Andrea Ferretti
9e6b4d7415
Merge branch 'devel-upflib' of gitlab.com:max-centre/components/q-e into devel-upflib
2021-02-22 11:56:08 +01:00
Andrea Ferretti
7147add67a
removal of uspp-gpum module.
2021-02-22 08:26:48 +01:00
Andrea Ferretti
b45f9439ed
upf_spinorb_gpum.f90 module made obsolete
2021-02-18 23:15:14 +01:00
Paolo Giannozzi
c26fbbbf48
Let's see if this fixes CMake once again
2021-02-18 19:52:39 +00:00
Paolo Giannozzi
f3ab093338
One more leftover deleted file
2021-02-18 17:19:24 +00:00
Paolo Giannozzi
8930961774
Forgot to remove deleted file
2021-02-18 17:08:04 +00:00
Paolo Giannozzi
ce83a72816
Merge branch 'devel-upflib' of https://gitlab.com/max-centre/components/q-e into free_unit
2021-02-18 17:01:49 +00:00
Andrea Ferretti
9eb0babe91
allocations and handling of variables in uspp_data (with special focus on GPU logic) simplified
2021-02-18 17:20:37 +01:00
Paolo Giannozzi
4ee9c76cf3
Unused variables removed, uspp_gpu renamed uspp_gpum for uniformity
2021-02-18 13:38:49 +00:00
Paolo Giannozzi
703ef26cd1
upf_erf deleted and replaced by standard erf
2021-02-18 10:22:29 +00:00
Paolo Giannozzi
8dfb9111df
Updated doc
2021-02-18 09:44:14 +00:00
Paolo Giannozzi
40a030ebb0
No need for a free_unit routine any longer
2021-02-18 09:08:47 +00:00
Andrea Ferretti
749499b12d
init_us_2*f90 moved to upflib and called via a wrapper in PW/src.
...
Module recvec updated with gpu variables taken from recvec_gpu.
2021-02-17 19:56:30 +01:00
Andrea Ferretti
f24ab5c10a
splinelib_gpu removed (splint_eq_gpu moved as splint_eq in splinelib module)
2021-02-16 08:19:03 +01:00
Andrea Ferretti
0c5a5922d3
dependencies regenerated
2021-02-15 19:10:36 +01:00
Andrea Ferretti
58233c75e8
more on renaming us_gpum -> uspp_data_gpum
2021-02-15 19:07:53 +01:00
Andrea Ferretti
53b3a5c6ff
us_gpum renamed to uspp_data_gpum (internal consistency)
2021-02-15 19:00:11 +01:00
Andrea Ferretti
3645861a13
compilation fixed
2021-02-15 18:48:15 +01:00
Andrea Ferretti
4c3ea7646f
GPU variables moved ordinary modules
2021-02-15 18:34:07 +01:00
Andrea Ferretti
d140ff04b5
updated
2021-02-15 18:33:52 +01:00
Andrea Ferretti
f1f8371430
Merge branch 'devel-upflib' of gitlab.com:max-centre/components/q-e into devel-upflib
2021-02-15 10:12:55 +01:00
Andrea Ferretti
9a69f149b9
more files related to PP moved from PW/src to upflib
2021-02-15 10:09:20 +01:00
Paolo Giannozzi
4a337f742b
Merge https://gitlab.com/QEF/q-e into devel-upflib
2021-02-15 08:47:55 +00:00
Andrea Ferretti
03c3c5209e
housekeeping
2021-02-13 17:41:20 +01:00
Paolo Giannozzi
52b4bdbd4e
Some compilers may write exceedingly long lines if an entire matrix is written
...
using free format. Some pseudopotential variables written in that way could thus
make an UPF file unreadable. Workaround: write one column at the time. Not nice
but at least it's readable. Noticed by Felix Goudreault, U. Montreal.
2021-02-10 22:05:28 +01:00
Paolo Giannozzi
995dea1f90
Oops again (cmake)
2021-02-09 08:28:55 +00:00
Paolo Giannozzi
76526538bb
Interpolation of radial atomic wavefunction moved into subroutines
...
interpol_at_wfc.f90 and interpol_at_wfc_gpu.f90
2021-02-08 21:06:16 +00:00
Paolo Giannozzi
c871456823
oops, CMake (more) broken (than usual)
2021-02-02 09:28:09 +00:00
Paolo Giannozzi
a333ee8d0c
Missing dependencies
2021-02-02 06:51:36 +00:00
Paolo Giannozzi
3e51878e87
Harmonization of modules upf_spinorb* in upflib after cpu-gpu merge
2021-02-01 21:54:09 +00:00
Paolo Giannozzi
1478cf4198
Merge https://gitlab.com/QEF/q-e into spinorb
2021-02-01 21:53:05 +00:00
Paolo Giannozzi
771b7da9f1
Cleanup of installation scripts and Makefiles
2021-01-29 15:59:09 +00:00
Paolo Giannozzi
2d3ea393db
Dumb horrible bug in PAW relativistic calculations
...
In the new PP-reading code, tag name PP_AEWFC_REL.* was incorrectly written
as PP_AEWFC_rel, leading to aewfc_rel array not being read and to a small but
nonnegligible error in the results. Guess what? there isn't a single test with
relativistic PAW in the test-suite, that would have allowed to spot the error
at once. Some cleanup as well.
Fixes issue 277. Noticed by Minkyu Park (Univ. of Ulsan, Rep. of Korea)
2021-01-28 11:45:04 +00:00
Paolo Giannozzi
229be57edb
Script for dependencies and make.depend updated
2021-01-23 22:32:58 +01:00
Paolo Giannozzi
69f3560fd0
Missing file added
2021-01-23 20:21:10 +00:00
Paolo Giannozzi
0a9f218755
Alignement to merged q-e and q-e-gpu sort of completed. Still some stuff to
...
move around but it seem to compile and run (for cpu: not sure about gpu)
2021-01-23 20:12:52 +00:00
Paolo Giannozzi
471da0ab46
Merge https://gitlab.com/QEF/q-e into devel-upflib
2021-01-23 19:28:44 +00:00
Daniele Cesarini
d462b73a6a
Fixed cmake files for cpu-gpu merge
2021-01-23 12:42:52 +01:00
Pietro Bonfa
0c285826d2
Merge branch 'develop' into syncqe8
2021-01-21 19:27:49 +01:00
Pietro
d1ec367d9d
Aligned gpu-develop branch with develop from QEF/q-e.
2021-01-19 22:16:25 +00:00
Paolo Giannozzi
671f25d895
Merge https://gitlab.com/QEF/q-e into devel-upflib
2021-01-19 17:44:14 +00:00
Paolo Giannozzi
3d7af97ec1
Take notice of what has happened in upflib
2021-01-19 14:20:43 +00:00
Daniele Cesarini
586f66aadf
Introduce CUDA support in CMake with some refactoring.
2021-01-18 14:50:50 +00:00
Andrea Ferretti
1ed5e62959
init_at_1 moved to upflib (same strategy as for init_us_1).
2021-01-17 23:05:14 +01:00
Paolo Giannozzi
70208df25e
CMake made, plus funny GWL error due to conflicting definitions of DP
2021-01-17 21:39:34 +00:00
Andrea Ferretti
504ce31a46
init_us_1 moved to upflib
2021-01-16 01:07:21 +01:00
Andrea Ferretti
dd000d90c9
init_us_0.f90 and init_us_b0.f90 disentangled and moved to upflib.
2021-01-15 23:41:27 +01:00
Paolo Giannozzi
fdbca7100b
Missing # in #ifdef
2021-01-05 08:19:23 +00:00
Pietro Bonfa
b0ef9261f0
Removed dependency from kinds and verbosity in modules for data duplication and suppressed debug messages from verbose output (only shown when compiling with _DEBUG flag)
2021-01-04 11:08:19 +01:00
Paolo Giannozzi
93ea75bbbc
Removal of several annoying printouts
2021-01-01 21:01:38 +01:00
Paolo Giannozzi
9d4a18d5b2
Add some missing initailizations to upf% in converters
2020-12-22 12:09:55 +00:00
Paolo Giannozzi
47953f0be5
PP mixing code for VCA wasn't (yet) working (once again)
2020-12-17 07:38:13 +00:00
Paolo Giannozzi
2e9ac6b320
[skip-CI] Move or remove the few residual files in upftools/
2020-12-15 09:57:37 +00:00
giannozz
20961fb9eb
Merge branch 'defox_upf' into 'develop'
...
FoX-based code removed from upflib
See merge request QEF/q-e!1233
2020-12-08 09:29:09 +00:00
Lorenzo Paulatto (naquite)
f634167b74
Added readtag_zt to interface
2020-12-08 09:09:07 +01:00
Lorenzo Paulatto (naquite)
5b88387c53
Added complex tensor i/o required by d3q
2020-12-08 08:53:04 +01:00
Paolo Giannozzi
ac3c36a31c
FoX-based code removed from upflib
2020-12-08 08:49:26 +01:00
Pietro Bonfa
4d484679ca
Moved GPU version of ylmr to upflib where the CPU implementation is
2020-11-24 13:47:23 +01:00
Pietro Bonfa
9529540d8d
Merge branch 'develop' into syncqe5
2020-11-24 13:44:01 +01:00
Daniele Cesarini
b2a4a6b89f
Fixed shared library compilation
2020-11-23 12:16:32 +01:00
Paolo Giannozzi
1f84b032e7
Forgot cMake ...
2020-11-22 14:19:05 +01:00
Paolo Giannozzi
15ef90850c
Circular dependency solved by moving ylmr2.f90 from Modules to upflib/
2020-11-22 12:59:30 +01:00
Pietro
499251fe25
In the HP postprocessing routine added some explanations
...
how to solve the problem of missing elements during the
reconstruction process of the response matrix chi.
2020-11-18 16:29:39 +00:00
Paolo Giannozzi
a689add63a
Some PP files do not have the mesh attribute in pp_r tag. Previously, this
...
went unnoticed because no message or error was issued in this case (mesh=0).
So, the small change of yesterday to deal with missing mesh_size in pp_header
was breaking a few tests using a pp file with missing mesh attribute...oh well.
2020-10-24 09:12:19 +02:00
Paolo Giannozzi
a571b3e4d5
Allow some UPF sloppiness
...
Make the behavior of new UPF reading code more tolerant:
- If an incorrect number is found in an attribute, do not crash: issue a
warning and set the attribute value as if the attribute was not found
- Do not check actual index vs the "index" attribute for v2 UPF files:
the previous code didn't do that, I think, didn't crash if index missing
2020-10-23 11:22:42 +00:00
Pietro Delugas
55f555af1e
corrects wrong logic of commit f68fc798
2020-10-21 13:08:23 +02:00
Pietro Delugas
f68fc79849
unread values from UPFv1 pseudos set to safe value
2020-10-19 11:55:46 +02:00
Pietro Bonfa
05c866e91f
Merge branch 'develop' into syncqe2
2020-10-18 17:50:41 +02:00
Paolo Giannozzi
096feece44
No reason to keep obsolete and confusing FIXME messages
2020-10-12 21:33:29 +02:00
giannozz
ab088df340
Merge branch 'fix_pp_schema_files' into 'develop'
...
qe_pp added number of gipaw orbitals to file.
See merge request QEF/q-e!1120
2020-10-12 15:45:50 +00:00
Pietro Delugas
7aab0f54a5
[CMake] adds forgotten targets in upflib directory
2020-10-09 11:50:46 +02:00
giannozz
5d42574c77
Merge branch 'python3' into 'develop'
...
Convert to Python3
See merge request QEF/q-e!1033
2020-10-07 18:53:59 +00:00
Pietro Bonfa
1a4df64ffe
Merge branch 'develop' into syncqe
2020-10-04 16:33:16 +02:00
giannozz
88165692d1
script for dependencies and make.depend updated
2020-10-02 08:46:22 +02:00
Federico Ficarelli
6f520240a1
Add alias upflib -> QE::UPF
2020-09-29 18:11:33 +02:00
Pietro Delugas
65567619a1
adapting Cmakelist.txt files to recent changes
...
and adding more files.
upflib and libbeef added as directories for now
2020-09-29 18:11:33 +02:00
Daniele Cesarini
cd4177a5e6
Replaced dgamma function with gamma because dgamma is only a GNU specific builtin
2020-09-29 18:11:32 +02:00
giannozz
adcbab61e3
Unlikely principal quantum number in UPF v.1, plus misspell
2020-09-29 13:00:09 +02:00
Pietro Delugas
a367fca243
qe_pp added number of gipaw orbitals to file.
...
two elements of the qe_pp schema were not written in the file.
2020-09-29 11:56:20 +02:00
Paolo Giannozzi
442c598514
Small upflib glitches: FoX no longer required, removed from Makefile;
...
upf_error was erroneous as well (didn't do anything due to bad IF)
2020-09-25 22:07:23 +02:00
Eisuke Kawashima
493d0e6340
Convert to Python3 and minor improvements
2020-09-23 23:50:24 +09:00
Pietro Bonfa
68477bc3d2
Merge branch 'develop' into gpu-develop
2020-09-16 09:51:07 +02:00
Paolo Giannozzi
a8941e63a4
Compiler-dependent xml write problem
...
Writing a long vector with free format (*) may result in very long lines,
exceeding the current 1024 character limit for reading, set in xmltools.f90.
Out od the two possible workarounds: limit the number of data per write, or
use an explicit format, the latter was chosen. The formats used for write,
I18 for integers, es24.15 for reals, should guarantee no loss of precision.
Fixes issue #222
2020-09-14 08:06:01 +00:00
Paolo Giannozzi
f90f4f7602
Dumb mistake in a recent commit, mysteriously not affecting Intel compiler
2020-09-11 07:56:41 +02:00
Paolo Giannozzi
159d7ebc6c
Once again, trouble with broadcast of "rinner": a recent fix for new xml broke
...
upf v.1 case. Now rinner is always dimensioned as "nqlc", except when nbeta=1 .
2020-09-09 11:52:32 +02:00
Paolo Giannozzi
333bac9827
Added comment explaining why the array is set to zero and what should be done
2020-09-01 13:59:09 +00:00
Paolo Giannozzi
db757fc720
Weird Q(r) roblems
...
Array qfuncl must be explicitly set to zero before reading, in order to
prevent weird problems in real-space treatment of augmentation problems
(coming from the presence, with some compilers, of nonzero numbers).
This is not needed for array qfunc, but it is done anyway for symmetry.
2020-09-01 13:40:07 +00:00
Paolo Giannozzi
d5c91fbaca
Pseudopotential file wasn't closed after an unsuccesfull attempt to read,
...
leading to an error when debug flags are activated
2020-09-01 08:34:05 +00:00
Paolo Giannozzi
fa701d1cdc
[skip-CI] Remove debug printout
2020-08-31 17:41:31 +00:00
Paolo Giannozzi
fd5f745ebe
FoX is no longer used to read and write pseudopotential files, unless
...
-D__use_fox is defined. The old code is still there and will be removed
only when the new code is more tested
2020-08-31 16:46:18 +00:00
Paolo Giannozzi
5c12d6c7ed
Improved handling of an error case, improved printout of debug information
2020-08-25 18:00:20 +02:00
Paolo Giannozzi
01b6d7a807
Workaround to enable the possibility to open another xml file while one is
...
already opened. I do not see any good reason for this but it seems to be the
simpler way to deal with the obscure I/O pattern of phonon and epw.
2020-08-12 10:50:49 +02:00
Paolo Giannozzi
b09a8e4f38
version.h renamed qe_version.h to prevent conflicts with other software
...
packages: version.h is too "generic"
2020-08-12 09:19:29 +02:00
Pietro Bonfa
b974b6dae4
Merge branch 'develop' into gpu-develop
2020-08-04 08:27:59 +02:00
Pietro Delugas
f33b123313
Revert "Add upf_msg to upflib and "
...
This reverts commit faa3dc58086c7e33b1f45dda69abe9444fb0b33b.
2020-08-03 11:23:33 +00:00
Pietro Bonfa
af515e4108
Merge branch 'develop' into gpu-develop
2020-07-15 14:47:33 +02:00
Paolo Giannozzi
11b30f49d7
Module modes brought back to where it belong. Dependencies updated
2020-07-15 07:27:17 +00:00
Pietro Bonfa
4e6cb3860c
Merge branch 'develop' into gpu-develop
...
(force_hub_gpu still to be aligned)
2020-07-09 15:19:05 +02:00
Paolo Giannozzi
8a3e92febe
Improvements to xml toolkit:
...
- can read tags not in order, although only in some specific cases (luckily
including those PP files that do have some tags not in the usual order)
- writes complex variables as two reals, not in fortran free format:
it makes reading data from other languages (e.g., python) easy
2020-07-06 12:51:40 +02:00
Pietro Bonfa
9645be8c42
Merge branch 'develop' into gpu-develop
2020-06-28 18:07:20 +02:00
Paolo Giannozzi
a72b129b2a
Simplify the xml output of atomic projections produced by projwfc.x
2020-06-24 13:29:48 +00:00
Pietro Bonfa
4ad10deccf
Merge branch 'develop' into gpu-develop
2020-06-23 23:39:19 +02:00
Paolo Giannozzi
bdb5330895
Pseudopotential converters reverted to use FoX for reading PPs while waiting
...
for more extensive testing
2020-06-16 20:13:24 +00:00
Paolo Giannozzi
5de7456585
Fixed "PAW as GIPAW" case, whatever it means - All PW tests passed
2020-06-10 16:51:09 +02:00
Paolo Giannozzi
2c11837478
Missing initialization of two "upf" variables (not read from pp file)
2020-06-10 12:34:04 +02:00
Paolo Giannozzi
f0ebe9549a
Spin-orbit case for v.2 UPF fixed (maybe)
2020-06-10 10:07:05 +02:00
Paolo Giannozzi
7eb44391b7
A few pieces were missing when reading PAW pseudopotentials
2020-06-09 23:06:23 +02:00
Paolo Giannozzi
c3215ed09b
"Protect" strings from troublesome characters
2020-06-05 17:35:34 +02:00
Paolo Giannozzi
436a7b6ef7
Better treatment for attributes, less susceptible to run into trouble
...
with attributes containing single quotes into double quotes, or vice versa
2020-06-05 14:39:08 +00:00
Paolo Giannozzi
01eb8dc1c4
Miscellaneous small glitches in new pseudopotential I/O
2020-06-05 13:44:25 +00:00
Paolo Giannozzi
e1a477ad6f
Treatment of attributes simplified
2020-06-05 10:43:46 +02:00
Paolo Giannozzi
b50ec06fcb
Treatment of attributes simplified and made similar for read and write
2020-06-05 08:33:48 +02:00
Paolo Giannozzi
b064c0ae9c
First not completely wrong UPF reading without FoX
2020-06-03 22:40:27 +02:00
Paolo Giannozzi
1b2aef5834
Some more steps towards a correct reading of UPF files
2020-06-03 17:27:38 +02:00
Paolo Giannozzi
af71552ec4
WIP: FoX replacement for pseudo-lib
2020-06-03 12:54:30 +02:00
Pietro Bonfa
cdd440a230
Merge branch 'develop' into gpu-develop
2020-05-28 15:19:12 +02:00
fabrizio22
2a29f075c4
additional argument removed in dvloc_ gpu routines
2020-05-26 17:12:57 +02:00
Paolo Giannozzi
fd27baa667
Small correction to pseudopotential converters
2020-05-04 17:28:10 +02:00
Pietro Bonfa
32d7b76240
Fixed make dependencies
2020-05-04 10:00:05 +02:00
Paolo Giannozzi
4fdcc195f4
[skip-CI] Small update, just to prompt buildbot to rebuild everything!
2020-05-03 22:35:19 +02:00
Pietro Bonfa
51119b3bca
Fix compilation of stress and upflib
2020-05-03 21:07:03 +02:00
fabrizio22
9630bd028a
stress local
2020-05-02 17:10:32 +02:00
Paolo Giannozzi
0db76c8efe
More missing initializations in old formats, leading to strange errors
2020-05-02 13:58:25 +00:00
Paolo Giannozzi
cd7b4b3adf
Unused routines removed
...
It sees to me that the radial grid is quite useless except for the atomic code
and maybe PAW: the rest of QE may use only the grid in upf% and variable "msh"
2020-05-02 09:18:00 +02:00
Paolo Giannozzi
0329ffafe1
Nonlocal UPF problem
...
Uninitialized variable when old PP formats are read was leading to a segfault
in parallel execution. It had gone unnoticed until now because upf% was a
pointer and was nullified at the beginning. Happens only on compilers that do
not set uninitialized logical variables to F (e.g., intel, but not gfortran)
2020-05-02 09:13:00 +02:00
giannozz
b8a3c2c4c0
Merge branch 'read_upf_simplify' into 'develop'
...
POINTER => ALLOCATABLE
See merge request QEF/q-e!918
2020-05-01 19:08:23 +00:00
Paolo Giannozzi
ff817b34f1
read_upf_v1 should always close the opened file on output
2020-05-01 10:19:29 +02:00
Paolo Giannozzi
6edef47848
POINTER => ALLOCATABLE
...
The old restriction on allocatable arrays in derived type no longer applies,
so all pointers in the derived-type upf variables are converted to allocatable
2020-04-30 15:43:04 +00:00
Paolo Giannozzi
2f349c1726
Some simplifications
...
"read_upf" replaced by "read_upf_new" that reads only upf v.2 and later;
"read_upf_v1" opens the file internally, removed useless optional argument.
2020-04-30 13:36:51 +00:00
Paolo Giannozzi
cc33771b46
Minor glitches
...
1. Some upf variables, unused with old formats, explicitly set to false when old formats are read
2. Leftover debug printout removed
3. Test updated to prevent yet another bogus error in test-suite
2020-04-30 07:10:37 +00:00
Pietro Bonfa
660b973a69
Merge branch 'develop' into gpu-develop (first attempt)
2020-04-29 18:54:24 +02:00
Paolo Giannozzi
2b2f7a9906
There was a radial grid in upf%, an additional one with more variables (of
...
questionable usefulness) in upf%grid%, and finally another one in module atom,
pointing to the former, or vice versa. The latter was set when the pseudo file
were read. I don't think this makes sense, so I removed upf%grid and set up the
grid in module atom after the upf variable is read.
BEWARE: not sure I correctly decyphered the (il)logic of grids in atomic/src/.
Additionally: fixed a few more glitches in the broadcast of upf%.
2020-04-28 08:15:50 +00:00
Paolo Giannozzi
3a5f2a8b1f
A few variables not read from upf v.1 format were leading to obscure crashes
...
when the upf variable was broadcast
2020-04-27 15:03:58 +02:00
giannozz
1b072fbc4f
First compiling version of QE with initial "pseudolib" in upflib/.
...
Most of the changes consist in small updates to Makefile and make.depend and
in the replacement of module "parameters" with "upf_params" whenever variables
lmaxx and lqmax are USEd. Routine "read_pseudo" moved back to Modules/;
"simpsn" moved to upflib/; a copy of "matches" temporarily added to upflib/
Note that everything compiles (except EPW) but nothing has been tested yet!
2020-04-25 19:13:15 +00:00