cheng/mpir
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
3,105 Commits 1 Branch 0 Tags 22 MiB
e508181a75
Commit Graph

1240 Commits

Author SHA1 Message Date
Alexander Kruppa
e508181a75 Version of mpn/x86_64/sandybridge/sub_n.as, super-optimized for Haswell 
New speed about 1.20c/l on Haswell, was 1.33c/l
 2016-11-28 19:43:46 +01:00
Brian Gladman
17b81f6006 add mpn_add_n and mpn_add_nc to x64 haswell build 2016-11-27 14:06:56 +00:00
Brian Gladman
d0d949835a Merge branch 'master' of https://github.com/akruppa/mpir 2016-11-27 10:56:14 +00:00
Alexander Kruppa
5d75ebc8bf Reduce number of registers used and use %defines for register names 2016-11-27 00:51:45 +01:00
Brian Gladman
d61bdcaf09 set release build of tests for foster linking 2016-11-26 22:52:26 +00:00
Brian Gladman
77b483e79f add more win64 assembler for haswell 2016-11-26 22:35:25 +00:00
Brian Gladman
a95556b926 Merge branch 'master' of https://github.com/akruppa/mpir 2016-11-26 18:35:20 +00:00
Brian Gladman
ee198165c9 prepare to add win64 assembler code with parameters in XMM/YMM registers 2016-11-26 09:41:07 +00:00
Alexander Kruppa
d11c3ca728 Bugfix: operand name macros were wrong 2016-11-25 18:11:38 +01:00
Alexander Kruppa
ea49db539e Revert "Temporarily removed due to bug" 
This reverts commit 38e8585c05.
 2016-11-25 18:11:21 +01:00
Brian Gladman
ed3aa00581 Merge branch 'master' of https://github.com/akruppa/mpir 2016-11-25 15:49:11 +00:00
Alexander Kruppa
38e8585c05 Temporarily removed due to bug 2016-11-25 15:27:21 +01:00
Alexander Kruppa
8100363a85 Version of mpn/x86_64/sandybridge/add_n.as, super-optimized for Haswell 
New speed about 1.21c/l on Haswell, was 1.33c/l
 2016-11-25 15:25:09 +01:00
Alexander Kruppa
f7f64a4ff2 Add missing colon 2016-11-25 14:55:31 +01:00
Alexander Kruppa
6316e39430 Increasing copy with AVX2 for Haswell 2016-11-25 11:51:54 +01:00
Alexander Kruppa
29577b5109 Decreasing copy with AVX2 for Haswell 2016-11-24 02:01:38 +01:00
Alexander Kruppa
4660be16f6 AVX-based rshift for 4-issue Intel cpus (Haswell and newer) 2016-11-22 23:18:52 +01:00
Alexander Kruppa
105c26c466 AVX-based lshift for 4-issue Intel cpus (Haswell and newer) 2016-11-22 21:58:43 +01:00
Alexander Kruppa
99a1f8d05b Add vzeroupper to avoid stall on Haswell if SSE2 code follows 2016-11-22 15:03:02 +01:00
Brian Gladman
0b48b8a843 add lshift1/rshiift1 to Visual Studio build 2016-11-18 22:03:24 +00:00
Alexander Kruppa
aa75752824 AVX-based lshift1 and rshift1 for 4-issue Intel cpus (Haswell and newer) 2016-11-18 21:54:07 +01:00
William Hart
8435273a1a Remove sb_div* small implementation (due to bug and due to being a very minor 
performance improvement).
 2015-11-13 14:47:44 +00:00
William Hart
0deca61277 Merge remote-tracking branch 'brian/master' 2015-11-13 13:06:17 +00:00
William Hart
a56b786513 Correct tuning values for x86/nehalem. 2015-10-03 12:29:33 +00:00
Brian Gladman
1d5c6f0508 correct bug in Windows assembler for k10/k102 2015-09-08 22:48:15 +01:00
William Hart
f541fce21f Removed slower sandybrige rshift1. 2015-06-12 11:27:42 +00:00
William Hart
de759a4525 Replace lahf and sahf with .byte declarations to support old coreutils such as 
that used by OpenBSD-5.6.
 2015-06-11 12:51:45 +00:00
William Hart
48bf24bdea Add missing rshift1.as/asm in Sandy Bridge. 2015-06-10 01:40:32 +00:00
William Hart
acd49c3818 INCLUDES -> AM_CPPFLAGS as per autoreconf advice. 2015-06-09 17:21:21 +00:00
William Hart
33c8ccc034 Ran autoreconf -fiv --no-recursive. 2015-06-09 15:50:53 +00:00
William Hart
240255708f Patch from Sage for fat binary build on 32 bit system. 2015-06-09 14:14:02 +00:00
William Hart
a6538ff19e Small improvement due to Paul Zimmermann. 2015-02-03 17:13:23 +00:00
William Hart
4a5686d5fb Added note about the revised version of Paul Zimmermann's article. 2015-02-03 16:59:31 +00:00
William Hart
24b7b6eab7 Small improvement, suggested by Paul Zimmermann. 2015-02-02 23:41:47 +00:00
William Hart
248e56c88d Removed unnecessary line of code. 2015-02-02 13:57:01 +00:00
William Hart
95d4eab65c Fixed bugs in invert.c and t-invert.c. The code was always computing correct 
results, but this seems to be more of an accident than by design.
 2015-02-02 12:30:50 +00:00
William Hart
a0e819627e Merge remote-tracking branch 'brian/master' 2014-10-01 13:11:30 +01:00
BrianGladman
277565fdb4 1. Correct error in last commit for mpir_config.py 
2. rebuild standard Windows builds
3. correct an error in win32 builds
 2014-08-07 22:07:24 +01:00
BrianGladman
9194070b24 remove faulty tuning files from the mpn\x86w assembler directories 2014-07-14 22:14:57 +01:00
William Hart
b5f8c19e58 Remove a broken assert in the generic build. 2014-06-29 21:05:59 +01:00
William Hart
f996e6343f Replace a gmp.h with mpir.h. 2014-06-26 14:33:32 +01:00
William Hart
bb15e496b6 Remove some compiler warnings. 2014-06-16 14:37:08 +01:00
William Hart
b67f620f08 Missing endline. 2014-06-16 14:19:48 +01:00
BrianGladman
d5f382c4f3 add the mpn_mul_fft function (needed by GMP-ECM) and update Windows builds accordingly 2014-06-12 14:34:20 +01:00
BrianGladman
206700ebe9 copy Atom tuning to Windows 2014-04-05 00:30:25 +01:00
William Hart
fedb3560cc Added 64 bit atom tuning values. Thank you JP Flori! 2014-04-04 21:40:13 +01:00
BrianGladman
84a0c677b1 Merge branch 'master' of https://github.com/wbhart/mpir 2014-04-01 08:24:12 +01:00
BrianGladman
e3e00705de add Ivybridge tuning on Windows 2014-04-01 08:23:35 +01:00
William Hart
af62ef2f1f Put netburst tuning back. 2014-04-01 01:13:01 +01:00
BrianGladman
053a6cd91a copy GCC tuning values over to Windows 2014-04-01 00:35:49 +01:00