amd/blis
1
0
Fork 0
mirror of https://github.com/amd/blis.git synced 2026-05-13 18:52:14 +00:00
Code Issues Packages Projects Releases Wiki Activity
617 Commits 10 Branches 88 Tags
49f85177f886f38889b60503a4e12fa7f04be1fd
Commit Graph

617 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Devin Matthews
49f85177f8 KNL ukernel compiles with gcc. 2016-04-18 10:14:11 -05:00
Devin Matthews
58b2c3cf04 Rewrite of KNL kernel in GNU extended asm syntax. 2016-04-16 16:12:24 -05:00
Devin Matthews
dd856c2cb7 Translated MIC kernel to KNL and cleaned up a bit. Only real change is lack of swizzle modifiers for FMA instructions (used bcast from memory instead). 2016-04-11 10:39:18 -05:00
Devin Matthews
7f27431d3f Copy mic kernel to knl for transliteration. 2016-04-08 10:04:39 -05:00
Devin Matthews
f8f02f0334 Merge branch 'master' into const_correctness 2016-04-06 11:37:05 -05:00
Devin Matthews
32c92d945c Merge branch 'master' into const_correctness 2016-04-06 11:36:02 -05:00
Field G. Van Zee
d1f8e5d9b2 Merge pull request #60 from esauvage/master 
sgemm µkernel for bulldozer : bug correction for k%4 != 0
 2016-04-05 12:21:27 -05:00
Etienne Sauvage
c11d28eed8 cgemm µkernel for bulldozer : bug correction for k%4 != 0 2016-04-02 21:15:48 +02:00
Field G. Van Zee
20af937b57 Merge pull request #59 from devinamatthews/fix_testsuite_makefile 
Fix testsuite makefile
 2016-03-31 14:37:30 -05:00
Devin Matthews
fc61a1143e Fix formatting in configure. 2016-03-31 10:53:01 -05:00
Devin Matthews
26379b14de Adjust paths in common.mk to support building from testsuite dir. 2016-03-31 10:45:48 -05:00
Field G. Van Zee
36c3abb05f Merge pull request #58 from esauvage/master 
cgemm & zgemm micro-kernels for FMA4 instruction set (bulldozer confi…
 2016-03-31 10:26:17 -05:00
Devin Matthews
356d854fc9 Make symlink to common.mk in build directory. 2016-03-30 16:33:15 -05:00
Devin Matthews
edbb847004 Refactor out some definitions which moved from make_defs.mk to Makefile for use in testsuite Makefile. 2016-03-30 16:27:11 -05:00
Etienne Sauvage
917ce75482 cgemm & zgemm micro-kernels for FMA4 instruction set (bulldozer configuration), based on x86_64/avx micro-kernel 2016-03-30 22:03:09 +02:00
Devin Matthews
62914ccbcd Merge branch 'master' into const_correctness 2016-03-29 15:24:25 -05:00
Field G. Van Zee
64b41fa554 Merge pull request #54 from devinamatthews/more_config_opts 
More config opts
 2016-03-29 15:19:41 -05:00
Field G. Van Zee
1b09e343df Updated gcc version from 4.8 to 4.9 in .travis.yml. 2016-03-29 12:55:28 -05:00
Devin Matthews
0171ad5899 Add icc and clang support for Intel architectures, fixes #47. 2bd036f fixes #49 BTW. 2016-03-28 13:55:06 -05:00
Field G. Van Zee
3090fff64c Merge pull request #44 from esauvage/master 
sgemm micro-kernel for FMA4 instruction set
 2016-03-28 12:36:25 -05:00
Devin Matthews
e6e566426a Merge branch 'master' into more_config_opts 2016-03-26 14:10:15 -05:00
Field G. Van Zee
8624e36543 Merge pull request #50 from devinamatthews/fix_noopt_avx 
Fix configuration issue where instruction set flags are not specified for debug builds.
 2016-03-26 13:56:28 -05:00
Devin Matthews
469429ec34 Fix LD_FLAGS -> LDFLAGS. 2016-03-25 20:45:41 -05:00
Devin Matthews
8442d65c9e Replace -march=native with specific architecture flags to support cross-compiling, and add icc support for Intel architectures. 2016-03-25 20:06:48 -05:00
Devin Matthews
76099f20be Add threading option to configure. 2016-03-25 17:22:58 -05:00
Devin Matthews
ad43eab4c7 Merge branch 'fix_noopt_avx' into more_config_opts 2016-03-25 15:00:02 -05:00
Devin Matthews
9452bdb3af Add options for verbose make output and static/shared linking to configure. 2016-03-25 14:59:50 -05:00
Devin Matthews
2bd036f1f9 Fix configuration issue where instruction set flags are not specified for debug builds. 2016-03-25 12:16:49 -05:00
Devin Matthews
bbf704bf75 Add missing const to bli_read_nway_from_env. 2016-03-25 09:55:35 -05:00
Field G. Van Zee
a315833f06 Merge pull request #48 from figual/master 
Updated and improved ARMv8 micro-kernels.
 2016-03-24 12:30:21 -05:00
figual
af92773f4f Updated and improved ARMv8 micro-kernels. 2016-03-23 22:07:02 +01:00
Devin Matthews
a4d7729776 Set default value for debug_type variable. 2016-03-21 09:55:21 -05:00
Devin Matthews
0e2447fa55 Add const correctness to auxinfo_t struct (microkernels need update theoretically). 2016-03-17 16:32:05 -05:00
Field G. Van Zee
1d1a426d18 Merge pull request #46 from devinamatthews/new-config-opts 
Add several changes to the build system.
 2016-03-07 15:17:53 -06:00
Devin Matthews
d226dfa051 Add several changes to the build system. 
1) Add -- options.
2) Add -d/--enable-debug option to enable debugging symbols with and without optimization.
3) Allow user to specify CC at configure time, and determine vendor (gcc/icc/etc.). For now configurations enforce a particular vendor.
4) Add make V=[0,1] option to control build verbosity.
 2016-03-05 16:18:14 -06:00
Field G. Van Zee
5a978fffdb Merge pull request #45 from devinamatthews/high_prec_timers 
Use clock_gettime(CLOCK_MONOTONIC) and mach_absolute_time instead of gettimeofday
 2016-03-04 17:26:58 -06:00
Devin Matthews
63e2642390 Make sure that -lrt is linked on Linux. 2016-03-04 13:17:50 -06:00
Devin Matthews
44fddd48dc Add missing \. 2016-03-04 12:36:38 -06:00
Devin Matthews
7cabd2131f Use clock_gettime(CLOCK_MONOTONIC) and mach_absolute_time instead of gettimeofday. 2016-03-03 11:43:07 -06:00
Tyler Smith
adb2b4e096 Fixing guard for non implemented partitioning through packed matrices 2016-03-02 14:48:12 -06:00
Etienne Sauvage
4ca5d5b1fd sgemm micro-kernel for FMA4 instruction set (bulldozer configuration), based on x86_64/avx micro-kernel 2016-03-01 21:33:01 +01:00
Etienne Sauvage
627d59b5ba symbolic link for bulldozer configuration to kernels baseline 2016-02-29 21:53:12 +01:00
Field G. Van Zee
2dc5c0ae03 Merge pull request #40 from tkelman/bulldozer-symlink 
Add symlink from config/bulldozer/kernels to kernels/x86_64/bulldozer
 2016-02-29 12:22:51 -06:00
Field G. Van Zee
f2809fc5f7 Merge pull request #39 from devinamatthews/fix_f2c_conflicts 
Devin's f2c type namespace update.

Details:
- Added "bla_" prefix to f2c type names to prevent conflicts with external user code.
- Removed most of the body of bli_f2c.h, which was unused.
 2016-02-27 13:06:03 -06:00
Tony Kelman
3d0fae810d Add symlink from config/bulldozer/kernels to kernels/x86_64/bulldozer 
to fix linking issue mentioned in #37 and https://groups.google.com/forum/#!topic/blis-devel/iypwljcaeEI
 2016-02-25 23:24:03 -08:00
Devin Matthews
8624a33ccc Fix remaining f2c conflicts. 2016-02-25 13:51:26 -06:00
Devin Matthews
372eef0b6c Fixed most conflicts after hack-n-slash ofr bli_f2c.h, cleanup in 
progress.
 2016-02-25 12:01:58 -06:00
Field G. Van Zee
f86b94f206 Included missing blas2blis integer def to CBLAS. 
Details:
- Added #include "bli_config_macro_defs" to all cblas_*.c files in
  compat/cblas/src. This has the effect of defining
  BLIS_BLAS2BLIS_INT_TYPE_SIZE to the default value if bli_config.h does
  not define it. Thanks to Tony Kelman for reporting this bug.
- In cblas_i?amax.c, changed the type of the variable 'iamax' from 'int'
  to 'f77_int'. This eliminates a compiler warning and a potential
  runtime bug and/or crash when the size of an int differs from the size
  of f77_int (as determined by BLIS_BLAS2BLIS_INT_TYPE_SIZE).
 2016-02-23 18:12:34 -06:00
Field G. Van Zee
0b126de134 Consolidated packm_blk_var1 and packm_blk_var2. 
Details:
- Consolidated the two blocked variants for packm into a single
  implementation (packm_blk_var1) and removed the other variant.
- Updated all induced method _cntl_init() functions in frame/cntl/ind/
  to use the new blocked variant 1.
- Defined two new macros, bli_is_ind_packed() and bli_is_nat_packed(),
  to detect pack_t schemas for induced methods and native execution,
  respectively.
 2015-11-13 16:29:12 -06:00
Field G. Van Zee
30e5eb29e0 Minor changes to treatment of rs, cs in bli_obj.c. 
Details:
- Applied a patch submitted by Devin Matthews that:
  - implements subtle changes to handling of somewhat unusual cases of
    row and column strides to accommodate certail tensor cases, which
    includes adding dimension parameters to _is_col_tilted() and
    _is_row_tilted() macros,
  - simplifies how buffers are sized when requested BLIS-allocated
    objects,
  - re-consolidates bli_adjust_strides_*() into one function, and
  - defines 'restrict' keyword as a "nothing" macro for C++ and pre-C99
    environments.
 2015-11-13 12:14:19 -06:00