2 ====================================================================
6 * Fixed a bug about installation. The header file "fblas77.h"
8 * Fixed #61 a building bug about setting TARGET and DYNAMIC_ARCH.
9 * Try to handle absolute path of shared library in OSX. (#57)
13 * Fixed #58 zdot/xdot SEGFAULT bug with GCC-4.6 on x86. According
14 to i386 calling convention, The callee should remove the first
15 hidden parameter.Thank Mr. John for this patch.
17 ====================================================================
22 * Added DTB_ENTRIES into dynamic arch setting parameters. Now,
23 it can read DTB_ENTRIES on runtime. (Refs issue #55 on github)
25 ====================================================================
30 * Fixed a building bug when DYNAMIC_ARCH=1 & INTERFACE64=1.
31 (Refs issue #44 on github)
33 ====================================================================
38 * Stop the build and output the error message when detecting
39 fortran compiler failed. (Refs issue #42 on github)
41 ====================================================================
46 * Fixed blasint undefined bug in <cblas.h> file. Other software
47 could include this header successfully(Refs issue #13 on github)
48 * Fixed the SEGFAULT bug on 64 cores. On SMP server, the number
49 of CPUs or cores should be less than or equal to 64.(Refs issue #14
51 * Support "void goto_set_num_threads(int num_threads)" and "void
52 openblas_set_num_threads(int num_threads)" when USE_OPENMP=1
53 * Added extern "C" to support C++. Thank Tasio for the patch(Refs
55 * Provided an error message when the arch is not supported.(Refs
57 * Fixed issue #23. Fixed a bug of f_check script about generating link flags.
58 * Added openblas_set_num_threads for Fortran.
59 * Fixed #25 a wrong result of rotmg.
60 * Fixed a bug about detecting underscore prefix in c_check.
61 * Print the wall time (cycles) with enabling FUNCTION_PROFILE
62 * Fixed #35 a build bug with NO_LAPACK=1 & DYNAMIC_ARCH=1
63 * Added install target. You can use "make install". (Refs #20)
67 * Fixed #28 a wrong result of dsdot on x86_64.
68 * Fixed #32 a SEGFAULT bug of zdotc with gcc-4.6.
69 * Fixed #33 ztrmm bug on Nehalem.
70 * Walk round #27 the low performance axpy issue with small imput size & multithreads.
73 * Fixed #28 a wrong result of dsdot on Loongson3A/MIPS64.
74 * Optimized single/double precision BLAS Level3 on Loongson3A/MIPS64. (Refs #2)
75 * Optimized single/double precision axpy function on Loongson3A/MIPS64. (Refs #3)
77 ====================================================================
82 * Support "make NO_LAPACK=1" to build the library without
84 * Fixed randomly SEGFAULT when nodemask==NULL with above Linux 2.6.34.
85 Thank Mr.Ei-ji Nakama providing this patch. (Refs issue #12 on github)
86 * Added DEBUG=1 rule in Makefile.rule to build debug version.
87 * Disable compiling quad precision in reference BLAS library(netlib BLAS).
88 * Added unit testcases in utest/ subdir. Used CUnit framework.
89 * Supported OPENBLAS_* & GOTO_* environment variables (Pleas see README)
90 * Imported GotoBLAS2 1.13 BSD version
93 * On x86 32bits, fixed a bug in zdot_sse2.S line 191. This would casue
94 zdotu & zdotc failures.Instead,Walk around it. (Refs issue #8 #9 on github)
95 * Modified ?axpy functions to return same netlib BLAS results
96 when incx==0 or incy==0 (Refs issue #7 on github)
97 * Modified ?swap functions to return same netlib BLAS results
98 when incx==0 or incy==0 (Refs issue #6 on github)
99 * Modified ?rot functions to return same netlib BLAS results
100 when incx==0 or incy==0 (Refs issue #4 on github)
101 * Detect Intel Westmere,Intel Clarkdale and Intel Arrandale
102 to use Nehalem codes.
103 * Fixed a typo bug about compiling dynamic ARCH library.
105 * Improve daxpy performance on ICT Loongson 3A.
106 * Supported ICT Loongson 3A CPU (Refs issue #1 on github)
107 ====================================================================