8908c2c027
Added support for "cn/msr" also known as "cn-fast".
2018-06-10 19:48:34 +07:00
1748a7bd57
Fix double hash mode too.
2018-06-03 05:42:10 +07:00
6cf24936df
#551 Fixed cn-heavy for ARMv8.
2018-06-03 02:56:49 +07:00
651637d637
#446 Better fix, second variable always aligned.
2018-06-03 00:22:26 +07:00
26ee5028e1
Use native checks instead of XMRIG_ARMv8.
2018-06-03 00:09:59 +07:00
d900a6d9dd
#446 Fixed SIGBUS error on 32 bit ARM CPUs.
2018-06-02 04:37:12 +07:00
a00024cf51
Fixed ARMv8 build.
2018-05-05 22:44:20 +07:00
3df99fbced
Add Stellite (XTL) support as cn/xtl.
2018-04-27 00:28:52 +07:00
41e8c4f887
Send supported algorithms to pool in login request.
2018-04-26 15:02:01 +07:00
ba5f92c6dd
Added support for cn-lite/ipbc.
2018-04-25 18:31:18 +07:00
ca149d2eed
Sync changes with proxy.
2018-04-25 14:48:32 +07:00
274992e50d
Basic cryptonight-ipbc definition.
2018-04-21 00:19:33 +07:00
98e7308597
Move keccak to common code.
2018-04-20 14:45:51 +07:00
2d2e60a197
Fix x86 build.
2018-04-20 10:14:33 +07:00
14576f599c
Fix ARMv7 build.
2018-04-19 19:44:17 +07:00
e119f7f402
Rearrange test vectors, for catch cn-heavy bug.
2018-04-19 13:50:06 +07:00
ad94e9a7d2
Simplify ARM implementation.
2018-04-19 11:54:11 +07:00
9e3f2ae9f9
Added x3 x4 x5 hashing modes.
2018-04-16 15:40:37 +07:00
6b4f2d0a91
Fixed ARM build.
2018-04-15 12:58:17 +07:00
9125b6c251
Rewrite memory allocation.
2018-04-15 11:08:47 +07:00
51422f4b1e
Move xmrig.h to common/xmrig.h.
2018-04-13 07:00:51 +07:00
b13640e4a1
Fixes for build without cn-lite and cn-heavy.
2018-04-12 11:38:43 +07:00
ccbb78d4e9
Improved performance for cryptonight v7.
2018-04-09 21:56:15 +07:00
3a67ee6d11
Small fixes.
2018-04-06 23:32:54 +07:00
4e8ef7c6ed
Added cryptonight-heavy support for ARM
2018-04-03 17:51:06 +07:00
26e1b14020
Added test hashes for self test.
2018-04-03 17:19:01 +07:00
dd6bc339bf
First working cryptonight-heavy.
2018-04-03 16:55:41 +07:00
5c6ec587ac
Move selfTest to threads, remove legacy CryptoNight.cpp.
2018-04-03 14:51:05 +07:00
903b243308
New style function selector.
2018-04-03 02:55:28 +07:00
a042cbf885
Added classes IThread, CpuThread and API endpoint "GET /1/threads".
2018-04-01 22:49:21 +07:00
aac7b0404a
Options class replaced to xmrig::Config.
2018-03-31 16:29:47 +07:00
1011fd4891
Partial fix for ARMv7
2018-03-24 10:14:21 +07:00
2de5d92d3a
Removed unnecessary pointers cast.
2018-03-18 05:48:19 +07:00
7adf30e326
Remove align.h.
2018-03-14 17:11:51 +07:00
125072a30e
Added test hashes for AEON.
2018-03-13 01:29:31 +07:00
aec31c43c0
Better v1 PoW implementation, added variant option.
2018-03-12 22:29:44 +07:00
83e5832bc1
Some small fixes.
2018-03-12 14:44:23 +07:00
79779b51da
Added reference hashes.
2018-03-07 15:32:44 +07:00
8b60585004
PoW changes WIP
2018-03-06 21:34:20 +07:00
df6a397e52
Update copyright and move version into Job class.
2018-03-06 17:06:07 +07:00
48b1de0b59
Changes for the Monero v1 PoW
2018-03-02 22:17:29 +00:00
184f79ad3f
Fix code style, replace tabs to space #2 .
2018-02-08 17:21:12 +07:00
e78e810cfe
Fix code style, replace tabs to space.
2018-02-08 17:02:32 +07:00
9a28ad590c
up to 20% perf increase with Cryptonight with non-AES CPU
...
This time, the performance increase is got with MSVC and GCC. On non-AES CPU, there were an useless load/store SSE2 register. The last MSVC "hack" is replaced by a portable code and he's more complete (a load is saved).
On my C2Q6600, with 3 thread, I have +16% with MSVC2015 and +20% with GCC 7.3, compared to official 2.4.4 version.
2018-01-28 12:58:19 +01:00
9bceb65ad8
+15% boost with non-AES CPU
...
Performance boost validated on Core 2 Quad processor under Windows 10.
But it's Windows/MS Visual C++ specific.
2018-01-20 10:43:56 +01:00
aa4f8b6fa7
#216 Added ARMv7 support.
2017-11-26 22:23:23 +03:00
6cc152e26f
Added ARMv8 (aarch64) support.
2017-11-06 03:11:35 +03:00
d403dcf95c
Optimized software aes.
2017-11-03 05:35:29 +03:00
4cf3bb9930
#111 Fixed build without AEON support.
2017-09-16 15:20:33 +03:00
a07b0e5953
Fix Visual Studio warnings.
2017-08-14 09:30:41 +03:00
XMRig
Lee Clagett
Foudge