On Wed, Mar 5, 2008 at 9:01 AM, MK <halfcountplus@xxxxxxxxxxxxx> wrote: > Kam Leo hypothesized: No hypothesis. Just applying basic knowledge regarding the kernel to the problem that was described. > > You might be able to get improved performance by changing from the > > kernel's default I/O scheduler (CFQ). > > I switched the default to the "anticipatory" scheduler (leaving CFQ and > deadline built-in) because of xine lock-ups and that DEFINATELY did the > trick (2.6.22.6). Why was building a custom version of the kernel necessary? Didn't appending "elevator=as" work for you? > I have not noticed a difference otherwise.
