Commits

Branch selector

 master

User selector

All users

All time

Commit History

Commits on Jan 9, 2018

• Disable CUDA JIT debug flags on ARM archs

  

  pradeep

  committed
  c5e5f87

  Copy full SHA for c5e5f87

Commits on Jan 1, 2018

• Fix assert message for debug build

  

  umar456

  authored and

  pradeep

  committed
  5cf32df

  Copy full SHA for 5cf32df

Commits on Dec 29, 2017

• Fixed variable name typo in vectorization.md (#2032)

  Show description for c847344
  

  HoneyPatouceul

  authored and

  umar456

  committed
  c847344

  Copy full SHA for c847344

• Update configuring_arrayfire_environment.md

  Show description for 78812bb
  

  HoneyPatouceul

  authored and

  umar456

  committed
  78812bb

  Copy full SHA for 78812bb

• FEAT: Gradient & Curvature Anisotropic Smoothing

  Show description for a160a1c
  

  pradeep

  authored and

  umar456

  committed
  a160a1c

  Copy full SHA for a160a1c

• Improve morph object file build times in CUDA/OpenCL

  Show description for fc18d44
  

  pradeep

  committed
  fc18d44

  Copy full SHA for fc18d44

• Fix morphology fns to handle even size kernels

  Show description for dd81051
  

  pradeep

  committed
  dd81051

  Copy full SHA for dd81051

• Support for <backend>_NOT_SUPPORTED macros to accept messages

  Show description for 2f7a1d7
  

  pradeep

  committed
  2f7a1d7

  Copy full SHA for 2f7a1d7

Commits on Dec 23, 2017

• BUGFIX: make af_lookup behavior consistent across backends

  Show description for 81d5817
  

  pradeep

  authored and

  umar456

  committed
  81d5817

  Copy full SHA for 81d5817

Commits on Dec 21, 2017

• Disable threaded sort test temporarily

  Show description for 94620e6
  

  pradeep

  authored and

  umar456

  committed
  94620e6

  Copy full SHA for 94620e6

• Test for checking boost compute's thread-safety

  

  pradeep

  authored and

  umar456

  committed
  5b99f39

  Copy full SHA for 5b99f39

• BUGFIX: Enable thread-safe flags for Boost::compute upstream

  

  pradeep

  authored and

  umar456

  committed
  3183473

  Copy full SHA for 3183473

Commits on Dec 19, 2017

• Move clamp utility fn to common location on all backends

  

  pradeep

  authored and

  umar456

  committed
  421df36

  Copy full SHA for 421df36

• BUGFIX: Fix OpenCL tests on Windows for CUDA9

  Show description for d7b34ff
  

  pradeep

  authored and

  umar456

  committed
  d7b34ff

  Copy full SHA for d7b34ff

• BUGFIX: Fix Boost compile definitions on Windows

  Show description for 0afd229
  

  pradeep

  authored and

  umar456

  committed
  0afd229

  Copy full SHA for 0afd229

Commits on Dec 15, 2017

• Prefix path for MKL and homebrew. Add platform specific CMake file

  Show description for 04d1bac
  

  umar456

  authored and

  pradeep

  committed
  04d1bac

  Copy full SHA for 04d1bac

• Remove NVVM related cautionary warnings.

  Show description for 2733d58
  

  pradeep

  committed
  2733d58

  Copy full SHA for 2733d58

• Add example to af_lookup/af::lookup func docs

  

  pradeep

  committed
  4f9d8fc

  Copy full SHA for 4f9d8fc

• BUGFIX: Hyperlink in indexing tutorial docs

  

  pradeep

  committed
  140fc40

  Copy full SHA for 140fc40

• Enable documentation build in CMake

  Show description for c683ac0
  

  pradeep

  committed
  c683ac0

  Copy full SHA for c683ac0

• BUGFIX: Fix RandomEngine class attributes

  Show description for 7399ff6
  

  pradeep

  committed
  7399ff6

  Copy full SHA for 7399ff6

• BUGFIX: correct assets path in examples

  Show description for dd067b4
  

  pradeep

  committed
  dd067b4

  Copy full SHA for dd067b4

• Fix fast for CUDA 9. Use CUB library for reductions

  Show description for b3e927e
  

  umar456

  authored and

  pradeep

  committed
  b3e927e

  Copy full SHA for b3e927e

Commits on Dec 2, 2017

• Allow tests to specify the backends to build

  Show description for fb341e6
  

  umar456

  authored and

  pradeep

  committed
  fb341e6

  Copy full SHA for fb341e6

Commits on Dec 1, 2017

• FEAT: Add support for batched matrix multiply

  

  pavanky

  authored and

  umar456

  committed
  9447258

  Copy full SHA for 9447258

• BUGFIX: ireduce with single value in reduced dimension

  Show description for d0e9408
  

  umar456

  authored and

  pradeep

  committed
  d0e9408

  Copy full SHA for d0e9408

• Updated CLBlast to 1.2.0 which includes the latest TRSM/TRSV bugfixes

  

  CNugteren

  authored and

  umar456

  committed
  e220001

  Copy full SHA for e220001

• Updated CLBlast to 1.1.0, CLBlast's herk only called for cfloat/cdouble

  

  CNugteren

  authored and

  umar456

  committed
  93cd4be

  Copy full SHA for 93cd4be

Commits on Nov 29, 2017

• Reduced the spatialSigma value to improve meanshift test runtimes

  

  pradeep

  committed
  e408163

  Copy full SHA for e408163

• Improve meanshift filter performance on CPU

  Show description for 9d7cf5f
  

  umar456

  authored and

  pradeep

  committed
  9d7cf5f

  Copy full SHA for 9d7cf5f

Commits on Nov 15, 2017

• PERF, OpenCL: Fixes the launch config for approx1

  

  pavanky

  authored and

  umar456

  committed
  360b0d6

  Copy full SHA for 360b0d6

Commits on Nov 2, 2017

• Link afcpu and afopencl with pthreads

  umar456

  committed
  28a057b

  Copy full SHA for 28a057b

• Fix target_link_library call when compiling with CUDA stubs

  umar456

  committed
  f332183

  Copy full SHA for f332183

• Add missing functional header in memory.hpp

  Show description for 41811f3
  umar456

  committed
  41811f3

  Copy full SHA for 41811f3

Commits on Oct 30, 2017

• Use Array<T> objects instead of Param objects in several functions

  Show description for c7fd7f8
  umar456

  committed
  c7fd7f8

  Copy full SHA for c7fd7f8