Paolo Giannozzi
7ca5d92615
Fix for weird MPI crash
...
Apparently on some MPI libraries you should not try to free a nonexistent
communicator. Should fix issue #532 .
2022-08-15 19:36:39 +02:00
Pietro Delugas
49e5ba9176
add function to compute workers of each block
...
as created by block_distribute
2022-07-22 16:43:53 +02:00
Ye Luo
f8e7b3699e
Remove eigensolver_gpu submodule and its CMake ref.
2022-02-04 13:05:03 -06:00
Paolo Giannozzi
45cc8753c3
Missing protection for CUDA variables
2022-02-04 16:49:47 +01:00
Paolo Giannozzi
d262aca3cb
Support for CUDA EIGENSOLVER removed
2022-02-04 15:39:13 +01:00
Paolo Giannozzi
35b51e4fc3
Cleanup of laxlib tests: use laxlib_parallel_include instead of module mpi,
...
not always compiled for the same compiler. Closes issue 380.
2022-01-06 22:34:27 +01:00
Ye Luo
ad86965d68
Merge branch 'test_LAXlib' into 'develop'
...
LAXlib comment
See merge request QEF/q-e!1621
2021-11-22 18:37:02 +00:00
Ye Luo
70ab5d42e1
Remove never used world_cntx in LAXlib
2021-11-12 19:45:23 -06:00
Ye Luo
13e00cc76f
Respect communicator when initializing laxlib.
2021-11-12 19:30:18 -06:00
Oscar Baseggio
52dc2dfa99
fix LAXlib/ptoolkit.f90
2021-11-10 11:53:17 +01:00
Oscar Baseggio
430bff52a2
patch for FORD comments
2021-11-10 10:46:54 +01:00
Paolo Giannozzi
5d74f827c4
Remove make.depend files from git
2021-10-24 21:29:29 +02:00
Oscar Baseggio
f1b2d557b2
fix segmentation fault for intel compiler, fix the tester algorithm
2021-10-12 12:04:04 +02:00
Oscar Baseggio
38543b4350
fix tests for LAXlib in LAXlib/tests/
2021-10-12 12:04:04 +02:00
Paolo Giannozzi
31a7d9a284
Some more Makefile simplification
2021-08-27 11:07:16 +02:00
Ye Luo
cae2e0fdca
Add a unit test for LAPACK zdotc.
2021-07-17 00:28:43 -05:00
giannozz
1760281527
Merge branch 'cusolver-fix' into 'develop'
...
Initialize CUSolver only once
Closes #283
See merge request QEF/q-e!1479
2021-07-09 17:24:11 +00:00
Paolo Giannozzi
90a0de8e13
Small fixes for "pedantic" compilation
2021-07-05 22:22:53 +02:00
Pietro Bonfa
2ad5b4244a
OMP safety check moved outside initialization
2021-07-05 20:33:51 +02:00
Pietro Bonfa
2a429c78ef
Saving CUSolver handle. Fixes #283 .
2021-07-05 18:12:41 +02:00
Ye Luo
54f46aa7f3
Fix missing cuda library dependencies
2021-06-17 16:28:56 -05:00
Paolo Giannozzi
963cccd595
Unneeded error replaced by warning
2021-06-15 18:47:27 +02:00
Ye Luo
ce8fec50c8
Test cusolver Zhegvdx instead of a warning blindly.
2021-05-25 09:02:57 -05:00
Ye Luo
6856f016d0
Better test_qe_lax test sizes.
2021-05-23 16:31:07 -05:00
Ye Luo
daef299bf2
Merge remote-tracking branch 'origin/develop' into add-unit-test-runner
2021-05-23 16:28:20 -05:00
Ye Luo
69fbdcb0b1
Add unit test runner.
2021-05-23 09:58:06 -05:00
Ye Luo
c11c70909b
Correct timer.
2021-05-22 19:00:01 -05:00
Ye Luo
004e8204b2
Fix LAXlib/test.f90 in non-MPI case.
2021-05-22 16:27:59 -05:00
Ye Luo
7fd7b86f14
Fix shared lib.
2021-05-16 01:10:45 -04:00
Ye Luo
0016a9fe0f
qe_lax should depend on qe_lapack.
2021-05-15 00:29:25 -04:00
Ye Luo
01a13df4b2
Relocate unit tests.
2021-05-02 14:43:11 -05:00
Daniele Cesarini
042a8ca1a0
CMake fixes for shared lib builds and added a CI build for NVHPC
2021-04-06 19:07:49 +00:00
Ye Luo
961f50b7f5
Reduce QE_ENABLE_LAXLIB_CUSOLVER requirement to CUDA 10.1
2021-03-10 22:42:07 -06:00
Victor Yu
351f4871fe
Clean up redundant ELPA preprocessing flags
...
__ELPA_2017, __ELPA_2018, __ELPA_2019, and __ELPA_2020 were removed. Now
we have three:
* __ELPA_2015: compatible with ELPA 2015.x, 2016.05
* __ELPA_2016: compatible with ELPA 2016.11, 2017.x, 2018.05
* __ELPA (default): compatible with ELPA 2018.11, 2019.x, 2020.x
2021-03-08 10:45:32 -08:00
Victor Yu
69d42a40eb
Improve support for recent versions of ELPA
...
* Always ask for error code when calling elpa_deallocate and elpa_uninit.
Depending on the version and configure options of ELPA, the error code
may or may not be optional.
* Update doc and configure to reflect that ELPA 2020.x is supported.
2021-03-07 21:26:19 -05:00
Ye Luo
35ad49ead2
Let nvfortran decide what cudatoolkit to use.
2021-01-31 18:08:57 -06:00
Paolo Giannozzi
229be57edb
Script for dependencies and make.depend updated
2021-01-23 22:32:58 +01:00
Daniele Cesarini
d462b73a6a
Fixed cmake files for cpu-gpu merge
2021-01-23 12:42:52 +01:00
Pietro Delugas
db0da8b0d9
Merge branch 'merge_qegpu' into HEAD
2021-01-22 17:20:50 +01:00
Paolo Giannozzi
b71754532d
Doc: one misspell fixed, reference to github repository added
...
Only needed libraries used in Makefile for LAXlib
2021-01-22 12:55:33 +00:00
Pietro Bonfa
0c285826d2
Merge branch 'develop' into syncqe8
2021-01-21 19:27:49 +01:00
Daniele Cesarini
586f66aadf
Introduce CUDA support in CMake with some refactoring.
2021-01-18 14:50:50 +00:00
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
9ea6196fc6
I think that ESSL (not LINUX_ESSL) is definitely obsolete
2020-10-28 16:45:44 +00:00
Pietro Bonfa
a53816c4de
Now QE and QEGPU have the same makefiles! No more conflicts, yuppie!
2020-10-19 16:00:37 +02:00
Pietro Bonfa
05c866e91f
Merge branch 'develop' into syncqe2
2020-10-18 17:50:41 +02:00
Pietro
afe89dfbc6
Using new devxlib global buffers in LAXlib. A comments describes the compiler flag that disentangles the two libraries.
2020-10-17 14:17:34 +00:00
Pietro Bonfa
1a4df64ffe
Merge branch 'develop' into syncqe
2020-10-04 16:33:16 +02:00
Ye Luo
ccb30bc8db
Fix unit tests.
2020-09-30 23:14:23 -05:00
Ye Luo
8d7b692508
Move executables and static archives to bin lib
2020-09-30 13:42:46 -05:00