Commit Graph

1194 Commits

Author SHA1 Message Date
wbhart
1350c07091 Hacked fft tuning code in to tuneup. Also added tuning values for k102. 2012-10-17 21:18:28 +00:00
gladman
2208dcbe51 remove multiple definition warning in win32 build 2012-10-09 07:25:35 +00:00
wbhart
b0f3391568 Fixed integer overflow warning in gcdext.c. 2012-09-28 15:57:37 +00:00
gladman
b3c032a269 add some missing win32 assembler files 2012-09-28 11:36:44 +00:00
gladman
a5d6cad0b0 remove some Visual Studio build files wrongly added to the repository 2012-09-28 08:23:15 +00:00
gladman
d981389b86 change mpn_invert_truncate to mpn_invert_trunc for naming compatibility 2012-09-27 21:12:13 +00:00
gladman
92a6cdf4bf update Visual Studio build files for file name changes 2012-09-27 20:40:31 +00:00
gladman
e8680c1ed9 revert invert_trunc back to invert_truncate 2012-09-27 20:16:48 +00:00
wbhart
05c51ddd91 Removed old FFT tuning params from gmp-mparam.h files. 2012-09-25 15:43:18 +00:00
gladman
5205d30dda Update to batch file build for MSVC provided by Case
Remove macro bug found by Case
2012-09-05 14:11:27 +00:00
gladman
4fc4fe0884 merge mpir-exp branch into trunk 2012-07-27 13:26:23 +00:00
gladman
c67430f35f commit changes to remove conflicts after merge from trunk 2012-07-27 08:16:21 +00:00
gladman
685a5726fd add some trunk changes to exp branch 2012-07-19 19:52:41 +00:00
gladman
1bd41a7b27 add minor changes from trunk to Windows x64 assembler files 2012-07-19 19:42:18 +00:00
gladman
f629c5ce4f convert mpn_mulmod_2expp1 to mpn_mulmod_Bexpp1 and remove test of the original version 2012-07-19 13:01:13 +00:00
gladman
2c05d7c98b complete new FFT and strip out old one 2012-07-18 11:02:59 +00:00
wbhart
b50f939a74 Added missing mpn_mul_fft function (now mpn_mulmod_Bexpp1_fft) and used it. 2012-07-17 21:33:24 +00:00
gladman
0e06c37a20 Add recent bug fixes to the FFT code 2012-07-14 14:39:23 +00:00
gladman
f790a80397 remove a few duplicate PROLOGUE lines 2012-07-12 12:47:59 +00:00
jasonmoxham
5241475f1a make sure x86_64w asm files are in dos format line feeds 2012-07-08 14:05:43 +00:00
jasonmoxham
5a9db35787 change global func to prologue 2012-07-08 13:56:43 +00:00
jasonmoxham
1128c43da4 add GLOBAL_FUNC to all x86_64w asm files 2012-06-27 04:35:20 +00:00
gladman
f43cb59137 further improvements to Windows build generator
allow generic mpn_sumdiff_n to work with zero limbs
2012-06-11 20:31:25 +00:00
gladman
6efedb9b50 Add latest bugfix to FFT code 2012-06-05 14:09:03 +00:00
gladman
8d35e726c2 2012-05-09 07:55:05 +00:00
gladman
6d1a7e93af rebuild all Visual Studio builds to remove win32 build errors 2012-03-29 11:42:49 +00:00
gladman
8d2d920902 correct 32-bit build error reporteed by Case 2012-03-19 09:46:25 +00:00
jasonmoxham
97f6f431fd asm redc_1 for sandybridge and bobcat for x86_64 and x86_64w 2012-03-19 07:10:43 +00:00
jasonmoxham
854897bb3c correct header file in redc_2 2012-03-13 07:21:30 +00:00
jasonmoxham
3eb1475068 correct assert in sumdiff 2012-03-11 11:11:31 +00:00
gladman
87b591f284 minor changes to add comments for symbols defined in macros 2012-03-10 21:55:33 +00:00
gladman
e4f7ccacc1 minor changes to add comments for symbols defined in macros 2012-03-10 21:54:15 +00:00
gladman
b5b117c93e add latest trunk changes to exp branch 2012-03-10 09:38:09 +00:00
jasonmoxham
03d14b7e45 correct pop order for karasub , add redc_2 , add generic addadd addsub subadd sumdiff 2012-03-10 08:27:37 +00:00
gladman
89fb944670 add bug fixes from trunk 2012-02-18 12:26:54 +00:00
jasonmoxham
00deb1f830 karasub for windows 2012-02-18 11:40:28 +00:00
jasonmoxham
935c0158ef karasub for all x64 2012-02-18 09:15:44 +00:00
jasonmoxham
b1d11d499b gmp-5.0.3 to gmp-5.0.4 bug fixed 2012-02-18 07:22:11 +00:00
gladman
49cbdf7a0b new karasub C code and a Windows postbuild tweak 2012-02-14 21:14:56 +00:00
jasonmoxham
22e1c60ce6 correct generic karasub for carry/borrow propagation 2012-02-14 20:06:42 +00:00
gladman
e572b67961 correct prebuild step 2012-02-13 17:27:02 +00:00
gladman
406c8e8b24 prepare for new Python build system 2012-02-13 15:18:13 +00:00
jasonmoxham
a9314ae1e5 correct k8 karasub 2012-02-13 08:54:17 +00:00
gladman
a30ee5b4ea update win32 builds 2012-02-12 23:40:14 +00:00
gladman
9f7e38798b Correct the Windows kara_sub assembler implementation. 2012-02-03 19:45:18 +00:00
gladman
5a508593a5 adopt Jason's six loop solution for carries in karasub assembler 2012-01-31 19:38:55 +00:00
gladman
67645bface correction to karasub assembler documentation 2012-01-28 09:08:54 +00:00
gladman
d1d2b839fe small change to Windows karasub asseembler 2012-01-27 22:51:07 +00:00
gladman
78e52a0525 improve documentation in karasub assembler 2012-01-27 16:25:51 +00:00
gladman
340449f492 correct the Windows versions of karasub assembler 2012-01-27 12:49:09 +00:00
gladman
7b6cec91a5 karasub documentation correction 2012-01-26 20:59:30 +00:00
gladman
ceb36b722d minor documentation update for karasub assembler on Windows 2012-01-26 18:33:47 +00:00
gladman
e625d10b9c correct bug in WIndows assembler code for karasub 2012-01-26 18:12:57 +00:00
gladman
fb616eed78 Change gmp_ui/gmp_si types to mpir_ui/mpir_si 2012-01-24 14:37:16 +00:00
gladman
f8048b4818 minor updates to FFT code 2012-01-21 17:32:30 +00:00
jasonmoxham
467c6db8d4 update makefiles and configuree for fft dir 2012-01-20 00:11:55 +00:00
jasonmoxham
30467ad7b7 change backslash to forward slash in fft_tuning.h 2012-01-19 23:54:20 +00:00
gladman
6880d37a07 set zero for output when bits = 0 2012-01-17 17:49:00 +00:00
gladman
7c55450bb2 correct call to old FFT from new code (for timing old FFT only) 2012-01-17 16:53:33 +00:00
gladman
98b7fc8c66 Bring FFT calling interfaces into line with GMP/MPIR conventions
Add const to declarations where necessary
2012-01-17 15:13:24 +00:00
gladman
af7be56a68 correct a bug in the new FFT code 2012-01-16 08:42:34 +00:00
gladman
2d7ae4b37f add missing file to SVN 2012-01-15 20:27:12 +00:00
gladman
1253dff171 Final integration of Bills new FFT code 2012-01-10 16:26:07 +00:00
gladman
e1ba31c102 2012-01-07 15:05:27 +00:00
gladman
4ed5a54b38 put FFT mpn_ files back in FFT directory 2012-01-06 13:38:56 +00:00
gladman
37d4f0ff5c update the FFT multiplication wrapper 2012-01-06 08:22:53 +00:00
jasonmoxham
86ab6589b3 generic addadd addsub and subadd 2012-01-05 23:06:24 +00:00
gladman
4d29441935 2012-01-04 23:48:12 +00:00
jasonmoxham
01e488b38b manual fix powerpc64/gmp-mparam mulhigh param to be>=3 2011-12-21 17:57:15 +00:00
gladman
338e5f2984 populate new branch 2011-12-16 19:47:56 +00:00
jasonmoxham
06779788c2 manual fix sparc32/v9/mulhigh param to be >=3 2011-12-13 17:57:04 +00:00
jasonmoxham
97d24b5ad2 crlf removal 2011-12-05 07:07:44 +00:00
jasonmoxham
dc5dc05709 powerpc64 tuneup 2011-12-04 10:37:18 +00:00
gladman
d090b3e7c5 copy Jason's latest tuning to Windows 2011-12-03 23:27:11 +00:00
jasonmoxham
46457ff76b core2/penryn tuning 2011-12-03 18:33:44 +00:00
jasonmoxham
d270f8e0a7 k8/k10/k102 tuning and copy to k8/k10 tuning 2011-12-03 17:33:10 +00:00
jasonmoxham
8ccae24c87 sparc32/v9 tuning 2011-12-03 17:06:34 +00:00
jasonmoxham
1487ddca4a sparc64 tuning 2011-12-03 15:15:19 +00:00
gladman
6323fbfb28 copy Jasons tuning to Windows 2011-11-23 22:15:43 +00:00
jasonmoxham
06d3e05d93 ia64 tuning 2011-11-23 15:02:55 +00:00
jasonmoxham
87e6fb66cf pentium4 tuning 2011-11-23 13:48:18 +00:00
jasonmoxham
47760fd81e core2 tuning 2011-11-23 12:21:42 +00:00
jasonmoxham
0f430bfca9 netburst tuning 2011-11-23 11:11:37 +00:00
gladman
833b850d8f Copy latest tuning to Windows builds 2011-11-22 16:14:06 +00:00
jasonmoxham
a047b009a7 k8 tuning 2011-11-22 10:57:58 +00:00
jasonmoxham
50fb23fd6e nehalem tuning 2011-11-22 09:41:10 +00:00
jasonmoxham
5c2f5ad253 bobcat tuning 2011-11-22 09:00:31 +00:00
jasonmoxham
3251323495 atom tuning 2011-11-22 08:33:46 +00:00
jasonmoxham
6bdb901899 sandybridge tuning 2011-11-22 07:48:41 +00:00
gladman
8db4ca244c 1. Correct errors in Visual Studio builds
2. add tow missing Windows assembler files to SVN
2011-10-14 14:26:09 +00:00
gladman
b3efc4e4dc commit new GCD code to trunk 2011-10-14 11:22:46 +00:00
jasonmoxham
f16f543573 bump version and so numbers ready for release 2011-09-27 23:04:16 +00:00
jasonmoxham
65b1ec121a removed all use HAVE_HOST , except for timing on the powerpc 2011-09-12 01:23:43 +00:00
jasonmoxham
c0f53686c5 remove all references to GMP_IEEE_FLOATS , thats the only type we support anyway 2011-09-11 19:04:40 +00:00
jasonmoxham
d09a1e4dd1 removed all have_host_x86 etc except for build,vc10 dir 2011-09-11 18:48:44 +00:00
jasonmoxham
9f34508f50 correct rounding mode tests and generalize, remove some more have_host's 2011-09-11 18:41:46 +00:00
jasonmoxham
0490047f68 disable yasm for *.asm and use gas as some functions are slightly faster 2011-09-11 13:38:18 +00:00
jasonmoxham
4c22eae4ed remove more have_host_cpu_* 2011-08-30 23:57:26 +00:00
jasonmoxham
f07959acd0 remove referances to some non existent paths 2011-08-30 22:06:14 +00:00
jasonmoxham
f585f30def remove some old powerpc32 minor optimiztions 2011-08-30 21:28:48 +00:00
jasonmoxham
5a21c6ff6a remove rest of special alpha code , very minor slowdowns 2011-08-30 20:56:10 +00:00
jasonmoxham
b503fcced2 remove pointless differences between old cpus ie k6's 2011-08-30 20:41:29 +00:00
jasonmoxham
1cf83265ac remove use of FORCE_DOUBLE ie 80bit floats which are slower now anyway 2011-08-17 15:27:10 +00:00
jasonmoxham
7c6376758a remove some HAVE_HOST_CPU_* and some x86 macro's in then x64 dirs 2011-08-17 15:09:34 +00:00
jasonmoxham
a6c2480527 removed some alpha specific stuff 2011-08-10 15:54:48 +00:00
jasonmoxham
fedabb4c9a replace old add_n_sub_n by sumdiff 2011-07-24 19:19:58 +00:00
jasonmoxham
1f133530f1 write new toom eval for +-1 using addadd and sumdiff 2011-07-24 19:09:36 +00:00
gladman
bb009d186b Add sumdiff_n for core2 and penryn and mod_1_1 for penryn on Windows 2011-07-23 15:57:46 +00:00
jasonmoxham
4a02d0ca94 New asm fn mpn_sumdiff_n for penryn 2011-07-23 14:32:28 +00:00
jasonmoxham
19b89f6bda New asm mpn_sumdiff_n for core2 2011-07-23 13:16:30 +00:00
gladman
01eba40795 add mpn_sumdiff_n for nehalem on Windows 2011-07-22 23:03:59 +00:00
jasonmoxham
30e7a89da6 New mpn_sumdiff_n for nehalem 2011-07-22 21:47:35 +00:00
gladman
427dead3d1 Update latest core2 assembler on Windows 2011-07-22 12:22:35 +00:00
gladman
c42eb996ca update latest assembler for nehalem on Windows 2011-07-22 12:13:14 +00:00
jasonmoxham
210aa44bfb New addadd addsub subadd for core2/penryn , note same as nehalem code 2011-07-22 11:34:39 +00:00
jasonmoxham
1a7e5f137f New nehalem asm functions for addadd addsub subadd 2011-07-22 10:07:54 +00:00
jasonmoxham
ecc41bd176 Use new half and double code 2011-07-20 23:32:23 +00:00
gladman
4d114dca5d add double and half assembler to the Windows builds 2011-07-13 13:52:06 +00:00
jasonmoxham
f1ec8a275e add half/double test code and correct return values 2011-07-13 12:54:47 +00:00
gladman
fc35327f0d add latest nehalem code to Windows build 2011-07-13 09:50:14 +00:00
jasonmoxham
5955fc2424 mpn_half mpn_double asm for K8 2011-07-13 09:48:02 +00:00
jasonmoxham
94cc9b7a7b karasub/add for nehalem 2011-07-13 08:35:13 +00:00
gladman
9171d38e78 1. rename umul.asm and umul.asm to align with the symbols they export
2. minor Windows build tweaks
2011-07-13 06:37:51 +00:00
gladman
8546be98a1 change labels from global to local in latest Windows assembler code 2011-07-07 13:48:55 +00:00
jasonmoxham
5825857cc6 remove old toom22 code 2011-07-06 19:42:16 +00:00
jasonmoxham
9caa565e90 remove thumb,m68k,sh cpu's 2011-07-06 00:26:36 +00:00
gladman
36a0e7e247 add latest assembler code to nehalem and core2 builds 2011-07-05 21:18:30 +00:00
jasonmoxham
466614556f copy k8 and karaadd/sub to the other cpu arches linux and windows 2011-07-05 20:16:41 +00:00
jasonmoxham
fd650a98cf karaadd/sub for atom , same as K8 2011-07-05 14:17:51 +00:00
jasonmoxham
0cdb4fe24b karaadd/sub for sandybridge , same as K10 2011-07-05 14:08:01 +00:00
jasonmoxham
bdfb643048 karaadd/sub for bobcat , same as K10 2011-07-05 14:07:31 +00:00
gladman
6cfea75175 add karaadd and karasub assembler code to k8 and k10 Windows builds 2011-07-05 13:10:25 +00:00
jasonmoxham
0573a58d96 tweek karaadd/sub , add k10 versions 2011-07-05 11:20:40 +00:00
jasonmoxham
543331fc68 karasub change mirrored to karaadd 2011-07-04 20:41:38 +00:00
jasonmoxham
9638ffae9d karasub fixed rcx simplifications 2011-07-04 20:34:42 +00:00
jasonmoxham
94cd105050 karasub swap rcx,rdx save one inst 2011-07-04 20:13:45 +00:00
jasonmoxham
73e2007369 add new karaadd/sub to speed 2011-07-04 18:14:25 +00:00
jasonmoxham
96bad6323d removr 2.3 branch 2011-07-04 16:53:52 +00:00
jasonmoxham
1572383093 modify kara_sqr for use karasub 2011-07-04 08:25:21 +00:00
jasonmoxham
38246bb348 kara tweek 2011-07-04 05:55:43 +00:00
jasonmoxham
a962f39ee1 karaadd karasub and toom22 2011-07-04 04:34:15 +00:00
jasonmoxham
5868f23cdd switch all x86_64 asm code from gas/yasm to just yasm 2011-06-30 13:48:51 +00:00
jasonmoxham
7e1657b342 x86_64w fat build also uses rip relative addressing 2011-06-30 13:39:11 +00:00
jasonmoxham
7e4a9913db x86_64 fat build uses rip relative addressing 2011-06-30 13:03:08 +00:00
jasonmoxham
fdd1dc256f CRLF for all *.h and *.c files not in build.vc10 or yasm 2011-06-08 10:46:43 +00:00
jasonmoxham
6077fd9b98 CRLF to LF on all c files except in build.vc10 dir 2011-06-08 07:15:12 +00:00
jasonmoxham
0104a83deb delete inc/declsh_n entry point from sandybridge add/submul_1 as in mingw64 fat build the symbols clash 2011-06-07 09:01:59 +00:00
jasonmoxham
13a63a7ef1 and rename symbol in sandybridge/mul_1 2011-06-07 09:00:24 +00:00
jasonmoxham
327f3f585a unique symbols requires for miongw64 fat build 2011-06-07 07:14:56 +00:00
jasonmoxham
fd66c7d027 bobcat gmp-mparams 2011-06-05 06:24:47 +00:00
jasonmoxham
13813b3901 sandybridge gmp-mparam.h 2011-06-05 06:11:01 +00:00
jasonmoxham
3746f2faa2 fixed useless warning with yasm 32msvc signed constant out of range 2011-05-08 22:55:00 +00:00
jasonmoxham
225cdd47cd comment out un-needed header files 2011-05-08 22:38:55 +00:00
gladman
f19cade86e correct windows byte swap intrinsics 2011-05-08 16:43:04 +00:00
jasonmoxham
bdad56d27e remove old asm x86w/p6/sqr_basecase.old.asm 2011-05-08 16:13:05 +00:00
jasonmoxham
e960733ff1 remove old code 2011-05-01 12:30:23 +00:00
gladman
a7a7c79be7 add bswap limb definitions to MSVC longlong_inc.h versions 2011-04-30 08:47:19 +00:00
jasonmoxham
3066b516d5 remove s390 pa32 pa64 explicit code 2011-04-30 07:16:39 +00:00
jasonmoxham
e4313f9a13 move asm code in gmp-impl into the arch specific dirs 2011-04-30 07:05:19 +00:00
jasonmoxham
1b0ca8d6f8 include some extra ifdef around inline asm 2011-04-29 10:26:41 +00:00
jasonmoxham
8e655622cf make x86 fat preinv_* general , no more bloody special cases..... 2011-04-29 07:46:44 +00:00
jasonmoxham
c8f630b6fb add missing fft params for x86 fat build 2011-04-29 07:16:21 +00:00
gladman
d3f4f67aaf correct name of longlong.h in mpn\x86_64w\ to longlong_inc.h 2011-04-28 16:03:31 +00:00
gladman
cf2a5ef3c7 make changes for building longlong.h 2011-04-28 09:34:30 +00:00
jasonmoxham
deb7d231da add bulldozer and sandybridge paths to the x86_64w fat path 2011-04-28 02:26:50 +00:00
jasonmoxham
96819f622b final bits of splitting longlong.h 2011-04-28 00:27:14 +00:00
jasonmoxham
1bb005f714 split longlong into arches , windows and possibly a bit of ifdef wrapping to do 2011-04-27 22:59:10 +00:00
jasonmoxham
33eaf3ddab add new bulldozer paths to the fat system 2011-04-27 22:51:30 +00:00
jasonmoxham
7184d3f5de add new sandybridge path to the fat system 2011-04-27 22:49:54 +00:00
jasonmoxham
e7b0887cc2 add in paths and the best asm we've got for sandybridge 2011-04-26 21:18:12 +00:00
jasonmoxham
996f7010b8 movq to movd in asm 2011-04-16 16:55:00 +00:00
jasonmoxham
28fea4c7af remove define UMUL_TIME SMUL_TIME as never used 2011-04-13 08:39:59 +00:00
jasonmoxham
d4995d0370 remove define UDIV_TIME as never used 2011-04-13 08:32:07 +00:00
jasonmoxham
059ea30181 split out windows longlong parts 2011-04-11 07:41:43 +00:00
jasonmoxham
08dc99d709 put license info in all the new longlong's 2011-04-10 23:42:50 +00:00
jasonmoxham
a18af002d1 split up longlong into arches 2011-04-10 23:06:16 +00:00
jasonmoxham
84ccff258b and for x64w 2011-04-07 14:22:50 +00:00
jasonmoxham
3ea0bbfc8b and for x64w 2011-04-07 14:21:29 +00:00
jasonmoxham
59dc133875 and for x64w 2011-04-07 14:21:03 +00:00
jasonmoxham
3f15c30522 make sure fat dir has every file in fat structure 2011-04-07 14:19:31 +00:00
jasonmoxham
99af8524fb copy some k8 files to x64 dir so that fat build on unknown cpu's work 2011-04-04 23:01:32 +00:00
jasonmoxham
75cc63f4c7 delete , for now k8's udiv.asm and umul.asm 2011-04-04 16:36:07 +00:00
jasonmoxham
0af26b5bf4 update cpuid.c to be a function rather than a code fragment 2011-04-04 00:40:39 +00:00
jasonmoxham
f2d19d0b29 delete empty dirs 2011-04-03 16:31:49 +00:00
jasonmoxham
4a10577a68 netburst and netburstlahf cpus now use the same directory 2011-04-03 15:59:49 +00:00
jasonmoxham
dd4ba4af65 nano uses the k8 path for assembler 2011-04-03 15:54:22 +00:00
jasonmoxham
170440861d identify k103 as k103 , not as just another k102 , but no separate path for it yet 2011-04-03 15:22:36 +00:00
jasonmoxham
acf341adde remove cpuid_availible function from fat detection 2011-04-03 09:00:11 +00:00
jasonmoxham
7c463d96f5 clean up fat stuff 2011-04-03 08:50:09 +00:00
jasonmoxham
bacdc85410 disable yasm from fat builds , temporary 2011-04-03 08:49:53 +00:00
jasonmoxham
57d37eea1d get yasm to assemble intel format asm code 2011-03-18 16:39:37 +00:00
jasonmoxham
9ead216342 autoreconf -f with the latest autotools automake 1.11.1 autoconf 2.68 libtool 2.4 2011-03-05 16:47:46 +00:00
jasonmoxham
2b4f1626bc bobcat params 2011-03-04 14:56:09 +00:00
jasonmoxham
d4b0827cc2 asm basecases for bobcat from k8 for linux and windows 2011-03-04 14:31:05 +00:00
jasonmoxham
4e6496d3bb select from existing dir's fastest functions for bobcat linux and windows 2011-03-04 11:55:38 +00:00
jasonmoxham
13009502e8 New asm directory for bobcat 2011-03-04 08:38:40 +00:00
gladman
2733006356 2011-02-28 22:52:56 +00:00
jasonmoxham
cd09dcf7c0 autotools 2011-02-23 08:34:08 +00:00
jasonmoxham
0f87f64b65 update version numbers to 2.3.0-RC1 2011-02-23 00:00:48 +00:00
jasonmoxham
7f6602a57f retune netburst params 2011-02-22 10:00:06 +00:00