summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Fix non-padded GPULink memory handling in FP3D kernelsWillem Jan Palenstijn2021-11-162-0/+11
| | | | | | | | | | This would fail silently if the output projection data object was not padded to a multiple of 32 pixels, potentially corrupting the start of projection rows. 3D GPU memory allocated by ASTRA itself is always padded by cudaMalloc3D and therefore not affected. GPULink allows bypassing this, possibly triggering this bug.
* Merge branch 'cufft'Willem Jan Palenstijn2021-11-161-66/+58
|\ | | | | | | | | This makes FFT/FDK more robust and faster by better sync handling, and cleans up error logging/handling.
| * Remove fft.cu custom cuda error handling macrosWillem Jan Palenstijn2021-11-161-42/+21
| |
| * Remove unnecessary costly syncs in FFTWillem Jan Palenstijn2021-11-161-2/+8
| |
| * Add missing synchronize before free in FFTWillem Jan Palenstijn2021-11-161-2/+10
| |
| * Add checkCufft functionWillem Jan Palenstijn2021-11-161-23/+22
|/
* Merge branch 'checkCuda'Willem Jan Palenstijn2021-11-1616-227/+133
|\ | | | | | | This cleans up error logging and handling for cuda calls.
| * Remove cudaTextForceKernelsCompletionWillem Jan Palenstijn2021-11-1616-121/+83
| |
| * Remove ASTRA_CUDA_ASSERTWillem Jan Palenstijn2021-11-163-86/+36
| |
| * Remove reportCudaError functionWillem Jan Palenstijn2021-11-163-25/+7
| |
| * Add new checkCuda function to simplify error handlingWillem Jan Palenstijn2021-11-163-0/+12
|/
* Add specialization to BP3D kernels for single slice volumesWillem Jan Palenstijn2021-11-162-13/+24
|
* Remove unused backprojectPointX/Y/Z functionsWillem Jan Palenstijn2021-11-0310-385/+0
|
* Display version in astra.test()Willem Jan Palenstijn2021-10-282-2/+6
|
* Merge pull request #306 from tomopy/python-versionWillem Jan Palenstijn2021-10-271-1/+1
|\ | | | | DOC: Update __version__ param in python module
| * DOC: Update __Version__ param in python moduleDaniel Ching2021-10-271-1/+1
|/
* Fix conda package meta.yaml for WindowsWillem Jan Palenstijn2021-10-271-1/+1
|
* Version 2.0.0Willem Jan Palenstijn2021-10-276-10/+10
|
* Update NEWSWillem Jan Palenstijn2021-10-271-1/+8
|
* Update copyright yearsWillem Jan Palenstijn2021-10-27439-877/+877
|
* Prepare for python 3.10Willem Jan Palenstijn2021-10-274-1/+15
| | | | We can't (easily) build for python 3.10 yet until cython is available in conda.
* Reorganize docker environmentsWillem Jan Palenstijn2021-10-2610-26/+111
| | | | Now using debian8, debian9, debian11 for cuda 8 - 11.4
* Update miniconda version used in docker build environmentsWillem Jan Palenstijn2021-10-263-6/+6
|
* Update READMEWillem Jan Palenstijn2021-10-112-18/+18
|
* Add brief conda build instructions for win64Willem Jan Palenstijn2021-10-113-0/+27
|
* Windows Python release updateWillem Jan Palenstijn2021-10-113-85/+10
|
* Fix Windows exportsWillem Jan Palenstijn2021-10-112-4/+4
|
* Update conda packages to python 3.9, cuda 11.1Willem Jan Palenstijn2021-07-017-20/+41
| | | | Since cudatoolkit=11.1 isn't in the main anaconda channel (yet?), the nvidia conda channel is used.
* Restrict conda host env libstdc++-ng to prevent build errors on docker envWillem Jan Palenstijn2021-07-011-0/+2
|
* Minor de-duplicationWillem Jan Palenstijn2021-02-251-2/+4
|
* Add note about Octave dev pkg requirementWillem Jan Palenstijn2021-02-252-2/+6
|
* Fix Windows conda package buildWillem Jan Palenstijn2021-02-251-1/+1
|
* Slightly improve configure logWillem Jan Palenstijn2021-02-191-2/+2
|
* Suppress nvcc deprecated-gpu-targets warningWillem Jan Palenstijn2021-02-191-1/+5
|
* Fix memory leak in CUDA 2D fan_bpFerdia Sherry2021-01-271-0/+2
|
* Add pkgconfig file to gitignoreWillem Jan Palenstijn2021-01-221-0/+1
|
* Fix supersampling version of cone_bp and add testWillem Jan Palenstijn2021-01-222-7/+28
|
* Split conda/cuda build environment in twoWillem Jan Palenstijn2020-11-136-57/+93
|
* Add conda packages for cudatoolkit-11.0Willem Jan Palenstijn2020-10-153-4/+13
|
* Merge pull request #256 from fsherry/fix_cuda2d_par_fp_memleakWillem Jan Palenstijn2020-09-171-0/+4
|\ | | | | Fix memory leak in CUDA 2D par_fp
| * Fix memory leak in CUDA 2D par_fpFerdia2020-09-111-0/+4
|/ | | Arrays were dynamically allocated in the function convertAndUploadAngles, but were not deallocated, resulting in a memory leak. At the end of this function I now delete [] these arrays, fixing the memory leak.
* Bump conda package versionWillem Jan Palenstijn2020-07-041-1/+1
|
* Fix missing importWillem Jan Palenstijn2020-07-041-0/+2
| | | | Thanks to @jleuschn for catching this.
* Merge branch 'direct_fpbp'Willem Jan Palenstijn2020-07-028-48/+159
|\
| * Add error checkWillem Jan Palenstijn2020-07-021-0/+4
| |
| * Add mode parameter to direct_FPBP3DAllard Hendriksen2020-07-021-6/+5
| | | | | | | | This change allows additive forward and backprojections.
| * Check numpy array typeWillem Jan Palenstijn2020-07-025-16/+25
| |
| * Add python astra.experimental.direct_FP3D/BP3DWillem Jan Palenstijn2020-07-021-0/+54
| |
| * Refactor cython data3d object creation to prepareWillem Jan Palenstijn2020-07-024-37/+82
|/
* Use numpy backward binary compatibility for conda packagesWillem Jan Palenstijn2020-02-078-46/+8
| | | | | | | | We now only build a single package built with numpy 1.11.x for numpy>=1.11,<2 . There is no guarantee that every future numpy version covered by this will be compatible with numpy 1.11, but numpy checks this internally at run-time, and should report this (noisily) when this occurs.
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-29 09:10:23 +0000