wbhart
4babcebbfa
Turned off unbalanced multiplications as they slow things down.
...
-This line, and those below, will be ignored--
M mpn/generic/mul.c
2009-05-11 11:06:38 +00:00
wbhart
102e4f59eb
I put the wrong version of this in. Here it is back again.
2009-05-11 11:02:51 +00:00
wbhart
9e56c61071
Added toom42 and code to handle unbalanced multiplication.
2009-05-11 10:09:09 +00:00
jasonmoxham
0a1d07af4e
New asm function mpn_sublsh1_n for K8/K10
2009-05-10 20:03:47 +00:00
jasonmoxham
574f3be308
New asm function mpn_divexact_byff for K8/K10/Core2/penryn/nehalem
2009-05-10 19:35:54 +00:00
jasonmoxham
359fab42b5
New asm functions mpn_rsh1add_n mpn_rsh1sub_n for K8/K10/Core2/penryn/nehalem
2009-05-10 18:46:48 +00:00
jasonmoxham
428e43b40e
New asm functions mpn_addadd_n mpn_addsub_n mpn_subadd_n for K8/K10
2009-05-10 16:25:01 +00:00
wbhart
4c442dbf94
Went back to old mul.c (again) :-(
2009-05-10 14:09:59 +00:00
wbhart
1b58a8b49e
Speed toom4 up by passing some arguments to the interpolate code in the
...
output space so they don't have to be moved at the end.
2009-05-10 13:45:27 +00:00
gladman
45e909b337
correct erroneous Windows addition to gmp-h.in
2009-05-10 13:15:12 +00:00
wbhart
e9bb26a854
Added missing #include "mpir.h".
2009-05-10 07:28:21 +00:00
wbhart
4f99bbe9fc
Added missing toom3 file.
2009-05-10 07:15:02 +00:00
wbhart
9c79e0a98b
Factored out mpn_toom3_sqr_n and mpn_toom3_mul_n and removed duplication
...
of mpn_toom3_interpolate. Rewrote mpn_toom3_sqr_n.
2009-05-10 07:12:38 +00:00
wbhart
44dadcf975
Reverted to earlier version of mul.c because the later version slows
...
down the 128x128 mul bench.
2009-05-10 04:39:49 +00:00
wbhart
c8aa69c789
Added toom3_mul_n with better memory usage.
2009-05-10 04:24:39 +00:00
jasonmoxham
90d8207a80
New functions mpn_sumdiff for core2/penryn/nehalem , or rather faster to do separate add and sub
2009-05-10 03:39:43 +00:00
jasonmoxham
b07549802a
New asm functions mpn_add_n mpn_sub_n for Core2/penryn/nehalem
2009-05-10 01:26:52 +00:00
jasonmoxham
0c3c909910
New asm functions for mpn_copyi mpn_copyd for k8,k10,core2,penryn,nehalem
2009-05-10 00:20:44 +00:00
wbhart
1cc8b35cfe
Another slight speedup.
2009-05-09 21:51:40 +00:00
wbhart
96e8e4e410
Added my copyright info.
2009-05-09 21:38:06 +00:00
wbhart
6ed1dd6474
Whoops I screwed up toom4 and toom7, putting them back now.
2009-05-09 21:23:15 +00:00
wbhart
7a0e036d36
Fixed toom4 and toom7 issues and added k8 tuning code.
2009-05-09 21:12:13 +00:00
wbhart
72f93a085c
Added new toom3 code.
2009-05-09 20:56:34 +00:00
gladman
d942415a1c
1. Update Windows Powershell scripts in mpirbench to refer to MPIR rather than GMP
...
2. Update MPIR version number in Windows config files to 1.1.2
3. Add an MSVC inline definition in in gmp-h.in
4. Correct locale test (as per GMP correction)
5. Add Windows x64 set/copy intrinsics to mul_fft.c (improves FFT speed score by 2%)
2009-05-09 13:26:27 +00:00
wbhart
0ba06242c6
Fixed some bugs in best_k code used by FFT.
2009-05-09 02:54:08 +00:00
wbhart
911916ce7e
Fixed a carry issue with tc*_addmul which created a requirement for extra memory in toom code.
2009-05-08 14:12:47 +00:00
wbhart
5624d9a6fc
New toom4 and toom7 code.
...
* Don't make copies before basecase multiplications
* Factor out interpolation code
* Convert interpolation code to twos complement
* Optimise code using new assembly functions where available
2009-05-08 13:21:14 +00:00
gladman
b55aae6dd5
Add Windows build files to bench in trunk
2009-05-06 18:35:56 +00:00
gladman
bd34c0bfc5
1. Update g2y.py, the GAS to YASM Python script
...
2. Provide tuning for new FFT code
3. Add some documentation to YASM assembler macros for Windows
2009-05-06 18:20:52 +00:00
wbhart
2ad5066cea
Tried to clean up a little.
2009-05-05 23:52:02 +00:00
wbhart
49441a5e20
Fixed bug in mul_fft.c
2009-05-05 22:18:16 +00:00
gladman
7f8174f08d
update g2y.py to allow command line parameters
2009-05-05 20:43:45 +00:00
gladman
eeaca671af
Remove C99 features in mul_fft.c
2009-05-05 20:41:29 +00:00
wbhart
041df82e0d
Added Zimmermann et al's FFT (after making a bug fix).
2009-05-05 12:27:29 +00:00
gladman
57f06bfe7e
2009-05-01 19:03:56 +00:00
gladman
d18fab116f
Adjust Windows build slightly
2009-05-01 18:33:21 +00:00
gladman
8692baeb3d
Make two 'ease of use' changes to the Windows build
2009-05-01 14:30:27 +00:00
jasonmoxham
8e7cb567a0
correct divexact_byBm1of speed measuring
2009-04-24 22:29:42 +00:00
wbhart
b399395e1b
Added -apple to support very recent apple machines.
2009-04-23 15:42:05 +00:00
jasonmoxham
b52aa5a59e
mpn_mul_2 doesn't do size=1
2009-04-22 22:05:08 +00:00
gladman
a2baca61f0
Change Windows test projects to provide full build in release mode
2009-04-22 19:50:54 +00:00
gladman
a4df93cf01
commit a missing windows test
2009-04-22 19:24:05 +00:00
jasonmoxham
6ca0ab2faf
speed update
2009-04-22 09:37:41 +00:00
jasonmoxham
71343403af
updated speed to accept column sums
2009-04-22 09:02:19 +00:00
gladman
c503ef2397
commit missing Windows assembler file to trunk
2009-04-22 08:01:40 +00:00
jasonmoxham
d33a113cd2
config.guess update for pathscale cc
2009-04-21 16:35:04 +00:00
jasonmoxham
a823bc7102
define gnu_source in configure.in
2009-04-21 12:50:03 +00:00
jasonmoxham
bd911ba13e
update speed options -r and -d to accept any column
2009-04-21 09:33:22 +00:00
gladman
0e21f1f351
Commit missing asm files
2009-04-17 20:36:47 +00:00
wbhart
fd32e5fb9c
Credit Bodrato in the way he requested.
2009-04-15 22:03:24 +00:00