Commit Graph

  • ab8631439f Update configure script sfilippone 2023-11-29 16:35:42 +0100
  • 6c9ca58282 Silly bug in coo insert sfilippone 2023-11-29 16:14:30 +0100
  • 8633e76cb0 Silly bug in coo insert sfilippone 2023-11-29 16:11:11 +0100
  • d3b2b7816d Fix coo insert OpenMP. Fix Make.inc.in sfilippone 2023-11-29 15:35:55 +0100
  • 492b28f342 Fix wrong insert without OpenMP sfilippone 2023-11-29 15:33:32 +0100
  • 655c86caed Updated docs. tcud rename-cuda sfilippone 2023-11-29 10:20:38 +0100
  • 9b713c177b Fix cuda interfaces for renaming sfilippone 2023-11-28 19:46:20 +0100
  • 6fa0bf7fe7 Complete cuda renaming sfilippone 2023-11-28 19:14:28 +0100
  • cce3103bb4 Fix CXXDEFINES sfilippone 2023-11-25 20:06:26 +0100
  • a082cdb1b6 Deactivate OpenMP in hash_g2lv_ins for the time being. v3.8.1-2 maint-3.8.1 sfilippone 2023-11-10 13:38:19 +0100
  • b850c0ef6a Fix typo sfilippone 2023-11-09 08:41:43 +0100
  • f7d2c856ec Fix gather with SERIAL v3.8.1-rc1-3 v3.8.1-rc1-2 sfilippone 2023-11-07 21:44:22 +0100
  • ae7fad95d4 Merge branch 'development' into non-diag sfilippone 2023-11-07 18:02:55 +0100
  • a6ec655a97 Prepare merge sfilippone 2023-11-07 17:57:32 +0100
  • a2788bdf0b New version with ND product sfilippone 2023-11-07 13:39:44 +0100
  • d718ef1e6d Always allocate szs in psb_gather sfilippone 2023-11-07 10:56:23 +0100
  • baf18cebd7 Further fix for gather. v3.8.1-rc1-1 sfilippone 2023-11-03 14:39:01 +0100
  • 5caee551e5 Fixed IN_PLACE option for collectives. sfilippone 2023-11-03 14:28:04 +0100
  • d82b090289 Fix makefile for psi_acx & friends sfilippone 2023-10-25 12:47:31 +0200
  • 25e9183e50 Fix SHFT implementation, step 2 Salvatore Filippone 2023-10-23 15:31:13 +0200
  • 250a6300ba Fix SHFT implementation Salvatore Filippone 2023-10-23 15:14:48 +0200
  • 0b184e4313 Merge branch 'shift' into development Salvatore Filippone 2023-10-23 14:21:13 +0200
  • d3fcd566d9 Define a SHIFT argument to compute ILU( A+shft I) Salvatore Filippone 2023-10-23 14:16:30 +0200
  • 38203f4d10 Linear combination of multivector columns anderson Cirdans-Home 2023-10-19 18:21:49 +0200
  • 6aa7987d52 Rename GPU into cuda, and merge SPGPU code. sfilippone 2023-10-19 14:22:32 +0200
  • 614dc53dbd Implementex axpby for single column of multivect Cirdans-Home 2023-10-19 13:42:06 +0200
  • 301528e9d9 Merged with development to fix Cirdans-Home 2023-10-19 13:41:13 +0200
  • 2732336915 Fix gpu/makefile repack-precuda sfilippone 2023-10-16 16:15:00 +0200
  • 81e9121c91 Add GPULDLIBS into Make.inc (and fix configry) sfilippone 2023-10-02 13:08:56 +0200
  • d0d4e45877 Fix for I gpu types from template: use psb_sizeof_ip sfilippone 2023-09-29 13:59:40 +0200
  • 3a9116bc42 Merge psblas-ext into psblas step 2: GPU formats and configry. sfilippone 2023-09-29 13:54:58 +0200
  • 1d5faa388d Merge psblas-ext into psblas, step 1: ext storage formats. sfilippone 2023-09-27 16:23:17 +0200
  • d1bf46b0b1 Merge branch 'dev-openmp' into development v3.8.1 master sfilippone 2023-09-07 09:50:59 +0200
  • e31dd52c41 Fixed CRITICAL in hash_mod sfilippone 2023-09-07 09:48:18 +0200
  • def0635c53 More OMP directives in cd_inloc sfilippone 2023-09-07 09:25:37 +0200
  • efa29bc3ad Fix configure for GCC when using --with-fcopt omp-walther sfilippone 2023-09-05 17:00:49 +0200
  • 3e29e603d2 Fix necessary flags for GCC sfilippone 2023-09-05 16:59:03 +0200
  • 41be1357c3 Set defaults for SPSPMM depending on OpenMP compilation. sfilippone 2023-08-30 13:45:51 +0200
  • 0d8a5d3dc2 New SPSPMM implementation Salvatore Filippone 2023-08-29 19:27:29 +0200
  • d0cacda995 Moved various modules related to RB around, into auxil, update Makefile. Salvatore Filippone 2023-08-29 16:19:56 +0200
  • 7b45994b70 Setter/getter for SPSPMM algorithm in base_mat_mod Salvatore Filippone 2023-08-29 05:07:55 -0400
  • 2322a9ce61 using end_idx to copy data from threads in gustavson and gustavson_1d wlthr 2023-08-28 11:06:30 +0200
  • 0185b79b2a added setter for d_csr_spspmm implementation wlthr 2023-08-28 11:04:00 +0200
  • 0fe95c3c76 added use statement wlthr 2023-08-28 10:37:41 +0200
  • 979a3da95f merged dev-openmp into omp-walther wlthr 2023-08-26 10:55:09 +0200
  • 1af76c067c added parallel double precision spspmm implementations wlthr 2023-08-26 10:53:00 +0200
  • 94c9060579 Merge branch 'dev-openmp' into development sfilippone 2023-08-24 09:16:43 +0200
  • f4f37924e2 Update contributor list sfilippone 2023-08-24 09:15:25 +0200
  • 815ecd5cc1 Merge branch 'dev-openmp' into development sfilippone 2023-08-22 10:27:28 +0200
  • f001ebbad3 Final fix for COO on OMP sfilippone 2023-08-22 10:22:46 +0200
  • 26bf4c5d69 Fixed COO csput for OMP/not OMP sfilippone 2023-08-21 17:26:37 +0200
  • 3aa748b0e3 Finish dual OMP/notOMP g2lv1_ins sfilippone 2023-08-21 17:22:16 +0200
  • 08c1ab0cd1 Fix tril/triu in COO for non-OMP paths. sfilippone 2023-08-21 16:32:55 +0200
  • ca82520b88 Reworked CSR TRIL/TRIU for OpenMP sfilippone 2023-08-01 10:16:22 +0200
  • 5e691d5bff Some improvements for openmp vector updates sfilippone 2023-07-24 13:21:20 +0200
  • bb9f213551 Define and implement OMP version of TRIL/TRIU sfilippone 2023-07-18 18:19:03 +0200
  • 4929991249 changed all unsigned longs to int, modified makefile for debug purposes, reallocate c matrix arrays in sp3mm_impl.f90 sp3mm-interface wlthr 2023-07-18 10:44:04 +0200
  • 5e252b77bb fixed issues with passing of arrays from fortran to c wlthr 2023-07-13 16:35:23 +0200
  • c252f16fd0 merged development into sp3mm-interface wlthr 2023-07-07 10:47:16 +0200
  • 7ea80e04e7 fixed some minor errors wlthr 2023-07-06 19:01:17 +0200
  • 981e40f689 compilation errors fix wlthr 2023-07-06 11:40:36 +0200
  • 9b6542d350 added using new spmm implementation wlthr 2023-07-05 16:39:35 +0200
  • 867ca0c1d0 added makefile to compile sp3mm source code wlthr 2023-07-05 16:22:55 +0200
  • bfd88ca2c1 finished interface for spmm with uperbound method wlthr 2023-07-05 14:20:31 +0200
  • 250163f1bc created fortran interface for C spmm code wlthr 2023-07-03 10:03:27 +0200
  • 0c88352530 added sp3mm4amg source code to the repository and created C wrapper routine to be used in the fortran interface wlthr 2023-06-29 19:07:23 +0200
  • 2f403e0df7 Rework cp_{from|to}_fmt for better OpenMP performance sfilippone 2023-06-28 11:16:53 +0200
  • 1a4ae1e973 created sp3mm module and implementation files as well as an interface, still needs C files and C binds wlther 2023-06-27 16:28:58 +0200
  • c093a6e11d added cbinds for C structs wlther 2023-06-27 11:05:40 +0200
  • 00df32c581 Added getelem_vec implementation Cirdans-Home 2023-06-22 15:20:33 +0200
  • 402354dbd8 Fix syntax issue Salvatore Filippone 2023-06-22 11:52:08 +0200
  • d378266f33 Fix synatx error sfilippone 2023-06-21 13:37:23 +0200
  • fd0aea4afa Merge branch 'dev-openmp' of github.com:sfilippone/psblas3 into dev-openmp sfilippone 2023-06-21 12:22:08 +0200
  • a66778f270 Improve coo and merge development sfilippone 2023-06-21 12:21:27 +0200
  • e72edf70a1 Merge branch 'development' into dev-openmp sfilippone 2023-06-21 12:16:58 +0200
  • 677789469c Reworked ensure_size sfilippone 2023-06-21 12:14:47 +0200
  • 347352fe1e Make spins work in OpenMP from either par or serial sfilippone 2023-06-02 11:07:23 +0200
  • db0e4db507 Minimize debug sttements in hash_ins sfilippone 2023-06-02 11:06:53 +0200
  • 1941affe7a Exposed error in AMG test when not parallelizing generation loop sfilippone 2023-05-30 12:26:58 +0200
  • 494e29dd2e Cosmetic adjustments to COO and BSRCH omp-threadsafe sfilippone 2023-05-23 15:56:48 +0200
  • 739dc78a75 Merge branch 'development' into omp-threadsafe sfilippone 2023-04-27 13:07:42 +0200
  • 7e5dc20e03 Define new options for BSRCH, clean interface sfilippone 2023-04-27 13:05:46 +0200
  • 40cc78854a Improve implementation of fix_coo using exscan sfilippone 2023-04-21 13:41:16 +0200
  • 91d3e66547 Merge branch 'omp-threadsafe' of github.com:sfilippone/psblas3 into omp-threadsafe sfilippone 2023-04-21 12:02:38 +0200
  • 74a8217520 Fixed silly bug in EXSCAN and usage in CSR_IMPL sfilippone 2023-04-21 12:02:18 +0200
  • 5bc02fb2e6 Take out redundant statements in SPINS Salvatore Filippone 2023-04-21 08:55:20 +0200
  • f3efea0a89 Take out IBASE from exscan, makes no sense. sfilippone 2023-04-20 12:31:32 +0200
  • 08ff37332a Enable OpenMP in test/pargen sfilippone 2023-04-19 14:06:40 +0200
  • 05b684ddbb Updated use of exscan in CSC sfilippone 2023-04-19 13:23:48 +0200
  • 9c248a31e2 Refactored EXSCAN and its OpenMP usage. sfilippone 2023-04-19 12:12:58 +0200
  • 02dd204351 Implement psi_exscan and use in _from_coo sfilippone 2023-04-18 17:00:22 +0200
  • dbd55321f8 Fixed CSR mv and cp _from_coo with OpenMP. sfilippone 2023-04-17 21:00:20 +0200
  • 6ba7d93159 Fix CRITICAL in LIST%G2L_INS sfilippone 2023-04-16 12:03:16 +0200
  • 5a5712b4f0 Rely on CRITICAL inside G2L_INS implementation sfilippone 2023-04-16 12:02:54 +0200
  • f068d73ef1 First working version sfilippone 2023-04-16 11:30:09 +0200
  • 8459ea28f5 Modified matrix build procedures with OpenMP sfilippone 2023-04-14 18:25:43 +0200
  • eb11e5e053 Put CRITICAL(name) in G2L_INS sfilippone 2023-04-14 17:48:15 +0200
  • 0f1603a2e9 The current version of test/omp seems to be working. To be completed sfilippone 2023-04-13 16:40:29 +0200
  • 98945f36b5 Fix nrm2 with overlap sfilippone 2023-03-31 13:37:20 +0200
  • e6a70410e0 Make sure to use install -p sfilippone 2023-03-30 15:50:21 +0200