Commit Graph

53 Commits

Author SHA1 Message Date
Jeff Roberson
0c0a98b231 - When stealing a kse in kseq_move() ignore the current kseq's min nice
value.  We want to steal any thread, even one that is not given a slice
   on its current queue.
2003-07-08 06:19:40 +00:00
Jeff Roberson
0ec896fd28 - Clean up an unused variable.
Submitted by:	Steve Kargl <skg@routmask.apl.washington.edu>
2003-07-07 21:08:28 +00:00
Jeff Roberson
749d01b011 - Parse the cpu topology map in sched_setup().
- Associate logical CPUs on the same physical core with the same kseq.
 - Adjust code that assumed there would only be one running thread in any
   kseq.
 - Wrap the HTT code with a ULE_HTT_EXPERIMENTAL ifdef.  This is a start
   towards HyperThreading support but it isn't quite there yet.
2003-07-04 19:59:00 +00:00
Jeff Roberson
7a20304f84 - Don't migrate to stopped cpus. 2003-06-28 09:09:33 +00:00
Jeff Roberson
86f8ae9663 - If smp is not started yet don't try to load balance or we'll put threads
on cpus that aren't running yet.
2003-06-28 08:24:42 +00:00
Jeff Roberson
a91172ade1 - Throttle the inherited sleep and run time in sched_fork_kseg(). This
allows us to learn the behavior of a thread much more quickly after it
   starts up.
2003-06-28 06:19:56 +00:00
Jeff Roberson
e493a5d90c - Adjust the default maximum slice value to ~140ms. This has improved the
nice distribution without significantly impacting interactive response.
   As a side effect it should also allow batch processes to run for a
   slightly longer period which will positively impact their performance.
2003-06-28 06:04:47 +00:00
Jeff Roberson
1a7a9d0ec2 - lticks was erroneously being updated in sched_pctcpu(). This was causing
us to skip the pctcpu_update() call which lead to inaccurate cpu usage
   statistics for processes that didn't run often.
2003-06-21 02:31:49 +00:00
Jeff Roberson
665cb285a8 - Don't allow nice to have such a large effect on priority. This was
causing poor interactive performance while unnice processes were running.
   The new scheme still allows nice to have an effect on priority but it is
   not as dramatic as the effect of the interactivity score.
2003-06-21 02:22:47 +00:00
Jeff Roberson
d07ac847ef - Use a more robust mechanism for determining whether or not a kse is on a
kseq.
2003-06-17 19:49:18 +00:00
Jeff Roberson
7cd0f83355 - Temporarily patch a problem where the interact score could be negative
because the run time exceeds the largest value a signed int can hold.
   The real solution involves calculating how far we are over the limit.
   To quickly solve this problem we loop removing 1/5th of the current value
   until it falls below the limit.  The common case requires no passes.
2003-06-17 10:21:34 +00:00
Jeff Roberson
4b60e3242e - Add a new function "sched_interact_update()" that scales back the sleep
and run time.
 - Scale the sleep and run time back via sched_interact_update() in more
   places.  This is to keep the statistic more accurate.
 - Charge a parent one tick for forking a child.
 - Add only the run time and not the sleep time to the parents kg when a
   thread exits.  This allows us to give a penalty for having an expensive
   thread exit but does not give a bonus for having an interactive thread
   exit.
 - Change the SLP_RUN_THROTTLE to limit us to 4/5th and not 1/2.
 - Change the SLP_RUN_MAX to two seconds.  This keeps bursty interactive
   applications like mozilla and openoffice in the interactive range even
   through expensive tasks.
 - Recalculate the slice after every sleep.  This ensures that once a task
   has been marked interactive it only has a slice of 1 at the risk of
   giving tasks that sleep for a very brief period a longer time slice.
2003-06-17 06:39:51 +00:00
Jeff Roberson
3c12473229 - Increase the ksegrp's cpu time history buffer to 250ms.
- Decrease the history buffer divisor to 2 so that we remember more of the
   old behavior.
2003-06-15 04:14:25 +00:00
Jeff Roberson
b41f3d22cc - Cap the growth of sleep and run time in sched_exit_kse(). 2003-06-15 02:52:29 +00:00
Jeff Roberson
210491d3d9 - Fix the maximum slice value. I accidentally checked in a value of '2'
which meant no process would run for longer than 20ms.
 - Slightly redo the interactivity scorer.  It follows the same algorithm but
   in a slightly more correct way.  Previously values above half were
   incorrect.
 - Lower the interactivity threshold to 20.  It seems that in testing non-
   interactive tasks are hardly ever near there and expensive interactive
   tasks can sometimes surpass it.  This area needs more testing.
 - Remove an unnecessary KTR.
 - Fix a case where an idle thread that had an elevated priority due to
   priority prop. would be placed back on the idle queue.
 - Delay setting NEEDRESCHED until userret() for threads that haad their
   priority elevated while in kernel.  This gives us the same context switch
   optimization as SCHED_4BSD.
 - Limit the child's slice to 1 in sched_fork_kse() so we detect its behavior
   more quickly.
 - Inhert some of the run/slp time from the child in sched_exit_ksegrp().
 - Redo some of the priority comparisons so they are more clear.
 - Throttle the frequency of sched_pctcpu_update() so that rounding errors
   do not make it invalid.
2003-06-15 02:18:29 +00:00
David Xu
0e2a4d3aeb Rename P_THREADED to P_SA. P_SA means a process is using scheduler
activations.
2003-06-15 00:31:24 +00:00
David E. O'Brien
677b542ea2 Use __FBSDID(). 2003-06-11 00:56:59 +00:00
Jeff Roberson
356500a306 - Add a simple CPU load balancing algorithm. This works by executing once a
second and equalizing the load between the two most imbalanced CPU.  This
   is intended to clear up long term load imbalances that would not be handled
   by the 'pull' method in sched_choose().
 - Pull out some bits of sched_choose() into a kseq_move() function that moves
   an arbitrary thread from one kseq to another.
2003-06-09 00:39:09 +00:00
Jeff Roberson
b90816f188 - When a new thread is added to a kseq the load is incremented prior to
adding it to the nice tables.  Therefore, in kseq_add_nice, we should
   keep in mind that the load will be 1 if we are the only thread, and not
   0.
 - Assert that the sched lock is held in all the appropriate places.
 - Increase the scope of the sched lock in sched_pctcpu_update().
 - Hold the sched lock in sched_runnable().  It is not held by the caller.
2003-06-08 00:47:33 +00:00
Julian Elischer
43fdafb1e1 Fix typo in last commit 2003-05-02 06:18:55 +00:00
Julian Elischer
b1ac98d8b2 Move the flag that indicates an idle thread from the KSE to the thread.
It was always referenced via the thread anyhow.

Reviewed by:	jhb (a LOOOOONG time ago)
2003-05-02 00:33:12 +00:00
John Baldwin
2056d0a168 Add lock assertions for various proc/thread/kse/ksegroup fields to the
scheduler functions.
2003-04-23 18:51:05 +00:00
John Baldwin
0b5318c81a - Assert that the proc lock and sched_lock are held in sched_nice().
- For the 4BSD scheduler, this means that all callers of the static
  function resetpriority() now always hold sched_lock, so don't lock
  sched_lock explicitly in that function.
2003-04-22 20:50:38 +00:00
John Baldwin
828e7683bf Protect p_swtime with the sched_lock. 2003-04-22 19:48:25 +00:00
Jeff Roberson
7cd650a972 - Set the ke_cpu field in sched_add() for interrupt and realtime threads
since they are going on the current cpu and not their previously assigned
   cpu.
 - sched_runnable() should only return true in the SMP case if the other
   processor has more than one thread that is runnable.  We can not steal
   curthread.
 - Change kseq_print() to accept the cpuid instead of a kseq pointer.  This
   makes use of this function in ddb much easier.
2003-04-18 05:24:10 +00:00
Jeff Roberson
a5f099d0c4 - Unbreak priority prop. for timeshare threads. Always place something on
the current queue if its priority is really elevated.  This needs more work
   as there are cases where a next queue kse could be holding up what would
   be a curr queue kse, and thus hurting interactivity.  Also, when a thread
   with an elevated priority has its priority lowered it should be placed
   back on the next queue.
2003-04-12 22:33:24 +00:00
Jeff Roberson
9bca28a703 - Clean up some debug code left over from my earlier megacommit. 2003-04-12 07:28:36 +00:00
Jeff Roberson
b5c4c4a7e5 - We only care about the base priority. Ignore the SCHED_FIFO_BIT so that
we dont get confused.

Reported and debugged by:	Steve Kargl <sgk@troutmask.apl.washington.edu>
2003-04-12 07:00:16 +00:00
Jeff Roberson
141ad61c78 - Add sched_exit_*
- Call sched_exit_kse() from sched_exit() instead of implementing it here.
2003-04-11 19:24:00 +00:00
Jeff Roberson
58177de2de - Only select kseqs with more than one kse to steal. The running kse
is reflected in the load now and you can't very well migrate that.
2003-04-11 18:40:34 +00:00
Jeff Roberson
c36ccfa22b - When migrating a kse from one kseq to the next actually insert it onto
the second kseq's run queue so that it is referenced by the kse when
   it is switched out.
 - Spell ksq_rslices properly.

Reported by:	Ian Freislich <ianf@za.uu.net>
2003-04-11 18:37:34 +00:00
Jeff Roberson
15dc847e52 - Add a SYSCTL node for the ule scheduler.
- Allow user adjustable min and max time slices (suggested by hiten).
 - Change the SLP_RUN_MAX to 100ms from 2 seconds so that we learn whether a
   process is interactive or not much more quickly.
 - Place a process on the current run queue if it is interactive or if it is
   running at an interrupt thread priority due to priority prop.
 - Use the 'current' timeshare queue for interrupt threads, realtime threads,
   and idle threads that are running at higher priority due to priority prop.
   This fixes problems where priorities would have been elevated but we would
   not check the timeshare run queue until other lower priority tasks were
   no longer runnable.
 - Keep an array of loads indexed by the priority class as well as a global
   load.
 - Keep an bucket of nice values with a count of the number of kses currently
   runnable with that nice value.
 - Keep track of the minimum nice value of any running thread.
 - Remove the unused short term sleep accounting.  I was attempting to use
   this for load balancing but it didn't work out.
 - Define a kseq_print() for use with debugging.
 - Add KTR debugging at useful places so we can easily debug slice and
   priority assignment.
 - Decouple the runq assignment from the kseq assignment.  kseq_add now keeps
   track of statistics.  This is done so that the nice and load is still
   tracked for the currently running process.  Previously if a niced process
   was added while a non nice process was running the niced process would
   still get a slice since it was not aware of the unnice process.
 - Make adjustments for the sched api changes.
2003-04-11 03:47:14 +00:00
Julian Elischer
060563ec50 Move the _oncpu entry from the KSE to the thread.
The entry in the KSE still exists but it's purpose will change a bit
when we add the ability to lock a KSE to a cpu.
2003-04-10 17:35:44 +00:00
Jeff Roberson
a8949de20e - Keep seperate statistics and run queues for different scheduling classes.
- Treat each class specially in kseq_{choose,add,rem}.  Let the rest of the
   code be less aware of scheduling classes.
 - Skip the interactivity calculation for non TIMESHARE ksegrps.
 - Move slice and runq selection into kseq_add().  Uninline it now that it's
   big.
2003-04-03 00:29:28 +00:00
Jeff Roberson
5053d272c2 - Make the interactivity calculator decay faster.
- Make the pcpu estimator update faster.
2003-04-02 08:22:33 +00:00
Jeff Roberson
98c9b132d1 - I meant divide by two and not shift by two in SCHED_PRI_NHALF. 2003-04-02 08:21:24 +00:00
Jeff Roberson
245f3abfd5 - Add in support for KSEs with 0 slice values on the run queue. If we try
to select a KSE with a slice of 0 we will update its slice and insert it
   onto the next queue.
 - Pass the KSE instead of the ksegrp into sched_slice().  This more
   accurately reflects the behavior of the code.  Slices are granted to kses.
 - Add a function kseq_nice_min() which finds the smallest nice value
   assigned to the kseg of any KSE on the queue.
 - Rewrite the logic in sched_slice().  Add a large comment describing the
   new slice selection scheme.  To summarize, slices are assigned based on
   the nice value.  Priorities are still calculated based on the nice and
   interactivity of a process.  Slice sizes of 0 may be granted for KSEs
   whos nice is 20 or futher away from the lowest nice on the run queue.
   Other nice values are scaled across the range [min, min+20].  This fixes
   ULEs bad behavior with positively niced processes.
2003-04-02 06:46:43 +00:00
Jeff Roberson
e1f89c222b - Create a function sched_interact_score() which decides on the
interactivity of a kseg and assigns it a value of 0 through 100.
 - Use sched_interact_score() to determine the dynamic priority.
 - Define SCHED_CURR() in terms of sched_interact_score().
 - Adjust the maximum slice back down to 100ms.
 - Remove redundant clearing of ke_runq in sched_wakeup()
 - Clean up #defines and comment them.
2003-03-04 02:45:59 +00:00
Jeff Roberson
65c8760dbf - Shift the tick count by 10 and back around sched_pctcpu_update()
calculations.  Keep this changes local to the function so the tick count
   is in its natural form otherwise.  Previously 1000 was added each time
   a tick fired and we divided by 1000 when it was reported.  This is done
   to reduce rounding errors.
2003-03-03 05:29:09 +00:00
Jeff Roberson
a6ed41865b - In sched_add() special case PRI_TIMESHARE and PRI_ITHD|PRI_REALTIME. We
always place ITHD & REALTIME threads on the current queue of the current
   cpu.  Prior to this change an interrupt thread would only ever run on one
   cpu.
2003-03-03 04:28:07 +00:00
Jeff Roberson
f1e8dc4a3b - Refrain from setting the td_priority in sched_wakeup(). It will be reset
before we return to user space.
2003-03-03 04:11:40 +00:00
Julian Elischer
ac2e415327 Change the process flags P_KSES to be P_THREADED.
This is just a cosmetic change but I've been meaning to do it for about a year.
2003-02-27 02:05:19 +00:00
Julian Elischer
4a338afd7a Move a bunch of flags from the KSE to the thread.
I was in two minds as to where to put them in the first case..
I should have listenned to the other mind.

Submitted by:	 parts by davidxu@
Reviewed by:	jeff@ mini@
2003-02-17 09:55:10 +00:00
Jeff Roberson
783caefbbf - Enable STRICT_RESCHED until code that dynamically decides on resched
strictness based on the current workload is finished.
2003-02-10 14:11:23 +00:00
Jeff Roberson
407b015791 - Add a new variable 'kg_runtime' that tracks the amount of time we've run.
- Use the ratio of kg_runtime / kg_slptime to determine our dynamic priority.
 - Scale kg_runtime and kg_slptime back when the sum of the two exceeds
   SCHED_SLP_RUN_MAX.  This allows us to slowly forget old behavior.
 - Scale back the runtime and slptime in fork so that the new process has the
   same ratio but much less accumulated time.  This causes new behavior to be
   noticed more quickly.
2003-02-10 14:03:45 +00:00
Jeff Roberson
5d7ef00cfe - Make some context switches conditional on SCHED_STRICT_RESCHED. This may
have some negative effect on interactivity but it yields great perf. gains.
   This also brings the conditions under which ULE context switches inline
   with SCHED_4BSD.
 - Define some new kseq_* functions for manipulating the run queue.
 - Add a new kseq member ksq_rslices and ksq_bload.  rslices is the sum of
   the slices of runnable kses.  This will be used for push load balance
   decisions.  bload is the number of threads blocked waiting on IO.
2003-02-03 05:30:07 +00:00
Jeff Roberson
cd6e33df1c - Stop abusing oncpu for our cpu binding. Define a scheduler local element
in the kse datastructure called ke_cpu.  This is the cpu which we are
   currently bound to.  Some flags may be added later to support hard binding.
2003-02-03 02:26:28 +00:00
Scott Long
7121cce58a Use hz if stathz is zero. Adopted from sched_4bsd. 2003-02-02 08:24:32 +00:00
Jeff Roberson
0a016a05a4 - Use ksq_load as the authoritive count of kses on the pair of kseqs for
sched_runnable() et all.
 - Remove some dead code in sched_clock().
 - Define two macros KSEQ_SELF() and KSEQ_CPU() for getting the kseq of the
   current cpu or some alternate cpu.
 - Start introducing kseq_() functions, such as kseq_choose() and kseq_setup().
2003-01-29 07:00:51 +00:00
Jeff Roberson
bf857e69a2 - Remove debugging code that didn't work on UP. 2003-01-29 00:26:47 +00:00