Improve rumtime statistics #298
Assignees
Comments
RonRahaman
added a commit
to RonRahaman/nekRS
that referenced
this issue
May 26, 2021
977e71b9d CHANGELOG adjustments 0ca69b7ed Update CHANGELOG 1b1ca7baf Fixed typo in ref manual 540a72b9f Sorted user manual references 0a82c6d41 One more update to CHANGELOG d380ba124 amg parameter numbers c05d97a62 more on GPU AMG 9d1a9d410 boomerAMG on GPUs 57d1288bd bug fix dfceea17d Merge branch 'master' of github.com:hypre-space/hypre 0bb67902b GPU build 3b65190b2 A few updates to the CHANGELOG 0fb7956ea Update version number for new release 426b83b64 Updated CHANGELOG for 2.21.0 release 3bc7d267e Gpu default (Nek5000#336) 8a41a42c8 Cmake cuda updates (Nek5000#349) 995969c9a fix reordering option for diagonal of square matrix transpose (Nek5000#370) ae362727b Change HYPRE_RELEASE_NUMBER to be int in CMake Config.h file (Nek5000#372) 3894857b1 Adding a new HYPRE_RELEASE_NUMBER macro (Nek5000#366) 391f4fd6e Fix hypre_CSRMatrixAddHost (Nek5000#369) 08b8cdbf7 bug fix 74482aef4 This branch fixes some of the profiling code in nvtx. All Push/Pop ranges never explicitly use PushColor version around device kernels. We allow the general method to determine the correct color. Next, there was a missing PopRange() in par_mod_lr_interp.c. Finally, I changed the profiling names in par_mod_lr_interp.c as they conflicted with names in another file. (Nek5000#368) c7175a987 Cuda11.2 (Nek5000#362) 1c979f338 Fix so this compiles with hip (Nek5000#365) c38527c45 Add OMP support to Mat/Mat add functions (Nek5000#341) 91e82e699 hybridSetAggInterpType (Nek5000#358) fbaf2b75d Fix a corner-case divide-by-zero in PFMG (see github issue Nek5000#332) (Nek5000#355) c272f6980 Fix MGR setup on GPU and MGR bugfixes. (#260) 3f12d4765 Add support for matrices with many zero rows (Nek5000#300) 521ef12ff Fix colmap (Nek5000#328) c5110654c add include in fei makefile (Nek5000#334) 4718207e9 hypre_strcpy (Nek5000#340) 9a28cc622 Config without args (Nek5000#338) 25646da90 Mat descr (Nek5000#331) 34b2f43ba Merge branch 'master' of github.com:hypre-space/hypre 0d730f374 Fixed seg-fault problem 3f6c0abc4 Merge branch 'master' of https://github.com/hypre-space/hypre e79d05d0b removed some missed col_offd_S_to_A d2ef5cf18 Fixed ExtInterpDevice 0a0ddd072 Fix s (Nek5000#312) 25d0498a3 Roctx (Nek5000#319) b3a4a76a5 Roc sparse (Nek5000#316) 366b80f89 2 stage gs (Nek5000#314) 8002200aa Revert "2 stage gs (Nek5000#308)" (Nek5000#313) f7fb2853f 2 stage gs (Nek5000#308) fcd2b2b4b Add missing utilities source file to CMakeList ce2a43821 removing device bindings in HYPRE_Init (Nek5000#296) 8223465ea More hip support (Nek5000#302) 67a0bdeab Hip support (Nek5000#298) 4630047ad HYPRE_USING_GPU (Nek5000#295) 9d45b1934 Remove unnecessary including of superlu_ddefs in ij test 2e97ffcc9 bug fix recvbufsize (Nek5000#294) c9848a5ed Fixed a small mistake in configure/configure.in b5070c14e configure fix for last merged PR 8c00f18a9 Hip build (Nek5000#293) 7859aeee3 Changed the rhs in several TEST_sstruct/periodic runs (Nek5000#291) 512219634 Adding filter to check-license test to ignore runtests-* files 68f510c11 Test jobs for enable-mixedint (Nek5000#280) 277b35785 Removing 'memory.h' include since it is not ISO standard (Nek5000#279) c4ca3e105 fixed wrong pointer checking; error handler in debug mode (Nek5000#275) 33da39f68 Fixed a typo in the sstruct.in.README file comment 51e5a4c6d SStruct interface without UVM (Nek5000#170) 3438132e1 GPU examples (Nek5000#268) 6eb66f869 Fixed a small mistake in the sludist.sh test be18e595a Remove the global partition code from hypre (Nek5000#273) 6f9260b67 Add saved-file extension to runtest (Nek5000#271) 414fa671b Umpire (Nek5000#243) 8462f60dc Hypre warp bitshift (Nek5000#267) 2186a8fb3 triangular solve on GPUs; runcheck (Nek5000#256) bd76daf12 Updated saved files to reflect change in NSH solve on ILU Schur system -- See PR#251. (Nek5000#254) 6a1caf899 Modification to fix error code warning for coarse level solver (#251) a6c852be5 fixed syntax error with --enable-gpu-aware-mpi (Nek5000#250) 950f9f250 fix GPU SpMV for zero matrices (#246) b49727f16 Cuda triangular smoothers (Nek5000#240) 763ea8a5a cmake: Optionally accept path to BLAS/LAPACK libraries (Nek5000#230) 804609b6c Reorg relax (Nek5000#237) 9c24f006a ILU updates (#239) eae9be29b bug fix cycle_param (Nek5000#236) 56012897e Ilu dev 2019 (Nek5000#160) 2bc4228ec Changed sludist.saved file to correct new superlu-dist autotest errors 796ab0af4 Use basename when checking compiler in configure (Nek5000#225) 22f4d3f8c Cuda 11 API (Nek5000#163) 641f7a4e3 Minor spelling fix (Nek5000#222) 2e1ccee24 Euclid fix (Nek5000#218) 5ac2b3a54 Improve portability for update-release script (for LINUX, UNIX and macOS). (Nek5000#227) 9fb1b351c MS-Windows OMP pragma (Nek5000#223) 636706acd Fixing compile issues --with-caliper (Nek5000#216) dd4ddba0f Added a filter to runtest for 'lrun warning' ff45ecef3 Set default convergence tolerance to 1.0e-6 (#206) 7b2379c0d optimization in hypre_CSRMatrixBigJtoJ and JtoBigJ (Nek5000#204) 1ddd69f27 Fixed problems when calling HYPRE_Finalize() multiple times (#207) 5988a506b Update CHANGELOG 54190a846 Update CHANGELOG 4a5c5aca4 Update CHANGELOG d257887cd Another CHANGELOG update for 2.20.0 00b826e84 Update version number and date for release 2.20.0 2fe718e11 Update CHANGELOG for release 2.20.0 aaf5aa564 Aggressive coarsening and 2- stage MM-ext Interpolations on GPUs (Nek5000#195) 0fcb67054 Fix AMGDD (Nek5000#190) 37f7a0a3f Epe gpu (Nek5000#187) 5f3141a64 Change issue reporting to use github's issue tracker (Nek5000#189) d5e4eb4bd Fixed a few minor autotest errors 36d0bfba4 Fixed a compile error. 0b80656ce AMG-DD implementation (Nek5000#145) 7f9d222ed run headers 2b2ea3920 should run `headers' to make sure _hypre_parcsr_mv.h is not directly changed 3ae6c7fec Merge pull request Nek5000#172 from hypre-space/PETScFix 1c0598626 Merge branch 'master' of https://github.com/hypre-space/hypre into PETScFix 8833bed15 add compile flags in GPU regression test scripts ffe35407a format change f6f98cb36 bug fix (hopefully...) 66b653a5e Merge pull request Nek5000#184 from hypre-space/fix-issue183 26e2f4f4d Add implementation of ffs for MSVC f4e128cab Merge pull request Nek5000#178 from hypre-space/fix-hopscotch 9a73fbbae Merge branch 'master' into fix-hopscotch 88c7a2d73 Add regression test for concurrent hopscotch 0b2f113c5 fixed leak 5da38fea6 Merge branch 'master' of https://github.com/hypre-space/hypre b599ea759 kokkos + gcc warnings 057b1fe0f Merge pull request Nek5000#179 from hypre-space/fix-regtest-error fd9d861c9 Complete renaming of superlu.* to dsuperlu.*. a96a763c4 Renamed superlu.* to dsuperlu.*. f31b8b7aa Merge branch 'master' into fix-regtest-error 7c663ef6c Fixes for check-header errors. 6b42e64d4 Bug fix in header file d2f566e5d Include superlu.h b222e7aea Move DSLUData structure to its own header file. d03a93b3e Add missing include guards 8a7617c6c Init Kokkos in the driver 7c0da1bee Edited to fix regression test errors. e42ab96f9 Move ifdefs to proper places 079402c27 Update private header file fba878903 Remove unnecessary ifdefs 2d9a37f9a Merge pull request Nek5000#176 from hypre-space/mgr-dev e9166cdaf Revert accidental deletion of EOL 9a268728e Truncate the coarse grid at each reduction level using a threshold. e75fb7bb2 Add new functions prototypes to protos.h. 07487b84c Merge branch 'master' into mgr-dev 7e6683786 Fixing another slurm filter issue in 'runtest.sh' 9b5f2730c changed some assertions; offset in matvec 6906e8e0f Merge branch 'master' of https://github.com/hypre-space/hypre into PETScFix efa4c2c85 minor changes, white spaces, comments, etc dfa31b5e0 Merge pull request Nek5000#157 from pbauman/gpu-fixes cea946c51 CXXFLAGS ba717c7ee Update example test. 13e2cad48 Merge pull request Nek5000#165 from hypre-space/fix-mixed-int e98ac7777 fixed a mixed-int in GPU code d8f683d11 Merge pull request Nek5000#141 from hypre-space/gpu_doc 4431a4acd Minor clean up. 079e99515 First implementation of direct solver option for coarse grid solve in MGR. b5d162e80 fixed some mixed-int issues. 6846fc28d Added a comment associated with previous --enable-mixedint commit 3ff4454aa Fixing a problem with --enable-mixedint on a Mac d20cef846 fixed compile flag issue for gpu 2370d7d68 Merge branch 'master' of https://github.com/hypre-space/hypre into PETScFix ffeed7229 fix cusparse matvec for ncol==0 26a1a8a97 result_type needed by clang a5594908f Parentheses for better readability 2e5ccb684 Critical portability fix in warp_prefix_sum 2f51e7501 Use HYPRE_WARP_SIZE for shared memory alloc in blockReduceSum 950a747ae Use HYPRE_WARP_SIZE instead of hardcoded numbers b075d6425 Merge pull request Nek5000#155 from hypre-space/fix-caliper d8a270c3f Merge branch 'master' into fix-caliper 5fca3a5f3 Annotate Setup levels of BoomerAMG 163ef941b Annotate ILU d2c4df7f6 Annotate MGR 33f423213 Merge pull request Nek5000#147 from hypre-space/fix-fortint 1079429f7 Merge branch 'master' into fix-fortint 175dd5912 description of GPU options of AMG 42448bf06 Adding filter for 'HYPRE_config.h.in' in AUTOTEST/check-license.sh f0b7c705e Merge pull request Nek5000#150 from pbauman/use-autoheader 2e4bc39f5 Updated configure from the bootstrap 24250fe92 Run bootstrap, regenerate HYPRE_config.h.in 5b7ca434c run autoheader after autoconf dd8e7ba37 Add descriptions to AC_DEFINE ca195740f Merge pull request Nek5000#144 from hypre-space/fix-bool 84423d95d Merge branch 'master' into fix-bool 2ba72ed41 Merge pull request Nek5000#149 from pbauman/fix-nvtx-build c10b129ce Compile hypre_nvtx.c with nvcc 9ff066b52 Move some preprocessor directives before extern C 14c9b050d Merge branch 'fix-bool' of https://github.com/hypre-space/hypre into fix-bool 4b8bb0f5f Erase ifdef statement 8dd210957 Moved the autotest run filter to 'runtest.sh' where it belongs c94ba5842 Merge branch 'fix-bool' of https://github.com/hypre-space/hypre into fix-bool 0be3e041f Use HYPRE_REAL_MIN de87515b0 add a warning msg in headers f7673f1a3 Set atol to correct value when HYPRE_SINGLE is defined 343e31c84 Merge pull request Nek5000#148 from hypre-space/mgr-dev 04c4e0601 Minor cleanup f145db623 Fix annotation for single level MG af38cf66f Merge branch 'master' into mgr-dev 24fb670c0 Remove SparseMSG c989919d8 Update struct fortran driver c9216c4f7 Update SetPrecond routines 105dae6b9 - Change option for using full AMG V-cycle for F-relaxation from 99 -> 2 96cd74c43 - Set tolerance of global smoother, F-relaxation, and coarse grid solvers to 0 to avoid non-convergence error return. 735df4bc3 Adding a run.filters file to AUTOTEST 93d011a3d Annotate PFMG 1277fbcb5 Annotate BoomerAMG levels 7e5b059b7 Support annotation of multigrid levels 61ebb5cee Annotate krylov solvers b71a8ad5c Update caliper calls according to new API 89095db97 Update Euclid private header 1738f970f removed option listings but added explanations 27d263927 Fix compiler warnings 1f1ec406c Use stdbool in Euclid 2ce9617b6 GPU build doc 3b22847de Merge pull request Nek5000#138 from hypre-space/headers d9d040ded minor changes 3a8c7f96f fixed one more typo ..... 837a40c95 fixed a typo b0a79cfbb further clean-up 5eb6e965e changes per Rob's advice 3f8ff325b bug fixes after regression tests aad671be8 separation of C/C++ headers, and all the adjustment from it ebfc63324 Merge pull request Nek5000#136 from hypre-space/Interp 870d530de fixed issue with hypre_BoomerAMGBuildModExtPIInterpHost and hypre_BoomerAMGBuildModExtHost 7fd7dfca5 added some regression tests for new 2-stage-ee interpolation 0f8aafcb8 Merge branch 'master' into Interp 7baf13204 Merge pull request Nek5000#135 from hypre-space/RL_FIX b0c0c9591 added fix for large number of threads febafc15b Removed experimental option e6a3b3512 fixed another issue with missed special fine points 082f4c4dc fixed some bugs caused by lacking inclusion of special fine points 9edf9d274 Merge branch 'master' of https://github.com/hypre-space/hypre into RL_FIX fb8e844f3 Merge branch 'master' into Interp f875c2a38 Merge pull request Nek5000#134 from hypre-space/RL_FIX_INTERP 59836adad uploaded a wrong file.... 6440948e6 bug fix (omp with large #threads) 933ceed52 Merge pull request Nek5000#128 from hypre-space/SpGeMMDist 5d5b75bc0 GPU regression tests 3d1a44167 a minor change in check-headers.sh 1d35b0f7d Merge branch 'master' of https://github.com/hypre-space/hypre into SpGeMMDist 4ab72fd3a Merge pull request Nek5000#129 from hypre-space/fix-headers b39a8455a minor bug fix 6f34c5a8e fixes for PETSc 0f3a1c613 cuda build/compile error fix e121789fc Adding header check to machine-tux.sh autotest script 0b322c996 Fixed a few additional header file issues in examples 4542bf094 Fixed header usage problems and added regression test aeb23b458 Removed lower-case header dependencies in examples ce7ef0849 new sparse mat-mat-dist, triple-mat-dist 8bc028e5a fixed a few bugs 2a76685fc added a new 2-stage interpolation routine and auxiliary routines 60212ab6f Minor changes to CHANGELOG 28a7abe9b Merge branch 'master' into Interp 72c8dc25a added changes 708f58b21 Updates for new release 2.19.0 5bf6396a2 added new 2 stage interpolation numbers cc7063a70 added support file b84f132be updated reference manual added support file 2e9e8747f fixed a bug when using rlx 7 with OpenMP e76dced04 Merge pull request Nek5000#122 from hypre-space/spack-autotest-update cc0dde722 Small changes to spack autotest script d08a0f35c Uninstall spack packages in hypre+superlu-dist test. a2090abe5 fixed EXT interp seg fault issue with non-GPU coarsenings b70859897 added Rob's interpolation (18) and some new 2stage versiona built on it (7) 9791056f0 Merge pull request Nek5000#121 from hypre-space/gpudev-bjorn b94555398 updated regression tests on lassen 0c37e65d5 a minor change 636725e4e regression tests for the new functionalities 23b0e1226 Merge branch 'gpudev-bjorn' of https://github.com/hypre-space/hypre into gpudev-bjorn 3e6aa4390 some good format changes were lost (how come?) fc8305e05 fix after regression tests fe5e811b9 do not change Makefile in test 7336eab09 Fixed a bug in autotest 'make.sh' script 5cebc9383 bug fix d632e6a8a ext/ext+i interp. device 6c0e7d241 Merge pull request Nek5000#119 from hypre-space/spack-autotest 1d014c27c Adding a Spack autotest script 62b61918c Merge pull request Nek5000#118 from hypre-space/superlu-dist-update 35386791f Updating autotest for SuperLU-dist version 6.3.1 506fc0a35 headers a129c4257 Merge branch 'master' of https://github.com/hypre-space/hypre into gpudev-bjorn 249383ad6 removed a duplicate header 0ea3f8da1 Merge pull request Nek5000#117 from hypre-space/parcsr_header d31f28b82 Merge branch 'master' of https://github.com/hypre-space/hypre into parcsr_header 8e9c5cebe Merge pull request Nek5000#111 from hypre-space/RL_GPU 535d01e01 Merge pull request Nek5000#114 from hypre-space/RAJA 8bf373b0e Merge pull request Nek5000#115 from hypre-space/fix-2s-interp 751925751 fixed a bug in two-stage interpolation 53912e72f Merge branch 'master' of https://github.com/hypre-space/hypre into parcsr_header 3c6cb66fb changes according to the latest RAJA d3cc09b0c Merge pull request Nek5000#113 from hypre-space/vm-fix 6f8f51316 added protos.h; bug fix ad7fb8700 header scripts in parcsr_mv/_ls and seq_mv 96bc109ec Build recv_rboxes in SStructAMRInterCommunication b718ebfda Merge branch 'master' of https://github.com/hypre-space/hypre into RL_GPU 435eab61c bug fix 0d4089bdf FFFC on device ff79fb0d7 optimization for fffc: squeeze out unused offd columns 15737275b added extended+i interpolation on GPU bd1f981c6 Merge pull request Nek5000#110 from hypre-space/fix-issue108 176b84eda Fix periodic.out 41df3db1d Merge branch 'master' into fix-issue108 786e25cd4 Add new regression test. 84ebcd0b0 Map RecvRBoxes to coarse grid 2f7349400 Check if rboxes are empty before send/recv b4c779c36 Merge pull request Nek5000#109 from hypre-space/mgr-dev 74d16716b Minor clean up. 83ff765c4 Cleanup to remove unused variables 7ab476182 Fixed dividing by 0 computing inv(a_diag) in building interpolation. 5ff2630da Merge with master. 4182f90cd Merge pull request Nek5000#106 from hypre-space/dslu_bugfix 5ee53e36c Edited comment to pass regression test. 02c04fff4 Merge branch 'master' into dslu_bugfix f289a8b5b Modified superlu.c to use preprocessor definitions to decide when to copy data for MixedInt case 9f4e1b273 Add option for setting local reordering in ILU 9d1fbbc4e Merge pull request Nek5000#104 from hypre-space/new-interp 1ae1c7778 added another file to CMakeLists.txt 4d8fe2042 added new files to CMakeLists.txt 6888e05b6 changed regression test and a few small edits 29cb50203 Fixed some memory bugs and updated hypre to use latest version of superlu_dist 212182668 fixed memory locations for GPU 035f2a7d9 added new extended(interptype 16), extended+i(iinterptype 17) and i 2stage extended (iagg_interptype 5) interpolation and supporting routines. 84df86f9b updated extended interpolation on device, fixed issue in extractsubmatrixFC 952a1477c Merge pull request Nek5000#102 from hypre-space/hypre_Handle_wo_HYPRE_Init dbc45fcb1 bug fix 3869b3196 minor changes 55475ea61 changed function arg `l1_norms' back to HYPRE_Real * 71f420863 minor changes after regression tests 445732424 access global var _hypre_handle via hypre_handle() 2e843c512 fixed typos 25eacb1d5 Merge pull request Nek5000#101 from jedbrown/jed/fix-complex 85c80d00a Fix for inclusion with complex.h e9bcf42db Merge pull request Nek5000#93 from hypre-space/amg-setup e8efde802 Merge branch 'master' of https://github.com/hypre-space/hypre 51f7834b0 remove license from cub 28f60b924 Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup 7b7aa2d0a Merge pull request Nek5000#98 from hypre-space/fix-issue96 aed4a3fdc Add test to treat S2 with zero size 62b15c742 Do not compute avg_nnzrow when S has global size 0 0bd0c5b15 Enable additional tests b921cd2b8 Add SetOffProc test eab7b7ea2 IJSetValues: tried to save some malloc, free, and memcpy, but the overall time seems not being improved 1e4dd38c4 ij_assembly 23fb795ca Merge branch 'RL_FIX' of https://github.com/hypre-space/hypre into amg-setup 97631a505 fixed compiler warning 432be1611 ij_assembly 0f102bd3e assembly driver b0000ee0b Merge branch 'RL_FIX' of https://github.com/hypre-space/hypre into amg-setup 3327f4fc9 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup 761e014ed Update with new memory model and test_Set 51f085b23 minor changes 02dad90b9 add exec_policy in more drivers f87ec504f new file in CMakefile c4e6cdc08 for running regression tests with GPU setup/assembly 3982a1e89 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup 4f8352e70 bug fix 2e4c8145e remove cub from check license filters b58c1f2f4 more bug fixes regarding CUB cf4d9b78b bug fix in ILU fbb703400 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup 43ad3d670 cub allocator 5b1d72cb2 bug fix 0f20a71d7 fix after pull in master f491d130f Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup 4ffbb00db Fix indentation and minor clean up. 84423747c New interface for setting Cpoints using a point marker array. Updated test for compositional flow with wells. 022d476ed bug fix 22b18ab98 bug fix in ams c9b0f47a3 bug fix 9ef162da5 HYPRE_Init in examples; cmake; ams 9e83173a8 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup 1afd4a09a memory model, exec policy, etc eb71948ed Merge branch 'master' into mgr-dev 9e2e1491a Merge pull request Nek5000#80 from hypre-space/docs 4bfbba62a added par_lr_interp_device.c f68cb7bdb Added copyright header to readthedocs.yml file 0b1f1f244 Fixed a few remaining issues in HYPRE_parcsr_ls.h 1e70d39cd Adding copyright header to docs/usr-manual/requirements.txt file 716ddf63d Updated ParCSR solver header file to conform to new documentation approach 96a23ae14 first version of extended interpolation on device 766504983 Merge branch 'master' into docs 10a4cd53f Merge pull request Nek5000#86 from hypre-space/lobpcg 086c56e5d Merge branch 'master' into lobpcg 3d6424fa5 Merge pull request Nek5000#72 from hypre-space/mgr-dev f3b79f23f Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup 1147d72e5 Added hypre_ILU solver info to user manual. 68c064eac Minor change to fix compile warning and error 33bb597ac Modified to remove clang warning and potential bug. aac486403 Merge branch 'master' into mgr-dev 3d0cd5064 Some benign edits to revert some recent changes. 0b0e66fae Merge branch 'gpu-assembly' of https://github.com/hypre-space/hypre into amg-setup 271d55ed7 First version of an interface to set Cpoints by a marker array. e22ae7f6a Add driver for testing IJMatrix assembly 7253e38ad Add missing include guards 687232e11 gpu regression test [clean regression tests on ray] 79aa36fc7 bug fix 9b9e44d83 bug fix 93261e436 bug fix c7b29020d Merge pull request Nek5000#91 from hypre-space/gpu-assembly 79b5d7bc1 wrong configure pushed... fc9e301a9 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into gpu-assembly eaecaf173 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into gpu-assembly d2e5108f3 cuda stream sync, prefetch stream, etc 039dc2833 Merge pull request Nek5000#90 from hypre-space/gpu-interp-new 94c0834c6 bug fix 2b8fd590b bug fix 3945f82e4 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into gpu-interp-new 60101610d 1. put l1-norms inside container hypre_vector 2. code restructure d160f09a9 commented out all CudaPrefetch 74ae74d52 cub name space --> hypre::cub 794a1fc12 Merge pull request Nek5000#87 from hypre-space/fix-pcg 336a7bee9 Merge branch 'master' into fix-pcg 0fa087949 Fix memory leak, a block mode check in BoomerAMG, and string literal type a88591509 pcg fix from Rob 03688671a new impl of direct interp and a new interp formula dd9a724f1 CMakelist change corresponding to the previous name change 11f56bde6 Added an extra check for subnormal values in pcg 9492af96d bug fix b239a5c57 change Makefile after the previous name change 2f412591b rename a file with a better name that describes 34a21aca5 Merge branch 'improve-coarsen' acac8fd66 All tux tests are passing now 33475ea51 Merge branch 'master' into improve-coarsen 39e4f836a nvcc flag -lineinfo 2cb1b8389 minor changes 4775c5ff9 Merge branch 'improve-coarsen' bae22bc5f Remaining fixes for two memory leaks fcab2e6c5 Merge branch 'master' into improve-coarsen b2f6d96e7 tux tests are passing 7fbe167e9 bug fix e59109ad1 bug fix 641a018a3 Fix memory leak 5134773fb Merge branch 'master' into mgr-dev fc4ccf513 Minor changes to docs readme file b9fb9345d Adding a README file for building documentation 97e981469 Simplified the docs makefiles a bit adde096bd Added a dependence to documentation makefile 45588b218 Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup 4e1185331 1. IJMatrix Set/Add values, allow ncols==NULL 2. changes in ij.c for non-um f4f2a54de Merge pull request Nek5000#85 from koning/bugfix/koning/fei_memory_bigint 2ec1031b8 Fixing a small include file problem in examples/vis.c c69a3cbb0 more changes on drivers etc 2e3d60e34 Cleaned up some lobpcg-related header file stuff 013f111eb Change AMSData structure to use HYPRE_BigInt , HYPRE_Int and HYPRE_Real data types. Use hypre_CTAlloc and hypre_TFree for allocations sent to the ams solver. Remove one AMSData free due to double free in hypre_AMSFEIDestroy. 16961497a Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into gpu-assembly 0b47276cd IJ vector assembly 0a0aaf472 Merge pull request Nek5000#83 from hypre-space/cmake-install-prefix-fix f454c6e96 cmake: Use default location to install if CMAKE_INSTALL_PREFIX is initialized to default value. 8f873ab25 Merge pull request Nek5000#82 from hypre-space/RL-FIX 85fecd552 fixed gpu memory leak 216cbba91 bug fix - part 2 7bfebfe7a bug fix 7ed934afd Adding a .readthedocs.yml file for readthedocs 69451499e Adding a requirements.txt file for readthedocs 10f4317ed Trying to get this to run on readthedocs 2a2c8107b Adding some readthedocs stuff to usr-manual/conf.py 5e9dda399 Updating a few informational items in user-manual zREADME file 93fe7d5aa First complete version of API in User Manual 015643413 Adding system interface reference to new API tab in User Manual c4f019bd7 add off-proc values 692fb1d4d Adding license header to new cmake file 94036101b Merge pull request Nek5000#78 from ZedThree/cmake-update c7b290ab8 changes on IJ set sizes d34f9f7b4 local assembly should work now a5072811d save this version 4bb6d3d7c CMake: Remove unused HYPRE_SOURCES a708f8b33 CMake: first pass at using target-based approach 023e10571 CMake: Install basic CMake config files 81b8f8abf CMake: Don't clobber standard CMake flags 5644ce88e CMake: Require more recent version of CMake 7eb73531a Add remaining BigInts 82d208c30 Change global indices to HYPRE_BigInt ef6e3b042 Merge branch 'master' into improve-coarsen 79c70b44c Bug fix when using aggressive coarsening. f0275d9eb Add support to set F/C points to the various solvers. 1ef19f2f6 Drop ToKeep from SetCPoints function and C_points variables. HYPRE_BoomerAMGSetCpointsToKeep is deprecated. 4409ee92e Merge pull request Nek5000#77 from hypre-space/RL-FIX c1175f50a Fixed memory issue for 2nd time solves e03a335f0 Merge branch 'master' into mgr-dev 776122b97 Passes GPU regression test 67376c1c7 hypre_IJMatrixSetAddValues (*not done*) save this impl 7fb119959 Fixes for mixed-int build. 01a462ad5 Merge pull request Nek5000#66 from hypre-space/pgi-gpu-fix 1f89f7702 bug fix aac4d2567 Merge branch 'master' into mgr-dev 885fe5aa4 Some bug fixes for BigInt and Global_Partition b8c05a5c3 Added some new sort routines 97b3a9652 cub allocator b25be1a83 Merge pull request Nek5000#71 from hypre-space/mempool 9395a69f1 fixes after regression tests f80db4bc3 cub 9fadef736 Merge branch 'mempool' of https://github.com/hypre-space/hypre into mempool 41d5a9a19 ij.c 4f3b23f65 fixed conflicts after merge aa61cc43d Merge branch 'mempool' of https://github.com/hypre-space/hypre into mempool 9be64bd29 added calls to the cub memory pool 213d2390c Merge branch 'gpu-assembly' of https://github.com/hypre-space/hypre into mempool 8c636c0df minor changes 3e111ca88 initial update 52d3c26ad some changes 5f9a2c513 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into mempool 0a51805aa ij driver option 2a1c2d8e3 cub allocator 5d0ffec33 Do not restrict error messages to proc 0 5923a6a3b Remove gitignore file db237b821 Throw warning message by using the proper API 6c51a90ee memory tracker; do not save the size in device malloc anymore; prototype of TRealloc changed! d2765b551 Merge branch 'pgi-gpu-fix' of https://github.com/hypre-space/hypre into pgi-gpu-fix 564b469a0 Merge branch 'master' into improve-coarsen 18eaa531a Bug fix - invalid free ef8a25258 Clean debug statements d5f8142a4 Add gitignore file 5b06704f6 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into mempool b1d56dcac Reserved coarse points are BigInt 452fd1f46 Pass global coarse indices to SetCpointsToKeep 3f88e69d8 Initialize CF_marker to NULL 04bb19ab7 Minor fix + valgrind clean bdad4e3e2 Add option to set a list of C points to keep in the AMG hierarchy 197a822f1 HYPRE_BoomerAMGSetCpointsToKeep works with global indices a7ba73679 Fix file permissions 2995470c0 Fix file permissons 385e935fe Improve support to pgi compilers df469267e Use deprecated-register flag only with xlc 0adffeead Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup d533ee4f8 bug fix in ams, and ij driver to read rhs and x0 81c876e87 Add options to read fine and isolated fine points from file d8ff006e6 Allocate CF_marker if it is a null pointer 61af69b37 Add BoomerAMGSetFPoints function 37aef4347 Move CF_marker allocation to par_amg_setup. c3dc01761 Merge pull request Nek5000#61 from hypre-space/HYPRE_DEBUG 2d3064277 Add BoomerAMGSetIsolatedFPoints for seting a list of points as SF_PT in the finest level 7d9bd0947 Fix typo 6351dd1c8 Fix minor style issues d50750314 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup ff6c49af1 hybrid timer f0e28b79f Add SetCoarsenCutFactor function d0723ca24 Add strategy to automatically treat dense rows as SF_PT 81d5200e7 Fix some memory leaks 0906874b1 Bug fixes f3b9891d9 passed tuxbox regression tests bae521e55 nothing changed 3e190f283 Merge pull request Nek5000#63 from hypre-space/lessUM 1168e7481 passed gpu regression tests c2d875ffc Add suport to read vectors from a single file and distribute them to multiple processes in the case of no_global_partition 80cf3ca14 Extend VectorToParVector to no_global_partition case fb543f356 Use GetGlobalRowPartitioning in BuildFuncsFromFile b85fbc895 Add function that computes a global row partitioning of a ParCSRMatrix b029be434 Extend HYPRE_CSRMatrixToParCSRMatrix and BuildParFromOneFile to no_global_partition case 3322ffff3 Extend BuildFuncsFromOneFile to no_global_partition case 9da1274f0 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into lessUM f2751c61b Merge branch 'master' of https://github.com/hypre-space/hypre into lessUM c9755f5d0 less UM a647d7208 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup 973acba80 sstruct driver 0333171fb bug fix ccbc5a7d7 Updates to fix valgrind errors, and some minor cleanup 05fa1b500 removed .swp file 6f713accf Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into lessUM 145c1359d change memory copy behavior from/to UM 39b4e0393 nvtx 37899993f change default spgemm to cusparse c54fab8c8 Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup 95c48458e util header, comment in hypre_memory 8b422d61f Merge branch 'HYPRE_DEBUG' of https://github.com/hypre-space/hypre into amg-setup 11b510828 oversimple memory pool [proof of concept] e0a7ecf22 some changes in sstruct driver, hybrid solver setup time, and nvtx ca08a9668 ptr casting to int problem in `hypre_error_assert` cdf51f96d hypre_printf.h 8843d7a83 bug fix 9029876ed Merge branch 'HYPRE_DEBUG' of https://github.com/hypre-space/hypre into amg-setup 17dc86f84 bug fix bf25c6ed6 Merge pull request Nek5000#62 from hypre-space/hybrid-fix 6246b9486 bug fix hybrid amg 00616067b Merge branch 'HYPRE_DEBUG' of https://github.com/hypre-space/hypre into amg-setup ad5ff7940 bug fix (need to use assert in device code) ce2fe70fd changed all assert to hypre_assert 4c15015b8 changes after regression tests 1f2a3171b Incorporate Quan's edits to fix possible memory leak. e1a600b6e bug fix 1d97ab76e Changed 'strings.h' to 'string.h' in hopscotch stuff. f0fc40255 Merge branch 'HYPRE_DEBUG' of https://github.com/hypre-space/hypre into amg-setup b98bf115e Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup dda5db244 bug fix db024c782 Fix to CMakeLists (elseif -> else). 76d581ea2 bug fixes c1a45d643 a small change 69b73a038 nvcc c++11 does not like cast void (really a pain) b117db0fb Casting to void is used to suppress compiler warnings. 61db186de fixes after regression test 684820607 Bug fix by Quan 1f07247da Cleanup to resolve issues in regression test 8cbdca2b7 HYPRE Debug mode b56fb2d51 bug fix in hypre_ParCSRMatrixSetDNumNonzeros e195422be a "better" but tricker impl of ptr2ind 2a2f992e7 bug fix e01dd0ba3 stream sync for mpi-gpu 3d9f90e8c revoke all format changes 0c564e148 format changes 78f5ccf5d Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup f053ea13a more format changes a667cdbe4 Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup abee12a09 only format changes [typos, trailing whitespaces, tabs] d46fae80c Commented out unused variable to avoid compile warnings 94fe8cebc Minor edits to fix potential bug in ILU when residual vectors are not initialized to zero 2191feedb - Add the option to reduce the reserved Cpoints after a specified level. - Remove building A_ff matrix if an external F-relaxation solver is given. 93baaa8c9 Fixing a couple of PETSc-related valgrind errors 0a00f18bb Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup c758e4c26 Fixed bug in updating coarse indices on each level for more than 3 levels when setNonCpointToF = 1. d56485b3e - Add option to set F-relax solver for the first level. - Rearrange codes for building restriction and non-Galerkin coarse grid more efficiently. 9a621e346 Edits to cleanup build warnings 09ebccbc2 rebuilt _hypre_utilties.h`` d21f6dd2d Some cleanup after merge from master ab0024110 Merge branch 'master' into mgr-dev d509a088b Merge branch 'master' of https://github.com/hypre-space/hypre into amg-setup e35d6268e need a sync stream for gpu-mpi [temporary fix] ee84ffa23 Add mgr and ILU to CMakeLists 8da315a4a Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup 0bf20cd66 Merge branch 'amg-setup' of https://github.com/hypre-space/hypre into amg-setup d3aa485dd commit after pull 3b38dbaa1 fixed conflicts after git pull 06e77adf0 minor changes to have less conflicts between device memory and host memory 94b70cff6 Merge branch 'master' into mgr-dev 41cc6f3ec Some edits to BoomerAMGGetGridHierarchy function e7ecd8054 Added code to print coarse grid hierarchy for AMG 4b9681b9e Updated saved files for ILU tests 0d7aa599e Updated tests to include ILU testing. Also added RCM local reordering. adf7962ea Edited to fix compilation issues after merging in master 139c08fb1 Merge branch 'master' into mgr-dev b7bc2151b Fixed bug for GPU compilers 6e9651149 Merge branch 'master' into mgr-dev 78f8d23bb Fix parallel bugs in build sub-blocks of a matrix. 203ef956f Fix bug for getting sub-blocks of a matrix. Clean up merge. 12647b851 Merge with BigInt changes from master 6b4acd98c Merge branch 'master' into mgr-dev e4e2560bf Merged BigInt changes and minor cleanup of unused variables 62fcc11b1 Merge branch 'mgr-dev' of github.com:hypre-space/hypre into mgr-dev b00f87412 Merged changes to use hypre ILU and approximate inverse smoothers and interpolation operators 45ece2bdc Merge branch 'master' into mgr-dev 7805fecaa Change the routine for computing the non-Galerkin coarse grid in MGR to ensure correct sorting using hypre_qsort2_abs (hypre_qsort2_abs now sorts descendingly instead of ascendingly). 80aa9dba9 Merge with orgin/mgr-dev. Add non-Galerkin coarse grid option. Also add approximate inverse for building interpolation and coarse grid. 4db1092e1 Minor cleanup and updates to header file to remove compile warnings af58accdb Modified saved files to reflect code updates and bug fixes 8fcad73a8 Fixed bug for multilevel with vcycle smoother 6b32feb05 Some minor bug fixes and code cleanup. 5a06c8dd8 Added functionality to perform dual truncation for parcsr (utilizing existing code). Also did some minor cleanup to use routines that already exist without having to rewrite them. 53087be58 Changes made in this version include - Added block Gauss-Seidel relaxation and change the function name from 'hypre_block_jacobi' to 'hypre_blockRelax_solve'. - Remove 'last_level' argument from hypre_MGRBuildInterp. Now users can specify the interpolation type at all levels. - Fixed some memory leaks when using full AMG for F-relaxation. 90da2ceea Merge branch 'master' into mgr-dev 9ef6800b3 Comment edit. testing push to hypre-space 0a9dba42a Merge branch 'ilu-dev' into mgr-dev 93d2e4e3d Merge branch 'master' into mgr-dev 8cc2eef72 Merge branch 'master' into ilu-dev f4cb34489 Merge branch 'master' into ilu-dev ea047fdf4 Fixed misleading indentation. 50f358389 Cleaned up some print statements. bbf55a077 Merge branch 'master' into mgr-dev 015d56e06 Changes included in this version: - Separate the functions for specifying interpolation, restriction, and F-relaxation methods using an integer (for all the levels) and an array (for each level) to keep the previous interface intact. - Revert the name of the function for setting block data to the previous version and rename the new function for setting block data (when the physical variables are ordered contiguously). - Set default values for the parameters to prevent the code from crashing even if the user does not specify any parameters. 89a35de55 Make functions that set parameters for MGR more consistent. Fixed some memory leaks. 377a5b79f New code from Tianshi Xu 5735f5192 Put back option for doing a full AMG V-cycle for F-relaxation for testing purposes. 6d087582a Fixed the routine building the strength connection matrix for a sub-matrix block. The previous version does not account for the off diagonal part. 574fc2a9a Do some clean up and fix indentation to make the code more readable. We use 2 spaces (not tab) for indentation. a7bc1f9b3 Changes for this update include: - Allow for different interpolation and restriction options for each MGR level. - Add the number of functions for the Frelax V-cycle. - Fixed a bug for Frelax V-cycle when used as a preconditioner. The RHS should be obtained from the Solve phase, not the Setup phase. - Optimize sparsity pattern of interpolation operator. Injection does not need non-zero mapping for zero block. - Set some default values for using AMG for F-relaxation to prevent crashes. 142dd5912 This update contains the following changes: - New interface for setting C-F splitting for matrices with block structure, i.e., the same variables are ordered contiguously (s_1,s_2,...,s_n,p_1,p_2,...,p_n,...) - Allow different methods for F-relaxation at different levels. - Added a test file for MGR to test flow matrices coming from geocentric. 660cce2df Add a new qsort function to utilities. 3701a7f2e Final push before leave. Add ddPQ-GMRES-ILU into HYPRE. Add ParILU as a smoother options for AMG. 48ee6a098 Add local ddPQ permutation into ilu-dev. 5c8b8278f Sort U part to have better scaling result for high number of processors. 3c9365e52 Add RAS-ILU solve to ilu-dev. e154ba332 Add NSH-ILU solve/precond to ilu-dev 1ab5433aa Update operator_complexity, now ILU-GMRES also consider the memory for the preconditioner on Schur Complement. d4891d10f Add iluk and ilut for Parallel ILU solver and preconditioner Add ILU-GMRES solve for Parallel ILU solver and preconditioner Update ij.c to support more tests 55087f868 Updated ij driver to include solver options for ILU. Currently tested with ILU(0) and everything is working fine. c98edc4d2 Finalized initial implementation of ILUSetup and ILUSolve routines. Also added setup for ILU(0). 13bde64ce Updated setup routine. Still incomplete. dba5d829b Added setup and solve functions for ILU. These functions are currently place holders and are empty for now. ae03e79ae Added more functions to set ILU preconditioner. 8829a2731 Added AIR option for restriction de0656db9 Merge branch 'master' into mgr-dev c24f9abfa Initial commit for new ilu-dev branch. Added ILU data struct and create/destroy functions. git-subtree-dir: 3rd_party/hypre git-subtree-split: 0b09a2238c307411bb4eb8a42f6f1870b5285c02
|
Note: we should also have some way of distinguishing between, e.g. velocity residuals and pressure residuals when running in verbose mode. |
|
Added in 7a55f45 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The text was updated successfully, but these errors were encountered: