 Raulo
|
a27e69aec5
Changes to make the 4-way code compile on Intel Compiler
|
15 years ago |
 Jeff Garzik
|
1083e15c47
Mark hash success as unlikely, using __builtin_expect() intrinsic
|
15 years ago |
|
Jeff Garzik
|
714c0fd7c9
Continue scanhash, even if high 32 bits are zero.
|
15 years ago |
|
Jeff Garzik
|
c68ffb30dd
Display proof-of-work hash when one is discovered
|
15 years ago |
|
Jeff Garzik
|
0b67740707
Pass max-nonce as arg to each sha256 algo.
|
15 years ago |
|
Jeff Garzik
|
8504cf4803
sha256_4way minor optimizations
|
15 years ago |
|
Jeff Garzik
|
0ee24c00ce
sha256_4way: work around gcc 4.5.x bug by always inlining ROTR and SHR
|
15 years ago |
|
Jeff Garzik
|
8ca6ffa8a9
sha256_4way: do not include unnecessary headers, when !4way
|
15 years ago |
|
Jeff Garzik
|
339ddf4d75
Move sha256_generic into its own .o build output. Const-ify midstate param.
|
15 years ago |
|
Jeff Garzik
|
6288d943ab
sha256_4way: patch proper nonce into data block
|
15 years ago |
|
Jeff Garzik
|
86eb37d631
Improve and modularize compile-time CPU detection.
|
15 years ago |
|
Jeff Garzik
|
500759cea1
s/__SSE__/__SSE2__/
|
15 years ago |
|
Jeff Garzik
|
eed9b3c7df
sha256_4way: perform 16m hashes, not 64k, per invocation
|
15 years ago |
|
Jeff Garzik
|
35ea649d97
Improve hash performance statistics.
|
15 years ago |
|
Jeff Garzik
|
c639149977
Add tcatm's 4way SSE2 sha256 implementation.
|
15 years ago |
