Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit history

Commits on Sep 29, 2016

• Remove string definition with no length in gpuarray_buffer_opencl, see #262

  adler-j

  committed
  466a346View commit details

  Copy full SHA for 466a346

  View code at this pointBrowse repository at this point

Commits on Jul 28, 2016

• Add gpu_ prefix and change signature to unsigned ints

  tsirif

  committed
  f49da95View commit details

  Copy full SHA for f49da95

  View code at this pointBrowse repository at this point

Commits on Jul 26, 2016

• Add cuda extension to return cuda device count

  tsirif

  committed
  5fce66fView commit details

  Copy full SHA for 5fce66f

  View code at this pointBrowse repository at this point

Commits on Jul 19, 2016

• Add a new preamble macro GA_WARP_SIZE (#217)

  Show description for 1369056
  

  huanzhang12

  authored and

  abergeron

  committed
  1369056View commit details

  Copy full SHA for 1369056

  View code at this pointBrowse repository at this point

Commits on Jul 15, 2016

• enable to compile in under C99 compiler

  NeutralCode

  committed
  e0d424aView commit details

  Copy full SHA for e0d424a

  View code at this pointBrowse repository at this point

Commits on Jul 14, 2016

• GA_USE_HALF should always work with OpenCL

  huanzhang12

  committed
  52eb618View commit details

  Copy full SHA for 52eb618

  View code at this pointBrowse repository at this point

Commits on Jul 11, 2016

• Merge pull request #212 from huanzhang12/master

  Show description for 2a0bfd5
  abergeron

  authored
  2a0bfd5View commit details

  Copy full SHA for 2a0bfd5

  View code at this pointBrowse repository at this point

• fix crash when dst == src in cl_move

  huanzhang12

  committed
  78a6359View commit details

  Copy full SHA for 78a6359

  View code at this pointBrowse repository at this point

Commits on Jul 10, 2016

• Local memory support for OpenCL kernels

  huanzhang12

  committed
  d601eb3View commit details

  Copy full SHA for d601eb3

  View code at this pointBrowse repository at this point

Commits on Jun 22, 2016

• Add collectives and nccl build to CMakeLists.txt

  Show description for c7882ca
  tsirif

  committed
  c7882caView commit details

  Copy full SHA for c7882ca

  View code at this pointBrowse repository at this point

• Add funct to init comm_ops properly

  Show description for 3c99b2b
  tsirif

  committed
  3c99b2bView commit details

  Copy full SHA for 3c99b2b

  View code at this pointBrowse repository at this point

Commits on Jun 9, 2016

• Add a flags to disable multi-stream for a context.

  abergeron

  committed
  41458ebView commit details

  Copy full SHA for 41458eb

  View code at this pointBrowse repository at this point

Commits on May 24, 2016

• Make the compile cache per-context and make it cache a bigger portion

  Show description for 3fa284b
  abergeron

  committed
  3fa284bView commit details

  Copy full SHA for 3fa284b

  View code at this pointBrowse repository at this point

• Move the extcopy functionality out of the backends.

  Show description for 89f4f4f
  abergeron

  committed
  89f4f4fView commit details

  Copy full SHA for 89f4f4f

  View code at this pointBrowse repository at this point

• Hide the definiton of gpuarray_buffer_ops and provide a new less

  Show description for 8e3bed4
  abergeron

  committed
  8e3bed4View commit details

  Copy full SHA for 8e3bed4

  View code at this pointBrowse repository at this point

• Define a struct for the contexts and use it everywhere.

  abergeron

  committed
  b045f72View commit details

  Copy full SHA for b045f72

  View code at this pointBrowse repository at this point

Commits on Apr 19, 2016

• Merge pull request #143 from abergeron/gen_elemwise

  Show description for c1ef8df
  nouiz

  committed
  c1ef8dfView commit details

  Copy full SHA for c1ef8df

  View code at this pointBrowse repository at this point

Commits on Apr 18, 2016

• Add properties to expose the individual limits on each of the dimensions for grid/block sizes.

  abergeron

  committed
  4fb6610View commit details

  Copy full SHA for 4fb6610

  View code at this pointBrowse repository at this point

Commits on Mar 31, 2016

• Add proper support for event scheduling around kernels with setarg().

  abergeron

  committed
  0c03fbeView commit details

  Copy full SHA for 0c03fbe

  View code at this pointBrowse repository at this point

• Add back a setarg api for kernels, but make it optional.

  abergeron

  committed
  376e3e1View commit details

  Copy full SHA for 376e3e1

  View code at this pointBrowse repository at this point

Commits on Mar 30, 2016

• Add cluda wrapper to load/store half on non-native plaforms.

  abergeron

  committed
  ae66e4fView commit details

  Copy full SHA for ae66e4f

  View code at this pointBrowse repository at this point

• Add the NATIVE_FLOAT16 property.

  abergeron

  committed
  a659ae4View commit details

  Copy full SHA for a659ae4

  View code at this pointBrowse repository at this point

Commits on Oct 26, 2015

• use memdup in place of strndup (not defined in Microsoft libraries)

  

  thebucc

  committed
  97b0836View commit details

  Copy full SHA for 97b0836

  View code at this pointBrowse repository at this point

Commits on Sep 9, 2015

• Resolved errors and fixed the ununsed warnings.

  

  Nirvik-D

  committed
  fcbd342View commit details

  Copy full SHA for fcbd342

  View code at this pointBrowse repository at this point

Commits on Aug 6, 2015

• Fix bugs in the OpenCL backend.

  abergeron

  committed
  ccf00a2View commit details

  Copy full SHA for ccf00a2

  View code at this pointBrowse repository at this point

• Add the implementation for ERRBUF to the opencl backend.

  abergeron

  committed
  7014a40View commit details

  Copy full SHA for 7014a40

  View code at this pointBrowse repository at this point

Commits on Jul 23, 2015

• Add preallocated small buffer for error returns from kernels.

  abergeron

  committed
  273b838View commit details

  Copy full SHA for 273b838

  View code at this pointBrowse repository at this point

Commits on Jul 21, 2015

• Add new type ga_ssize.

  abergeron

  committed
  92e75bfView commit details

  Copy full SHA for 92e75bf

  View code at this pointBrowse repository at this point

Commits on May 14, 2015

• Rework of the call protocol for kernel to lower overhead and allow

  Show description for 921ba76
  abergeron

  committed
  921ba76View commit details

  Copy full SHA for 921ba76

  View code at this pointBrowse repository at this point

Commits on Feb 19, 2015

• Increased buffer size from 16 to 64 to accommodate long driver strings from vendors

  paolodedios

  committed
  40a4f04View commit details

  Copy full SHA for 40a4f04

  View code at this pointBrowse repository at this point

Commits on Nov 28, 2014

• Comment made less like actual code

  mdda

  committed
  b50e833View commit details

  Copy full SHA for b50e833

  View code at this pointBrowse repository at this point

Commits on Nov 25, 2014

• len now size_t, and (possibly) final touches

  mdda

  committed
  656a48fView commit details

  Copy full SHA for 656a48f

  View code at this pointBrowse repository at this point

Commits on Nov 24, 2014

• Tidying code

  mdda

  committed
  de8ae27View commit details

  Copy full SHA for de8ae27

  View code at this pointBrowse repository at this point

• Line-numbered source in error message for CUDA kernels too

  mdda

  committed
  c51f7a0View commit details

  Copy full SHA for c51f7a0

  View code at this pointBrowse repository at this point

Commits on Nov 23, 2014

• Append an extra \0 for safety

  mdda

  committed
  25e485cView commit details

  Copy full SHA for 25e485c

  View code at this pointBrowse repository at this point