kernel-ark/kernel
Martin Hicks 753ee72896 [PATCH] VM: early zone reclaim
This is the core of the (much simplified) early reclaim.  The goal of this
patch is to reclaim some easily-freed pages from a zone before falling back
onto another zone.

One of the major uses of this is NUMA machines.  With the default allocator
behavior the allocator would look for memory in another zone, which might be
off-node, before trying to reclaim from the current zone.

This adds a zone tuneable to enable early zone reclaim.  It is selected on a
per-zone basis and is turned on/off via syscall.

Adding some extra throttling on the reclaim was also required (patch
4/4).  Without the machine would grind to a crawl when doing a "make -j"
kernel build.  Even with this patch the System Time is higher on
average, but it seems tolerable.  Here are some numbers for kernbench
runs on a 2-node, 4cpu, 8Gig RAM Altix in the "make -j" run:

			wall  user   sys   %cpu  ctx sw.  sleeps
			----  ----   ---   ----   ------  ------
No patch		1009  1384   847   258   298170   504402
w/patch, no reclaim     880   1376   667   288   254064   396745
w/patch & reclaim       1079  1385   926   252   291625   548873

These numbers are the average of 2 runs of 3 "make -j" runs done right
after system boot.  Run-to-run variability for "make -j" is huge, so
these numbers aren't terribly useful except to seee that with reclaim
the benchmark still finishes in a reasonable amount of time.

I also looked at the NUMA hit/miss stats for the "make -j" runs and the
reclaim doesn't make any difference when the machine is thrashing away.

Doing a "make -j8" on a single node that is filled with page cache pages
takes 700 seconds with reclaim turned on and 735 seconds without reclaim
(due to remote memory accesses).

The simple zone_reclaim syscall program is at
http://www.bork.org/~mort/sgi/zone_reclaim.c

Signed-off-by: Martin Hicks <mort@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-21 18:46:14 -07:00
..
irq [PATCH] drop note_interrupt() for per-CPU for proper scaling 2005-05-28 11:14:00 -07:00
power [PATCH] smp_processor_id() cleanup 2005-06-21 18:46:13 -07:00
acct.c
audit.c AUDIT: Unify auid reporting, put arch before syscall number 2005-05-23 21:35:28 +01:00
auditsc.c AUDIT: Record working directory when syscall arguments are pathnames 2005-05-27 12:17:28 +01:00
capability.c
compat.c
configs.c
cpu.c
cpuset.c [PATCH] cpuset exit NULL dereference fix 2005-05-27 08:07:26 -07:00
dma.c
exec_domain.c
exit.c [PATCH] timer exit cleanup 2005-06-17 10:03:50 -07:00
extable.c
fork.c
futex.c
intermodule.c
itimer.c
kallsyms.c
kfifo.c
kmod.c
kprobes.c
ksysfs.c
kthread.c
Makefile
module.c [PATCH] smp_processor_id() cleanup 2005-06-21 18:46:13 -07:00
panic.c
params.c [PATCH] sysfs: (rest) if show/store is missing return -EIO 2005-06-20 15:15:03 -07:00
pid.c
posix-cpu-timers.c
posix-timers.c [PATCH] timer exit cleanup 2005-06-17 10:03:50 -07:00
printk.c
profile.c [PATCH] profile.c: `schedule' parsing fix 2005-05-17 07:59:21 -07:00
ptrace.c
rcupdate.c
resource.c
sched.c [PATCH] smp_processor_id() cleanup 2005-06-21 18:46:13 -07:00
seccomp.c
signal.c Merge with master.kernel.org:/pub/scm/linux/kernel/git/torvalds/linux-2.6.git 2005-06-02 16:39:11 +01:00
softirq.c
spinlock.c [PATCH] spin_unlock_bh() and preempt_check_resched() 2005-05-21 10:46:48 -07:00
stop_machine.c [PATCH] smp_processor_id() cleanup 2005-06-21 18:46:13 -07:00
sys_ni.c [PATCH] VM: early zone reclaim 2005-06-21 18:46:14 -07:00
sys.c
sysctl.c
time.c
timer.c
uid16.c
user.c
wait.c
workqueue.c