     &  call errquit(pname//': ga_destroy failed g_tmp2',0, GA_ERR)
+
+c     transform MO energies. We assume that the Fock matrix was diagonal
+c     in the basis of input MOs. We also assume that the loop dimensions
+c     do not exceed the capacity of array c(:,:)
+
+      do jmo = 1,n2
+        call ga_get(g_tmp1, 1, n1, jmo, jmo, c(1,1), 1)
+        rtemp = 0.0d0
+        do imo = 1,n1
+          rtemp = rtemp + eval(imo+noff) * c(imo,1)**2
+        end do ! jmo
+        c(jmo,2) = rtemp ! transformed MO energies
+      end do ! imo
+      do imo = 1,n2
+        eval(imo+noff) = c(imo,2)
+      end do
+
+      if (debug) write(luout,*) 'transformed MO energies'
+
+c     sort IBOs by energy. We'll have to do this on the master
+c     node, or otherwise avoid having multiple nodes working on the
+c     same data
+
+      if (master) then
+      do imo = 1,n2 -1
+        minval = eval(imo+noff)
+c       find lowest eval(kmo) below eval(imo) (with offsets)
+        kmo = 0
+        do jmo = imo+1, n2
+          if (eval(jmo+noff).lt.minval) then
+            kmo = jmo
+            minval = eval(jmo+noff)
+          end if
+        end do
+c       if kmo > 0 we swap imo and jmo
+        if (kmo.ne.0) then
+          if (debug) write(luout,*) 'swap ',imo,kmo
+c         swap eigenvalues
+          swap = eval(imo+noff)
+          eval(imo+noff) = eval(kmo+noff)
+          eval(kmo+noff) = swap
+c         swap columns of LMO transformation
+          call ga_get(g_tmp1, 1, n1, imo, imo, c(1,1), 1)
+          call ga_get(g_tmp1, 1, n1, kmo, kmo, c(1,2), 1)
+          call ga_put(g_tmp1, 1, n1, imo, imo, c(1,2), 1)
+          call ga_put(g_tmp1, 1, n1, kmo, kmo, c(1,1), 1)
+c         swap LMOs in IAO basis, for the analysis below
+          call ga_get(g_cib, 1, mnbf, imo, imo, c(1,1), 1)
+          call ga_get(g_cib, 1, mnbf, kmo, kmo, c(1,2), 1)
+          call ga_put(g_cib, 1, mnbf, imo, imo, c(1,2), 1)
+          call ga_put(g_cib, 1, mnbf, kmo, kmo, c(1,1), 1)
+        end if
+      end do
+      end if ! master
+
+      call ga_sync
+
+c     write transformation to scratch file
+
+      call util_file_name('lmotrans',.true.,.true.,lmotrans)
+      if(.not.file_write_ga(lmotrans,g_tmp1)) call errquit
+     &  (pname//': could not write lmotrans',0, DISK_ERR)
+
+      if (debug) write (luout,*)
+     &  pname//': lmos reordered and lmotrans written'
+
+c     transform CMOs to re-ordered LMOs
+
+      if (.not. ga_create(MT_DBL, nbf, n2 , 'loc:tmp2',
+     &  nbf, 0, g_tmp2)) call errquit(pname//': tmp2',0, GA_ERR)
+
+      call ga_dgemm('n', 'n', nbf, n2, n1,
+     &  1.0d0, g_mo, g_tmp1, 0.0d0, g_tmp2)
+
+
+      if (debug) write (luout,*) pname//': g_tmp2 now has LMOs'
+
+      if (.not. ga_destroy(g_tmp1))
+     &  call errquit(pname//': ga_destroy failed g_tmp1',0, GA_ERR)
+
+c     copy the IBOs into the relevant part of movecs
+
+      if (ltyp.eq.'occ') then
+        call ga_copy_patch('n',
+     &    g_tmp2,   1, nbf, 1, nocc,
+     &    g_movecs, 1, nbf, 1, nocc)
+
+      else if (ltyp.eq.'vir') then
+        call ga_copy_patch('n',
+     &    g_tmp2,   1, nbf, 1, nsing,
+     &    g_movecs, 1, nbf, nocc+1, nocc+nsing)
+      end if
+
+      if (debug) write (luout,*) pname//': IBOs copied to g_movecs'
+
+      if (.not. ga_destroy(g_tmp2))
+     &  call errquit(pname//': ga_destroy failed g_tmp2',0, GA_ERR)
+
+c     Analyze localization of each MO: per LMO, a list of atomic
+c     populations is printed in decreasing magnitude, with the
+c     polulations in parentheses.  This code is equivalent to the on
+c     found in the Pipek-Mezey localization routine
+
+      if (master) then
+        write(luout,'(/1x,a/)') 'IAO-IBO localized orbitals'
+        do s = 1, n2
+          call ga_get(g_cib,  1, mnbf, s, s, c(1,1), 1)
+          nlist = 0
+          do a = 1, natoms
+            if (.not. bas_ce2bfr(minbas, a, bflo, bfhi))
+     &        call errquit('localized: basis ', 3,
+     &        BASIS_ERR)
+            qas  = 0.0d0
+            do u = bflo, bfhi
+              qas  = qas  + c(u,1)*c(u,1)
+            end do
+            if (abs(qas) .gt. 0.01d0) then
+              nlist = nlist + 1
+              list(nlist) = a
+              pop(nlist) = qas
+            end if
+          end do
+          do u = 1, nlist
+            do t = 1, u-1
+              if (abs(pop(t)).lt.abs(pop(u))) then
+                rtemp = pop(u)
+                pop(u) = pop(t)
+                pop(t) = rtemp
+                tt = list(u)
+                list(u) = list(t)
+                list(t) = tt
+              end if
+            end do
+          end do
+          write(luout,9002) s, eval(s+noff),
+     &      occ(s+noff),(list(a), pop(a), a=1,nlist)
+ 9002     format(i5, 1x, f14.6,1x, f5.3, 1x,100(2x,i4,'(',f5.2,')'))
+        end do
+        write(luout,*)
+        call util_flush(luout)
+      end if
+
+      if (.not. ga_destroy(g_cib))
+     &  call errquit(pname//': ga_destroy failed g_cib',0, GA_ERR)
+
+c     deallocate remaining arrays that are no longer needed
+
+      if (.not. ga_destroy(g_iao)) call errquit(
+     &  pname//': error destroying g_iao',0, GA_ERR)
+
+      if (.not. ga_destroy(g_mo)) call errquit(
+     &  pname//': error destroying g_mo',0, GA_ERR)
+
+c     smat not needed anymore
+      if (.not. ga_destroy(g_smat)) call errquit(
+     &  pname//': error destroying g_smat',0, GA_ERR)
+
+c     destroy minimal basis (iao basis)
+      if (.not.bas_destroy(minbas))
+     &  call errquit(pname//'iao bas_destroy failed',0,BASIS_ERR)
+
+      if (ltyp.eq.'occ') then
+        n1 = 1
+        n2 = nocc
+      else if (ltyp.eq.'vir') then
+        n1 = nocc + 1
+        n2 = nocc + nsing
+      end if
+
+      if (master) write(luout,
+     &  '(/1x,a,a,a/1x,a/1x,i10,2x,a,1x,i10/)')
+     &  'IBO localization (',ltyp,'): IBOs will be stored',
+     &  'in file locorb.movecs, number ',n1, 'to', n2
+
+      return
+
+      end
+
+
+
+c     =================================================================
+
+      subroutine sym_orth(g_c, g_s, n1, n2)
+      
+      implicit none
+      
+#include "global.fh"
+#include "mafdecls.fh"
+#include "errquit.fh"
+#include "stdio.fh"
+#include "numerical_constants.fh"
+      
+c     Given input vectors collected in matrix C, and a metric S,
+c     calculate
+c     C' = C times [transpose(C) S C]**(-1/2) such that
+c     transpose(C') S C' = unit matrix
+      
+c     matrix S is square symmetric, dimension n1, n1
+c     matrix C has dimension n1, n2
+      
+c     in the calling routine, matrix C may be allocated with more than
+c     n2 columns, which is the reason we use ga_copy_patch below
+      
+      integer g_c !  in/out
+      integer g_s !  input
+      
+      integer n1, n2 ! input
+      
+      integer g_sc, g_csc
+      
+      character*(8) pname
+      
+      logical debug
+      
+c     -----------------------------------------------------------------
+      
+      pname = 'sym_orth'
+      
+      debug = .false. .and. ga_nodeid().eq.0 !during development
+      
+      if (debug) write(luout,*) 'entering '//pname
+      
+      if(.not.ga_create(mt_dbl,n1,n2,'g_sc',0,0,g_sc))
+     &  call errquit(pname//': error creating g_sc',141, GA_ERR)
+      
+      if(.not.ga_create(mt_dbl,n2,n2,'g_csc',0,0,g_csc))
+     &  call errquit(pname//': error creating g_csc',142, GA_ERR)
+      
+      call ga_dgemm('n','n',n1, n2, n1, 1.0d0,
+     &  g_s, g_c, 0.0d0, g_sc)
+      
+      call ga_dgemm('t','n',n2, n2, n1, 1.0d0,
+     &  g_c, g_sc, 0.0d0, g_csc)
+      
+      call ga_matpow(g_csc, -FP_Half, 1d-12)
+      
+      call ga_copy_patch('n',
+     &  g_c,  1, n1, 1, n2,
+     &  g_sc, 1, n1, 1, n2)
+      
+      call ga_dgemm('n','n',n1, n2, n2, 1.0d0,
+     &  g_sc, g_csc, 0.0d0, g_c)
+      
+      
+      if (.not. ga_destroy(g_sc))
+     &  call errquit(pname//': error destroy g_sc',141, GA_ERR)
+      
+      if (.not. ga_destroy(g_csc))
+     &  call errquit(pname//': error destroy g_csc',142, GA_ERR)
+      
+      if (debug) write(luout,*) 'leaving '//pname
+      
+      return
+      
+      end
+      
+c     =================================================================
+
+      subroutine wrap_dgesvd(g_a, g_u, g_vt, values)
+      
+      implicit none
+      
+#include "errquit.fh"
+#include "global.fh"
+#include "mafdecls.fh"
+      
+      integer g_a, g_u, g_vt
+      double precision values(*)
+
+c     wrapper routine for call to dgesvd. This is similar to ga_svd_seq
+c     in util/ga_it2.F but some of the array dimensions had to be
+c     adjusted to match the usage case in the virtual IBO construction
+c     where m < n
+      
+      integer n, m, type, l_a, k_a, l_u, k_u, l_vt, k_vt,
+     &  l_work, k_work, lwork, info, nsing
+
+      character*(11) pname
+
+c     -----------------------------------------------------------------
+
+      pname = 'wrap_dgesvd'
+      
+      call ga_inquire(g_a, type, m, n)
+
+      if (type.ne.MT_DBL) then
+        call errquit (pname//': array type not MT_DBL',
+     &    type, UNKNOWN_ERR)
+      end if
+      
+      nsing = min(m,n)
+
+      if (ga_nodeid() .eq. 0) then
+        
+        lwork = 10*max(m,n)
+        if (.not. ma_push_get(type, m*n, 'svd', l_a, k_a))
+     &    call errquit(pname//': a',m*n, MA_ERR)
+        if (.not. ma_push_get(type, m*nsing, 'svd', l_u, k_u))
+     &    call errquit(pname//': u',m*nsing, MA_ERR)
+        if (.not. ma_push_get(type, nsing*n, 'svd', l_vt, k_vt))
+     &    call errquit(pname//': vt',nsing*n, MA_ERR)
+        if (.not. ma_push_get(type, lwork, 'svd', l_work, k_work))
+     &    call errquit(pname//': work',lwork, MA_ERR)
+        
+        call ga_get(g_a, 1, m, 1, n, dbl_mb(k_a), m)
+
+        call dgesvd('s','s',m,n,dbl_mb(k_a),m,values,
+     &    dbl_mb(k_u),m,dbl_mb(k_vt),nsing,
+     &    dbl_mb(k_work),lwork,info)
+           
+        if (info.ne.0) call errquit(pname//':svd failed',info,MEM_ERR)
+        
+        call ga_put(g_u,  1, m,     1, nsing, dbl_mb(k_u),  m)
+        call ga_put(g_vt, 1, nsing, 1, n,     dbl_mb(k_vt), nsing)
+
+        if (.not. ma_chop_stack(l_a)) call errquit(pname//': svd ma',0,
+     &    MA_ERR)
+         
+      end if
+      
+      call ga_sync()
+      call ga_brdcst(1,values,nsing*8,0)
+      call ga_sync()
+
+      end      

src/property/localization_driver.F

@@ -34,8 +34,8 @@ c     ==================================================================
 #include "apiP.fh"
 #include "prop.fh"
 #include "bgj.fh"
-#include "util.fh"    
-
+#include "util.fh"
+#include "nwc_const.fh"
 
 c     subroutine arguments:
       integer rtdb    ! [input] rtdb handle
@@ -46,17 +46,18 @@ c     subroutine arguments:
 c     local GA handles:
       integer g_uc(4), g_smat, g_sc, g_tran
       integer g_movecs(2), g_cmo(2), g_temp, g_tmp1, g_tmp2
-      integer g_s2, g_s12, g_p12, g_p21
-      integer g_ctilde, g_iao, g_mo
 
 c     MA variables:
       integer l_c, k_c, l_sc, k_sc, l_eval, k_eval, l_occ, k_occ
       integer l_dip(3), k_dip(3)
+      integer l_pop, k_pop, l_list, k_list
 
 c     other local variables:
 
-      integer nclosed(2), nopen(2), nvirt(2), ncore,
-     &  ndens, nbf, nmo, nloc, i
+      integer loc_opt
+
+      integer nclosed(2), nopen(2), nvirt(2), ncore, nocc, nvir,
+     &  ndens, nbf, nmo, nloc, i, natoms
 
       integer info, dbg, ispin
       
@@ -92,13 +93,8 @@ c     other local variables:
       logical int_normalize
       external int_normalize
 
-c     variables for IAO basis
-      integer
-     &  nbf_iao, nbf_iao_mxprim, nbf_iao_mxang, nbf_iao_mxcont,
-     &  nbf_iao_mxnbf_cn, nbf_iao_mxnbf_ce, nshells_iao,
-     &  minbas      
-
       character*(19) pname
+      character*(3) ltyp
 
 c     ==================================================================
 
@@ -124,20 +120,52 @@ c     ----------------------
         write (luout,9000) ! general header
         if (loctype.eq.1) then
           call util_print_centered(luout,
-     &       'Pipek-Mezey localization', 32, .true.)
+     &       'Pipek-Mezey localization', 25, .true.)
         elseif (loctype.eq.2) then
           call util_print_centered(luout,
      &      'Boys localization', 25, .true.)
         elseif (loctype.eq.3) then
           call util_print_centered(luout,
-     &       'IAO-IBO localization', 20, .true.)
+     &      'IAO-IBO localization', 25, .true.)
         else
           write (luout,*) 'loctype =',loctype
           write (luout,*) 'LocType value not supported. Exiting.'
           return
         end if
       end if
-      
+
+c     retrieve localization option from rtdb if it exists
+
+      if (.not. rtdb_get(rtdb, 'prop:loc_opt', mt_int, 1,
+     &                               loc_opt  ))
+     &  loc_opt = 0
+
+c     option is currently only implemented for IBOs:
+
+      if (loc_opt.gt.0 .and. (loctype.ne.3)) then
+        if (oprint) write(luout,'(/1x,a/1x,a)')
+     &    'localization option not implemented for chosen loc. type',
+     &    'setting option to 0 ...'
+        loc_opt = 0
+      end if
+
+      if (oprint) then
+        if (loc_opt .eq.0) then
+          call util_print_centered(luout,
+     &      '(occupied orbitals)', 25, .false.)
+        else if (loc_opt.eq.1) then
+          call util_print_centered(luout,
+     &      '(virtual orbitals)', 25, .false.)
+        else if (loc_opt.eq.2) then
+          call util_print_centered(luout,
+     &      '(occ and virt orbitals)', 25, .false.)
+        else
+          call errquit(pname//': localization option out of range',
+     &      loc_opt, INPUT_ERR)
+        end if
+      end if
+
+
 
 c     -------------------------------------------------------
 c     assemble some data:
@@ -160,8 +188,19 @@ c     allocate dbl_mb(leval) = orbital energies:
 
       if (debug)
      &   write (luout,*) 'loc driver: nclosed,nopen,nvirt',nclosed(1),
-     &   nopen(1), nvirt(1)
-      
+     &  nopen(1), nvirt(1)
+
+      natoms = 0
+
+      if (.not. geom_ncent(geom, natoms)) call errquit
+     &  (pname//': geom',0, GEOM_ERR)
+
+      if (natoms.lt.1) call errquit
+     &  (pname//': could not get natoms from rtdb', 911, UNKNOWN_ERR)
+
+      if (natoms.gt.nw_max_atom) call errquit
+     &  (pname//': nw_max_atom too small ', 911, UNKNOWN_ERR)
+
 c     Skip localization if we have an open-shell system
 c     (to be extended eventually, localizing alpha and
 c     beta spin orbitals separately):
@@ -190,9 +229,15 @@ c     dynamically but for now let's make sure we don't get out of
 c     bounds:
 
       if (nmo.gt.maxnloc) call
-     &   errquit('loc_driver: maxnloc too small',0, BASIS_ERR)
+     &  errquit('loc_driver: maxnloc too small',0, BASIS_ERR)
+
+c     for the time being, we set these variables here:
+
+      ispin = 1
+      nocc = nclosed(1)
+      nvir = nvirt(ispin)
+
 
-      
 c     --------------------------------------------
 c     setup complete. Now proceed to the localization
 c     --------------------------------------------
@@ -471,335 +516,72 @@ c        smat not needed anymore
          if (.not. ga_destroy(g_smat)) call errquit(
      &     'loc_driver: error destroying g_smat',0, GA_ERR)
 
-c      ================================================================
+
+c      ===============================================================
 
        else if (loctype.eq.3) then
 
-c       ====================
-c       IAO-IBO localization
-c       ====================
-        
-c       AO Overlap Matrix S1 -> g_smat:
-
-        g_smat  = ga_create_atom_blocked(geom, basis, 'loc:smat')
-        call ga_zero(g_smat)
-        call int_1e_ga(basis, basis, g_smat, 'overlap', .false.)
-
-c       auxiliary basis overlap S2 -> g_s2 basis needs to defined in the
-c       input as "iao basis". we will now create the basis here and then
-c       calculate the overlap and the mixed ao-iao basis overlap S12
-        
-        if (.not. bas_create(minbas, 'iao basis'))
-     &    call errquit(pname//': cannot create iao bas', 86, BASIS_ERR)
-        
-        if (.not. bas_rtdb_load(rtdb, geom, minbas, 'iao basis'))
-     &    call errquit(pname//': iao basis not present', 86, BASIS_ERR)
-        
-        if (.not. int_normalize(rtdb,minbas))
-     &    call errquit(pname//': int_normalize failed iao', 86, INT_ERR)
-        
-        if (.not. bas_numbf(minbas, nbf_iao))
-     &    call errquit(pname//': bas error nbf_iao', 86, BASIS_ERR)
-        
-        if (.not. bas_nprim_cn_max(minbas,nbf_iao_mxprim))
-     &    call errquit(pname//': bas error iao mxprim', 86, BASIS_ERR)
-        
-        if (.not. bas_high_angular(minbas,nbf_iao_mxang))
-     &    call errquit(pname//': bas error iao mxang', 86, BASIS_ERR)
-        
-        if (.not. bas_ncontr_cn_max(minbas,nbf_iao_mxcont))
-     &    call errquit(pname//': bas error iao mxcont', 86, BASIS_ERR)
-        
-        if (.not. bas_nbf_cn_max(minbas,nbf_iao_mxnbf_cn))
-     &    call errquit(pname//': bas error iao mcnbf_cn', 86, BASIS_ERR)
-        
-        if (.not. bas_nbf_ce_max(minbas,nbf_iao_mxnbf_ce))
-     &    call errquit(pname//': bas error iao mxnbf_ce', 86, BASIS_ERR)
-        
-        if (.not. bas_numcont(minbas,nshells_iao))
-     &    call errquit(pname//': bas error iao nshells', 86, BASIS_ERR)
-
-c       create overlap for minbas, and the mixed basis-minbas
-c       overlap S12.
-c       Then we calculate the projectors P12 and P21
-c       via solving S1 P12 = S12 for P12 etc.
-        
-        g_s2  = ga_create_atom_blocked(geom, minbas, 'loc:s2')
-        call ga_zero(g_s2)
-        call int_1e_ga(minbas, minbas, g_s2, 'overlap', .false.)
-
-        if (.not. ga_create(MT_DBL, nbf, nbf_iao, 'loc:s12',
-     &    nbf, 0, g_s12)) call errquit(pname//': S12',61, GA_ERR)
-
-        call ga_zero(g_s12)
-        call int_1e_ga(basis, minbas, g_s12, 'overlap', .false.)
-
-        if (.not.ga_duplicate(g_s12, g_p12, 'g_p12')) call
-     &    errquit(pname//': ga_dup P12',62,GA_ERR)
-
-        call ga_copy(g_s12, g_p12)
-        
-#ifdef SCALAPACK
-        info = ga_solve(g_smat, g_p12) 
-        if (debug) write(luout,*)
-     &    'info from ga_solve P12:', info
-#else
-        if (debug) write(luout,*) 'call 1 ma_solve' 
-        call ma_solve(g_smat, g_p12)
-#endif
-
-        if (.not. ga_create(MT_DBL, nbf_iao, nbf, 'loc:p21',
-     &    nbf_iao, 0, g_p21)) call errquit(pname//': P21',63, GA_ERR)
-
-        call ga_transpose(g_s12, g_p21)
-        if (debug) write(luout,*) 'P12 assembled' 
-
-c       S12 no longer needed
-        if (.not. ga_destroy(g_s12))
-     &    call errquit(pname//': ga_destroy failed g_s12',61, GA_ERR)  
-
-c : scalapack throws a 'ga_to_SL: rows/cols error' so we're using
-c       ma_solve for the time being
-c#ifdef SCALAPACK
-c        info = ga_solve(g_s2, g_p21) 
-c        if (debug) write(luout,*)
-c     &    'info from ga_solve P21:', info
-c#else
-        if (debug) write(luout,*) 'call 2 ma_solve' 
-        call ma_solve(g_s2, g_p21)
-c#endif
-
-c       g_s2 no longer needed
-        if (.not. ga_destroy(g_s2)) call errquit(
-     &    pname//': error destroying g_s2',0, GA_ERR)
-
-        if (dbg>1) then
-          call ga_print(g_p12)
-          call ga_print(g_p21)
-        end if
-
-c       store P12 * P21 in array g_temp; we'll need it soon
-
-        if (.not.ga_duplicate(g_smat, g_temp, 'g_temp')) call
-     &    errquit(pname//': ga_dup temp', 71, GA_ERR)
-        
-        call ga_dgemm('n', 'n', nbf, nbf, nbf_iao,
-     &    1.0d0, g_p12, g_p21, 0.0d0, g_temp)
-
-         if (debug) write (luout,*) 'p12 p21 -> temp'        
-
-c        P21 is no longer needed
-         
-         if (.not. ga_destroy(g_p21))
-     &     call errquit(pname//': ga_destroy failed g_p21',61, GA_ERR)
-        
-c        we will localize the doubly occupied orbitals, assuming a
-c        closed-shell calculation and no designated core
-c        corbitals. Eventually, this will need to be extended to a loop
-c        over spins, for unrestricted SCF, and we may want to add
-c        localization of the virtual space, too
-
-        if (.not. geom_num_core(rtdb, geom, 'ddscf', ncore)) ncore = 0
-
-        if (debug) write (luout,*) 'ncore = ',ncore
-
-        ispin = 1
-        
-         do i = 1, nclosed(ispin)
-            iloc(i) = i
-         end do
-         nloc = nclosed(ispin)
-
-c        copy the relevant CMOs to g_mo
-         
-         if (.not. ga_create(MT_DBL, nbf, nloc, 'loc:g_mo',
-     &     nbf, 0, g_mo)) call errquit(pname//': g_mo',0, GA_ERR)
-
-         call ga_copy_patch('n',
-     &     g_movecs(ispin), 1, nbf, 1, nloc,
-     &     g_mo,            1, nbf, 1, nloc)
-
-         if (debug) write (luout,*) 'movecs -> mo'         
-
-c        create C-tilde from Appendix C of Knizia's IBO paper.  g_temp
-c        holds P12 * P21; we won't need it after the next matrix
-c        multiplication
-         
-         if (.not. ga_create(MT_DBL, nbf, nloc, 'loc:ctilde',
-     &     nbf, 0, g_ctilde)) call errquit('loc_driver: sc',0, GA_ERR)
-
-         call ga_dgemm('n', 'n', nbf, nloc, nbf, 
-     &     1.0d0, g_temp, g_mo, 0.0d0, g_ctilde)
-
-         if (.not. ga_destroy(g_temp))
-     &     call errquit(pname//': ga_destroy failed g_temp',71, GA_ERR)
-
-         if (debug) write (luout,*) 'pre sym_orth 1'
-         call sym_orth(g_ctilde, g_smat, nbf, nloc)
-
-         if (dbg>1) call ga_print(g_ctilde)
-
-c        now assemble the matrix products to form matrix A from Appendix
-c        C of the IBO paper. The result will be accumulated in array
-c        g_iao which are the IAOs in basis 1. The shape of the array is
-c        the same as P12
-
-         if (.not.ga_duplicate(g_smat, g_tmp1, 'g_tmp1')) call
-     &     errquit(pname//': ga_dup tmp1', 72, GA_ERR)
-         
-         if (.not.ga_duplicate(g_smat, g_tmp2, 'g_tmp2')) call
-     &     errquit(pname//': ga_dup tmp2', 72, GA_ERR)
-
-         if (.not.ga_duplicate(g_smat, g_temp, 'g_temp')) call
-     &     errquit(pname//': ga_dup temp', 72, GA_ERR)
-
-         if (.not.ga_duplicate(g_p12, g_iao, 'g_iao')) call
-     &     errquit(pname//': ga_dup iao', 72, GA_ERR)
-
-
-c        C * C(trans) 
-         call ga_dgemm('n', 't', nbf, nbf, nloc, 
-     &     1.0d0, g_mo, g_mo, 0.0d0, g_temp)
-
-c        C * C(trans) * S1 -> tmp1
-         call ga_dgemm('n', 'n', nbf, nbf, nbf, 
-     &     1.0d0, g_temp, g_smat, 0.0d0, g_tmp1)
-
-c        tilde-C * tilde-C(trans) 
-         call ga_dgemm('n', 't', nbf, nbf, nloc, 
-     &     1.0d0, g_ctilde, g_ctilde, 0.0d0, g_temp)
-
-c        tilde-C * tilde-C(trans) * S1 -> tmp2
-         call ga_dgemm('n', 'n', nbf, nbf, nbf, 
-     &     1.0d0, g_temp, g_smat, 0.0d0, g_tmp2)         
-
-c        C * C(trans) * S1  * tilde-C * tilde-C(trans) * S1
-
-         call ga_dgemm('n', 'n', nbf, nbf, nbf, 
-     &     1.0d0, g_tmp1, g_tmp2, 0.0d0, g_temp)
-
-c        multiply prev. result by P12 and store in g_iao
-
-         call ga_dgemm('n', 'n', nbf, nbf_iao, nbf, 
-     &     1.0d0, g_temp, g_p12, 0.0d0, g_iao)
-         
-c        form C * C(trans) -1
-         call ga_add_todiag(g_tmp1, -1.0d0)
-         
-c        form tilde-C * tilde-C(trans) -1
-         call ga_add_todiag(g_tmp2, -1.0d0)
-
-c        multiply the two prev. results with each other, then multiply
-c        the result with P12 and add to g_iao
-         
-         call ga_dgemm('n', 'n', nbf, nbf, nbf, 
-     &     1.0d0, g_tmp1, g_tmp2, 0.0d0, g_temp)
-
-         call ga_dgemm('n', 'n', nbf, nbf_iao, nbf, 
-     &     1.0d0, g_temp, g_p12, 1.0d0, g_iao)
-
-c        g_ctilde no longer needed
-         
-         if (.not. ga_destroy(g_ctilde))
-     &     call errquit(pname//': ga_destroy failed ctilde',62, GA_ERR)
-
-c        delete temp arrays before calling sym_orth, which allocates its
-c        own temp memory
-
-         if (.not. ga_destroy(g_tmp1))
-     &     call errquit(pname//': ga_destroy failed g_tmp1',71, GA_ERR)
-         
-         if (.not. ga_destroy(g_tmp2))
-     &     call errquit(pname//': ga_destroy failed g_tmp2',71, GA_ERR)
-         
-         if (.not. ga_destroy(g_temp))
-     &     call errquit(pname//': ga_destroy failed g_temp',71, GA_ERR)         
-c        last step: orthonormalize g_iao
-
-         if (debug) write (luout,*) 'pre sym_orth 2'
-         call sym_orth(g_iao, g_smat, nbf, nbf_iao)
-
-         call ga_sync
-         
-         if (dbg>1) call ga_print(g_iao)
-
-c        check orthogonality of IAOs w.r.t. metric S1
-c        g_p12 is used for scratch
-
-        if (.not. ga_create(MT_DBL, nbf_iao, nbf_iao, 'g_temp',
-     &    nbf_iao, 0, g_temp)) call errquit(pname//': temp',73, GA_ERR)
-         
-         call ga_dgemm('n', 'n', nbf, nbf_iao, nbf, 
-     &     1.0d0, g_smat, g_iao, 0.0d0, g_p12)
-
-         call ga_dgemm('t', 'n', nbf_iao, nbf_iao, nbf, 
-     &     1.0d0, g_iao, g_p12, 0.0d0, g_temp)
-
-         if (dbg>1) call ga_print(g_temp)
-
-         call ga_add_todiag(g_temp, -1.0d0)
-
-         call ga_maxelt(g_temp, rtemp)
-         if (master) then
-           write(luout,'(/1x,a,1x,f18.12)')
-     &       'IBO loc: largest element in C(iao, trans) S C(iao) -1: ',
-     &     rtemp
-           write(luout,'(1x,a/1x,a/)')
-     &       'Significant deviations from zero may indicate',
-     &       'elevated numerical noise in the IAO generation'
-         end if
-
-         if (.not. ga_destroy(g_temp))
-     &     call errquit(pname//': ga_destroy failed g_temp',73, GA_ERR)
-
-c        g_p12 no longer needed
-         
-         if (.not. ga_destroy(g_p12))
-     &     call errquit(pname//': ga_destroy failed g_p12',61, GA_ERR)
-         
-c        allocate some memory used in the localization routine:
-
-         if (.not. ma_push_get(mt_dbl, 2*nbf_iao, 'c', l_c, k_c))
-     &        call errquit(pname//': ma for c', 0, MA_ERR)
-
-
-c        jochen: comment
-c        IBO localization was modeled on the pre-existing PM
-c        localization code
-         call localizeIBO(basis, minbas,
-     &     dbl_mb(k_c),
-     &     nloc, iloc, nbf, nmo, g_mo, g_smat, g_iao, nbf_iao,
-     &     dbl_mb(k_eval+(ispin-1)*nbf), dbl_mb(k_occ+(ispin-1)*nbf))
-
-c        g_mo now holds the localized MOs in the AO basis
-
-c        assuming occupied MO localization, we copy the LMOs into the
-c        relevant part of movecs
-
-         call ga_copy_patch('n',
-     &     g_mo,        1, nbf, 1, nloc,
-     &     g_movecs(1), 1, nbf, 1, nloc)
-
-
-c        deallocate remaining arrays that are no longer needed
-         
-         if (.not. ma_pop_stack(l_c)) call errquit(
-     &      pname//': error MA pop c',0, MA_ERR)
-
-         if (.not. ga_destroy(g_iao)) call errquit(
-     &     pname//': error destroying g_iao',0, GA_ERR)
-
-         if (.not. ga_destroy(g_mo)) call errquit(
-     &     pname//': error destroying g_mo',0, GA_ERR)         
-         
-c        smat not needed anymore
-         if (.not. ga_destroy(g_smat)) call errquit(
-     &     pname//': error destroying g_smat',0, GA_ERR)
-
-c        destroy minimal basis (iao basis)
-         if (.not.bas_destroy(minbas))
-     &     call errquit(pname//'iao bas_destroy failed',0,BASIS_ERR)
-
+c        ===================================================
+c        IAO-IBO localization (occupied or virtual orbitals)
+c        ===================================================
+
+
+         ispin = 1
+         nocc = nclosed(ispin)
+         nvir = nvirt(ispin)
+
+c        allocate MA array for pairs of MOs. The actual dimensions
+c        needed are less or equal to 2*nbf
+
+         if (.not. ma_push_get(mt_dbl, 2*nbf, 'c mo', l_c, k_c))
+     &     call errquit(pname//': ma for c', 0, MA_ERR)
+
+         if (.not. ma_push_get(mt_dbl, natoms, 'pop', l_pop, k_pop))
+     &     call errquit(pname//': loc:pop', 0, MA_ERR)
+
+         if (.not. ma_push_get(mt_int, natoms, 'list', l_list, k_list))
+     &     call errquit(pname//': loc:list', 0, MA_ERR)
+
+         if (loc_opt.eq.0) then
+           ltyp = 'occ'
+           call ibo_localization(rtdb, geom, ltyp, basis,
+     &       g_movecs(ispin),nocc,nvir, nmo, nbf, natoms,
+     &       dbl_mb(k_eval+(ispin-1)*nbf),
+     &       dbl_mb(k_occ+(ispin-1)*nbf),
+     &       dbl_mb(k_c),
+     &       dbl_mb(k_pop), int_mb(k_list))
+
+         else if (loc_opt.eq.1) then
+           ltyp = 'vir'
+           call ibo_localization(rtdb, geom, ltyp, basis,
+     &       g_movecs(ispin),nocc,nvir, nmo, nbf, natoms,
+     &       dbl_mb(k_eval+(ispin-1)*nbf),
+     &       dbl_mb(k_occ+(ispin-1)*nbf),
+     &       dbl_mb(k_c),
+     &       dbl_mb(k_pop), int_mb(k_list))
+
+         else if(loc_opt.eq.2) then
+           ltyp = 'occ'
+           call ibo_localization(rtdb, geom, ltyp, basis,
+     &       g_movecs(ispin),nocc,nvir, nmo, nbf, natoms,
+     &       dbl_mb(k_eval+(ispin-1)*nbf),
+     &       dbl_mb(k_occ+(ispin-1)*nbf),
+     &       dbl_mb(k_c),
+     &       dbl_mb(k_pop), int_mb(k_list))
+
+           ltyp = 'vir'
+           call ibo_localization(rtdb, geom, ltyp, basis,
+     &       g_movecs(ispin),nocc,nvir, nmo, nbf, natoms,
+     &       dbl_mb(k_eval+(ispin-1)*nbf),
+     &       dbl_mb(k_occ+(ispin-1)*nbf),
+     &       dbl_mb(k_c),
+     &       dbl_mb(k_pop), int_mb(k_list))
+
+         else
+           call errquit(pname//': loc_opt out of range',loc_opt,
+     &       UNKNOWN_ERR)
+
+         end if ! loc_opt
 
        end if ! loctype
 
@@ -849,84 +631,6 @@ c     ==================================================================
      &  10x,28(1h*),/,10x,'*** ORBITAL LOCALIZATION ***',/,
      &  10x,28(1h*)/)
 
- 9020 format(/,10x,'Open shell calculations not yet implemented',
+ 9020 format(/,10x,'Spin-unrestricted calculations not yet implemented',
      &      ' for the Localization module. Exiting')
       end
-
-c     =================================================================
-
-      subroutine sym_orth(g_c, g_s, n1, n2)
-
-      implicit none
-
-#include "global.fh"
-#include "mafdecls.fh"
-#include "errquit.fh"
-#include "stdio.fh"      
-#include "numerical_constants.fh"      
-
-c     Given input vectors collected in matrix C, and a metric S,
-c     calculate
-c     C' = C times [transpose(C) S C]**(-1/2) such that
-c     transpose(C') S C = unit matrix
-
-c     matrix S is square symmetric, dimension n1, n1
-c     matrix C has dimension n1, n2
-      
-c     in the calling routine, matrix C may be allocated with more than
-c     n2 columns, which is the reason we use ga_copy_patch below
-      
-      integer g_c !  in/out
-      integer g_s !  input
-
-      integer n1, n2  ! input
-
-      integer g_sc, g_csc
-
-      character*(8) pname
-
-      logical debug
-
-c     -----------------------------------------------------------------
-
-      pname = 'sym_orth'
-
-      debug = .false. .and. ga_nodeid().eq.0 !during development
-
-      if (debug) write(luout,*) 'entering '//pname
-
-       if(.not.ga_create(mt_dbl,n1,n2,'g_sc',0,0,g_sc))
-     &  call errquit(pname//': error creating g_sc',141, GA_ERR)
-
-       if(.not.ga_create(mt_dbl,n2,n2,'g_csc',0,0,g_csc))
-     &    call errquit(pname//': error creating g_csc',142, GA_ERR) 
-
-       call ga_dgemm('n','n',n1, n2, n1, 1.0d0,
-     &   g_s, g_c, 0.0d0, g_sc)
-
-       call ga_dgemm('t','n',n2, n2, n1, 1.0d0,
-     &   g_c, g_sc, 0.0d0, g_csc)
-
-       call ga_matpow(g_csc, -FP_Half, 1d-12)
-
-       call ga_copy_patch('n', 
-     &   g_c,  1, n1, 1, n2,
-     &   g_sc, 1, n1, 1, n2)
-
-       call ga_dgemm('n','n',n1, n2, n2, 1.0d0,
-     &   g_sc, g_csc, 0.0d0, g_c)       
-
-
-       if (.not. ga_destroy(g_sc))
-     &   call errquit(pname//': error destroy g_sc',141, GA_ERR)
-
-       if (.not. ga_destroy(g_csc))
-     &   call errquit(pname//': error destroy g_csc',142, GA_ERR)       
-
-       if (debug) write(luout,*) 'leaving '//pname
-       
-       return
-
-       end
-
-      

src/property/prop_input.F

@@ -68,7 +68,7 @@ c ... jochen:
       integer aoresponse, nfreq, l_fr, k_fr, 
      &        velocity, magneticpert, aor_giao, aor_damp, aor_conv,
      &        aor_orbeta, aor_gprime, aor_quad, aor_analyze,
-     &        loc_boys, loc_pm, loc_ibo
+     &        loc_boys, loc_pm, loc_ibo, loc_opt, itmp
       double precision gamwidth, convergence
 c     
 c>>>  declarations from NWdft
@@ -144,7 +144,8 @@ c ... jochen:
       loc_boys          = 1
       loc_pm            = 1
       loc_ibo           = 1
-c     
+      loc_opt           = 0
+c
 c     read inputs; store in rtdb
 c     
       do iline = 1,infinity     !  infinite loop over input lines
@@ -315,15 +316,29 @@ c ... jochen: localization input
         if (inp_a(test)) then
           if ( inp_compare(.false., 'boys', test)) then
             loc_boys = 0
+            itmp = 0
+            if (inp_i(itmp)) then ! localization option
+              loc_opt = itmp
+            end if
           elseif ( inp_compare(.false., 'ibo', test)) then
             loc_ibo = 0
+            itmp = 0
+            if (inp_i(itmp)) then ! localization option
+              loc_opt = itmp
+            end if
           elseif ( inp_compare(.false., 'pm', test)) then
             loc_pm = 0
+            itmp = 0
+            if (inp_i(itmp)) then ! localization option
+              loc_opt = itmp
+            end if
           else ! default is PM
             loc_pm = 0
+            loc_opt = 0
           end if
         else
           loc_pm = 0
+          loc_opt = 0
         end if
 c ... jochen end        
       elseif ( inp_compare(.false., 'stoneanalysis', test)) then
@@ -687,7 +702,12 @@ c
 c      
       if (.not. rtdb_put(rtdb, 'prop:loc_ibo', mt_int, 1,
      $                               loc_ibo  ))
-     $      call errquit('prop_input: rtdb_put failed', 0, RTDB_ERR)      
+     $      call errquit('prop_input: rtdb_put failed', 0, RTDB_ERR)
+c
+      if (.not. rtdb_put(rtdb, 'prop:loc_opt', mt_int, 1,
+     $                               loc_opt  ))
+     $      call errquit('prop_input: rtdb_put failed', 0, RTDB_ERR)
+c
       if (.not. rtdb_put(rtdb, 'prop:spinspin', mt_int, 1,
      $                               spinspin         ))
      $      call errquit('prop_input: rtdb_put failed', 0, RTDB_ERR)