[clfft] branch master updated (3a2ae95 -> 3cd1154)
Ghislain Vaillant
ghisvail-guest at moszumanska.debian.org
Thu Oct 22 14:54:31 UTC 2015
This is an automated email from the git hooks/post-receive script.
ghisvail-guest pushed a change to branch master
in repository clfft.
from 3a2ae95 updated performance measurement scripts and instruction manual
adds d1b800d bumping version to 2.7
adds 5a10046 changing from doc to docs
adds 834e7d5 merging from master
adds df0cd9d Merge branch 'master' into develop
adds bb880d9 changes to fix ECC on performance issue
adds ccff4c9 fixing ECC issue in inverse C2R tranform; pow2 only
adds 5faf9ed fixing event leak issues that caused seg fault of client with 14.502 driver, this problem did not affect standalone library usage
adds cf16403 minor fixes to event handling from previous checkin
adds a67ea0a optimizing EnqueueTransform API timing measurement
adds 9cbda54 fixing duplicate timing displays
adds 6ee3759 adding missing plans in timer
adds dc663b6 changing algorithm for 2^19 and 2^20 sizes
adds 81be7cd bug fix for length 1
adds 0d5eeed fixing version number
adds 874cf45 fix #93: use ${CMAKE_DL_LIBS} for libdl linkage in libclfft
adds 8489d30 Merge pull request #97 from ghisvail/fix/libdl-linkage
adds c03d7bb Merge branch 'develop' of https://github.com/clMathLibraries/clFFT-private into develop
adds 1ca97a4 Merge branch 'develop'
new 8339cd3 adjusting the const multipliers
new 8db381c fixing a C2R issue that produces global work-items not multiple of local
new e4063ef timer implementation for OS X
new 80d6721 Merge pull request #100 from John-Colvin/develop
new 6a3ba77 Install cmake configuration to lib/cmake/clFFT
new 5b7abb0 Merge pull request #88 from marbre/develop
new 45807c8 Merge branch 'develop' of https://github.com/clMathLibraries/clFFT into develop
new cde5e6d add missing include on stdlib
new 4938fca Merge pull request #102 from ghisvail/bugfix/undeclared-getenv
new e232087 adding cuda client
new e8f9952 using host side gflops measurment
new 37ccac0 minor fixes
new 6da80e3 minor
new 92c72f1 minor
new 0d86ca4 initial commit for generator.inplace
new 0646887 replacing transpose name INPLACE->Inplace
new 7d914b5 fixing cmaklists to build inplace transpose
new 15be25d fixing INPLACE addition framework code, gen work to follow next
new ea46f22 adding kernel dump feature for INPLACE as well
new abba627 eliminating non-const global variables that can be thread unsafe
new 4423f3d removing some unneeded branch blocks for small size transposes
new 9ae2529 fixing sub-64 size transpose issues with real transforms
new 7b2ea61 adding new inplace transpose algorithm to test c2c functionality
new 7f5a6b9 added support for planar complex to generator inplace transpose
new e299c39 adding twiddle multiplies to inplace generator, non-branch case only
new 462b2bb changing inplace name to square to better reflect implementation feature
new 0d55f62 renaming the inplace files to square
new cecbeb2 fixing file path
new 59fa54f Fix redefinition warnings when using with clBLAS
new 714e25d Merge pull request #103 from shehzan10/redef_fixes
new d0379cc fixing install and package issues, keeping install files to minimum
new 4018ace removing call to client in travis build
new c1b0b27 added appropriate deallocate statements before merging pull request
new 8a4c20c fixed leaking of memory
new e4e1b86 Merge pull request #107 from VictorCarlquist/develop
new 8e34e31 Install examples to share/clFFT/examples
new b55121f Merge pull request #108 from marbre/examples
new 55281a4 typo in the loop to manage 3d fft for the hermitian to real case
new 96991eb Merge pull request #109 from BenjaminCoquelle/develop
new c1f779c Merge branch 'develop' of https://github.com/clMathLibraries/clFFT into develop
new a35ea0e adding offset calc in square transpose, minor fixes
new 1e661fb enabling inplace transposition for select conditions
new 3b8afc5 precallback-initialversion
new 6401e29 callbackclient-check input is 1D for now
new 9228c06 Precallback-verify with fftw output, user data as an array
new a45ed36 Pre-callback - Readme for client and minor update in client code
new 0b23880 Pre-callback - Readme update
new 8154963 Precallback - support for sizes > 4096
new 377e665 Precallback - skip redundant pointer declaration in transpose kernel if precallback set
new bac0f46 Precallback - C2C 2D and 3D support
new 5e9b96b Precallback - C2C double precision updates
new 8f07253 Precallback - Complex-Real 1D single kernel SP
new 82ba436 CMakeLists merge with master
new fec5292 Precallback - C2R 1D out-place updates
new 37b5a1f Precallback - Handle C2R Hermitian Planar and single pass use cases
new 0c57638 Precallback - C2R 2D and 3D SP
new 2a3e9bb Precallback - Added few Google Test test cases
new e10d1cd Precallback - R2C large 1D and few more GTests
new 8092c99 Precallback - R2C 2D and 3D implementation plus few GTests
new 16b6cf2 Precallback - Large 1D C2R, callback LDS updates and GTests
new 3a1625c Precallback - Merge from upstream develop branch and fix broken C2C/C2R precallback
new 7395828 Precallback - Fix for mixed radix C2R test case failure
new b8587a4 fix compile errors on linux
new 7abe254 Precallback - performance callback client first version
new 6dbd399 Precallback - fix client compile error
new f3d3417 Precallback - client updates
new 40cec4e Precallback - exclude R2C function from client
new 954a270 Precallback - Tests restructuring and bug fix for odd planar C2R
new 83bba5d Remove deleted file
new 3177750 Precallback - Round Trip tests
new 5512e2c ignore batch file from tracking
new 9f7f15b Precallback documentation in mainpage.h
new 0967583 Precallback - Few review comment fixes
new 857c7b2 Precallback - Remove userdatastructString parameter from SetPlanCallback API
new fab73e7 precallback-remove old client code
new 5cc4f20 Precallback - documentation update
new 9ac208d Precallback - client sample for converting 24bit input to 32bit format using precallback
new c55b715 merging from pradeep fork to pull in pre-callback feature
new c80becc fixing compile warnings
new 425ad98 Precallback - support precallback in the new generator TransposeSquare - Mul16 cases
new 66a8d16 fixing compile problems, accuracy issues remain
new ed99fdd fixing accuracy issues
new 88b39d2 Merge branch 'develop' of https://github.com/clMathLibraries/clFFT-private into pradeep
new b8151dc Support precallback for in-place square transpose non mul_16 cases and fix few compiler warnings
new 900dc03 Precallback - Documentation update and compiler warning fixes in client code
new d6c6127 adding radix7 functionality
new 16a788c Merge branch 'rad7prep' into develop
new ddab8a1 Merge branch 'develop' of https://github.com/bragadeesh/clFFT into develop
new 46fcd03 adding coverage for rad7 in 1 through 4096 range
new 91efde4 Merge branch 'develop' of https://github.com/clMathLibraries/clFFT-private into pradeep
new ed43c5f Mixed radices with Radix 7 values tests for Pre-callback
new da3965c Double Precision pragma enable in Square Transpose generator
new d98c247 enable double precision pragma in square transpose generator
new bb9cb39 Adding radix 7 test cases.
new 95ab622 Added parameterized test support for radix 7 test cases.
new 7e6a2df simplifying transform kernels for 3,5,7 combined sizes
new 0993a32 fixing some bugs in real2hermitian, debug continues
new 618cdd6 fixing more bugs
new 36cd6ba fixing many failures caused by deeper recursions, added more visibility of plan info in stattimer
new 53176bc Merge remote-tracking branch 'refs/remotes/origin/develop' into santanu
new 1b70609 changes and bug fixes in power7 test suite.
new e74c43f changed normal7 value to 343 from 2401.
new 436988c adding env variable to control in-place FFTs
new 84c846f fixing assertion failure found with rad7 test
new 040e512 Merge remote-tracking branch 'refs/remotes/origin/develop' into santanu
new 044fe01 Bug fix for dimensionality and batch size increase for radix 7 parameterized test.
new eca70cc Precallback - Radix 7 precallback unit tests
new b00f044 Precallback - Radix 7 with scale factor test
new ace0d18 Merge branch 'develop' of https://github.com/clMathLibraries/clFFT-private into pradeep
new 1dac82e Merge branch 'develop' of https://github.com/clMathLibraries/clFFT-private into pradeep
new 7fb5f78 merging pradeep branch after conflic fixes
new 4ed5802 Support FreeBSD in shared library switch. This is identical to the Linux case.
new ca6f0e9 Need to import sys/timespec header on FreeBSD to get access to the timespec structure.
new 0ece5e5 Add .h for the include.
new c61da98 Merge pull request #112 from iotamudelta/develop
new 091f3b7 Refactoring the .travis.yml file; adding OSX support Adding an appveyor.yml file to build for visual studio
new 7e2ec73 Readme changes to add build badges for master/develop and appveyor
new a4ad8b7 Merge pull request #113 from kknox/develop
new 5184be3 Merge branch 'develop' of https://github.com/clMathLibraries/clFFT into develop
new 4ea7539 fixing build fail in travis
new 76de7c5 updating header file with documentation changes in the comments, no change in code
new 433dbc1 fixing naming consistency for callback enum
new 6de3223 making readme updates
new 92997f5 documentation updates
new f34c68b Merge branch 'develop'
new 6738f98 fixing missed items
new ad15050 merging from develop with fixes
new 3cd1154 doc fix
The 128 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails. The revisions
listed as "adds" were already present in the repository and have only
been added to this reference.
Summary of changes:
.gitignore | 3 +
.travis.yml | 150 +-
README.md | 74 +-
ReleaseNotes.txt | 41 +-
appveyor.yml | 116 +
src/CMakeLists.txt | 23 +-
src/{client => callback-client}/CMakeLists.txt | 31 +-
src/callback-client/callback-client.cpp | 569 ++
src/callback-client/client.h | 119 +
src/{client => callback-client}/openCL.misc.cpp | 7 +-
src/{client => callback-client}/openCL.misc.h | 2 +-
src/{client => callback-client}/stdafx.cpp | 0
src/client/CMakeLists.pack | 182 -
src/client/CMakeLists.txt | 6 -
src/client/client.cpp | 105 +-
src/client/client.h | 23 +
src/client/openCL.misc.cpp | 7 +-
src/cuFFT-client/CMakeLists.txt | 41 +
src/cuFFT-client/cuFFT-client.cpp | 2748 +++++++
src/examples/CMakeLists.txt | 3 +-
src/examples/fft1d.c | 4 +-
src/include/clFFT.h | 405 +-
src/include/sharedLibrary.h | 8 +
src/library/CMakeLists.txt | 48 +-
src/library/accessors.cpp | 47 +
src/library/action.cpp | 26 +-
src/library/action.h | 57 +
src/library/fft_binary_lookup.cpp | 29 +-
src/library/fft_binary_lookup.h | 8 +-
src/library/generator.copy.cpp | 126 +-
src/library/generator.h | 1 +
src/library/generator.stockham.cpp | 1126 ++-
src/library/generator.stockham.h | 350 +
src/library/generator.transpose.gcn.cpp | 211 +-
src/library/generator.transpose.square.cpp | 1005 +++
...ranspose.gcn.h => generator.transpose.square.h} | 0
src/library/lifetime.cpp | 1 +
src/library/mainpage.h | 691 +-
src/library/plan.cpp | 522 +-
src/library/plan.h | 43 +-
src/library/private.h | 5 +
src/library/transform.cpp | 149 +-
src/scripts/perf/CMakeLists.txt | 11 +-
src/scripts/perf/fftPerformanceTesting.py | 21 +-
src/scripts/perf/measurePerformance.py | 6 +-
src/statTimer/statisticalTimer.CPU.h | 3 +
src/statTimer/statisticalTimer.GPU.cpp | 70 +-
src/statTimer/statisticalTimer.GPU.h | 18 +-
src/tests/CMakeLists.txt | 23 +-
src/tests/accuracy_test_common.h | 434 ++
src/tests/accuracy_test_mixed_precallback.cpp | 314 +
src/tests/accuracy_test_mixed_radices.cpp | 32 +-
...pow2.cpp => accuracy_test_pow2_precallback.cpp} | 1431 ++--
src/tests/accuracy_test_pow3.cpp | 2 +-
...pow3.cpp => accuracy_test_pow3_precallback.cpp} | 1529 ++--
src/tests/accuracy_test_pow5.cpp | 2 +-
...pow5.cpp => accuracy_test_pow5_precallback.cpp} | 1529 ++--
src/tests/accuracy_test_pow7.cpp | 7485 ++++++++++++++++++++
src/tests/accuracy_test_pow7_precallback.cpp | 1085 +++
src/tests/accuracy_test_random.cpp | 1 +
src/tests/buffer.h | 125 +
src/tests/cl_transform.h | 92 +-
src/tests/fftw_transform.h | 34 +
src/tests/test_constants.h | 89 +
src/tests/unit_test.cpp | 12 +-
65 files changed, 19799 insertions(+), 3661 deletions(-)
create mode 100644 appveyor.yml
copy src/{client => callback-client}/CMakeLists.txt (61%)
create mode 100644 src/callback-client/callback-client.cpp
create mode 100644 src/callback-client/client.h
copy src/{client => callback-client}/openCL.misc.cpp (99%)
copy src/{client => callback-client}/openCL.misc.h (98%)
copy src/{client => callback-client}/stdafx.cpp (100%)
delete mode 100644 src/client/CMakeLists.pack
create mode 100644 src/cuFFT-client/CMakeLists.txt
create mode 100644 src/cuFFT-client/cuFFT-client.cpp
create mode 100644 src/library/generator.transpose.square.cpp
copy src/library/{generator.transpose.gcn.h => generator.transpose.square.h} (100%)
create mode 100644 src/tests/accuracy_test_mixed_precallback.cpp
copy src/tests/{accuracy_test_pow2.cpp => accuracy_test_pow2_precallback.cpp} (73%)
copy src/tests/{accuracy_test_pow3.cpp => accuracy_test_pow3_precallback.cpp} (70%)
copy src/tests/{accuracy_test_pow5.cpp => accuracy_test_pow5_precallback.cpp} (70%)
create mode 100644 src/tests/accuracy_test_pow7.cpp
create mode 100644 src/tests/accuracy_test_pow7_precallback.cpp
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-science/packages/clfft.git
More information about the debian-science-commits
mailing list