fdurastante
/
psblas3
mirror of https://github.com/sfilippone/psblas3.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • e18de650f2 Take out debug print sfilippone 2024-04-17 13:53:04 +0200
  • ccb4f73dca Choose version of HLG-multivect kernel Salvatore Filippone 2024-04-17 05:24:20 -0400
  • 19aa3f56b6 Merge branch 'repackage' into cuda-multivect sfilippone 2024-04-15 12:42:03 +0200
  • 148a5e5e14 Check for shmemsize sfilippone 2024-04-15 12:38:22 +0200
  • eed4c574bc Take out obsolete ilu_fct files sfilippone 2024-04-15 12:19:16 +0200
  • 6f92a5c37a Merge branch 'repackage' of github.com:sfilippone/psblas3 into repackage sfilippone 2024-04-15 12:18:09 +0200
  • 553531eefb Take out obsolete ilu_fct source files sfilippone 2024-04-15 12:16:43 +0200
  • 2f575894fc Fix --with-cudacc in configure sfilippone 2024-04-15 11:18:51 +0200
  • 378c126055 Use MMBSZ=8, prepare to check shmem size Salvatore Filippone 2024-04-13 07:10:14 -0400
  • 897cfb4028 Multicolumn HLG product Salvatore Filippone 2024-04-12 12:05:53 -0400
  • cf315660e1 Updated tests gabrielequatrana 2024-04-12 10:49:54 +0200
  • 5389ff245e Merge branch 'psblas-bgmres' of https://github.com/sfilippone/psblas3 into psblas-bgmres gabrielequatrana 2024-04-11 22:35:04 +0200
  • 77b1e03b39 Added test gabrielequatrana 2024-04-11 22:33:44 +0200
  • 5b95f1920c Regenerate configure, fix typo in hlg_vect_mv Salvatore Filippone 2024-04-10 10:22:27 -0400
  • 08984619dc Fixed SpMM for HLG gabrielequatrana 2024-04-10 15:34:57 +0200
  • beb418e00b Fixed SpMM for ELG (AXPBY GPU not working) gabrielequatrana 2024-04-10 14:57:15 +0200
  • c807d88c57 SpMM using Cusparse dedicated routine (CSRG) gabrielequatrana 2024-04-10 11:59:52 +0200
  • 399818a482 Never do arithmetic on a (void *) Salvatore Filippone 2024-04-10 05:08:12 -0400
  • 1ecf36c9c3 Merge branch 'psblas-bgmres' of github.com:sfilippone/psblas3 into psblas-bgmres Salvatore Filippone 2024-04-10 05:01:51 -0400
  • 7d74ebf5c4 Make multivectors work Salvatore Filippone 2024-04-10 04:46:18 -0400
  • 82715fec9b Fixed SpMM for CSRG gabrielequatrana 2024-04-10 00:04:32 +0200
  • 409b51e609 Try to fix SpMM gabrielequatrana 2024-04-09 23:21:17 +0200
  • ee140bc8dd Read/Write multivect fixed (SpMM bug) gabrielequatrana 2024-04-09 10:39:50 +0200
  • a624b7098b Cuda multivect methods implementation gabrielequatrana 2024-04-06 21:29:49 +0200
  • c1e4f9c2b1 Merge branch 'repackage' into psblas-bgmres, fixes to resolve merge sfilippone 2024-04-04 14:57:02 +0200
  • 0760e4d553 Fix C function declarations for compilation with LLVM/clang in CUDA repack-llvm sfilippone 2024-04-04 10:53:20 +0200
  • 4347c663c2 Change conftest **argv to recognize CUDA_VERSION. sfilippone 2024-04-04 10:52:41 +0200
  • a2f92e616f Put VOLATILE under ifdef for FLANG sfilippone 2024-04-01 15:58:32 +0200
  • 59e6df73a4 Make sure configure recognizes FLANG sfilippone 2024-04-01 15:57:44 +0200
  • ef179f12af Modified resolution of method names for NV to compile repack-nvid sfilippone 2024-03-29 10:10:00 +0100
  • 02fb43ba82 Fixed convergence gabrielequatrana 2024-03-28 13:36:02 +0100
  • 6b785d66c6 Check convergence now working gabrielequatrana 2024-03-25 20:10:44 +0100
  • 0839165bdc Added convergence check gabrielequatrana 2024-03-25 19:56:50 +0100
  • 0023b8ac78 Compile adjcncy_fnd_owner sfilippone 2024-03-25 15:41:47 +0100
  • 3a25d7b04a Fixes for LLVM compilation sfilippone 2024-03-22 17:10:30 +0100
  • 373d841bce Don't need renaming of psi_gth and psi_sct sfilippone 2024-03-22 15:44:51 +0100
  • 472f16f0df Fix compilation with --enable-serial sfilippone 2024-03-19 14:00:03 +0100
  • 1b79939255 Fixed some bugs (QR_fact serial) gabrielequatrana 2024-03-17 22:22:51 +0100
  • 676652fcff Working parallel (QR_fact serial) gabrielequatrana 2024-03-16 23:25:20 +0100
  • 9ec01020db Take out spurious fprintf repack-track Salvatore Filippone 2024-03-15 12:01:12 -0400
  • 84981b9135 Fix track cuda memory for matrix types Salvatore Filippone 2024-03-15 07:16:15 -0400
  • d94b3dae58 Fix tracking of vector allocations. Salvatore Filippone 2024-03-15 06:43:16 -0400
  • ef82b975e3 Track CUDA allocation sfilippone 2024-03-15 08:26:49 +0100
  • d10631530f Init Parallelize gabrielequatrana 2024-03-14 23:34:35 +0100
  • e0a4d362fa Define flag TRACK_CUDA_MALLOC sfilippone 2024-03-14 12:08:57 +0100
  • 6987582c30 Done SERIAL gabrielequatrana 2024-03-09 19:22:40 +0100
  • b5f1442ac8 Merge branch 'nond-rep' into repackage Salvatore Filippone 2024-03-05 19:46:14 +0100
  • 48455190ec Add GPU version of XYZW nond-rep sfilippone 2024-03-05 13:57:03 +0100
  • a11f328e62 Added CUDA version of XYZW sfilippone 2024-03-05 12:42:21 +0100
  • 86be8ebcd0 New method W%XYZW() sfilippone 2024-03-04 16:29:47 +0100
  • b5d5f97661 Improve cuda%zero() sfilippone 2024-02-20 14:04:23 +0100
  • 0e269ed641 typo in Cabgdxyz sfilippone 2024-02-20 13:57:32 +0100
  • d95077ffd6 Fix typo in vectordev_mod Salvatore Filippone 2024-02-20 13:46:58 +0100
  • 2d3773df98 CUDA kernels for ABGDXYZ Salvatore Filippone 2024-02-20 13:05:16 +0100
  • 2a75d677d0 ABGDXYZ in vectordev_mod Salvatore Filippone 2024-02-20 13:04:40 +0100
  • 2391f64df6 X_cuda_vect%abgdxyz sfilippone 2024-02-20 12:30:07 +0100
  • 93c71c4316 Fix %ZERO() on cuda sfilippone 2024-02-20 10:25:31 +0100
  • 0568a83734 Fix ifdef and old code sfilippone 2024-02-19 10:53:23 +0100
  • 35d68aa4e3 Reuse calls to getDeviceProperties done at init time Salvatore Filippone 2024-02-18 16:44:37 +0100
  • 1ba8dfc7b7 Switch FOR and IF in AXPBY Salvatore Filippone 2024-02-18 10:31:32 +0100
  • f9677bc892 Enabled new CUDA version of ABGDXYZ Salvatore Filippone 2024-02-17 18:42:56 +0100
  • 4681767ef8 New implementation for ABGDXYZ in CUDA Salvatore Filippone 2024-02-17 18:20:12 +0100
  • 105aa3c570 Intermediate impl of ABGDXYZ Salvatore Filippone 2024-02-17 17:46:09 +0100
  • 864872ecac Intermediate implementation of abgdxyz on cuda Salvatore Filippone 2024-02-17 17:28:32 +0100
  • a41b209144 Better AXPBY implementation in CUDA. Salvatore Filippone 2024-02-17 17:18:59 +0100
  • f4c7604f61 Fix base implementation of abgdxyz to call set_host Salvatore Filippone 2024-02-17 09:40:09 +0100
  • b8f9badf95 Fix interface between vect and base_vect%ABGD Salvatore Filippone 2024-02-14 20:05:52 +0100
  • 2a40b82b58 Fix typo in base_vect_mod Salvatore Filippone 2024-02-14 16:01:16 +0100
  • 4e611bb078 Enable psi_abgdxyz Salvatore Filippone 2024-02-14 15:55:55 +0100
  • 9ced67634d Fix KIND for NR in axpby Salvatore Filippone 2024-02-14 08:52:38 +0100
  • 3121c43582 Silly bug in abgdxyz implementation Salvatore Filippone 2024-02-13 16:16:13 +0100
  • 5c3d5f0235 Silly bug in abgdxyz implementation Salvatore Filippone 2024-02-13 16:13:06 +0100
  • 29669b56a2 Implementation of psb_abgdxyz Salvatore Filippone 2024-02-13 16:07:06 +0100
  • a942b47f7c Merge branch 'nond-rep' of github.com:sfilippone/psblas3 into nond-rep Salvatore Filippone 2024-02-13 15:49:22 +0100
  • 6c53b6ec79 Fix typo in interface for psb_abgdxyz Salvatore Filippone 2024-02-13 15:48:43 +0100
  • 83ededd02b Implementatino of abgd_xyz sfilippone 2024-02-13 12:54:37 +0100
  • 92a95699ba Merge branch 'nond-rep' of github.com:sfilippone/psblas3 into nond-rep Salvatore Filippone 2024-02-12 16:30:28 +0100
  • ebc7c6b3b4 Fix call to base%abgdxyz Salvatore Filippone 2024-02-12 16:29:48 +0100
  • 45f00e6e19 Fixed comments sfilippone 2024-02-12 15:10:58 +0100
  • 14c4ff0f32 Added new methd for two combined axpbys Salvatore Filippone 2024-02-12 14:15:19 +0100
  • b49ce6b610 Merge branch 'repackage' into nond-rep Salvatore Filippone 2024-02-09 18:53:50 +0100
  • 6433dc797e Fix CUDA implementation of %set_scal and %zero sfilippone 2024-02-02 15:27:42 +0100
  • 097d63147a Fix cuda dir makefile sfilippone 2024-01-24 17:58:39 +0100
  • 161dd21d20 Fixed bug psblas-anderson gabrielequatrana 2024-01-17 12:24:36 +0100
  • e1e7aa58d2 Fixed bug gabrielequatrana 2024-01-16 15:36:18 +0100
  • 3aa3c795e9 Refactor assembly and cnv non-diag sfilippone 2023-12-23 13:15:01 +0100
  • 4d051c777d Fix makefile and test program sfilippone 2023-12-22 12:01:59 +0100
  • 49e99a3e82 Fix conversion and product to enable overlap with GPU sfilippone 2023-12-22 12:01:41 +0100
  • 74cf138a6c Merge branch 'repackage' into non-diag sfilippone 2023-12-22 10:31:48 +0100
  • be7571f568 Fix missing directive sfilippone 2023-12-20 13:45:43 +0100
  • e9d1238b43 Add detailed measurements. sfilippone 2023-12-20 13:30:09 +0100
  • 20a01d4d71 Attempt at fixing CSRG in CUDA 10.2. Not complete yet. Salvatore Filippone 2023-12-19 11:54:42 -0500
  • 1bc2a884e2 Adjust conditional compilation on CUDA version sfilippone 2023-12-15 10:52:48 +0100
  • 62db7c0449 Fix spsv with CSRG handling of descriptors. Salvatore Filippone 2023-12-15 04:11:01 -0500
  • d28ea462d9 Modified CSRG to work with latest versions; cusparse docs are unclear Salvatore Filippone 2023-12-15 03:41:19 -0500
  • 6b65199afb Check CUDA version for -dopt=on only from 11.7 sfilippone 2023-12-11 13:21:23 +0100
  • 0230fbb7af Identufied problems with CSRG. Will fix in a branch sfilippone 2023-12-07 12:09:00 +0100
  • 41491f7b9c Fix HAVE_CUDA in test programs sfilippone 2023-12-01 15:14:50 +0100
  • b2b7b074df Fix usage of HAVE_CUDA/HAVE_GPU (mostly disappeared) sfilippone 2023-12-01 15:11:17 +0100
  • e373ed7e0b Modify configry to only use HAVE_CUDA, since SPGU is recompiled. sfilippone 2023-12-01 15:10:39 +0100