Hi everyone... Sorry, a very very late reply. Was buried under tons of jobs... > Do you want to measure the latency on kernel end ? Honestly, I didn't know which context I tried to measure. But I think, what I did was measuring latency from user space point of view. CMIIW. > If yes, then one should not actually, rely on the system timer tick. > In fact the time should be measured by other clock unit. > It may be RTC, timer unit (if your system has otherthan system timer > ) or the existing timer unit - just by reading the time remaining > factor of that timer unit. This way you get the exact time latency in > kernel. Hm...another time source. How about something like TSC (i.e calling rdtsc())? Maybe a more general question. Do you think changing HZ=100 to HZ=1000 affects things like mp3 playback, video (let's say DivX type) playback and gaming (Quake?). Are the differences sensible? regards, Mulyadi -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
