* Don't fill the timing tests with stdio shit, gives much better results. Now the only overhead is a tiny fraction of the tested time. 140MB/s in both directions in line buffered mode on my lightly loaded Athlon 3000. SVN revision: 19472 |
||
---|---|---|
legacy |
* Don't fill the timing tests with stdio shit, gives much better results. Now the only overhead is a tiny fraction of the tested time. 140MB/s in both directions in line buffered mode on my lightly loaded Athlon 3000. SVN revision: 19472 |
||
---|---|---|
legacy |