fdurastante/psblas3
1
0
Fork 0
mirror of https://github.com/sfilippone/psblas3.git synced 2026-10-16 19:02:09 +00:00
Code Issues Packages Projects Releases Wiki Activity
3,286 Commits 105 Branches 48 Tags
cd0fa59647dfa07823ddfc8531c5fa060da3d927
Commit Graph
1563 Commits
Author SHA1 Message Date
Stack-1 cd0fa59647 [UPDATE] Extended the 4-accumulator SpMV unrolling to s/c/z 
Generated from template-psblas, where the unrolled csmv_inner kernel was
added to the type-generic X_csr_impl.P90. The d version produced by the
template is byte-identical to the hand-written one in f7db723f, so this
commit only carries s/c/z.

Same dispatch as d: mean row length >= 16 nonzeros/row uses four
accumulators to break the FMA latency chain, shorter rows keep the
original loop. Applies to the beta==0/alpha==1 and beta==1/alpha==1
branches of psb_?_csr_csmv_inner.

Checked with mpif90 -O3 -fopenmp on all three types; the unrolled and
reference loops agree to rounding on rows of every length mod 4,
including empty rows.
 2026-08-07 12:40:46 +02:00
Stack-1 f7db723fdb [ADD] Added loop unrolling using 4 accumulators in csr SpMV 2026-07-12 17:07:54 +02:00
sfilippone 8254b66140 Overzealous make clean 2026-06-11 15:20:29 +02:00
sfilippone 1c51cab1b5 Make desc_is_asb always a method 2026-06-10 13:52:23 +02:00
sfilippone 847e638123 Fix detail in base/modules/makefile 2026-06-10 13:51:18 +02:00
sfilippone 09308b5c7d various changes, including test matrix generators 2026-05-05 15:49:13 +02:00
sfilippone 19ce4a0942 Fix multivect assembly and comments in base_vect 2026-04-29 14:06:35 +02:00
sfilippone 1cb7634d1f Fix GEMIN 2026-04-29 14:06:23 +02:00
sfilippone 3f83bc2918 Improve psb_exit handling of MPI communicators 2026-04-16 14:58:05 +02:00
sfilippone 19ffaaa915 Update dependencies for i2_vect_mod 2026-04-15 09:06:56 +02:00
sfilippone 97299b8b3f Update VERSION file 2026-04-10 13:37:10 +02:00
Salvatore Filippone eae1311c99 Improved dependencies. 2026-03-26 15:47:02 +01:00
Salvatore Filippone 0a97ea2924 Fix makefile 2026-03-26 15:22:23 +01:00
Salvatore Filippone bc9adfeb3e Fix CMakeLists 2026-03-24 16:44:28 +01:00
Salvatore Filippone e717d2d228 Fix makefile dependencies 2026-03-24 16:11:02 +01:00
sfilippone a223c7eaf3 Rebuild I2 supprot 2026-03-24 15:42:12 +01:00
sfilippone 80c02af47a Fix CMakeLists i2 2026-03-24 11:31:13 +01:00
sfilippone 769c2f65bf Prepare i2 2026-03-24 09:51:10 +01:00
Salvatore Filippone eeeb450f0b Simplified build fix 2026-03-24 09:22:08 +01:00
sfilippone e1ab2bddf6 Fix import and build dependencies 2026-03-24 09:05:51 +01:00
Stack-1 2d388ce45e [FIX] Fixed compilation dependencies resulting in race condition when compiling using many threads. The problem was encountered usingg make -j30, now it compiles. 2026-03-24 07:51:09 +01:00
sfilippone c990dc6840 Fix info in %reinit 2026-03-20 13:55:33 +01:00
sfilippone ca458870b1 Fix intent in precbld to align with AMG 2026-03-20 13:55:17 +01:00
fdurastante 1ee1b83341 Added routine to extract single element from sparse matrix 2026-03-19 14:58:00 +01:00
sfilippone d2fba9e952 Fix licensing text 2026-03-18 14:49:55 +01:00
Marco Feder 303edfedc7 Fix return value 2026-03-18 10:48:38 +01:00
sfilippone ab2c5edcd5 Fix module subroutine declaration in swapdata. 2026-03-16 12:31:29 +01:00
sfilippone 884f7cb0b9 Merge branch 'development' of github.com:sfilippone/psblas3 into development 2026-03-13 14:09:24 +01:00
sfilippone 356bdba5c7 Fix intent on base_vect%reinit 2026-03-13 14:07:57 +01:00
sfilippone 7151c728db Final fixes 2026-03-13 12:17:50 +01:00
sfilippone 724345ab27 Rework use of icomm in halo and friends, add submodules 2026-03-13 11:27:32 +01:00
sfilippone bdee97e991 Fix use of MPI communicator 2026-03-12 17:30:21 +01:00
sfilippone e8c5c0b8b0 Add clear argument to reinit method in vectors. 2026-03-12 16:29:18 +01:00
sfilippone e2ddc08aea Merge branch 'development' of github.com:sfilippone/psblas3 into development 2026-03-10 10:15:11 +01:00
sfilippone 9c3084985a Cosmetic change in desc_const_mod 2026-03-10 10:14:08 +01:00
Salvatore Filippone 3d92668973 Change name of desc%get_list into desc%get_list_p 2026-02-12 12:54:50 +01:00
sfilippone 96ab382ce8 Fix vect_clone and define check_addr 2026-01-27 12:17:39 +01:00
sfilippone 897d466013 Define DESC%check_addr for debugging 2026-01-26 11:26:30 +01:00
sfilippone 3965c3a526 Modified psb_config and machinery for better C interfacing 2026-01-22 13:44:28 +01:00
sfilippone fe7a4174e0 Fix build for CBIND 2026-01-16 10:29:16 +01:00
sfilippone d7b22f0538 Fix string_c2f 2026-01-15 11:35:05 +01:00
sfilippone ac5512974b Fix vector allocation etc. 2026-01-13 16:57:06 +01:00
sfilippone 3337a12e59 Fix multivect handling 2026-01-13 15:26:11 +01:00
sfilippone 82d7f0fe7e Bump patchlevel 2026-01-09 13:26:43 +01:00
sfilippone 9552bda6de Merge branch 'fix-build' into development 2026-01-09 12:05:25 +01:00
sfilippone 01610305b2 Rename base/internals into base/auxil. Fix aplusat in CMakeLists 2025-12-12 22:44:03 +01:00
sfilippone fd27654d93 Fix DUPL_DEF value in psb_config.h 2025-12-04 14:22:25 +01:00
sfilippone c1ca1665e7 Fixes to vector status and build processing for dealii 2025-12-02 17:34:10 +01:00
sfilippone 5333117010 Fix vect_clone, expose to C interface 2025-11-27 15:13:29 +01:00
sfilippone 43fcc53344 Final (hopefully) version of Makefile dependencies 2025-11-24 17:00:39 +01:00