|
|
1 | |
|
|
2 | - fixed hash calculation in C♯ version (Tiago Freitas Leal). |
|
|
3 | - unroll copy for small sizes, use memcpy for larger sizes, |
|
|
4 | greatly speeding up decompression in most cases. |
|
|
5 | - finally disable rep movsb - it's a big loss on modern intel cpus, |
|
|
6 | and only a small win on amd cpus. |
|
|
7 | - improve c++ compatibility of the code. |
|
|
8 | |
|
|
9 | 3.5 Fri May 1 02:28:42 CEST 2009 |
|
|
10 | - lzf_compress did sometimes write one octet past the given output |
|
|
11 | buffer (analyzed and nice testcase by Salvatore Sanfilippo). |
|
|
12 | |
| 1 | 3.4 Tue Sep 2 06:45:00 CEST 2008 |
13 | 3.4 Tue Sep 2 06:45:00 CEST 2008 |
| 2 | - the fix from 3.3 introduced a compression bug, which is fixed in |
14 | - the fix from 3.3 introduced a compression bug, which is fixed in |
| 3 | this release (which explains the mysterious prerelease...). thanls |
15 | this release (which explains the mysterious prerelease...). Thanks |
| 4 | once more to Clément Calmels. |
16 | once more to Clément Calmels. |
| 5 | |
17 | |
| 6 | 3.3 Mon Aug 25 03:17:42 CEST 2008 |
18 | 3.3 Mon Aug 25 03:17:42 CEST 2008 |
| 7 | - lzf_compress could access memory after the given input buffer |
19 | - lzf_compress could access memory after the given input buffer |
| 8 | when outputting back references. reported with nice testcase |
20 | when outputting back references. reported with nice testcase |
