> My point with raising the PR also is that clock_getres() et al are poorly > documented (and as you have discovered), just plain wrong. I just setn a diff to tech-kern where someone might actually know what for sure. Jan