>Has there been any discussion of including this in the mainstream >kernel? It would be wonderful to finally have full POSIX nanosleep() >and clock_nanosleep() support for everyone.
linus nixed it at the 2.5 stage. the POSIX API was added to the kernel, but not the APIC-based high-res implementation. he just didn't like the way it messed with the "tick". jesse chappell has played with a previous version of the patch and has had problems with X stalling for seconds (?) when in use. the resolution also didn't seem anywhere close to what was claimed. --p
