[arrayfire] annotated tag upstream/v3.2.0+dfsg1 created (now 92ba6ae)

Ghislain Vaillant ghisvail-guest at moszumanska.debian.org
Tue Nov 17 15:55:39 UTC 2015


This is an automated email from the git hooks/post-receive script.

ghisvail-guest pushed a change to annotated tag upstream/v3.2.0+dfsg1
in repository arrayfire.

        at  92ba6ae   (tag)
   tagging  62696f13a0db2458a37810d21c64061cd21c1a82 (commit)
  replaces  upstream/v3.1.3+dfsg1
 tagged by  Ghislain Antony Vaillant
        on  Mon Nov 16 16:34:49 2015 +0000

- Log -----------------------------------------------------------------
Upstream version v3.2.0+dfsg1

Brian Kloppenborg (13):
      Update installation documentation to match current methods.
      Include special instructions for Windows.
      Add OpenGL requirements for Forge.
      Update documentation.
      Add Ubuntu 14.04 installation quirk
      Update INSTALL.md
      Update CMake and Make examples.
      Update section titles, helloworld exe.
      Update link to example projects.
      Add XCode instructions, assets.
      Add link to ArrayFire project template repo.
      Add PPA for glfw3 on Ubuntu 14.04
      Update INSTALL.md

Ghislain Antony Vaillant (1):
      Merge tag 'upstream/v3.2.0' into dfsg-clean

Pavan Yalamanchili (20):
      Merge pull request #979 from shehzan10/approx-batch
      Renaming src/api/hapi to src/api/unified
      Fixing CMakeFiles for unified backend
      Changes required to make unified library build the cpp bindings
      Changes to examples and test CMakeLists to build *_unified binaries
      Merge pull request #977 from arrayfire/heterogeneous_api
      Merge branch 'arrayfire/master' into 'arrayfire/devel'
      Merge pull request #1001 from arrayfire/hotfixes-3.1.2
      Merge pull request #1007 from shehzan10/unified_doc
      Merge pull request #1014 from shehzan10/16bit
      Merge pull request #1053 from shehzan10/cudalapack
      Merge pull request #1065 from umar456/swe
      Merge pull request #1070 from bkloppenborg/devel
      Merge pull request #1069 from umar456/devel
      Merge pull request #1064 from shehzan10/devel
      Merge pull request #1089 from shehzan10/devel2
      Merge pull request #1087 from shehzan10/devel
      Merge pull request #1097 from shehzan10/unified_checks
      Merge pull request #1099 from syurkevi/maniparr_docupdate
      Merge pull request #1100 from shehzan10/docs-3.2

Peter Andreas Entschev (32):
      Added GLOH function prototypes
      Added C API for GLOH
      Added C++ API for GLOH
      Added CPU implementation of GLOH
      Added CUDA implementation of GLOH
      Added OpenCL implementation of GLOH
      Added GLOH documentation
      Added GLOH unit tests
      Merge branch 'devel' into gloh
      Added missing 'AFAPI' to C++ API
      Added unified API for GLOH
      Updated test data
      Changed std::sort to std::stable_sort in CPU SIFT
      Updated SIFT/GLOH test thresholds
      Merge remote-tracking branch 'upstream/devel' into sift_fixes
      Fixed out-of-bounds memory access in CUDA/OpenCL SIFT
      Added homography function prototype and API
      Added CPU backend for homography
      Added CUDA backend for homography
      Added OpenCL backend for homography
      Added homography documentation
      Added homography unit tests
      Updated test data
      Merge remote-tracking branch 'upstream/devel' into homography
      Fixed homography for Intel OpenCL
      Disabled homography LMedS unit tests
      Split vision.h prototypes into multiple lines
      Fixed __syncthreads() calls in homography
      Added AF_HOMOGRAPHY prefix to af_homography_t enum
      Fixed homography documentation
      Removed unnecessary __syncthreads() on homography
      Removed unnecessary barrier from homography

Pradeep (8):
      Windows specific changes to HAPI Symbol Manager
      Cleaned up function call in hapi functions
      Fixed typo in data, device & index wrapper source files
      Corrected BUILD_ALL cmake macro arguments
      fix: opencl backend alone build fails due to this missing header
      Fixed CMake source bugs for windows platform in unified api sources
      Removed AFAPI attribute declaration where no needed for func definitions
      Another cmake fix for windows platform in unified api project

Pradeep Garigipati (2):
      Merge pull request #994 from vakopian/fi-leak-fix
      Merge pull request #1031 from shehzan10/unified_fixes

Shehzan Mohammed (105):
      FEAT Added batch support for approx1 and approx2
      Changing int to dim_t in approx kernels
      Added any dimension batching and gfor support for approx1 and approx2
      Change condition structure in approx
      Fix sizes for approx batch tests
      Use af_print_array_gen in unified basic example
      Change unified backend priority. Add af/backend.h to arrayfire.h
      Changed unified/basic.cpp to use C++ api
      Add unified backend details to using on pages and cmake.in file
      Documentation for unified backend
      Reduced size of approx1 batched linear test
      Change output of DOG to floating type
      Increment version to 3.2.0
      Added AF_MSG macro
      Added short (s16) and ushort (u16) types for CPU
      Added short, ushort support for CUDA backend
      Fix memory alloc for fast opencl
      Added short and ushort support for CUDA backend
      Remove ushort redifinition from imageio
      Change ushort to unsigned short in cpp
      Add typedef for ushort in tests
      Corrections in unified backend doc
      Add 16-bit enums to docs
      Merge branch 'master' into devel
      Fix dlopen string for OSX
      Documentation fixes
      Merge pull request #1034 from 9prady9/set_native_device
      Merge remote-tracking branch 'upstream/hotfixes-3.1.3' into devel
      Merge pull request #1052 from pentschev/gloh
      Added CPU fallback for CUDA LU when CUDA older than 7
      Added CPU fallback for CUDA QR when CUDA older than 7
      Added CPU fallback for CUDA QR when CUDA older than 7
      Added CPU fallback for CUDA Solve when CUDA older than 7
      Added CPU fallback for CUDA Inverse when CUDA older than 7
      Added CPU fallback for CUDA SVD when CUDA older than 7
      Call deviceGC before solve tests to minimize memory (tegra)
      Default CPU fallback for CUDA LAPACK to OFF. Use CUDA_LAPACK_CPU_FALLBACK=ON
      Change condition when nonfree are removed from ctest
      Fix comparison warnings
      Call submodule update if submodules are missing
      Merge pull request #7 from pentschev/sift_fixes
      Updated SIFT/GLOH test thresholds
      Merge pull request #1062 from arrayfire/master
      Provide option for MKL use for CUDA lapack cpu fallback
      Fix compilation fixes for VS2015
      Add return type docs for functions with varying return type
      Fix warnings
      Change clBLAS tag to the corrected commit
      Added function to get available backends
      Optimizations to backends available computation
      Optimization for JPEG, cleanup
      Moved common functions from imageio into header file
      FEAT add loadImageT and saveImageT. Provides loading in different types
      Change loop in surface example
      Fix enum value conversion in image
      Fix imageio load order in case of bitmap and not bitmap
      Add s16 and u16 types to image (graphics)
      Add s16 and u16 types to surface (graphics)
      Add s16 and u16 types to histogram (graphics)
      Add s16 and u16 types to plot (graphics)
      Add s16 and u16 types to plot3 (graphics)
      Update forge build tag
      Add load_image_t and save_image_t to unified
      Doc for loadImageT and saveImageT
      Fixes for ushort on windows
      Update test data
      Add intl/uintl to sort, sort_index, sort_by_key
      use cl_long and cl_ulong in sort functions
      Add intl/uintl to lookup
      Add intl/uintl to histogram and histeq
      Add intl/uintl to convolve and fftconvolve
      Add intl/uintl to set functions
      Add intl/uintl to meanshift
      Fix cuda shared memory instantiation for s64 and u64
      Fix comparison warning
      Compilation fix for non-imageio builds
      Add install page to layout
      More documentation updates for tutorials
      API Change loadImageT -> loadImageNative
      Add support for c32/c64 for isInf, isNaN, iszero
      Update links
      Fix iota dims check
      Fix af_device_array dims check
      Typo AFF_ERR_NONFREE -> AF_ERR_NONFREE
      Add version guards for v3.2
      Encode backend info into ArrayInfo::devId
      Added array/backend checks to unified backend
      Add getBackendId function to get backend info of an array
      Update unified api docs
      CHECK_ARRAYS lets C-API return errors in case of arr = 0
      Merge branch 'doc-updates' of https://github.com/bkloppenborg/arrayfire into unified_checks
      Merge branch 'devel' of https://github.com/shehzan10/arrayfire into unified_checks
      Added version checks for getBackendId
      Fix triangle test failures
      Merge branch 'homography' of https://github.com/pentschev/arrayfire into unified_checks
      Merge pull request #1096 from 9prady9/susan_fixes
      Merge branch 'homography' of https://github.com/pentschev/arrayfire into unified_checks
      Moved det to rank test file. Removed rank and det from missing test
      Removed gfor unsupported functionality
      Added new examples
      Added release notes for 3.2.0
      Update forge tag for af3.2.0
      Add Tegra X1 badges to readme
      Transpose build table in readme
      Added groups for graphics func documentation

Umar Arshad (6):
      Merge pull request #1056 from shehzan10/devel
      Port shallow water eq example from 2.1
      Add unified backend binaries to the OSX installer
      Merge pull request #1076 from 9prady9/cuda_memcpy_stream_fixes
      Merge pull request #1077 from shehzan10/imageio
      Merge pull request #1102 from arrayfire/devel

Vardan Akopian (1):
      use RAII to avoid freeimage bitmap resource leaks

pradeep (57):
      Additional operator* overloads for cfloat, cdouble
      Added mean<T> instantiations for int64 and uint64 in C++ API
      Clean up mean helper functions & typo fix in af_mean_all_weighted
      Additional unit tests for mean
      Updating assets commit tag
      Merge branch 'devel' into stats_tests
      Merge branch 'devel' into stats_tests
      Corrected path typo in mean tests
      Merge branch 'devel' into stats_tests
      Merge branch 'devel' into stats_tests
      Initial commit for heterogenous api for ArrayFire backends
      Heterogeneous API for arith and algorithm header functions
      Merge branch 'devel' into stats_tests
      Merge branch 'devel' into heterogeneous_api
      Updated copyright year in hapi source files
      backend-independent api wrapper for image & vision headers
      backend-independent wrapper for arrayfire funtions
      Renamed cmake file hapi build-identifier
      Cleaned up symbol manager class in HAPI wrapper
      Changed default backend enum to point to zero
      Documentation for runtime backend selection functions
      Wrapper work around for af_make_seq function in hapi
      set_backend and get_backend_count functions
      Moved HAPI examples into standard examples location
      Merge branch 'devel' into heterogeneous_api
      Utility functions for generating af_index_t array objects
      Moved indexing utility functions to common location
      Added missing functions hapi wrapper
      changed unified api to load libraries using prioritized list of paths
      Fixed cmake bug in examples also
      Added error display strings for unified api error codes
      fix in unified api for af_save_image
      Disabled Sort1000 & SortMed tests for sort_by_key and sort_index
      Merge branch 'devel' into stats_tests
      typo fixes in mean unit test
      Merge branch 'devel' into stats_tests
      Corrected a typo in statistics functions documentation
      basic unit tests for `af::cov` and `af_cov`
      unit tests for standard deviation function
      unit tests for correlation coefficient function
      Enabled integral types to float/double reduction
      statistics functions fixes
      Merge branch 'devel' into stats_tests
      function to set active cuda device using native id
      function to set active opencl device using cl_device_d
      Fixed template specilization for MSVC compiler in mean function
      Added check to verify f64 support in covariance unit test
      Specilization for Binary functor for cdouble type in cpu backend
      Merge branch 'surface_plot' of git://github.com/syurkevi/arrayfire into syurkevi-surface_plot
      Fix for cuda backend surface rendering function
      Merge branch 'syurkevi-surface_plot' into devel
      Replaced deviceSychronize calls with async versions
      Removed uncessary stream synchronizes in device pointer functions
      Style fixes
      Memory leak fix in SUSAN feature detector
      specilizations for abs math function for int & char
      Indexing test for out of bounds access

syurkevi (3):
      3D line plot feature
      3d surface rendering features
      updates matrix manipulation documentation

-----------------------------------------------------------------------

No new revisions were added by this update.

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-science/packages/arrayfire.git



More information about the debian-science-commits mailing list