Lines Matching full:we
6 * figure. Its a silly number but people think its important. We go through
15 * We take a distributed and async approach to calculating the global load-avg
32 * serious number of CPUs, therefore we need to take a distributed approach
38 * So assuming nr_active := 0 when we start out -- true per definition, we
43 * across the machine, we assume 10 ticks is sufficient time for every
51 * to the wakeup path. Instead we increment on whatever CPU the task ran
106 * we find: x^n := x^(\Sum n_i * 2^i) := \Prod x^(n_i * 2^i), which is
173 * entering NO_HZ state such that we can include this as an 'extra' CPU delta
174 * when we read the global state.
178 * - When we go NO_HZ idle during the window, we can negate our sample
181 * We avoid this by keeping two NO_HZ-delta counters and flipping them
192 * This ensures we'll fold the old NO_HZ contribution in this window while
195 * - When we wake up from NO_HZ during the window, we push up our
196 * contribution, since we effectively move our sample point to a known
205 * When making the ILB scale, we should try to pull this in as well.
215 * See calc_global_nohz(), if we observe the new index, we also in calc_load_write_idx()
221 * If the folding window started, make sure we start writing in the in calc_load_write_idx()
250 * We're going into NO_HZ mode, if there's any pending delta, fold it in calc_load_nohz_start()
270 * If we're still before the pending sample window, we're done. in calc_load_nohz_stop()
277 * We woke inside or after the sample window, this means we're already in calc_load_nohz_stop()
299 * calc_load_nohz per calc_load_nohz_start(), all we need to do is fold
302 * Once we've updated the global active value, we need to apply the exponential
313 * Catch-up, fold however many we are behind still in calc_global_nohz()
331 * Make sure we first write the new time then flip the index, so that in calc_global_nohz()
377 * In case we went to NO_HZ for multiple LOAD_FREQ intervals in calc_global_load()