Commit Graph

  • fd838c75bc add optimized cdot- and zdot-kernel for haswell Werner Saar 2015-04-09 15:13:52 +0200
  • b57a60dac8 updated cdot and zdot for piledriver Werner Saar 2015-04-09 10:33:46 +0200
  • 5c51163972 added optimized cdot- and zdot-kernel for steamroller Werner Saar 2015-04-09 09:45:23 +0200
  • 9299d8cfd6 added optimized cdot- and zdot-kernels for bulldozer Werner Saar 2015-04-08 16:29:55 +0200
  • 0a3d3b945d Refs #535. Fix the wrong vector instruction in sgemm sandy bridge kernel. Zhang Xianyi 2015-04-08 03:55:49 +0800
  • 4f680a7d61 Merge pull request #534 from wernsaar/develop Zhang Xianyi 2015-04-07 12:48:11 -0500
  • ba926e807c added cdot- and zdot benchmark Werner Saar 2015-04-07 11:56:06 +0200
  • 60c6dec6e6 updated some lines for bulldozer Werner Saar 2015-04-06 18:47:16 +0200
  • 47898cca35 added optimized saxpy- and daxpy-kernel for sandybridge Werner Saar 2015-04-06 16:05:16 +0200
  • 53bb924287 added optimized saxpy- and daxpy-kernel for haswell Werner Saar 2015-04-06 12:33:16 +0200
  • 1e80b8b0d3 Merge pull request #531 from wernsaar/develop Zhang Xianyi 2015-04-05 16:42:39 -0500
  • a901b065d3 added optimized ddot-kernel for sandybridge Werner Saar 2015-04-05 20:19:38 +0200
  • 3937e2a0a0 add optimized sdot-kernel for sandybridge Werner Saar 2015-04-05 19:47:05 +0200
  • 9707d608d5 removed double definition line Werner Saar 2015-04-05 18:35:34 +0200
  • 701b9d7556 added optimized sdot- and ddot-kernel for HASWELL Werner Saar 2015-04-05 17:57:53 +0200
  • 8977b3f235 Refs #529. Support Intel Broadwell by Haswell kernels. Zhang Xianyi 2015-04-02 11:08:03 -0500
  • f6426395ea Merge pull request #527 from xantares/patch-1 Zhang Xianyi 2015-03-30 10:16:11 -0500
  • 0ac787eefe fix mingw install xantares 2015-03-30 09:30:55 +0200
  • e5b96e55a7 Fix build bug for ARM64. Zhang Xianyi 2015-03-24 15:27:17 -0500
  • d0c51c4de9 Merge branch 'develop' v0.2.14 Zhang Xianyi 2015-03-24 15:07:07 -0500
  • a3491e1e88 Update the doc for 0.2.14. Zhang Xianyi 2015-03-24 15:05:59 -0500
  • e81a5d61e4 Merge branch 'develop' of github.com:xianyi/OpenBLAS into develop Zhang Xianyi 2015-03-24 12:17:12 -0500
  • c674fa32be Add ARM targets. Zhang Xianyi 2015-03-24 12:17:04 -0500
  • e34911a73d Fix compiling bug for ARM with setting BINARY. Zhang Xianyi 2015-03-24 17:15:33 +0000
  • 76dcaf2281 Merge pull request #521 from maxlevesque/patch-1 Zhang Xianyi 2015-03-21 12:26:35 -0500
  • 770fac92eb Correct typo /proc/ instead of /pros/ Maximilien Levesque 2015-03-20 23:25:11 +0100
  • e95d64333a Refs #519. Avoid calling strncpy. Zhang Xianyi 2015-03-19 15:57:22 -0500
  • 75c40bcc48 Refs #520. Fixed ONLY_CBLAS=1 compiling bug on OSX. Zhang Xianyi 2015-03-19 11:51:36 -0500
  • b62f9f4120 Merge pull request #518 from ton/issue-508 Zhang Xianyi 2015-03-18 13:00:07 -0500
  • b6438dedea Fix issue #508 Ton van den Heuvel 2015-03-18 13:22:43 +0100
  • 1d183dcda8 Added lapacke sources. Hank Anderson 2015-02-25 16:51:08 -0600
  • cdefdb21cd Refs #492. Fixed c/zsyr bug with negative incx. Zhang Xianyi 2015-02-26 06:37:03 +0800
  • e19bf3a28b Removed MSVC cpuid func when using clang. Hank Anderson 2015-02-25 14:44:49 -0600
  • 3649cfbd7b Fixed EPILOGUE for clang. Hank Anderson 2015-02-25 12:23:26 -0600
  • 5ae8993752 Added intrinsics for MSVC. Hank Anderson 2015-02-25 11:52:51 -0600
  • 84d90d6ed8 Fixed some compiler errors/warnings for clang. Hank Anderson 2015-02-25 11:52:25 -0600
  • 518e2424a8 Fixed bad filename for cpuid.S compile. Hank Anderson 2015-02-25 11:51:29 -0600
  • ea7f9dacf4 Refs #509. Fixed geadd building bug with DYNAMIC_ARCH=1. Zhang Xianyi 2015-02-26 01:47:11 +0800
  • bf5dbb7e2a Refs#509. Merge branch 'grisuthedragon-develop' into develop Zhang Xianyi 2015-02-26 01:44:19 +0800
  • 00e373aea6 Added LAPACK sources directly to add_library call instead of OBJECT. Hank Anderson 2015-02-25 10:18:18 -0600
  • 9eaea02f33 Added additional gemm defines for complex types. Hank Anderson 2015-02-25 09:39:11 -0600
  • ab7043373f Fixed bug generating trmv complex source names. Hank Anderson 2015-02-24 15:18:16 -0600
  • 504cdb10ed Added check for MSVC before enabling fortran. Hank Anderson 2015-02-24 14:31:45 -0600
  • a8002b0c5f Separated getarch ASM file when using MSVC. Hank Anderson 2015-02-24 14:31:18 -0600
  • 0553476fba Added TRANS defines for complex sources in lapack. Hank Anderson 2015-02-24 14:30:35 -0600
  • 2416d9dbac Fixed TRANSA defines for complex sources in driver/level2. Hank Anderson 2015-02-24 13:18:07 -0600
  • 0d8e227ea7 Changed strategy for setting preprocessor definitions. Hank Anderson 2015-02-24 12:26:33 -0600
  • 12d1fb2e40 Fixed incorrect object name in kernel CMakeLists.txt Hank Anderson 2015-02-24 10:30:16 -0600
  • 1b7f427401 Added conj gemv objects for complex build. Hank Anderson 2015-02-23 10:24:31 -0600
  • b2284647a3 More complex objects. Hank Anderson 2015-02-23 07:51:05 -0600
  • a6116e5859 Added some more complex-only objects. Hank Anderson 2015-02-22 17:49:28 -0600
  • fb5d5bb971 Added defines for complex trmv. Hank Anderson 2015-02-21 12:39:03 -0600
  • 371071d461 Added CONJ defines for trmm/trsm. Hank Anderson 2015-02-21 10:59:02 -0600
  • 8a143516e3 Added alternate_name to a couple of the name mangling schemes. Hank Anderson 2015-02-20 17:03:33 -0600
  • e5897ecb9b Added zherk_kernel.c objects to driver/level3. Hank Anderson 2015-02-19 16:19:56 -0600
  • 714638c187 Added some TRMM objects for complex types. Hank Anderson 2015-02-19 16:11:51 -0600
  • e27c372e53 Fixed reuse of float_char from parent loop. Hank Anderson 2015-02-19 13:53:29 -0600
  • f3f2b3d768 Added complex and single netlib-lapack fortran sources to lapack.cmake. Hank Anderson 2015-02-19 12:26:11 -0600
  • 9492298048 Added other float types to Makefile.L3. Hank Anderson 2015-02-18 13:01:05 -0600
  • 43725b82c5 ParseMakefileVars now replaces Makefile vars with CMake vars. Hank Anderson 2015-02-18 12:23:17 -0600
  • 14fd3d35de Added checks for missing defines in kernel. Hank Anderson 2015-02-18 10:25:01 -0600
  • cebc07cebd ParseMakefileVars now recursively parses included makefiles. Hank Anderson 2015-02-17 22:09:41 -0600
  • 33c5e8db7f Added a helper function for setting the L1 kernel defaults. Hank Anderson 2015-02-17 21:36:23 -0600
  • 67e39bd8fb Added mangled complex filenames to interface and lapack CMakeLists.txt. Hank Anderson 2015-02-17 13:12:30 -0600
  • 9eb1499095 Added another param to GenerateNamedObjects to mangle complex source names. Hank Anderson 2015-02-17 10:30:28 -0600
  • 39cc6b21d3 Add ATLAS-style ?geadd function Martin Koehler 2015-02-16 13:46:20 +0100
  • 4662a0b13a Changed generate functions to iterate through a list of float types. Hank Anderson 2015-02-15 17:44:37 -0600
  • e74462a3f5 Moved declarations to start of functions to satisfy MSVC C89 implementation. Hank Anderson 2015-02-11 11:16:57 -0600
  • 056ba26755 Changed a number of inline calls to use __inline. Hank Anderson 2015-02-11 11:13:17 -0600
  • a0d9a7fd83 Changed _Complex types in common_level1.h to use the typedef. Hank Anderson 2015-02-11 11:11:47 -0600
  • 5d3fc092e9 Added MSVC defines to common.h. Hank Anderson 2015-02-11 11:10:45 -0600
  • c94fe71278 Removed incoming-stack-boundary for MSVC. Hank Anderson 2015-02-11 10:54:14 -0600
  • d60b49e5c5 Turned off uninizialized variable warning when compiling lapack-netlib. Hank Anderson 2015-02-10 14:36:43 -0600
  • 64b5a0ef84 Added AUX files from lapack-netlib. Hank Anderson 2015-02-10 14:29:05 -0600
  • 162791e30e Added common objects from kernel Makefile. Hank Anderson 2015-02-10 12:42:05 -0600
  • 8743093bd7 Added aux files from lapack-netlib. Hank Anderson 2015-02-10 11:47:46 -0600
  • 96cf6779ca Added DLA sources from lapack-netlib. Hank Anderson 2015-02-10 11:01:01 -0600
  • 3b20b62423 Fixed trti2 name. Hank Anderson 2015-02-09 15:29:28 -0600
  • 6ddbfea700 Added generic laswp object. Hank Anderson 2015-02-09 15:15:58 -0600
  • c0624a26be Fixed some dgemm_copy function names. Hank Anderson 2015-02-09 14:34:29 -0600
  • 4bfaf1ce66 Removed some list appends I missed. Hank Anderson 2015-02-09 12:56:55 -0600
  • e8c39138c6 Removed return value from GenerateNamedObjects. Hank Anderson 2015-02-09 12:28:09 -0600
  • f992799226 Added the rest of Makefile.L3. Hank Anderson 2015-02-09 10:47:35 -0600
  • 4c65afcce1 Changed kernel filenames to vars. These will need to be read from KERNEL. Hank Anderson 2015-02-09 09:52:14 -0600
  • 7fa5c4e2fd Fixed some case issues with ARCH. Hank Anderson 2015-02-08 15:29:18 -0600
  • 771b18ae9c Detect the wrong combined flags of USE_OPENMP=1 and USE_THREAD=0. Zhang Xianyi 2015-02-08 01:42:48 -0600
  • cfa9392ffa Fix openblas_get_num_threads and openblas_get_num_procs bug with single thread. Zhang Xianyi 2015-02-08 01:30:12 -0600
  • fa0e6a6c93 Added the rest of the L1 kernel makefile. Hank Anderson 2015-02-07 21:37:46 -0600
  • 2f59135eb6 Added gemv to level2 CMakeLists.txt. Hank Anderson 2015-02-07 21:15:21 -0600
  • 38681fb1c6 Added more kernel files. Hank Anderson 2015-02-07 12:54:30 -0600
  • 6b5d26e07b Added SMP sources to level2 CMakeLists.txt. Hank Anderson 2015-02-06 16:52:19 -0600
  • 13d2d48e67 Added yet another naming scheme for lapack functions. Hank Anderson 2015-02-06 13:42:20 -0600
  • 189fadfde0 Started implementing kernel/Makefile in cmake. Hank Anderson 2015-02-05 21:05:11 -0600
  • 627d5e7401 Added SMP objects to driver/level3. Hank Anderson 2015-02-05 12:22:48 -0600
  • 943fa2fb58 Fixed object names in level2. Hank Anderson 2015-02-05 10:49:11 -0600
  • 1b62a4f3c9 Changed some function parameters to optional. Hank Anderson 2015-02-05 09:39:40 -0600
  • 461e691127 Codes when define is absent are now a parameter to AllCombinations. Hank Anderson 2015-02-05 09:23:47 -0600
  • cfaf1c678f Added option to append define codes with an underscore. Hank Anderson 2015-02-05 09:17:18 -0600
  • 0d7bad1f35 Changed GenerateObjects to append combination codes (e.g. dtrmm_TU). Hank Anderson 2015-02-05 09:02:54 -0600
  • 373a1bdadb Converted lapack/Makefile to cmake. Hank Anderson 2015-02-04 15:47:10 -0600