Paolo Giannozzi
9af0317f8d
Once again, a problem with modules in Makefiles. In order to simplify things,
...
after the proliferarion of modules and libraries, I have collected into a
single variable BASEMOD_FLAGS, defined in make.inc, the list of directories
for modules, to be included by all codes. IMPORTANT NOTICE: re-run "configure",
the new Makefiles work in conjunction with the new make.inc
2018-01-08 22:28:48 +01:00
Carlo Cavazzoni
f36f848678
- adding helper function for the smallbox as well
2018-01-08 20:26:08 +01:00
Carlo Cavazzoni
0567d49b76
Merge branch 'master' of gitlab.com:QEF/q-e
2018-01-08 12:02:50 +01:00
Paolo Giannozzi
0187484c1b
Compilation problem with old xml
2018-01-07 21:26:35 +01:00
Carlo Cavazzoni
f5b066cb89
Merge branch 'master' of gitlab.com:QEF/q-e
2018-01-07 20:51:43 +01:00
Paolo Giannozzi
d5a5682bd9
Missing deallocation
2018-01-07 09:55:31 +01:00
Paolo Giannozzi
416bf19138
The calculation of the charge density can be simplified by adding augmentation
...
terms to charge density in G-space. Not a big deal, but two FFTs on rho can be
spared at each scf iteration.
2018-01-06 22:54:04 +01:00
Carlo Cavazzoni
6e499b9b2f
Merge branch 'master' of gitlab.com:QEF/q-e
2018-01-06 12:02:29 +01:00
Paolo Giannozzi
f950a749d8
Finally removed last remains of "custom" grid. The same two routines are used
...
to initialize G-vectors for all grids: one (ggen) to generate a new set of
G-vectors, one (ggens) to generate a subset of G-vectors with the same
ordering as in the original set.
2018-01-05 22:26:39 +01:00
Paolo Giannozzi
4ea5c0bda1
ggen no longer uses variables from module gvect: all g-vector related variables
...
are passed as arguments. Note that ngm is likely redundant since it is inside
the fft descriptor. Nothing should change apart from the call to ggen
2018-01-05 19:06:07 +01:00
Carlo Cavazzoni
63d200e081
- more ngm cleanup (FFT data type member replacing module variable)
2018-01-05 19:03:04 +01:00
Paolo Giannozzi
87890fb1c4
More "decustomization" of G-vectors: routine ggenx replaced by ggens
2018-01-05 10:23:42 +01:00
Paolo Giannozzi
bd615e906a
Compilation error in serial case
2018-01-04 21:34:21 +01:00
Paolo Giannozzi
4f67392dd5
Second step in G-vector generation reordering: subroutine "ggen" split into
...
two subroutines: "ggen" takes care of G-vectors for the FFT grid only,
"ggens" takes care of the subgrid only, with exactly the same ordering.
Seems to work, please verify
2018-01-04 11:42:42 +01:00
Paolo Giannozzi
ed40bdbb31
Merge branch 'master' of https://gitlab.com/QEF/q-e
2018-01-04 08:52:36 +01:00
Carlo Cavazzoni
b028caedef
- more code merge in fft_rho
2018-01-03 22:07:27 +01:00
Paolo Giannozzi
54a1567bfa
First attempt to unify the various ggen*: FFT descriptors are taken out from
...
modules and put in the call of ggen; case with no global sorting simplified
2018-01-03 19:23:22 +01:00
Paolo Giannozzi
1cd21f1d1f
EPW aligned with the rest of QE: indices nl* moved to dfft*%nl
...
No warranty, but tests seem to pass
2018-01-03 17:55:55 +01:00
Stefano de Gironcoli
86d479c907
bug fix for meta_gga w/o doublegrid
2018-01-03 13:04:57 +01:00
Carlo Cavazzoni
abba737d95
- rho_r2g and smooth_rho_r2g, merged in a single subroutine,
...
the fft descriptor is used in place of "smooth" prefix
2018-01-03 11:57:15 +01:00
Paolo Giannozzi
524e7193a6
Release-notes and make.depend updated, misspells corrected
2018-01-03 09:23:48 +01:00
Paolo Giannozzi
70e154e7b1
Merge branch 'master' of https://gitlab.com/QEF/q-e
2018-01-03 08:42:30 +01:00
Stefano de Gironcoli
9ed8602b5a
mispelled variable desc$have_task_groups corrected to desc%has_task_group
2018-01-03 01:38:48 +01:00
Stefano de Gironcoli
a121f57263
call sequence of fft_interpolate_real and fft_interpolate_complex made equal
2018-01-03 01:13:20 +01:00
Stefano de Gironcoli
407a1c444a
routine cinterpolate replaced by fft_interpolate_complex
...
similar to fft_interpolate_real
2018-01-03 00:50:24 +01:00
Stefano de Gironcoli
72532727af
subroutines interpolate and exx_interpolate unified into fft_interpolate_real
...
syntax changed
2018-01-02 23:42:06 +01:00
Paolo Giannozzi
b21f20023f
Structure exx_fft deleted. To be done: ggenx and ggent must be merged with ggen.
2018-01-02 22:53:20 +01:00
Stefano de Gironcoli
f1efd650da
Merge branch 'master' into fft_new_interface
...
need to update branch before pulling to gitlab
2018-01-02 19:20:20 +01:00
Stefano de Gironcoli
41e91c0dac
new interface to fft calls
...
three types of calls are possibles : 'Rho', 'Wave', 'tgWave'
In order to enable an fft-type for a given grid the corresponding clock_labels must be set.
One gives a name to desc%rho_clock_lable for 'Rho' type fft and a name to
desc%wave_clock_lable for 'Wave' and 'tgWave' types. Whether tg is
possible depends of the already defined value of desc%have_task_groups variable (mispell to be corrected soon).
definining
dffts%rho_clock_label='ffts', dffts%wave_clock_label='fftw',
dfftp%rho_clock_label='fft', dfftt%rho_clock_label='fftc' and
dfftt%wave_clock_label='fftcw'
and changing
'Dense'->'Rho', 'Smooth'->'Rho', 'Custom'->'Rho', 'CustomWave'->'Wave'
the same clock names and the same overall behavior as with the old interface is obtained.
2018-01-02 17:45:45 +01:00
Carlo Cavazzoni
0af66108d9
Merge branch 'master' of gitlab.com:QEF/q-e
2018-01-02 12:42:16 +01:00
Carlo Cavazzoni
b72df6b381
- cleanup
...
- removing redundant references to gvecs gvecr
- substituting ngm with %ngm
2018-01-02 12:40:57 +01:00
Stefano de Gironcoli
761dd6f6bf
call sequence to c2psi_gamma made consistent with soubroutine definition
2018-01-02 11:18:48 +01:00
Paolo Giannozzi
02db055030
More removal of variables from exx_fft
2018-01-01 18:58:31 +01:00
Paolo Giannozzi
997c8b06eb
A few more variables removed from exx_fft
2018-01-01 18:35:25 +01:00
Paolo Giannozzi
914ecdd19b
More removal of variables from exx_fft: ngmt_g
2018-01-01 18:05:02 +01:00
Paolo Giannozzi
3d0eecd236
Replicated variable ngmt removed from exx_fft (is alrady in dfftt)
2018-01-01 13:59:37 +01:00
Carlo Cavazzoni
0488d218df
- fix compilation of FFTXlib/test.f90
...
- replace ugly code lines with a pure funciton
2018-01-01 10:17:34 +01:00
Paolo Giannozzi
3f0d0348f9
Merge branch 'master' of https://gitlab.com/QEF/q-e
2017-12-30 19:08:54 +01:00
Paolo Giannozzi
aa3f5b1a19
More decustomizatio of custom FFT's: the FFT descriptor dfftt, together with
...
all unused or uselss variables, taken out of the exx_fft structure. Should
have no side effects but beware! especially for TDDFPT
2017-12-30 19:06:17 +01:00
Carlo Cavazzoni
f7a51b4fef
- drop _gamma suffix from helper subroutines where there is no ambiguity
...
- remove %nl indexes from fft_rho
- remove reference to gvec module from fft_rho (descriptor and helper sub. are enough)
2017-12-30 12:30:54 +01:00
Paolo Giannozzi
3f0e68bc79
Removal of unused routines, merge of gvec_init into ggent, make.depend updated
2017-12-30 08:54:07 +01:00
Carlo Cavazzoni
8ee416c373
finally all %nl indexes have been removed from CPV
2017-12-30 00:39:03 +01:00
Paolo Giannozzi
3b5d9bc249
"custom" fft used in EXX made a little bit less custom: indices nl[m] are
...
computed as in all other cases in routines fft_set_nl[m] of module fft_ggen.
Everything seems to work. Unused routine removed, some comments updated.
NB: differently customized "custom" ffts of GWW are left unchanged.
2017-12-29 21:20:53 +01:00
Paolo Giannozzi
6d358504c2
Merge branch 'master' of https://gitlab.com/QEF/q-e
2017-12-29 15:36:02 +01:00
Paolo Giannozzi
60e032239a
Band parallelization for EXX was (sometimes) broken by last changes in fft
...
descriptors, due to unpredictable behavior when allocated arrays are copied
into unallocated ones or vice versa. The current workaround must disappear
soon together with the current code that copies fft descriptors and other
stuff back and forth
2017-12-29 15:33:15 +01:00
Carlo Cavazzoni
400d7ae617
- using fft helper functions to remove explicit ref. to %nl and %nlm
2017-12-29 14:41:52 +01:00
Paolo Giannozzi
70ea0c149c
Merge branch 'master' of https://gitlab.com/QEF/q-e
2017-12-29 12:26:02 +01:00
Carlo Cavazzoni
5157165150
ome more helper subroutines to copy 1D array into 3D array
2017-12-29 11:44:01 +01:00
Paolo Giannozzi
6e695350a7
XMerge branch 'master' of https://gitlab.com/QEF/q-e
2017-12-29 10:57:42 +01:00
Paolo Giannozzi
752eac2948
Miller indices removed from exx_fft structure and made local
2017-12-28 21:40:36 +01:00