Yes, I think this is correct. Mostly I am wanting a more detailed
descritption of how "response time" in the scheduler is defined
and can be measured for both an I/O bound process and a CPU bound
process. So if I am calculating prime numbers for example, how would
I get an acuurate measurment for evaluating the response time for
this process. Not the time it takes to complete the process, but the
time it takes for (whom) to start... (doing what?)
The definition I have seems vague. maybe someone can fill in those
words and tell me what data I can collect to measure?
On Nov 13, 2005, at 12:40 AM, Mulyadi Santosa wrote:
Dear Filip...
If you wanted to measure the response time (the time between input
and the first response by process, not necessarily the first output)
how would you do it? Hopefully Ive defined resonse time correctly.
Would this be the first context switch or first interrupt thats
triggered, or is it the time it takes for a process to respond to an
interupt?
You mean, something like this scenario?
An user start vi. You want to measure the time how fast Vi response
keypress (let's say character "a"), from the time it is pressed right
until it is showed on Vi display?
If yes, you are measuring --> interrupt acknowledgment + bottom halves
execution time + putting back Vi process to run queue + Vi re
scheduling
More detailed information on what you want to achieve would be helpful
for us....
regards
Mulyadi
--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive: http://mail.nl.linux.org/kernelnewbies/
FAQ: http://kernelnewbies.org/faq/
