search

richard-evans / vampire

Atomistic simulator for magnetic materials
GNU General Public License v2.0
124 stars 94 forks source link

Anisotropy #24

Closed samwestmoreland closed 7 years ago

samwestmoreland commented 7 years ago

Hi Richard,

Here's the latest on the anisotropy module. It's not integrated with the input constants properly yet, and I'm still working out the kinks with the indexing for the larger tensors. It all works on paper with the tensors written out in 2D, I'm just trying to remember how higher order tensor multiplication works!

richard-evans commented 7 years ago

Hi Sam,

Thanks very much, though I think you issued a pull request against the wrong branch(!). Anyway I will pick this up today and tomorrow and see how things go. The main thing we need to sort is the local Neel anisotropy - the higher order stuff we can add later.

Cheers,

Richard

On 29 Jun 2017, at 07:01, samwestmoreland notifications@github.com wrote:

Hi Richard,

Here's the latest on the anisotropy module. It's not integrated with the input constants properly yet, and I'm still working out the kinks with the indexing for the larger tensors. It all works on paper with the tensors written out in 2D, I'm just trying to remember how higher order tensor multiplication works!

You can view, comment on, or merge this pull request online at:

https://github.com/richard-evans/vampire/pull/24

Commit Summary

Added Heun corrector step Make num_atoms a compile time parameter Make dt a compile time parameter Add beginnings of exchange calculation Lots of additions Use CSR matmul for now while I work on CSR to DIA conversion Don't need this division Various additions Fix stats initialization Allow calculation of stats on cpu created new shape cone --> defined in src/create/cone.cpp created new shape cone --> defined in src/create/cone.cpp Merge branch 'cone' into dipole Add class for 3d buffers, will use for eg {x,y,z}_spin_arrays Added timing components for speed tests and tuned. Make use of new Buffer3D class Free memory for Device3D objects modified cone shape creator: add parameter create:cone-angle = angle of the cone used to shape the cylinder Merge branch 'cone' into dipole Removed relocatable device code from cuda makefile. Merge pull request #18 from am1808/dipole Merge branch 'dipole' into develop Fixed minor compilation error, spurious output to screen and added Andrea Meo as contributor to vampire Merge branch 'develop' into spin-torque Merge branch 'develop' into vio Added Rory Pond and Sarah Jenkins as contributors Completed implementation of voronoi substructure Changed default laser pump power to sensible value Merge branch 'unitcell' into develop Merge branch 'cuda' of https://github.com/odarbelaeze/vampire into cuda Removed random junk output to screen and moved timer to simulation part of code Merge branch 'develop' into cuda Added missing header file and added required std compliance for cuda makefile Corrected makefile to allow cuda compilation Fixed segmentation fault in vectorial code corrected allocation of memory when resizing cells::atoms_in_cell_coords_array_x,y,z in src/cells/initialize.cpp Added support for user selectable device id in cuda Added compile time control of cuda matrix format and fixed compilation errors Merge branch 'dipole' of https://github.com/am1808/vampire into dipole Reinstated missing option for setting non-magnetic atom behaviour (keep/remove) Added flag to explicitly enable core shell generation Added Andrea Meo fix for cell initialisation memory bug Merge branch 'develop' of github.com:richard-evans/vampire into develop Merge branch 'develop' into capped-grains modified src/config/output.cpp included version from am1808->dipole branch src/utility/vcongif.cpp used to print cells configuration files that has been omitted in new develop branch Merge pull request #21 from am1808/develop Add a header file for OpenCL source code constants, types etc Attempt to improve readability Add necessary pragma Correct typos Typedef struct Don't need this file Check for errors in kernel building and setting args Make directory for OpenCL source files to keep things tidy Don't need these files anymore Add some comments Add single precision make targets Fix many stupid mistakes Make c volatile. Otherwise it segfaults on compilation for some reason Bugfix: fixed parallel determination of particle origin centred on atom and refactored c array to std::vector Use cl::NullRange to make implementation choose work group size Remove volatile. Add lots of consts Only set kernel arguments once Don't use OpenCL exceptions and declare functions noexcept Make pointers const and restrict + small fixes As previous commit Forgot to add this file before. Xorshift* + Box Muller transform Fix to loop in prng Don't call the pass kernel args each time and set args where possible at initialisation Only need N/2 work items for prng as each generates two numbers Add function for zeroing buffers if enqueueFillBuffer isn't available Simplify zeroing of buffers Silence some warnings Move function templates to class template for Buffer3D Fix for single precision Use vector OpenCL datatype to pack data for fewer global reads Bugfix: fixed erroneous call to MPI_Gather walking all over memory If dependant .hpp is changed, recompile .o Better use of vector types Do not seed xorshift with 0 Add compile option to use a cl vector type for data Small fixes Rename vcl_real_t to real_t to stop vcl::vcl_real_t Move kernel building and argument setting to the initialize stage. Remove unused files Fix for when not using vector type Fix heun params buffer size Add config option for number of threads to use Fallback to 64 threads if not specified in input file Fix variable name Remove unneeded sync Remove unneeded function calls Make sure std::stoi is supported Remove unneeded sync Simplify makefile and give description for compile options Kernels are now built before being initialised so that they can be tested Add script for timing vampire for different numbers of threads Correct debug flags and make opencl-debug actually work Fix dipole calculations for vector types Make code more consistent with the comment Use shorter form of kernel_call function License under GPLv2 Use inbuilt select function to allow vectorisation Add optimisation flags for building kernels Use OpenCL intrinsic functions More use of intrisic functions Use vector types to hopefully improve read/write speeds Feature: added new bubble particle shape Merge branch 'develop' into capped-grains Merge branch 'dipole' into develop Add python script to check results against analytic solution Correctly initialise materials Add scripts for comparing output to analytic results Pass or fail tests based on maximum absolute error Fix cell array buffer size Make use of constant memory for small arrays Organise physical tests and add beginnings of thermal test Remove unused buffers Use a function to create buffers instead of doing it manually Whitespace changes More whitespace changes Matrix multiplication tests are unneeded so have been removed. Moved topology initialisation to init_kernels init_exchange where it is needed. Use a function to create cl::Buffers to make sure it is being done correctly each time. Also add some asserts to make sure things are working Don't actually run vampire until this test is ready to be used Add make target to clean only OpenCL objects and dependencies (useful for only recompiling the OpenCL modules with different compiler flags) Simplify initialisation message and add extra information Correct typo Use create_device_buffer instead of write buffer for easier debugging Add checking for host to device memory copying Use std::chrono to time simulation Fix #ifdefs, always time initialisation Added check for unphysical negative material parameters Merge branch 'develop' of https://github.com/richard-evans/vampire into develop Merge branch 'develop' into config2 Fix: anisotopy was being overwritten by exchange Rewrite testing script, makes it easier to add more tests/options Parrallel IO update. Read functions still missing. -cl-fast-relaxed-math turns on -cl-mad-enable and -cl-no-signed-zeros anyway Go back to ternary operator from select function. Add ability to time kernels Time kernels if enabled. Add Thermal test. Initialised anisotropy module. Whitespace changes Use C++11 random instead of C-style rand Fix init with non-blocking writes Merge branch 'opencl' of https://github.com/srhm501/vampire into opencl Add options for selecting OpenCL platform and device Allow logging OpenCL information in non-debug build Remove template and rename variable Merge branch 'cuda' of https://github.com/richard-evans/vampire into opencl Add program for numbering OpenCL platforms and devices Make size of matrix and vector compile time constant for matmul kernel Put random number generation in a seperate queue as it is independant of the rest of the code. Enable multiple platforms and devices Do device to device copy without a host vector if in the same context Remove unused file Make multiple devices a compile time option Add explanation of -DENABLE_MULTIPLE_DEVICES to makefile Make use of vload3 and vstore3 and also get rid of -DUSE_VECTOR_TYPE option as it no longer gives any speed up Use fast_normalize if using single precision MPI_io working, gather issue still not fixed sincos seems to be faster than the other options Removed deprecated C++ MPI bindings Whitespace changes Moved sundry config variables to internal namespace Readd the option to use vector types for storage Add comments to cl_defs.h Add comments to initialize Use host memory flags as well as device memory flags Need to add to total field value, not overwrite Access the arrays as two component vectors. Completely refactored implementation of config io for atomic spins Removed superfluous headers from spins output Changed excessive number of io-nodes to warning rather than error Initial implementation of coordinate output Fix metafile output Corrected timings for spins output Reinstated support for legacy atoms output with timing information Deleted superfluous file in config module Added initial implementation of new vdc utility for new config module Added xyz output function Fixed byte offsets for mpi-io output Added povray output to vdc utility Removed defunct converter utility Added timer for spin i/o including overheads Added new archer target for makefile for compiling on ARCHER Modified formatting for total io timer Bugfix: Enabled system magnetisation calculation be default to prevent segfault in config output Moved program start message to avoid mangled output from module initialisation Bugfix: Added missing MPICF flags for serial compile Feature: Added Monte Carlo preconditioner to generate realistic spin state for thermalised LLG simulations Tidied get_string() function and used function to read all input text files Added vampire manual to source repository Manual: Added option descriptions for new config output Fixed compilation on older GNU compilers by reverting c++0x option and older MPI libraries requiring non-const arguments to functions Modified user interface and added timing information to precondition algorithm Manual: added entry for new preconditioning algorithm Bugfix: main output file now only opened on master process Added support for cell output Bugfix: fixed error in data file names for config output Added improvements to vdc utility display output and minor bug fixes Tidied gitignore file to make things more organised Added missing vampire manual tex file Added new options to main makefile to compile vdc utility and added install and uninstall options for the code Added povray spins.ini file to enable rendering of all files with 'povray spins' command Added default colour space to vdc utility and stub for command line interface (to be implemented) Added timing info to cells output Added support for average magnetisation vector Added manual reference for time averaged magnetisation vector Removed superfluous .s .p characters from executable name Added support for material susceptibility Added basic framework for supporting CSG operations for system generation with multiple materials in a unit cell. Material min and max parameters moved to create namespace under internal mp class Added option for rocksalt (NaCl) structure Added support for CSG operations with multi-atom unit cells Bugfix: now initializing material susceptibility with material magnetization Bugfix: Now correctly initializing atom unit cell category with uc_material instead of uc_lattice_category to allow proper generation of normal unit cells with only one material type Added support for un normalized mean magnetization and mean magnetization length for heights in statistics module Bugfix: fixed CSG for simpler materials by changing default material Bugfix: forced initialization of create::internal::mp material Bugfix: removed erroneous compile flag in gcc debug compile option Bugfix: sent correct statistic to mean material susceptibility calculation Bugfix: Added missing factor 0.5 in exchange energy calculation Merge branch 'develop' into opencl Added support for user definable precision output of magnetisation data to screen and file as well as fixed width formatting Merge commit 'ed3eeaf958e84c5a4b328feb026fb6cfc74555c6' into anisotropy Couple of typos Just stripped some whitespace Added initialise function in anisotropy module. Added fields calculation function in fields.cpp Revert "Added fields calculation function in fields.cpp" Added fields calculation in fields.cpp Cleaning up for git pull Merge branch 'anisotropy' of git://github.com/richard-evans/vampire into anisotropy Added scalar uniaxial fields calculation Added call to anisotropy::initialise() function in spin_fields calculation Corrected anisotropy::initialise call Deleted superfluous header file ALMOST compiles. External arrays are correctly referenced now Defined internal data structures in data.cpp Now compiles. All internal files added into internal makefile Anisotropy tensors declared Anisotropy calculated changed to always enabled. Second order tensor populated for uniaxial anisotropy. Added Néel anisotropy calculation Changed tensor declaration so that every atom has a tensor Changed tensors to multi-dimensional for ease of indexing. Amalgamated the tensor resize and tensor zero intialisation for all tensors Energy calculation for second order tensor executed in secondorder.cpp Moved energy calculation into energy.cpp. Tensor calculations now split into separate files. File Changes

A .editorconfig (20) M .gitignore (82) M Co.mat (6) A hdr/anisotropy.hpp (57) M hdr/atoms.hpp (26) M hdr/cells.hpp (116) A hdr/config.hpp (43) M hdr/create.hpp (174) A hdr/cub-1.5.2/.cproject (1103) A hdr/cub-1.5.2/.project (27) A hdr/cub-1.5.2/.settings/org.eclipse.cdt.codan.core.prefs (72) A hdr/cub-1.5.2/.settings/org.eclipse.cdt.core.prefs (177) A hdr/cub-1.5.2/.settings/org.eclipse.cdt.ui.prefs (3) A hdr/cub-1.5.2/.settings/org.eclipse.core.runtime.prefs (4) A hdr/cub-1.5.2/CHANGE_LOG.TXT (270) A hdr/cub-1.5.2/LICENSE.TXT (24) A hdr/cub-1.5.2/README.md (128) A hdr/cub-1.5.2/common.mk (212) A hdr/cub-1.5.2/cub/agent/agent_histogram.cuh (783) A hdr/cub-1.5.2/cub/agent/agent_radix_sort_downsweep.cuh (767) A hdr/cub-1.5.2/cub/agent/agent_radix_sort_upsweep.cuh (449) A hdr/cub-1.5.2/cub/agent/agent_reduce.cuh (465) A hdr/cub-1.5.2/cub/agent/agent_reduce_by_key.cuh (701) A hdr/cub-1.5.2/cub/agent/agent_rle.cuh (831) A hdr/cub-1.5.2/cub/agent/agent_scan.cuh (498) A hdr/cub-1.5.2/cub/agent/agent_segment_fixup.cuh (374) A hdr/cub-1.5.2/cub/agent/agent_select_if.cuh (698) A hdr/cub-1.5.2/cub/agent/agent_spmv_csrt.cuh (638) A hdr/cub-1.5.2/cub/agent/agent_spmv_orig.cuh (924) A hdr/cub-1.5.2/cub/agent/agent_spmv_row_based.cuh (470) A hdr/cub-1.5.2/cub/agent/single_pass_scan_operators.cuh (783) A hdr/cub-1.5.2/cub/block/block_discontinuity.cuh (1148) A hdr/cub-1.5.2/cub/block/block_exchange.cuh (1135) A hdr/cub-1.5.2/cub/block/block_histogram.cuh (415) A hdr/cub-1.5.2/cub/block/block_load.cuh (1250) A hdr/cub-1.5.2/cub/block/block_radix_rank.cuh (485) A hdr/cub-1.5.2/cub/block/block_radix_sort.cuh (865) A hdr/cub-1.5.2/cub/block/block_raking_layout.cuh (150) A hdr/cub-1.5.2/cub/block/block_reduce.cuh (607) A hdr/cub-1.5.2/cub/block/block_reduce_by_key.cuh (1139) A hdr/cub-1.5.2/cub/block/block_scan.cuh (0) M hdr/cub-1.5.2/cub/block/block_shuffle.cuh (0) M hdr/cub-1.5.2/cub/block/block_store.cuh (0) M hdr/cub-1.5.2/cub/block/specializations/block_histogram_atomic.cuh (0) M hdr/cub-1.5.2/cub/block/specializations/block_histogram_sort.cuh (0) M hdr/cub-1.5.2/cub/block/specializations/block_reduce_raking.cuh (0) M hdr/cub-1.5.2/cub/block/specializations/block_reduce_raking_commutative_only.cuh (0) M hdr/cub-1.5.2/cub/block/specializations/block_reduce_warp_reductions.cuh (0) M hdr/cub-1.5.2/cub/block/specializations/block_scan_raking.cuh (0) M hdr/cub-1.5.2/cub/block/specializations/block_scan_warp_scans.cuh (0) M hdr/cub-1.5.2/cub/cub.cuh (0) M hdr/cub-1.5.2/cub/device/device_histogram.cuh (0) M hdr/cub-1.5.2/cub/device/device_partition.cuh (0) M hdr/cub-1.5.2/cub/device/device_radix_sort.cuh (0) M hdr/cub-1.5.2/cub/device/device_reduce.cuh (0) M hdr/cub-1.5.2/cub/device/device_run_length_encode.cuh (0) M hdr/cub-1.5.2/cub/device/device_scan.cuh (0) M hdr/cub-1.5.2/cub/device/device_segmented_radix_sort.cuh (0) M hdr/cub-1.5.2/cub/device/device_segmented_reduce.cuh (0) M hdr/cub-1.5.2/cub/device/device_select.cuh (0) M hdr/cub-1.5.2/cub/device/device_spmv.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_histogram.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_radix_sort.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_reduce.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_reduce_by_key.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_rle.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_scan.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_select_if.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_spmv_csrt.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_spmv_orig.cuh (0) M hdr/cub-1.5.2/cub/device/dispatch/dispatch_spmv_row_based.cuh (0) M hdr/cub-1.5.2/cub/grid/grid_barrier.cuh (0) M hdr/cub-1.5.2/cub/grid/grid_even_share.cuh (0) M hdr/cub-1.5.2/cub/grid/grid_mapping.cuh (0) M hdr/cub-1.5.2/cub/grid/grid_queue.cuh (0) M hdr/cub-1.5.2/cub/host/mutex.cuh (0) M hdr/cub-1.5.2/cub/iterator/arg_index_input_iterator.cuh (0) M hdr/cub-1.5.2/cub/iterator/cache_modified_input_iterator.cuh (0) M hdr/cub-1.5.2/cub/iterator/cache_modified_output_iterator.cuh (0) M hdr/cub-1.5.2/cub/iterator/constant_input_iterator.cuh (0) M hdr/cub-1.5.2/cub/iterator/counting_input_iterator.cuh (0) M hdr/cub-1.5.2/cub/iterator/tex_obj_input_iterator.cuh (0) M hdr/cub-1.5.2/cub/iterator/tex_ref_input_iterator.cuh (0) M hdr/cub-1.5.2/cub/iterator/transform_input_iterator.cuh (0) M hdr/cub-1.5.2/cub/thread/thread_load.cuh (0) M hdr/cub-1.5.2/cub/thread/thread_operators.cuh (0) M hdr/cub-1.5.2/cub/thread/thread_reduce.cuh (0) M hdr/cub-1.5.2/cub/thread/thread_scan.cuh (0) M hdr/cub-1.5.2/cub/thread/thread_search.cuh (0) M hdr/cub-1.5.2/cub/thread/thread_store.cuh (0) M hdr/cub-1.5.2/cub/util_allocator.cuh (0) M hdr/cub-1.5.2/cub/util_arch.cuh (0) M hdr/cub-1.5.2/cub/util_debug.cuh (0) M hdr/cub-1.5.2/cub/util_device.cuh (0) M hdr/cub-1.5.2/cub/util_macro.cuh (0) M hdr/cub-1.5.2/cub/util_namespace.cuh (0) M hdr/cub-1.5.2/cub/util_ptx.cuh (0) M hdr/cub-1.5.2/cub/util_type.cuh (0) M hdr/cub-1.5.2/cub/warp/specializations/warp_reduce_shfl.cuh (0) M hdr/cub-1.5.2/cub/warp/specializations/warp_reduce_smem.cuh (0) M hdr/cub-1.5.2/cub/warp/specializations/warp_scan_shfl.cuh (0) M hdr/cub-1.5.2/cub/warp/specializations/warp_scan_smem.cuh (0) M hdr/cub-1.5.2/cub/warp/warp_reduce.cuh (0) M hdr/cub-1.5.2/cub/warp/warp_scan.cuh (0) M hdr/cub-1.5.2/docs/.gitignore (0) M hdr/cub-1.5.2/docs/.nojekyll (0) M hdr/cub-1.5.2/docs/Doxyfile (0) M hdr/cub-1.5.2/docs/DoxygenLayout.xml (0) M hdr/cub-1.5.2/docs/device_perf.xls (0) M hdr/cub-1.5.2/docs/download_cub.html (0) M hdr/cub-1.5.2/docs/doxygen.exe (0) M hdr/cub-1.5.2/docs/extra_stylesheet.css (0) M hdr/cub-1.5.2/docs/footer.html (0) M hdr/cub-1.5.2/docs/header.html (0) M hdr/cub-1.5.2/docs/html/.nojekyll (0) M hdr/cub-1.5.2/docs/html/CHANGE_LOG.TXT (0) M hdr/cub-1.5.2/docs/html/annotated.html (0) M hdr/cub-1.5.2/docs/html/argindexinputiterator_8cuh.html (0) M hdr/cub-1.5.2/docs/html/argindexinputiterator_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockdiscontinuity_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockdiscontinuity_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockexchange_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockexchange_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockhistogram_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockhistogram_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockload_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockload_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockradix__sort_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockradixsort_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockreduce_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockreduce_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockscan_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockscan_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/blockstore_8cuh.html (0) M hdr/cub-1.5.2/docs/html/blockstore_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/cachemodifiedinputiterator_8cuh.html (0) M hdr/cub-1.5.2/docs/html/cachemodifiedinputiterator_8cuh_source.html (0)
M hdr/cub-1.5.2/docs/html/cachemodifiedoutputiterator_8cuh.html (0) M hdr/cub-1.5.2/docs/html/cachemodifiedoutputiterator_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_arg_index_input_iterator-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_arg_index_input_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_discontinuity-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_discontinuity.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_exchange-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_exchange.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_histogram-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_histogram.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_load-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_load.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_radix_sort-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_radix_sort.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_reduce-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_reduce.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_scan-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_scan.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_store-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_block_store.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_cache_modified_input_iterator-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_cache_modified_input_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_cache_modified_output_iterator-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_cache_modified_output_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_constant_input_iterator-members.html (0)
M hdr/cub-1.5.2/docs/html/classcub_1_1_constant_input_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_counting_input_iterator-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_counting_input_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_swizzle_scan_op-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_swizzle_scan_op.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_tex_obj_input_iterator-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_tex_obj_input_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_tex_ref_input_iterator-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_tex_ref_input_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_transform_input_iterator-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_transform_input_iterator.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_warp_reduce-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_warp_reduce.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_warp_scan-members.html (0) M hdr/cub-1.5.2/docs/html/classcub_1_1_warp_scan.html (0) M hdr/cub-1.5.2/docs/html/classes.html (0) M hdr/cub-1.5.2/docs/html/constantinputiterator_8cuh.html (0) M hdr/cub-1.5.2/docs/html/constantinputiterator_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/countinginputiterator_8cuh.html (0) M hdr/cub-1.5.2/docs/html/countinginputiterator_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/cub_8cuh.html (0) M hdr/cub-1.5.2/docs/html/cub_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/device__histogram_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicehistogram_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/devicepartition_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicepartition_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/deviceradix__sort_8cuh.html (0) M hdr/cub-1.5.2/docs/html/deviceradixsort_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/devicereduce_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicereduce_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/devicerunlengthencode_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicerunlengthencode_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/devicescan_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicescan_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/devicesegmentedradixsort_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicesegmentedradixsort_8cuh_source.html (0)
M hdr/cub-1.5.2/docs/html/devicesegmentedreduce_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicesegmentedreduce_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/device__select_8cuh.html (0) M hdr/cub-1.5.2/docs/html/deviceselect_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/device__spmv_8cuh.html (0) M hdr/cub-1.5.2/docs/html/devicespmv_8cuh_source.html (0) M hdr/cub-1.5.2/docs/html/dir_011e1c944d88f71be72e1e24a5fda7cf.html (0) M hdr/cub-1.5.2/docs/html/dir_18fc672d63781b5a743137aee24ff656.html (0) M hdr/cub-1.5.2/docs/html/dir_80932b4cec52750ff92b1a1912314cf5.html (0) M hdr/cub-1.5.2/docs/html/dir_bb50a5ef59f19d030d06415663184d05.html (0) M hdr/cub-1.5.2/docs/html/dir_cb3a671affffe7eeb3fdf5ae58e42cc8.html (0) M hdr/cub-1.5.2/docs/html/dir_d583f216f1aafe19404e836b0c097ad2.html (0) M hdr/cub-1.5.2/docs/html/download_cub.html (0) M hdr/cub-1.5.2/docs/html/doxygen.css (0) M hdr/cub-1.5.2/docs/html/dynsections.js (0) M hdr/cub-1.5.2/docs/html/example_block_radix_sort_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_block_reduce_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_block_scan_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_histogram_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_partition_flagged_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_partition_if_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_radix_sort_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_reduce_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_scan_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_select_flagged_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_select_if_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/example_device_select_unique_8cu-example.html (0) M hdr/cub-1.5.2/docs/html/examples.html (0) M hdr/cub-1.5.2/docs/html/extra_stylesheet.css (0) M hdr/cub-1.5.2/docs/html/favicon.ico (0) M hdr/cub-1.5.2/docs/html/files.html (0) M hdr/cub-1.5.2/docs/html/functions.html (0) M hdr/cub-1.5.2/docs/html/functions_0x62.html (0) M hdr/cub-1.5.2/docs/html/functions_0x63.html (0) M hdr/cub-1.5.2/docs/html/functions_0x64.html (0) M hdr/cub-1.5.2/docs/html/functions_0x65.html (0) M hdr/cub-1.5.2/docs/html/functions_0x66.html (0) M hdr/cub-1.5.2/docs/html/functions_0x68.html (0) M hdr/cub-1.5.2/docs/html/functions_0x69.html (0) M hdr/cub-1.5.2/docs/html/functions_0x6c.html (0) M hdr/cub-1.5.2/docs/html/functions_0x6d.html (0) M hdr/cub-1.5.2/docs/html/functions_0x6e.html (0) M hdr/cub-1.5.2/docs/html/functions_0x6f.html (0) M hdr/cub-1.5.2/docs/html/functions_0x70.html (0) M hdr/cub-1.5.2/docs/html/functions_0x72.html (0) M hdr/cub-1.5.2/docs/html/functions_0x73.html (0) M hdr/cub-1.5.2/docs/html/functions_0x74.html (0) M hdr/cub-1.5.2/docs/html/functions_0x75.html (0) M hdr/cub-1.5.2/docs/html/functions_0x76.html (0) M hdr/cub-1.5.2/docs/html/functions_0x77.html (0) M hdr/cub-1.5.2/docs/html/functions_0x7e.html (0) M hdr/cub-1.5.2/docs/html/functions_func.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x62.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x63.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x64.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x65.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x66.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x68.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x69.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x6c.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x6d.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x6e.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x6f.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x72.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x73.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x74.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x75.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x77.html (0) M hdr/cub-1.5.2/docs/html/functions_func_0x7e.html (0) M hdr/cub-1.5.2/docs/html/functions_rela.html (0) M hdr/cub-1.5.2/docs/html/functions_type.html (0) M hdr/cub-1.5.2/docs/html/functions_vars.html (0) M hdr/cub-1.5.2/docs/html/globals.html (0) M hdr/cub-1.5.2/docs/html/globalsdefs.html (0) M hdr/cub-1.5.2/docs/html/groupblock_module.html (0) M hdr/cub-1.5.2/docs/html/group_collectivemodule.html (0) M hdr/cub-1.5.2/docs/html/groupdevice_module.html (0) M hdr/cub-1.5.2/docs/html/group_segmentedmodule.html (0) M hdr/cub-1.5.2/docs/html/groupsingle_module.html (0) M hdr/cub-1.5.2/docs/html/group_utilio.html (0) M hdr/cub-1.5.2/docs/html/grouputil_iterator.html (0) M hdr/cub-1.5.2/docs/html/group_utilmgmt.html (0) M hdr/cub-1.5.2/docs/html/grouputil_module.html (0) M hdr/cub-1.5.2/docs/html/group_utilptx.html (0) M hdr/cub-1.5.2/docs/html/groupwarp_module.html (0) M hdr/cub-1.5.2/docs/html/hierarchy.html (0) M hdr/cub-1.5.2/docs/html/index.html (0) M hdr/cub-1.5.2/docs/html/jquery.js (0) M hdr/cub-1.5.2/docs/html/modules.html (0) M hdr/cub-1.5.2/docs/html/namespacecub.html (0) M hdr/cub-1.5.2/docs/html/namespacemembers.html (0) M hdr/cub-1.5.2/docs/html/namespacemembers_enum.html (0) M hdr/cub-1.5.2/docs/html/namespacemembers_eval.html (0) M hdr/cub-1.5.2/docs/html/namespacemembers_func.html (0) M hdr/cub-1.5.2/docs/html/namespaces.html (0) M hdr/cub-1.5.2/docs/html/search/all_5f.html (0) M hdr/cub-1.5.2/docs/html/search/all_5f.js (0) M hdr/cub-1.5.2/docs/html/search/all_61.html (0) M hdr/cub-1.5.2/docs/html/search/all_61.js (0) M hdr/cub-1.5.2/docs/html/search/all_62.html (0) M hdr/cub-1.5.2/docs/html/search/all_62.js (0) M hdr/cub-1.5.2/docs/html/search/all_63.html (0) Patch Links:

https://github.com/richard-evans/vampire/pull/24.patch https://github.com/richard-evans/vampire/pull/24.diff — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.