Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
971 commits
Select commit Hold shift + click to select a range
9a15846
Minor cleanup to remove unneeded protos in header file.
oseikuffuor1 Feb 19, 2025
2c7188a
Update modified directories after merge from master.
oseikuffuor1 Feb 19, 2025
f713f5b
Additional updates after merge from master.
oseikuffuor1 Feb 20, 2025
d9d1109
Introduce vector and matvec optimizations that help to reduce copies …
rfalgout Feb 22, 2025
7cf2e31
Fixes to allow C++ linkage for multiprecision functions.
oseikuffuor1 Feb 25, 2025
c7f8069
Fix caliper region
victorapm Feb 26, 2025
4dd745c
Merge pull request #1239 from hypre-space/recmat-caliper
victorapm Feb 26, 2025
663ad55
Renamed script to be more descriptive and minor updates to comments.
oseikuffuor1 Mar 25, 2025
2db7495
Merged support for mixed-precision preconditioners for struct interface
oseikuffuor1 Mar 26, 2025
420fe9a
Merge branch 'master' into recmat
victorapm Mar 26, 2025
aefe728
Merge branch 'master' into recmat
victorapm Apr 9, 2025
bd3e329
Recmat partial device support (#1251)
victorapm Apr 27, 2025
fef3db0
Fix regressions from machine-tux
victorapm Apr 28, 2025
77edbdb
Fix regressions from machine-tux
victorapm Apr 28, 2025
edfa644
Merge branch 'recmat' of github.com:hypre-space/hypre into recmat
victorapm Apr 29, 2025
45a762d
Merge branch 'master' into recmat
victorapm Apr 29, 2025
60c1571
Mixed int fix
waynemitchell May 2, 2025
d6b9fca
Device porting for struct and sstruct (#1272)
victorapm May 14, 2025
ecdd9f9
Merge branch 'master' into recmat
victorapm May 14, 2025
278ab71
Fix bug causing sstruct matmult to take incorrect branch if num nonze…
waynemitchell May 14, 2025
9ac68e7
Merge branch 'recmat' of github.com:hypre-space/hypre into recmat
waynemitchell May 14, 2025
479c376
Recmat - Adds support for symmetric matrices in the StructMatmult rou…
rfalgout May 15, 2025
e401bb1
Merge branch 'master' into recmat
victorapm May 22, 2025
1d841da
Use exit instead of MPI_Abort to finish the driver gracefully
victorapm May 28, 2025
1e80935
Added an example script to generate multiprecision methods
oseikuffuor1 Jun 2, 2025
a61a63b
Recmat - fix for part boundary box coarsening and SSAMG convergence (…
rfalgout Jun 4, 2025
4ad36d0
Adding scripts for writing multiprecision _pre functions
rfalgout Jun 9, 2025
3389968
All of the drivers build cleanly with mixed precision (struct_mp.c is…
rfalgout Jun 11, 2025
6cecc5c
Added initial code for def/undef in MUP krylov headers and changed so…
rfalgout Jun 12, 2025
a70731b
Added a headers script to the krylov directory (still using name 'kry…
rfalgout Jun 13, 2025
5ab556a
Updated krylov/mup_headers and related files
rfalgout Jun 13, 2025
3fab6b4
Minor changes to multiprecision scripts
rfalgout Jun 13, 2025
20b1215
Adding FIXEDPRECISION macro for multiprecision build
rfalgout Jun 13, 2025
72c16f9
Updated scripts and defined function pointer typedefs in krylov direc…
rfalgout Jun 14, 2025
20b44d4
Updated mup scripts and extended auto code generation
rfalgout Jun 14, 2025
66402ef
This version compiles krylov/mup_pre.c correctly
rfalgout Jun 14, 2025
1c2ceff
Fixed a bunch of multiprecision lobpcg issues in krylov (clean commit)
rfalgout Jun 15, 2025
626815a
Fixed issue with normal compile without mixed precision
rfalgout Jun 15, 2025
60f5b35
Last commit wasn't quite correct
rfalgout Jun 15, 2025
dc5136f
More portable test makefile for multiprecision drivers
rfalgout Jun 15, 2025
90406b4
Modifed mup_pre.c code to avoid compiler warnings
rfalgout Jun 15, 2025
ad8cf93
More compiler warning fixes in mup_pre.c
rfalgout Jun 15, 2025
ebacdb8
Added build type MP_BUILD_DEFAULT to the Makefiles
rfalgout Jun 16, 2025
64721c1
Added prototypes for multiprecision functions
rfalgout Jun 16, 2025
e5246bb
Add 3D version of miller problem
victorapm Jun 19, 2025
d8e2863
Add missing extern C directives.
oseikuffuor1 Jun 26, 2025
5db9a09
Merge branch 'master' into mp-build-all
oseikuffuor1 Jun 26, 2025
0de42b8
Updated saved files after merge from master.
oseikuffuor1 Jun 26, 2025
8f7a1a0
Merge branch 'mp-build-all' into mup-functions
rfalgout Jun 30, 2025
0d8ad50
Merge branch 'master' into mp-build-all
rfalgout Jun 30, 2025
a006122
Merge branch 'mp-build-all' into mup-functions
rfalgout Jun 30, 2025
969deb9
Adding the summary option to checktest.sh as in the recmat branch
rfalgout Jun 30, 2025
079c550
Fixed some awk warnings on tux (hopefully this still works on the mac)
rfalgout Jul 1, 2025
2b4ff00
Temporary fix for an issue with FlexGMRES and ModifyPC for multipreci…
rfalgout Jul 1, 2025
0ebbe7f
Adding an autotest script for mixed precision
rfalgout Jul 1, 2025
80dbd66
Fixed a bug in the ij driver that showed up in TEST_ij/matrix.sh
rfalgout Jul 1, 2025
14fe9ee
Fixed some strict checking issues
rfalgout Jul 1, 2025
d482ab5
Merge branch 'master' into recmat
victorapm Jul 1, 2025
b24efaa
[recmat]: GPU porting for StructMatMult routines (#1285)
victorapm Jul 1, 2025
7c6724f
Fix regressions from 07-02-25
victorapm Jul 2, 2025
d5bf914
Changing name mup.none to mup.fixed
rfalgout Jul 2, 2025
5929b73
Fixed a memory issue with --enable-single by using the timing files f…
rfalgout Jul 2, 2025
86f3891
Updated the check-license.sh script to ignore function list files (e.…
rfalgout Jul 2, 2025
02b3623
Naming convention changes in the krylov/Makefile
rfalgout Jul 2, 2025
586589c
Changed the krylov directory mixed precision build to use suffixes
rfalgout Jul 2, 2025
551cea8
Fixing some mixed precision regression tests
rfalgout Jul 3, 2025
816952b
Minor fix
victorapm Jul 3, 2025
aa10319
Reworked the code generation approach - it's faster and easy to transfer
rfalgout Jul 4, 2025
012d3d1
Cleaning up a few things
rfalgout Jul 4, 2025
076f876
Cleaning up some more
rfalgout Jul 4, 2025
42c2fb7
More cleaning up
rfalgout Jul 4, 2025
af4c158
Minor script update to further increase speed and sort functions
rfalgout Jul 4, 2025
30e79f7
Forgot to add the mup_lobpcg files
rfalgout Jul 4, 2025
449f2d7
Moved the main code generation script to 'config'
rfalgout Jul 4, 2025
57851bd
Removing mup_methods.c file since it is not needed yet
rfalgout Jul 4, 2025
013e48d
Updated comments in config/gen_* scripts
rfalgout Jul 4, 2025
c13ea66
Fixing a small issue in pcg.c
rfalgout Jul 14, 2025
69adae8
Removing all the mup*.h headers from 'krylov'
rfalgout Jul 14, 2025
1d3d698
Minor changes to HYPRE_krylov.h and HYPRE_lobpcg.h prototype renaming
rfalgout Jul 15, 2025
d9062aa
Minor name changes to better conform to coding style conventions
rfalgout Jul 15, 2025
ada0c18
Fixing a compiler warning
rfalgout Jul 16, 2025
c4d799a
Minor header file change
rfalgout Jul 16, 2025
034b472
Caliper fix
waynemitchell Jul 16, 2025
78bbf69
Adding a top-level 'mup_code' script to write all multiprecision code…
rfalgout Jul 16, 2025
895f496
Changing name of multiprecision.h to _hypre_mup_def.h
rfalgout Jul 16, 2025
087810c
Renamed 'lobpcg.h' to '_hypre_lobpcg.h'
rfalgout Jul 16, 2025
8e55636
Changed 'krylov.h' to '_hypre_krylov.h'
rfalgout Jul 16, 2025
ee1e6f4
Moved some of '_hypre_mup_def.h' to 'HYPRE_utilities.h'
rfalgout Jul 17, 2025
ee8b705
Adding a global runtime precision variable
rfalgout Jul 17, 2025
a4a82f9
Fixing the regular compile
rfalgout Jul 17, 2025
c23eba5
Removing krylov.h
rfalgout Jul 17, 2025
06827de
Moving krylov/mup_headers script into generic config/gen_code.sh script
rfalgout Jul 17, 2025
1c0a855
Changed krylov mup.functions to be only external user API routines
rfalgout Jul 17, 2025
67ba363
Updating sstruct_ls to multiprecision and removing FAC and Maxwell so…
rfalgout Jul 18, 2025
90c553d
Updating top-level mup_code script
rfalgout Jul 18, 2025
6a25be5
Updating scripts for checking function lists for multiprecision
rfalgout Jul 18, 2025
003b7ab
Updating sstruct_mv to multiprecision
rfalgout Jul 18, 2025
86035ee
Removing mistake file commit
rfalgout Jul 18, 2025
2152978
Updating struct_mv to multiprecision
rfalgout Jul 19, 2025
b3e861c
Removing SparseMSG (it is already removed on recmat branch)
rfalgout Jul 19, 2025
9f0a75f
Updating struct_ls to multiprecision
rfalgout Jul 19, 2025
ff2c3b8
Updating parcsr_mv and parcsr_ls to multiprecision
rfalgout Jul 20, 2025
95a6f53
Update the parcsr_mv/mup.fixed list
rfalgout Jul 20, 2025
75104f2
Updating multiprecision code after code generation script changes
rfalgout Jul 20, 2025
18fe741
Updating IJ_mv to multiprecision
rfalgout Jul 20, 2025
6de4f65
Updating seq_mv to multiprecision
rfalgout Jul 20, 2025
9cc4a3e
Cleaned up a few things
rfalgout Jul 20, 2025
9542c54
Updated mup code generation scripts to handle Foo(void) functions cor…
rfalgout Jul 21, 2025
aef4620
Adding mup_functions.c implementations and removing '_def' symbols
rfalgout Jul 21, 2025
4c6b1bd
Updating blas and lapack to multiprecision
rfalgout Jul 22, 2025
c7bc774
Updating seq_block_mv to multiprecision
rfalgout Jul 22, 2025
04212d6
Updating check-license.sh autotest script to ignore the mup_*.c files
rfalgout Jul 22, 2025
c0a29c7
Updating ParaSails to multiprecision
rfalgout Jul 22, 2025
6722e6c
Merge branch 'mup-functions' of github.com:hypre-space/hypre into mup…
rfalgout Jul 22, 2025
6e10411
Updating utilities to multiprecision
rfalgout Jul 22, 2025
4ace2d8
Fixing an autotest error with check-double
rfalgout Jul 22, 2025
226e9b6
Fixing check-license autotest error
rfalgout Jul 23, 2025
aaf2220
Cleaning up some header file stuff
rfalgout Jul 23, 2025
0baabb9
Adding a draft of a runtime multiprecision sstruct.c driver
rfalgout Jul 24, 2025
10ed28e
Fixing a bug in sstruct.c and adding a -precision option
rfalgout Jul 24, 2025
044f42e
[recmat]: GPU support for SStruct interface and SSAMG (#1308)
victorapm Jul 25, 2025
456b3c3
Merge branch 'master' into recmat
victorapm Jul 25, 2025
3bcc19c
Merge branch 'master' into mp-build-all
rfalgout Jul 25, 2025
9a23fe6
Merge branch 'mp-build-all' into mup-functions
rfalgout Jul 25, 2025
ce9d41c
Fixed machine-tux-mixed regression test errors and added a C++ test
rfalgout Jul 25, 2025
c9f2023
Updating Euclid to multiprecision
rfalgout Jul 25, 2025
3792d61
Updating mup_check and mup_code scripts
rfalgout Jul 25, 2025
2db59fa
Merge pull request #1315 from hypre-space/recmat
rfalgout Jul 25, 2025
904c346
Updating parcsr_block_mv to multiprecision
rfalgout Jul 25, 2025
712547f
Updating pilut to multiprecision
rfalgout Jul 25, 2025
527af40
Updating distributed_matrix to multiprecision
rfalgout Jul 25, 2025
d6a780b
Updating matrix_matrix to multiprecision
rfalgout Jul 25, 2025
78067a8
Updating multivector to multiprecision
rfalgout Jul 25, 2025
a1b1311
Merge pull request #1296 from hypre-space/mup-functions
rfalgout Jul 28, 2025
97a1d81
Merge branch 'hypre-3.0' into mp-build-all
rfalgout Jul 28, 2025
caac44e
Fixing some sstruct.c runtime precision errors
rfalgout Jul 28, 2025
c031c06
Cleaning up the machine-tux-mixed.sh autotest runs
rfalgout Jul 28, 2025
e8f6459
Changed saved file for TEST_error - line numbers and sort order changed
rfalgout Jul 29, 2025
0ec946d
Removing some unnecessary prototypes from sstruct_ls/protos.h
rfalgout Jul 29, 2025
91cc646
Minor include file change
rfalgout Jul 29, 2025
6c0c9a0
Minor include file change
rfalgout Jul 29, 2025
2e5385c
Removing unused variables from par_lr_restr.c
rfalgout Jul 29, 2025
e12c46f
Removing unnecessary file
rfalgout Jul 29, 2025
b8b4907
Fix CMake build + Resolve compilation warnings with Wextra
victorapm Jul 29, 2025
420436d
Removing original multiprecision lists, scripts, and headers
rfalgout Jul 29, 2025
a934c6a
Removed writeHeader.sh script and updated usage comment for write_hea…
rfalgout Jul 30, 2025
ceb5b93
Merge pull request #1317 from hypre-space/mp-build-all
rfalgout Jul 30, 2025
ec94128
HYPRE 3.0 multiprecision macros and astyle (#1319)
rfalgout Jul 31, 2025
c625c73
Initial version of tux tests for hypre-3.0 (#1320)
rfalgout Jul 31, 2025
34c0e79
[hypre 3.0]: update mup scripts (#1321)
victorapm Aug 1, 2025
0aa1be3
[hypre 3.0]: Improve `mup_code` (#1322)
victorapm Aug 1, 2025
3fc1bf7
Removing FEI, MLI, FAC, Maxwell from documentation
rfalgout Aug 4, 2025
1109fa5
Initial draft of mixed precision user documentation
rfalgout Aug 5, 2025
381ed78
Minor updates to mixed precision documentation
rfalgout Aug 5, 2025
81a6427
First draft of multiprecision developer Wiki
rfalgout Aug 5, 2025
18f8876
Minor updates to mixed precision Wiki
rfalgout Aug 5, 2025
1f84345
More minor changes to mixed precision developer Wiki
rfalgout Aug 5, 2025
1745af7
[hypre 3.0]: CMake build - mixed precision (#1325)
victorapm Aug 11, 2025
bea1e0f
[hypre 3.0]: Add back APIs (#1327)
victorapm Aug 11, 2025
6c0782f
Run mup_code
victorapm Aug 12, 2025
edf3b17
Minor updates to multiprecision documentation
rfalgout Aug 14, 2025
ff7a45f
Another minor update to multiprecision documentation
rfalgout Aug 14, 2025
9b3ebb3
Minor updates to mixed precision developer wiki
rfalgout Aug 14, 2025
c07b4d2
Another minor update to mixed precision wiki
rfalgout Aug 14, 2025
ea51a0f
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 14, 2025
0eaa8fe
First draft of a utilities API for user manual
rfalgout Aug 15, 2025
4ef48bc
Fixing an issue with the draft utilities API update
rfalgout Aug 15, 2025
99e7335
[hypre 3.0]: Struct optimizations (#1323)
victorapm Aug 15, 2025
59533cd
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 15, 2025
2a3ef2b
Some cleaning up in the utilities API documentation
rfalgout Aug 15, 2025
323b772
Add SSAMG docs (#1330)
victorapm Aug 15, 2025
f763cdb
Remove unused code (#1331)
victorapm Aug 15, 2025
e2902cb
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 15, 2025
4d27894
Adding SSAMG page to docs
rfalgout Aug 15, 2025
59100c0
Hypre 3.0 unused code and minor cleanup (#1333)
oseikuffuor1 Aug 18, 2025
2167e56
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 18, 2025
f6995ee
HYPRE 3.0 Tux Regression Tests (#1335)
rfalgout Aug 21, 2025
3f01f17
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 21, 2025
5057b97
[hypre 3.0] TEST_struct GPU regression fixes (#1334)
waynemitchell Aug 22, 2025
bf7669d
Merge branch 'master' into hypre-3.0
victorapm Aug 22, 2025
3fbb183
[hypre 3.0] TEST_struct GPU .saved file changes (#1340)
waynemitchell Aug 22, 2025
217d80e
Hypre 3.0 ij (#1338)
liruipeng Aug 22, 2025
56a893c
Adding ParCSR matrix/vector interface to API docs
rfalgout Aug 26, 2025
80bca4c
Forgot to update a file for ParCSR matrix/vector API docs
rfalgout Aug 26, 2025
e223d28
Forgot one other thing for the ParCSR docs
rfalgout Aug 26, 2025
ec622f0
[hypre 3.0] GPU sstruct regression fixes (#1345)
waynemitchell Aug 26, 2025
9b9f718
Fix Wextra compilation warnings on CUDA/HIP builds (#1342)
victorapm Aug 26, 2025
fe06834
Updates of HYPRE_parcsr_mv.h for API documentation
rfalgout Aug 26, 2025
9b041ad
mv routines in HYPRE header
liruipeng Aug 27, 2025
cf01486
add matmat routines
liruipeng Aug 28, 2025
35d333e
[hypre 3.0] struct regression fixes on Tuolumne (#1347)
waynemitchell Aug 28, 2025
c96eca1
Changes to matrix/vector API documentation
rfalgout Aug 28, 2025
9afbe61
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 28, 2025
df74c57
More cleaning up of matrix-vector interface and API docs
rfalgout Aug 28, 2025
967e90c
Minor matrix-vector API changes and update of multiprecision
rfalgout Aug 28, 2025
211f602
First draft of hypre 3.0 CHANGELOG
rfalgout Aug 29, 2025
b6a2417
Update of CHANGELOG for hypre 3.0
rfalgout Aug 29, 2025
a3a1b36
[hypre 3.0]: GPU build fixes (#1348)
victorapm Aug 29, 2025
b5cc3f1
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 29, 2025
ff883fa
Change API of HYPRE_StructVectorSetValues and AddToValues (#1349)
rfalgout Aug 29, 2025
a86c269
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 29, 2025
bd6fea2
HYPRE 3.0 - Fixing compiler errors when complex is enabled (#1350)
rfalgout Aug 30, 2025
57a8368
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Aug 30, 2025
583f178
Add a cmake-compile regression test for mixed precision (#1351)
rfalgout Aug 30, 2025
fa8c792
[hypre-3.0]: Fix regressions with rocm 5.4.3 (#1352)
victorapm Sep 1, 2025
511ec7b
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Sep 1, 2025
4f31a1c
[hypre 3.0]: CUDA 13 support (#1353)
victorapm Sep 2, 2025
0dc95c1
Remove error checking from hypre_gselim (#1355)
rfalgout Sep 2, 2025
4e16b8f
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Sep 2, 2025
d9d97e5
Bug fix found through PETSc tests
victorapm Sep 3, 2025
0a76142
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Sep 3, 2025
6f657d3
Throw an error message in obsolete function HYPRE_AssumedPartitionCheck
rfalgout Sep 3, 2025
154514a
Add support to dane on runtest.sh (#1356)
victorapm Sep 3, 2025
4ab1f9e
Hypre 3.0 mp tests (#1357)
oseikuffuor1 Sep 3, 2025
3347fa4
CMake and autotools install the same include files (except for mup he…
victorapm Sep 3, 2025
d6a9f31
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Sep 3, 2025
fab3845
HYPRE 3.0 replace hypre_printf with hypre_error_w_msg in _pre functio…
rfalgout Sep 3, 2025
4ffc4ef
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Sep 3, 2025
0918e14
Fixing typo in cmake file
rfalgout Sep 3, 2025
a746fac
Merge branch 'hypre-3.0' into hypre-3.0-docs
rfalgout Sep 5, 2025
1547b80
Merge pull request #1329 from hypre-space/hypre-3.0-docs
rfalgout Sep 5, 2025
8cd7a43
Exclude the wiki-dev directory from the check-license.sh test
rfalgout Sep 6, 2025
424fa0d
[hypre 3.0]: Umpire default (#1360)
victorapm Sep 8, 2025
44a173a
Disable umpire explicitly in some tests
victorapm Sep 8, 2025
b6db3ce
Fix memory issue evidenced by AMReX tests
victorapm Sep 9, 2025
5e9af21
Moose fix: include HYPRE_config
victorapm Sep 9, 2025
47fd6f1
Minor fix for CMake build with cuda enabled
victorapm Sep 10, 2025
03cdcb1
Bug fix: git hash retrieval in cmake build
victorapm Sep 11, 2025
b950aec
mixed precision conversion (#1369)
ulrikeyang Sep 12, 2025
0d0f079
[hypre 3.0]: Moose Fixes (#1363)
victorapm Sep 15, 2025
0493b2c
Machine lassen fix (#1364)
waynemitchell Sep 18, 2025
905f9ee
Add cxx standard 11 to machine lassen cmake tests
waynemitchell Sep 19, 2025
7f47463
Updating release number and date for hypre-3.0.0
rfalgout Sep 19, 2025
b82140a
[hypre 3.0]: Fix MSVC build (#1371)
victorapm Sep 20, 2025
8d84ce9
[hypre 3.0]: Fix hypre headers (#1374)
victorapm Sep 20, 2025
d33a66b
Fix tests for PFMG with constant coefficients (#1372)
victorapm Sep 21, 2025
46e284d
[hypre 3.0]: Fix Struct I/O tests (#1373)
victorapm Sep 21, 2025
6fb426b
Separate minor fix
victorapm Sep 21, 2025
2cfe2cc
[hypre 3.0]: Fix sstruct tests using host in device builds (#1375)
victorapm Sep 22, 2025
c1822ae
Fix HYPRE_Int/HYPRE_BigInt usage (#1376)
victorapm Sep 22, 2025
0d281bb
Fixes for basic-cuda-um tests
victorapm Sep 22, 2025
899aac6
Run mup_code
victorapm Sep 22, 2025
517070b
[hypre 3.0]: Fix MGR issues (#1377)
victorapm Sep 22, 2025
de6c47f
[hypre 3.0]: Update regression tests baselines (#1378)
victorapm Sep 22, 2025
5df3647
Updating check-headers autotest filters to use new header names
rfalgout Sep 22, 2025
8bb935c
Fix unused variable warning
victorapm Sep 22, 2025
3b784da
Correct a residual norm from #1378
victorapm Sep 22, 2025
004de93
[hypre 3.0]: Fix CMake's generated soversion (#1379)
victorapm Sep 23, 2025
8a2a785
Correct a residual norm from #1378
victorapm Sep 23, 2025
530ce72
Fix CMake build with -DHYPRE_BUILD_EXAMPLES=ON
victorapm Sep 23, 2025
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
25 changes: 20 additions & 5 deletions .gitignore
Original file line number Diff line number Diff line change
@@ -1,5 +1,11 @@
*.o
*.obj
*.o_flt
*.o_dbl
*.o_ldbl
*.obj_flt
*.obj_dbl
*.obj_ldbl
*.a
*.so
*.out.*
Expand All @@ -18,9 +24,11 @@ config.status
Makefile.config
HYPRE_config.h
autom4te.cache
*.TVD.v4breakpoints*
src/TAGS
hypre/
install/
build*/
install*/
AUTOTEST/*.dir
out/
.vscode
Expand Down Expand Up @@ -52,18 +60,25 @@ src/test/ij_mv
src/test/ij_mv.exe
src/test/ij_mm
src/test/ij_mm.exe
src/test/maxwell_unscaled
src/test/maxwell_unscaled.exe
src/test/ij_mp
src/test/sstruct
src/test/sstruct.exe
src/test/sstruct_fac
src/test/sstruct_fac.exe
src/test/sstructmat
src/test/sstructmat.exe
src/test/struct
src/test/struct.exe
src/test/structmat
src/test/structmat.exe
src/test/struct_migrate
src/test/struct_migrate.exe
src/test/test_mp
src/test/test_mp_pcg
src/test/test_mp_pcg_3d
src/test/zboxloop
src/test/zboxloop.exe
src/test/test_mp
src/test/test_mp_pcg
src/test/test_mp_pcg_3d
src/examples/ex[0-9]
src/examples/ex[0-9].exe
src/examples/ex1[0-8]
Expand Down
7 changes: 6 additions & 1 deletion AUTOTEST/check-double.sh
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,12 @@ find . -type f -print | egrep '[.]*[.](c|cc|cpp|cxx|C|h|hpp|hxx|H)$' |
egrep -v '/docs' |
egrep -v '/examples' |
egrep -v '/FEI_mv' |
egrep -v '/hypre/include' > check-double.files
egrep -v '/hypre/include' |
egrep -v '/test/ij_mp.c' |
egrep -v '/test/struct_mp.c' |
egrep -v '/test/test_mp.c' |
egrep -v '/test/test_mp_pcg.c' |
egrep -v '/test/test_mp_pcg_3d.c' > check-double.files

egrep '(^|[^[:alnum:]_-]+)double([^[:alnum:]_-]+|$)' `cat check-double.files` >&2

Expand Down
8 changes: 4 additions & 4 deletions AUTOTEST/check-headers.filters
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
mpicc
mpif77
include/fortran_matrix.h
include/multivector.h
include/interpreter.h
include/temp_multivector.h
include/_hypre_fortran_matrix.h
include/_hypre_lobpcg_multivector.h
include/_hypre_lobpcg_interpreter.h
include/_hypre_lobpcg_temp_multivector.h
fatal: No names found, cannot describe anything.
7 changes: 6 additions & 1 deletion AUTOTEST/check-int.sh
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,12 @@ find . -type f -print | egrep '[.]*[.](c|cc|cpp|cxx|C|h|hpp|hxx|H)$' |
egrep -v '/docs' |
egrep -v '/examples' |
egrep -v '/FEI_mv' |
egrep -v '/hypre/include' > check-int.files
egrep -v '/hypre/include' |
egrep -v '/test/ij_mp.c' |
egrep -v '/test/struct_mp.c' |
egrep -v '/test/test_mp.c' |
egrep -v '/test/test_mp_pcg.c' |
egrep -v '/test/test_mp_pcg_3d.c' > check-int.files

egrep '(^|[^[:alnum:]_]+)int([^[:alnum:]_]+|$)' `cat check-int.files` >&2

Expand Down
2 changes: 0 additions & 2 deletions AUTOTEST/check-license.filters
Original file line number Diff line number Diff line change
Expand Up @@ -4,8 +4,6 @@
./AUTOTEST/.*.log
./AUTOTEST/runtests-.*
./src/test/struct_migrate
./src/test/maxwell_unscaled
./src/test/sstruct_fac
./src/test/ams_driver
./src/test/zboxloop
./src/test/ij
Expand Down
6 changes: 5 additions & 1 deletion AUTOTEST/check-license.sh
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,10 @@ egrep -LR "$LicStr" . | egrep -v '[.](o|obj|a|filters|pdf|svg|gif|png)$' |
egrep -v '[.]/[.]git' |
egrep -v '[.]/.*HYPRE_config[.]h' |
egrep -v '[.]/src/(blas|lapack)/.*[.]c' |
egrep -v '[.]/src/.*functions[.]saved$' |
egrep -v '[.]/src/.*mup[.](exclude|fixed|functions|methods)$' |
egrep -v '[.]/src/examples/docs' |
egrep -v '[.]/src/docs/wiki-dev' |
egrep -v '[.]/src/test/TEST_.*' > check-license.files

# Add these file back to the list
Expand Down Expand Up @@ -89,4 +92,5 @@ rm -f check-license.remove check-license.files
### Next check for files that should not have the license, but do

# blas and lapack '.c' files should not have an LLNL license
egrep -lR "$LicStr" ./src/blas ./src/lapack | egrep '[.]/src/(blas|lapack)/.*[.]c' >&2
egrep -lR "$LicStr" ./src/blas ./src/lapack | egrep '[.]/src/(blas|lapack)/.*[.]c' |
egrep -v 'mup.*[.]c' >&2
7 changes: 6 additions & 1 deletion AUTOTEST/check-mem.sh
Original file line number Diff line number Diff line change
Expand Up @@ -39,7 +39,12 @@ find . -type f -print | egrep '[.]*[.](c|cc|cpp|cxx|C|h|hpp|hxx|H)$' |
egrep -v '/utilities/memory_tracker.c' |
egrep -v '/utilities/memory.c' |
egrep -v '/utilities/general.c' |
egrep -v '/utilities/device_utils.c' > check-mem.files
egrep -v '/utilities/device_utils.c' |
egrep -v '/test/ij_mp.c' |
egrep -v '/test/struct_mp.c' |
egrep -v '/test/test_mp.c' |
egrep -v '/test/test_mp_pcg.c' |
egrep -v '/test/test_mp_pcg_3d.c' > check-mem.files

egrep '(^|[^[:alnum:]_]+)malloc[[:space:]]*\(' `cat check-mem.files` >&2
egrep '(^|[^[:alnum:]_]+)calloc[[:space:]]*\(' `cat check-mem.files` >&2
Expand Down
7 changes: 6 additions & 1 deletion AUTOTEST/check-mpi.sh
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,12 @@ find . -type f -print | egrep '[.]*[.](c|cc|cpp|cxx|C|h|hpp|hxx|H)$' |
egrep -v '/docs' |
egrep -v '/examples' |
egrep -v '/FEI_mv' |
egrep -v '/hypre/include' > check-mpi.files
egrep -v '/hypre/include' |
egrep -v '/test/ij_mp.c' |
egrep -v '/test/struct_mp.c' |
egrep -v '/test/test_mp.c' |
egrep -v '/test/test_mp_pcg.c' |
egrep -v '/test/test_mp_pcg_3d.c' > check-mpi.files

egrep '(^|[^[:alnum:]_]+)MPI_' `cat check-mpi.files` |
egrep -v 'MPI_Comm([^_]|$)' >&2
Expand Down
8 changes: 8 additions & 0 deletions AUTOTEST/cmake.filters
Original file line number Diff line number Diff line change
@@ -1,3 +1,11 @@
fatal: No names found, cannot describe anything.
Additional optimization may be attained by recompiling and specifying MAXMEM option with a value greater than 8192.
warning: 1501-264 The options qsmp and O0 are incompatible. Option O0 is ignored.
CMake Warning at config/cmake/HYPRE_SetupGPUToolkit.cmake:97 (message):
===============================================================

Umpire is explicitly disabled while building hypre for GPUs.
This is not recommended for performance reasons!
Enable it with -DHYPRE_ENABLE_UMPIRE=ON.
Call Stack (most recent call first):
CMakeLists.txt:400 (include)
2 changes: 1 addition & 1 deletion AUTOTEST/cmake.sh
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@

testname=`basename $0 .sh`

drivers="ij sstruct struct ams_driver maxwell_unscaled sstruct_fac struct_migrate ij_assembly"
drivers="ij sstruct struct ams_driver struct_migrate ij_assembly"

# Echo usage information
case $1 in
Expand Down
5 changes: 5 additions & 0 deletions AUTOTEST/configure.filters
Original file line number Diff line number Diff line change
@@ -1,2 +1,7 @@
fatal: No names found, cannot describe anything.
WARNING: unknown Fortran name-mangling scheme
configure: WARNING: ***************************************************************
configure: WARNING: Umpire is explicitly disabled while building hypre for GPUs.
configure: WARNING: This is not recommended for performance reasons!
configure: WARNING: Enable it with --with-umpire.
configure: WARNING: ***************************************************************
28 changes: 14 additions & 14 deletions AUTOTEST/machine-lassen.sh
Original file line number Diff line number Diff line change
Expand Up @@ -55,56 +55,56 @@ module -q load cuda
module -q load xl

# CUDA with UM in debug mode [ij, ams, struct, sstruct]
co="--with-cuda --enable-unified-memory --enable-persistent --enable-debug --with-gpu-arch=70 --with-memory-tracker --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-unified-memory --enable-persistent --enable-debug --with-gpu-arch=70 --with-memory-tracker --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
ro="-ij-gpu -ams -struct -sstruct -rt -mpibind -save ${save} -rtol ${rtol} -atol ${atol}"
eo="-gpu -rt -mpibind -save ${save} -rtol ${rtol} -atol ${atol}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro -eo: $eo
./renametest.sh basic $output_dir/basic-cuda-um

# CUDA with UM in debug mode [ij, ams, struct, sstruct]
co="--with-cuda --enable-unified-memory --enable-persistent --enable-debug --with-print-errors --with-gpu-arch=70 --with-memory-tracker --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
# CUDA with UM in debug mode with print errors [error]
co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-unified-memory --enable-persistent --enable-debug --with-print-errors --with-gpu-arch=70 --with-memory-tracker --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
ro="-error -rt -mpibind -save ${save} -rtol ${rtol} -atol ${atol}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-cuda-um-with-errors

# CUDA with UM and mixed-int
co="--with-cuda --enable-unified-memory --enable-mixedint --enable-debug --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-unified-memory --enable-mixedint --enable-debug --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
ro="-ij-mixed -ams -struct -sstruct-mixed -rt -mpibind -save ${save} -rtol ${rtol} -atol ${atol}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-cuda-um-mixedint

# CUDA with UM with shared library
co="--with-cuda --enable-unified-memory --with-openmp --enable-hopscotch --enable-shared --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-unified-memory --with-openmp --enable-hopscotch --enable-shared --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
ro="-gpumemcheck -rt -mpibind -cudamemcheck -save ${save}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-cuda-um-shared

# CUDA with UM and single precision
co="--with-cuda --enable-unified-memory --enable-single --enable-cusolver --enable-debug --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-unified-memory --enable-single --enable-cusolver --enable-debug --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
ro="-single -rt -mpibind -save ${save}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: ${ro}
./renametest.sh basic $output_dir/basic-cuda-um-single

# CUDA with UM without MPI [no run]
#co="--with-cuda --enable-unified-memory --without-MPI --with-gpu-arch=70 --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
#co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-unified-memory --without-MPI --with-gpu-arch=70 --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
#./test.sh basic.sh $src_dir -co: $co -mo: $mo
#./renametest.sh basic $output_dir/basic-cuda-um-without-MPI

# CUDA without UM with device memory pool [struct]
co="--with-cuda --enable-device-memory-pool --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-device-memory-pool --with-gpu-arch=70 --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
ro="-struct -rt -mpibind -save ${save}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-cuda-nonum

# CUDA without UM with umpire [benchmark]
UMPIRE_DIR=/usr/workspace/hypre/ext-libs/Umpire/install_umpire-2025.03.0_nvcc11.2-sm_70-xl2023.06.28-cuda-11.2.0-gcc-8.3.1
co="--with-cuda --with-gpu-arch=70 --with-umpire --with-umpire-include=${UMPIRE_DIR}/include --with-umpire-lib-dirs=${UMPIRE_DIR}/lib --with-umpire-libs=\\'camp umpire\\' --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
co="--with-cuda --with-cxxstandard=11 --with-gpu-arch=70 --with-umpire --with-umpire-include=${UMPIRE_DIR}/include --with-umpire-lib-dirs=${UMPIRE_DIR}/lib --with-umpire-libs=\\'camp umpire\\' --with-extra-CFLAGS=\\'-qsuppress=1500-029\\' --with-extra-CXXFLAGS=\\'-qsuppress=1500-029\\'"
ro="-bench -rt -mpibind -save ${save}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-cuda-bench

# run on CPU
co="--with-cuda --with-test-using-host --with-memory-tracker --enable-debug --with-gpu-arch=70"
co="--with-cuda --with-cxxstandard=11 --without-umpire --with-test-using-host --with-memory-tracker --enable-debug --with-gpu-arch=70"
ro="-ij-noilu -ams -struct -sstruct -rt -mpibind -save lassen_cpu"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-cuda-cpu
Expand Down Expand Up @@ -133,17 +133,17 @@ module list cmake/3.22 |& grep "None found"
mo="-j"

# CUDA with UM + CMake
co="-DCMAKE_C_COMPILER=$(which xlc) -DCMAKE_CXX_COMPILER=$(which xlc++) -DCMAKE_CUDA_COMPILER=$(which nvcc) -DMPI_C_COMPILER=$(which mpicc) -DMPI_CXX_COMPILER=$(which mpicxx) -DHYPRE_ENABLE_CUDA=ON -DHYPRE_ENABLE_UNIFIED_MEMORY=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_ENABLE_PERSISTENT_COMM=ON -DHYPRE_ENABLE_DEVICE_POOL=ON -DHYPRE_WITH_EXTRA_CFLAGS="\'"-qsuppress=1500-029"\'" -DHYPRE_WITH_EXTRA_CXXFLAGS="\'"-qsuppress=1500-029"\'" -DCMAKE_CUDA_ARCHITECTURES=70"
co="-DCMAKE_C_COMPILER=$(which xlc) -DCMAKE_CXX_COMPILER=$(which xlc++) -DCMAKE_CUDA_COMPILER=$(which nvcc) -DMPI_C_COMPILER=$(which mpicc) -DMPI_CXX_COMPILER=$(which mpicxx) -DCMAKE_CXX_STANDARD=11 -DHYPRE_ENABLE_CUDA=ON -DHYPRE_ENABLE_UMPIRE=OFF -DHYPRE_ENABLE_UNIFIED_MEMORY=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_ENABLE_PERSISTENT_COMM=ON -DHYPRE_ENABLE_DEVICE_POOL=ON -DHYPRE_WITH_EXTRA_CFLAGS="\'"-qsuppress=1500-029"\'" -DHYPRE_WITH_EXTRA_CXXFLAGS="\'"-qsuppress=1500-029"\'" -DCMAKE_CUDA_ARCHITECTURES=70"
./test.sh cmake.sh $root_dir -co: $co -mo: $mo
./renametest.sh cmake $output_dir/cmake-cuda-um-ij

# CUDA with UM [shared library] + CMake
co="-DCMAKE_C_COMPILER=$(which xlc) -DCMAKE_CXX_COMPILER=$(which xlc++) -DCMAKE_CUDA_COMPILER=$(which nvcc) -DMPI_C_COMPILER=$(which mpicc) -DMPI_CXX_COMPILER=$(which mpicxx) -DHYPRE_ENABLE_CUDA=ON -DHYPRE_ENABLE_UNIFIED_MEMORY=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_ENABLE_OPENMP=ON -DHYPRE_ENABLE_HOPSCOTCH=ON -DBUILD_SHARED_LIBS=ON -DHYPRE_WITH_EXTRA_CFLAGS="\'"-qsuppress=1500-029"\'" -DHYPRE_WITH_EXTRA_CXXFLAGS="\'"-qsuppress=1500-029 "\'" -DCMAKE_CUDA_ARCHITECTURES=70"
co="-DCMAKE_C_COMPILER=$(which xlc) -DCMAKE_CXX_COMPILER=$(which xlc++) -DCMAKE_CUDA_COMPILER=$(which nvcc) -DMPI_C_COMPILER=$(which mpicc) -DMPI_CXX_COMPILER=$(which mpicxx) -DCMAKE_CXX_STANDARD=11 -DHYPRE_ENABLE_CUDA=ON -DHYPRE_ENABLE_UMPIRE=OFF -DHYPRE_ENABLE_UNIFIED_MEMORY=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_ENABLE_OPENMP=ON -DHYPRE_ENABLE_HOPSCOTCH=ON -DBUILD_SHARED_LIBS=ON -DHYPRE_WITH_EXTRA_CFLAGS="\'"-qsuppress=1500-029"\'" -DHYPRE_WITH_EXTRA_CXXFLAGS="\'"-qsuppress=1500-029 "\'" -DCMAKE_CUDA_ARCHITECTURES=70"
./test.sh cmake.sh $root_dir -co: $co -mo: $mo
./renametest.sh cmake $output_dir/cmake-cuda-um-shared

# CUDA w.o UM + CMake
co="-DCMAKE_C_COMPILER=$(which xlc) -DCMAKE_CXX_COMPILER=$(which xlc++) -DCMAKE_CUDA_COMPILER=$(which nvcc) -DMPI_C_COMPILER=$(which mpicc) -DMPI_CXX_COMPILER=$(which mpicxx) -DHYPRE_ENABLE_CUDA=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_WITH_EXTRA_CFLAGS="\'"-qsuppress=1500-029"\'" -DHYPRE_WITH_EXTRA_CXXFLAGS="\'"-qsuppress=1500-029"\'" -DCMAKE_CUDA_ARCHITECTURES=70"
co="-DCMAKE_C_COMPILER=$(which xlc) -DCMAKE_CXX_COMPILER=$(which xlc++) -DCMAKE_CUDA_COMPILER=$(which nvcc) -DMPI_C_COMPILER=$(which mpicc) -DMPI_CXX_COMPILER=$(which mpicxx) -DCMAKE_CXX_STANDARD=11 -DHYPRE_ENABLE_CUDA=ON -DHYPRE_ENABLE_UMPIRE=OFF -DCMAKE_BUILD_TYPE=Debug -DHYPRE_WITH_EXTRA_CFLAGS="\'"-qsuppress=1500-029"\'" -DHYPRE_WITH_EXTRA_CXXFLAGS="\'"-qsuppress=1500-029"\'" -DCMAKE_CUDA_ARCHITECTURES=70"
./test.sh cmake.sh $root_dir -co: $co -mo: $mo
./renametest.sh cmake $output_dir/cmake-cuda-nonum-struct

Expand All @@ -160,7 +160,7 @@ mo="-j test"
module -q load gcc/8.3.1
module -q load cuda/12.0
module list cuda/12.0 |& grep "None found"
co="--with-cuda --enable-unified-memory --enable-thrust-nosync --enable-debug --with-gpu-arch=70 CC=mpicc CXX=mpicxx"
co="--with-cuda --with-cxxstandard=11 --without-umpire --enable-unified-memory --enable-thrust-nosync --enable-debug --with-gpu-arch=70 CC=mpicc CXX=mpicxx"
ro="-ij-gpu -ams -struct -sstruct -rt -mpibind -save ${save} -rtol ${rtol} -atol ${atol}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-cuda12_0
Expand Down
12 changes: 6 additions & 6 deletions AUTOTEST/machine-tioga.sh
Original file line number Diff line number Diff line change
Expand Up @@ -53,18 +53,18 @@ save="tioga"
module -q load rocm/6.2.1

# HIP without UM [benchmark, struct, ams]
co="--with-hip --with-MPI-include=${MPICH_DIR}/include --with-MPI-lib-dirs=${MPICH_DIR}/lib --with-MPI-libs=mpi --with-gpu-arch='gfx90a' CC=cc CXX=CC"
co="--with-hip --without-umpire --with-MPI-include=${MPICH_DIR}/include --with-MPI-lib-dirs=${MPICH_DIR}/lib --with-MPI-libs=mpi --with-gpu-arch='gfx90a' CC=cc CXX=CC"
ro="-ams -bench -struct -rt -save ${save}"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-hip-nonum

#HIP with UM and single precision [no run]
co="--with-hip --enable-unified-memory --enable-single --enable-debug --with-MPI-include=${MPICH_DIR}/include --with-MPI-lib-dirs=${MPICH_DIR}/lib --with-MPI-libs=mpi --with-gpu-arch='gfx90a' CC=cc CXX=CC"
co="--with-hip --without-umpire --enable-unified-memory --enable-single --enable-debug --with-MPI-include=${MPICH_DIR}/include --with-MPI-lib-dirs=${MPICH_DIR}/lib --with-MPI-libs=mpi --with-gpu-arch='gfx90a' CC=cc CXX=CC"
./test.sh basic.sh $src_dir -co: $co -mo: $mo
./renametest.sh basic $output_dir/basic-hip-um-single

# run on CPU
co="--with-hip --with-test-using-host --with-memory-tracker --enable-debug --with-MPI-include=${MPICH_DIR}/include --with-MPI-lib-dirs=${MPICH_DIR}/lib --with-MPI-libs=mpi --with-gpu-arch='gfx90a' CC=cc CXX=CC"
co="--with-hip --without-umpire --with-test-using-host --with-memory-tracker --enable-debug --with-MPI-include=${MPICH_DIR}/include --with-MPI-lib-dirs=${MPICH_DIR}/lib --with-MPI-libs=mpi --with-gpu-arch='gfx90a' CC=cc CXX=CC"
ro="-ij-noilu -ams -struct -sstruct -rt -D HYPRE_NO_SAVED"
./test.sh basic.sh $src_dir -co: $co -mo: $mo -ro: $ro
./renametest.sh basic $output_dir/basic-hip-cpu
Expand All @@ -79,17 +79,17 @@ module -q load cmake/3.24.2
mo="-j all check"

# HIP without UM + CMake (no full run, but with basic "make check")
co="-DCMAKE_C_COMPILER=cc -DCMAKE_CXX_COMPILER=CC -DMPI_C_COMPILER=cc -DMPI_CXX_COMPILER=CC -DHYPRE_ENABLE_HIP=ON -DCMAKE_HIP_ARCHITECTURES=gfx90a -DCMAKE_BUILD_TYPE=Debug -DHYPRE_BUILD_TESTS=ON"
co="-DCMAKE_C_COMPILER=cc -DCMAKE_CXX_COMPILER=CC -DMPI_C_COMPILER=cc -DMPI_CXX_COMPILER=CC -DHYPRE_ENABLE_HIP=ON -DHYPRE_ENABLE_UMPIRE=OFF -DCMAKE_HIP_ARCHITECTURES=gfx90a -DCMAKE_BUILD_TYPE=Debug -DHYPRE_BUILD_TESTS=ON"
./test.sh cmake.sh $root_dir -co: $co -mo: $mo
./renametest.sh cmake $output_dir/cmake-hip-nonum

# HIP without UM + Single precision + CMake (no full run, but with basic "make check")
co="-DBUILD_SHARED_LIBS=ON -DCMAKE_C_COMPILER=cc -DCMAKE_CXX_COMPILER=CC -DMPI_C_COMPILER=cc -DMPI_CXX_COMPILER=CC -DHYPRE_ENABLE_HIP=ON -DCMAKE_HIP_ARCHITECTURES=gfx90a -DHYPRE_ENABLE_SINGLE=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_BUILD_TESTS=ON"
co="-DBUILD_SHARED_LIBS=ON -DCMAKE_C_COMPILER=cc -DCMAKE_CXX_COMPILER=CC -DMPI_C_COMPILER=cc -DMPI_CXX_COMPILER=CC -DHYPRE_ENABLE_HIP=ON -DHYPRE_ENABLE_UMPIRE=OFF -DCMAKE_HIP_ARCHITECTURES=gfx90a -DHYPRE_ENABLE_SINGLE=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_BUILD_TESTS=ON"
./test.sh cmake.sh $root_dir -co: $co -mo: $mo
./renametest.sh cmake $output_dir/cmake-hip-nonum-single

# HIP with UM + Shared library + CMake (no full run, but with basic "make check")
co="-DBUILD_SHARED_LIBS=ON -DCMAKE_C_COMPILER=cc -DCMAKE_CXX_COMPILER=CC -DMPI_C_COMPILER=cc -DMPI_CXX_COMPILER=CC -DHYPRE_ENABLE_HIP=ON -DCMAKE_HIP_ARCHITECTURES=gfx90a -DHYPRE_ENABLE_UNIFIED_MEMORY=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_BUILD_TESTS=ON"
co="-DBUILD_SHARED_LIBS=ON -DCMAKE_C_COMPILER=cc -DCMAKE_CXX_COMPILER=CC -DMPI_C_COMPILER=cc -DMPI_CXX_COMPILER=CC -DHYPRE_ENABLE_HIP=ON -DHYPRE_ENABLE_UMPIRE=OFF -DCMAKE_HIP_ARCHITECTURES=gfx90a -DHYPRE_ENABLE_UNIFIED_MEMORY=ON -DCMAKE_BUILD_TYPE=Debug -DHYPRE_BUILD_TESTS=ON"
./test.sh cmake.sh $root_dir -co: $co -mo: $mo
./renametest.sh cmake $output_dir/cmake-hip-um-shared

Expand Down
Loading