2006-06-30 10:55:32 +02:00
|
|
|
/*
|
|
|
|
* linux/mm/vmstat.c
|
|
|
|
*
|
|
|
|
* Manages VM statistics
|
|
|
|
* Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds
|
2006-06-30 10:55:33 +02:00
|
|
|
*
|
|
|
|
* zoned VM statistics
|
|
|
|
* Copyright (C) 2006 Silicon Graphics, Inc.,
|
|
|
|
* Christoph Lameter <christoph@lameter.com>
|
2006-06-30 10:55:32 +02:00
|
|
|
*/
|
2008-10-06 02:13:52 +02:00
|
|
|
#include <linux/fs.h>
|
2006-06-30 10:55:32 +02:00
|
|
|
#include <linux/mm.h>
|
2007-07-30 00:36:13 +02:00
|
|
|
#include <linux/err.h>
|
2006-06-30 10:55:33 +02:00
|
|
|
#include <linux/module.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 09:04:11 +01:00
|
|
|
#include <linux/slab.h>
|
2006-09-01 06:27:35 +02:00
|
|
|
#include <linux/cpu.h>
|
2008-07-24 06:27:03 +02:00
|
|
|
#include <linux/vmstat.h>
|
Detach sched.h from mm.h
First thing mm.h does is including sched.h solely for can_do_mlock() inline
function which has "current" dereference inside. By dealing with can_do_mlock()
mm.h can be detached from sched.h which is good. See below, why.
This patch
a) removes unconditional inclusion of sched.h from mm.h
b) makes can_do_mlock() normal function in mm/mlock.c
c) exports can_do_mlock() to not break compilation
d) adds sched.h inclusions back to files that were getting it indirectly.
e) adds less bloated headers to some files (asm/signal.h, jiffies.h) that were
getting them indirectly
Net result is:
a) mm.h users would get less code to open, read, preprocess, parse, ... if
they don't need sched.h
b) sched.h stops being dependency for significant number of files:
on x86_64 allmodconfig touching sched.h results in recompile of 4083 files,
after patch it's only 3744 (-8.3%).
Cross-compile tested on
all arm defconfigs, all mips defconfigs, all powerpc defconfigs,
alpha alpha-up
arm
i386 i386-up i386-defconfig i386-allnoconfig
ia64 ia64-up
m68k
mips
parisc parisc-up
powerpc powerpc-up
s390 s390-up
sparc sparc-up
sparc64 sparc64-up
um-x86_64
x86_64 x86_64-up x86_64-defconfig x86_64-allnoconfig
as well as my two usual configs.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-05-20 23:22:52 +02:00
|
|
|
#include <linux/sched.h>
|
2010-05-24 23:32:26 +02:00
|
|
|
#include <linux/math64.h>
|
2010-10-26 23:21:36 +02:00
|
|
|
#include <linux/writeback.h>
|
2010-10-26 23:22:04 +02:00
|
|
|
#include <linux/compaction.h>
|
2013-09-11 23:22:36 +02:00
|
|
|
#include <linux/mm_inline.h>
|
|
|
|
|
|
|
|
#include "internal.h"
|
2006-06-30 10:55:32 +02:00
|
|
|
|
2006-06-30 10:55:45 +02:00
|
|
|
#ifdef CONFIG_VM_EVENT_COUNTERS
|
|
|
|
DEFINE_PER_CPU(struct vm_event_state, vm_event_states) = {{0}};
|
|
|
|
EXPORT_PER_CPU_SYMBOL(vm_event_states);
|
|
|
|
|
2010-08-10 02:18:59 +02:00
|
|
|
static void sum_vm_events(unsigned long *ret)
|
2006-06-30 10:55:45 +02:00
|
|
|
{
|
2008-02-05 07:29:22 +01:00
|
|
|
int cpu;
|
2006-06-30 10:55:45 +02:00
|
|
|
int i;
|
|
|
|
|
|
|
|
memset(ret, 0, NR_VM_EVENT_ITEMS * sizeof(unsigned long));
|
|
|
|
|
2010-08-10 02:18:59 +02:00
|
|
|
for_each_online_cpu(cpu) {
|
2006-06-30 10:55:45 +02:00
|
|
|
struct vm_event_state *this = &per_cpu(vm_event_states, cpu);
|
|
|
|
|
|
|
|
for (i = 0; i < NR_VM_EVENT_ITEMS; i++)
|
|
|
|
ret[i] += this->event[i];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Accumulate the vm event counters across all CPUs.
|
|
|
|
* The result is unavoidably approximate - it can change
|
|
|
|
* during and after execution of this function.
|
|
|
|
*/
|
|
|
|
void all_vm_events(unsigned long *ret)
|
|
|
|
{
|
2008-05-12 23:02:06 +02:00
|
|
|
get_online_cpus();
|
2010-08-10 02:18:59 +02:00
|
|
|
sum_vm_events(ret);
|
2008-05-12 23:02:06 +02:00
|
|
|
put_online_cpus();
|
2006-06-30 10:55:45 +02:00
|
|
|
}
|
2006-07-10 13:44:31 +02:00
|
|
|
EXPORT_SYMBOL_GPL(all_vm_events);
|
2006-06-30 10:55:45 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Fold the foreign cpu events into our own.
|
|
|
|
*
|
|
|
|
* This is adding to the events on one processor
|
|
|
|
* but keeps the global counts constant.
|
|
|
|
*/
|
|
|
|
void vm_events_fold_cpu(int cpu)
|
|
|
|
{
|
|
|
|
struct vm_event_state *fold_state = &per_cpu(vm_event_states, cpu);
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < NR_VM_EVENT_ITEMS; i++) {
|
|
|
|
count_vm_events(i, fold_state->event[i]);
|
|
|
|
fold_state->event[i] = 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* CONFIG_VM_EVENT_COUNTERS */
|
|
|
|
|
2006-06-30 10:55:33 +02:00
|
|
|
/*
|
|
|
|
* Manage combined zone based / global counters
|
|
|
|
*
|
|
|
|
* vm_stat contains the global counters
|
|
|
|
*/
|
2011-11-01 01:09:46 +01:00
|
|
|
atomic_long_t vm_stat[NR_VM_ZONE_STAT_ITEMS] __cacheline_aligned_in_smp;
|
2006-06-30 10:55:33 +02:00
|
|
|
EXPORT_SYMBOL(vm_stat);
|
|
|
|
|
|
|
|
#ifdef CONFIG_SMP
|
|
|
|
|
2011-01-14 00:45:43 +01:00
|
|
|
int calculate_pressure_threshold(struct zone *zone)
|
mm: page allocator: adjust the per-cpu counter threshold when memory is low
Commit aa45484 ("calculate a better estimate of NR_FREE_PAGES when memory
is low") noted that watermarks were based on the vmstat NR_FREE_PAGES. To
avoid synchronization overhead, these counters are maintained on a per-cpu
basis and drained both periodically and when a threshold is above a
threshold. On large CPU systems, the difference between the estimate and
real value of NR_FREE_PAGES can be very high. The system can get into a
case where pages are allocated far below the min watermark potentially
causing livelock issues. The commit solved the problem by taking a better
reading of NR_FREE_PAGES when memory was low.
Unfortately, as reported by Shaohua Li this accurate reading can consume a
large amount of CPU time on systems with many sockets due to cache line
bouncing. This patch takes a different approach. For large machines
where counter drift might be unsafe and while kswapd is awake, the per-cpu
thresholds for the target pgdat are reduced to limit the level of drift to
what should be a safe level. This incurs a performance penalty in heavy
memory pressure by a factor that depends on the workload and the machine
but the machine should function correctly without accidentally exhausting
all memory on a node. There is an additional cost when kswapd wakes and
sleeps but the event is not expected to be frequent - in Shaohua's test
case, there was one recorded sleep and wake event at least.
To ensure that kswapd wakes up, a safe version of zone_watermark_ok() is
introduced that takes a more accurate reading of NR_FREE_PAGES when called
from wakeup_kswapd, when deciding whether it is really safe to go back to
sleep in sleeping_prematurely() and when deciding if a zone is really
balanced or not in balance_pgdat(). We are still using an expensive
function but limiting how often it is called.
When the test case is reproduced, the time spent in the watermark
functions is reduced. The following report is on the percentage of time
spent cumulatively spent in the functions zone_nr_free_pages(),
zone_watermark_ok(), __zone_watermark_ok(), zone_watermark_ok_safe(),
zone_page_state_snapshot(), zone_page_state().
vanilla 11.6615%
disable-threshold 0.2584%
David said:
: We had to pull aa454840 "mm: page allocator: calculate a better estimate
: of NR_FREE_PAGES when memory is low and kswapd is awake" from 2.6.36
: internally because tests showed that it would cause the machine to stall
: as the result of heavy kswapd activity. I merged it back with this fix as
: it is pending in the -mm tree and it solves the issue we were seeing, so I
: definitely think this should be pushed to -stable (and I would seriously
: consider it for 2.6.37 inclusion even at this late date).
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Reported-by: Shaohua Li <shaohua.li@intel.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Nicolas Bareil <nico@chdir.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: <stable@kernel.org> [2.6.37.1, 2.6.36.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-01-14 00:45:41 +01:00
|
|
|
{
|
|
|
|
int threshold;
|
|
|
|
int watermark_distance;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* As vmstats are not up to date, there is drift between the estimated
|
|
|
|
* and real values. For high thresholds and a high number of CPUs, it
|
|
|
|
* is possible for the min watermark to be breached while the estimated
|
|
|
|
* value looks fine. The pressure threshold is a reduced value such
|
|
|
|
* that even the maximum amount of drift will not accidentally breach
|
|
|
|
* the min watermark
|
|
|
|
*/
|
|
|
|
watermark_distance = low_wmark_pages(zone) - min_wmark_pages(zone);
|
|
|
|
threshold = max(1, (int)(watermark_distance / num_online_cpus()));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Maximum threshold is 125
|
|
|
|
*/
|
|
|
|
threshold = min(125, threshold);
|
|
|
|
|
|
|
|
return threshold;
|
|
|
|
}
|
|
|
|
|
2011-01-14 00:45:43 +01:00
|
|
|
int calculate_normal_threshold(struct zone *zone)
|
2006-09-01 06:27:35 +02:00
|
|
|
{
|
|
|
|
int threshold;
|
|
|
|
int mem; /* memory in 128 MB units */
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The threshold scales with the number of processors and the amount
|
|
|
|
* of memory per zone. More memory means that we can defer updates for
|
|
|
|
* longer, more processors could lead to more contention.
|
|
|
|
* fls() is used to have a cheap way of logarithmic scaling.
|
|
|
|
*
|
|
|
|
* Some sample thresholds:
|
|
|
|
*
|
|
|
|
* Threshold Processors (fls) Zonesize fls(mem+1)
|
|
|
|
* ------------------------------------------------------------------
|
|
|
|
* 8 1 1 0.9-1 GB 4
|
|
|
|
* 16 2 2 0.9-1 GB 4
|
|
|
|
* 20 2 2 1-2 GB 5
|
|
|
|
* 24 2 2 2-4 GB 6
|
|
|
|
* 28 2 2 4-8 GB 7
|
|
|
|
* 32 2 2 8-16 GB 8
|
|
|
|
* 4 2 2 <128M 1
|
|
|
|
* 30 4 3 2-4 GB 5
|
|
|
|
* 48 4 3 8-16 GB 8
|
|
|
|
* 32 8 4 1-2 GB 4
|
|
|
|
* 32 8 4 0.9-1GB 4
|
|
|
|
* 10 16 5 <128M 1
|
|
|
|
* 40 16 5 900M 4
|
|
|
|
* 70 64 7 2-4 GB 5
|
|
|
|
* 84 64 7 4-8 GB 6
|
|
|
|
* 108 512 9 4-8 GB 6
|
|
|
|
* 125 1024 10 8-16 GB 8
|
|
|
|
* 125 1024 10 16-32 GB 9
|
|
|
|
*/
|
|
|
|
|
2013-02-23 01:33:52 +01:00
|
|
|
mem = zone->managed_pages >> (27 - PAGE_SHIFT);
|
2006-09-01 06:27:35 +02:00
|
|
|
|
|
|
|
threshold = 2 * fls(num_online_cpus()) * (1 + fls(mem));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Maximum threshold is 125
|
|
|
|
*/
|
|
|
|
threshold = min(125, threshold);
|
|
|
|
|
|
|
|
return threshold;
|
|
|
|
}
|
2006-06-30 10:55:33 +02:00
|
|
|
|
|
|
|
/*
|
2006-09-01 06:27:35 +02:00
|
|
|
* Refresh the thresholds for each zone.
|
2006-06-30 10:55:33 +02:00
|
|
|
*/
|
2011-05-25 02:11:33 +02:00
|
|
|
void refresh_zone_stat_thresholds(void)
|
2006-06-30 10:55:33 +02:00
|
|
|
{
|
2006-09-01 06:27:35 +02:00
|
|
|
struct zone *zone;
|
|
|
|
int cpu;
|
|
|
|
int threshold;
|
|
|
|
|
2009-04-01 00:19:31 +02:00
|
|
|
for_each_populated_zone(zone) {
|
2010-09-10 01:38:17 +02:00
|
|
|
unsigned long max_drift, tolerate_drift;
|
|
|
|
|
2011-01-14 00:45:43 +01:00
|
|
|
threshold = calculate_normal_threshold(zone);
|
2006-09-01 06:27:35 +02:00
|
|
|
|
|
|
|
for_each_online_cpu(cpu)
|
2010-01-05 07:34:51 +01:00
|
|
|
per_cpu_ptr(zone->pageset, cpu)->stat_threshold
|
|
|
|
= threshold;
|
2010-09-10 01:38:17 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Only set percpu_drift_mark if there is a danger that
|
|
|
|
* NR_FREE_PAGES reports the low watermark is ok when in fact
|
|
|
|
* the min watermark could be breached by an allocation
|
|
|
|
*/
|
|
|
|
tolerate_drift = low_wmark_pages(zone) - min_wmark_pages(zone);
|
|
|
|
max_drift = num_online_cpus() * threshold;
|
|
|
|
if (max_drift > tolerate_drift)
|
|
|
|
zone->percpu_drift_mark = high_wmark_pages(zone) +
|
|
|
|
max_drift;
|
2006-09-01 06:27:35 +02:00
|
|
|
}
|
2006-06-30 10:55:33 +02:00
|
|
|
}
|
|
|
|
|
2011-01-14 00:45:43 +01:00
|
|
|
void set_pgdat_percpu_threshold(pg_data_t *pgdat,
|
|
|
|
int (*calculate_pressure)(struct zone *))
|
mm: page allocator: adjust the per-cpu counter threshold when memory is low
Commit aa45484 ("calculate a better estimate of NR_FREE_PAGES when memory
is low") noted that watermarks were based on the vmstat NR_FREE_PAGES. To
avoid synchronization overhead, these counters are maintained on a per-cpu
basis and drained both periodically and when a threshold is above a
threshold. On large CPU systems, the difference between the estimate and
real value of NR_FREE_PAGES can be very high. The system can get into a
case where pages are allocated far below the min watermark potentially
causing livelock issues. The commit solved the problem by taking a better
reading of NR_FREE_PAGES when memory was low.
Unfortately, as reported by Shaohua Li this accurate reading can consume a
large amount of CPU time on systems with many sockets due to cache line
bouncing. This patch takes a different approach. For large machines
where counter drift might be unsafe and while kswapd is awake, the per-cpu
thresholds for the target pgdat are reduced to limit the level of drift to
what should be a safe level. This incurs a performance penalty in heavy
memory pressure by a factor that depends on the workload and the machine
but the machine should function correctly without accidentally exhausting
all memory on a node. There is an additional cost when kswapd wakes and
sleeps but the event is not expected to be frequent - in Shaohua's test
case, there was one recorded sleep and wake event at least.
To ensure that kswapd wakes up, a safe version of zone_watermark_ok() is
introduced that takes a more accurate reading of NR_FREE_PAGES when called
from wakeup_kswapd, when deciding whether it is really safe to go back to
sleep in sleeping_prematurely() and when deciding if a zone is really
balanced or not in balance_pgdat(). We are still using an expensive
function but limiting how often it is called.
When the test case is reproduced, the time spent in the watermark
functions is reduced. The following report is on the percentage of time
spent cumulatively spent in the functions zone_nr_free_pages(),
zone_watermark_ok(), __zone_watermark_ok(), zone_watermark_ok_safe(),
zone_page_state_snapshot(), zone_page_state().
vanilla 11.6615%
disable-threshold 0.2584%
David said:
: We had to pull aa454840 "mm: page allocator: calculate a better estimate
: of NR_FREE_PAGES when memory is low and kswapd is awake" from 2.6.36
: internally because tests showed that it would cause the machine to stall
: as the result of heavy kswapd activity. I merged it back with this fix as
: it is pending in the -mm tree and it solves the issue we were seeing, so I
: definitely think this should be pushed to -stable (and I would seriously
: consider it for 2.6.37 inclusion even at this late date).
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Reported-by: Shaohua Li <shaohua.li@intel.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Nicolas Bareil <nico@chdir.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: <stable@kernel.org> [2.6.37.1, 2.6.36.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-01-14 00:45:41 +01:00
|
|
|
{
|
|
|
|
struct zone *zone;
|
|
|
|
int cpu;
|
|
|
|
int threshold;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < pgdat->nr_zones; i++) {
|
|
|
|
zone = &pgdat->node_zones[i];
|
|
|
|
if (!zone->percpu_drift_mark)
|
|
|
|
continue;
|
|
|
|
|
2011-01-14 00:45:43 +01:00
|
|
|
threshold = (*calculate_pressure)(zone);
|
|
|
|
for_each_possible_cpu(cpu)
|
mm: page allocator: adjust the per-cpu counter threshold when memory is low
Commit aa45484 ("calculate a better estimate of NR_FREE_PAGES when memory
is low") noted that watermarks were based on the vmstat NR_FREE_PAGES. To
avoid synchronization overhead, these counters are maintained on a per-cpu
basis and drained both periodically and when a threshold is above a
threshold. On large CPU systems, the difference between the estimate and
real value of NR_FREE_PAGES can be very high. The system can get into a
case where pages are allocated far below the min watermark potentially
causing livelock issues. The commit solved the problem by taking a better
reading of NR_FREE_PAGES when memory was low.
Unfortately, as reported by Shaohua Li this accurate reading can consume a
large amount of CPU time on systems with many sockets due to cache line
bouncing. This patch takes a different approach. For large machines
where counter drift might be unsafe and while kswapd is awake, the per-cpu
thresholds for the target pgdat are reduced to limit the level of drift to
what should be a safe level. This incurs a performance penalty in heavy
memory pressure by a factor that depends on the workload and the machine
but the machine should function correctly without accidentally exhausting
all memory on a node. There is an additional cost when kswapd wakes and
sleeps but the event is not expected to be frequent - in Shaohua's test
case, there was one recorded sleep and wake event at least.
To ensure that kswapd wakes up, a safe version of zone_watermark_ok() is
introduced that takes a more accurate reading of NR_FREE_PAGES when called
from wakeup_kswapd, when deciding whether it is really safe to go back to
sleep in sleeping_prematurely() and when deciding if a zone is really
balanced or not in balance_pgdat(). We are still using an expensive
function but limiting how often it is called.
When the test case is reproduced, the time spent in the watermark
functions is reduced. The following report is on the percentage of time
spent cumulatively spent in the functions zone_nr_free_pages(),
zone_watermark_ok(), __zone_watermark_ok(), zone_watermark_ok_safe(),
zone_page_state_snapshot(), zone_page_state().
vanilla 11.6615%
disable-threshold 0.2584%
David said:
: We had to pull aa454840 "mm: page allocator: calculate a better estimate
: of NR_FREE_PAGES when memory is low and kswapd is awake" from 2.6.36
: internally because tests showed that it would cause the machine to stall
: as the result of heavy kswapd activity. I merged it back with this fix as
: it is pending in the -mm tree and it solves the issue we were seeing, so I
: definitely think this should be pushed to -stable (and I would seriously
: consider it for 2.6.37 inclusion even at this late date).
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Reported-by: Shaohua Li <shaohua.li@intel.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Nicolas Bareil <nico@chdir.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: <stable@kernel.org> [2.6.37.1, 2.6.36.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-01-14 00:45:41 +01:00
|
|
|
per_cpu_ptr(zone->pageset, cpu)->stat_threshold
|
|
|
|
= threshold;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-06-30 10:55:33 +02:00
|
|
|
/*
|
2014-06-05 01:09:51 +02:00
|
|
|
* For use when we know that interrupts are disabled,
|
|
|
|
* or when we know that preemption is disabled and that
|
|
|
|
* particular counter cannot be updated from interrupt context.
|
2006-06-30 10:55:33 +02:00
|
|
|
*/
|
|
|
|
void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item,
|
|
|
|
int delta)
|
|
|
|
{
|
2010-12-06 18:16:20 +01:00
|
|
|
struct per_cpu_pageset __percpu *pcp = zone->pageset;
|
|
|
|
s8 __percpu *p = pcp->vm_stat_diff + item;
|
2006-06-30 10:55:33 +02:00
|
|
|
long x;
|
2010-12-06 18:16:20 +01:00
|
|
|
long t;
|
|
|
|
|
|
|
|
x = delta + __this_cpu_read(*p);
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2010-12-06 18:16:20 +01:00
|
|
|
t = __this_cpu_read(pcp->stat_threshold);
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2010-12-06 18:16:20 +01:00
|
|
|
if (unlikely(x > t || x < -t)) {
|
2006-06-30 10:55:33 +02:00
|
|
|
zone_page_state_add(x, zone, item);
|
|
|
|
x = 0;
|
|
|
|
}
|
2010-12-06 18:16:20 +01:00
|
|
|
__this_cpu_write(*p, x);
|
2006-06-30 10:55:33 +02:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(__mod_zone_page_state);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Optimized increment and decrement functions.
|
|
|
|
*
|
|
|
|
* These are only for a single page and therefore can take a struct page *
|
|
|
|
* argument instead of struct zone *. This allows the inclusion of the code
|
|
|
|
* generated for page_zone(page) into the optimized functions.
|
|
|
|
*
|
|
|
|
* No overflow check is necessary and therefore the differential can be
|
|
|
|
* incremented or decremented in place which may allow the compilers to
|
|
|
|
* generate better code.
|
|
|
|
* The increment or decrement is known and therefore one boundary check can
|
|
|
|
* be omitted.
|
|
|
|
*
|
2006-09-01 06:27:35 +02:00
|
|
|
* NOTE: These functions are very performance sensitive. Change only
|
|
|
|
* with care.
|
|
|
|
*
|
2006-06-30 10:55:33 +02:00
|
|
|
* Some processors have inc/dec instructions that are atomic vs an interrupt.
|
|
|
|
* However, the code must first determine the differential location in a zone
|
|
|
|
* based on the processor number and then inc/dec the counter. There is no
|
|
|
|
* guarantee without disabling preemption that the processor will not change
|
|
|
|
* in between and therefore the atomicity vs. interrupt cannot be exploited
|
|
|
|
* in a useful way here.
|
|
|
|
*/
|
2007-02-10 10:43:01 +01:00
|
|
|
void __inc_zone_state(struct zone *zone, enum zone_stat_item item)
|
2006-06-30 10:55:33 +02:00
|
|
|
{
|
2010-12-06 18:16:20 +01:00
|
|
|
struct per_cpu_pageset __percpu *pcp = zone->pageset;
|
|
|
|
s8 __percpu *p = pcp->vm_stat_diff + item;
|
|
|
|
s8 v, t;
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2010-12-06 18:40:02 +01:00
|
|
|
v = __this_cpu_inc_return(*p);
|
2010-12-06 18:16:20 +01:00
|
|
|
t = __this_cpu_read(pcp->stat_threshold);
|
|
|
|
if (unlikely(v > t)) {
|
|
|
|
s8 overstep = t >> 1;
|
2006-09-01 06:27:35 +02:00
|
|
|
|
2010-12-06 18:16:20 +01:00
|
|
|
zone_page_state_add(v + overstep, zone, item);
|
|
|
|
__this_cpu_write(*p, -overstep);
|
2006-06-30 10:55:33 +02:00
|
|
|
}
|
|
|
|
}
|
2006-06-30 10:55:44 +02:00
|
|
|
|
|
|
|
void __inc_zone_page_state(struct page *page, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
__inc_zone_state(page_zone(page), item);
|
|
|
|
}
|
2006-06-30 10:55:33 +02:00
|
|
|
EXPORT_SYMBOL(__inc_zone_page_state);
|
|
|
|
|
2007-02-10 10:43:01 +01:00
|
|
|
void __dec_zone_state(struct zone *zone, enum zone_stat_item item)
|
2006-06-30 10:55:33 +02:00
|
|
|
{
|
2010-12-06 18:16:20 +01:00
|
|
|
struct per_cpu_pageset __percpu *pcp = zone->pageset;
|
|
|
|
s8 __percpu *p = pcp->vm_stat_diff + item;
|
|
|
|
s8 v, t;
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2010-12-06 18:40:02 +01:00
|
|
|
v = __this_cpu_dec_return(*p);
|
2010-12-06 18:16:20 +01:00
|
|
|
t = __this_cpu_read(pcp->stat_threshold);
|
|
|
|
if (unlikely(v < - t)) {
|
|
|
|
s8 overstep = t >> 1;
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2010-12-06 18:16:20 +01:00
|
|
|
zone_page_state_add(v - overstep, zone, item);
|
|
|
|
__this_cpu_write(*p, overstep);
|
2006-06-30 10:55:33 +02:00
|
|
|
}
|
|
|
|
}
|
2007-02-10 10:43:01 +01:00
|
|
|
|
|
|
|
void __dec_zone_page_state(struct page *page, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
__dec_zone_state(page_zone(page), item);
|
|
|
|
}
|
2006-06-30 10:55:33 +02:00
|
|
|
EXPORT_SYMBOL(__dec_zone_page_state);
|
|
|
|
|
2012-01-13 02:17:30 +01:00
|
|
|
#ifdef CONFIG_HAVE_CMPXCHG_LOCAL
|
2010-12-14 17:28:46 +01:00
|
|
|
/*
|
|
|
|
* If we have cmpxchg_local support then we do not need to incur the overhead
|
|
|
|
* that comes with local_irq_save/restore if we use this_cpu_cmpxchg.
|
|
|
|
*
|
|
|
|
* mod_state() modifies the zone counter state through atomic per cpu
|
|
|
|
* operations.
|
|
|
|
*
|
|
|
|
* Overstep mode specifies how overstep should handled:
|
|
|
|
* 0 No overstepping
|
|
|
|
* 1 Overstepping half of threshold
|
|
|
|
* -1 Overstepping minus half of threshold
|
|
|
|
*/
|
|
|
|
static inline void mod_state(struct zone *zone,
|
|
|
|
enum zone_stat_item item, int delta, int overstep_mode)
|
|
|
|
{
|
|
|
|
struct per_cpu_pageset __percpu *pcp = zone->pageset;
|
|
|
|
s8 __percpu *p = pcp->vm_stat_diff + item;
|
|
|
|
long o, n, t, z;
|
|
|
|
|
|
|
|
do {
|
|
|
|
z = 0; /* overflow to zone counters */
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The fetching of the stat_threshold is racy. We may apply
|
|
|
|
* a counter threshold to the wrong the cpu if we get
|
2011-04-15 00:21:58 +02:00
|
|
|
* rescheduled while executing here. However, the next
|
|
|
|
* counter update will apply the threshold again and
|
|
|
|
* therefore bring the counter under the threshold again.
|
|
|
|
*
|
|
|
|
* Most of the time the thresholds are the same anyways
|
|
|
|
* for all cpus in a zone.
|
2010-12-14 17:28:46 +01:00
|
|
|
*/
|
|
|
|
t = this_cpu_read(pcp->stat_threshold);
|
|
|
|
|
|
|
|
o = this_cpu_read(*p);
|
|
|
|
n = delta + o;
|
|
|
|
|
|
|
|
if (n > t || n < -t) {
|
|
|
|
int os = overstep_mode * (t >> 1) ;
|
|
|
|
|
|
|
|
/* Overflow must be added to zone counters */
|
|
|
|
z = n + os;
|
|
|
|
n = -os;
|
|
|
|
}
|
|
|
|
} while (this_cpu_cmpxchg(*p, o, n) != o);
|
|
|
|
|
|
|
|
if (z)
|
|
|
|
zone_page_state_add(z, zone, item);
|
|
|
|
}
|
|
|
|
|
|
|
|
void mod_zone_page_state(struct zone *zone, enum zone_stat_item item,
|
|
|
|
int delta)
|
|
|
|
{
|
|
|
|
mod_state(zone, item, delta, 0);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(mod_zone_page_state);
|
|
|
|
|
|
|
|
void inc_zone_state(struct zone *zone, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
mod_state(zone, item, 1, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
void inc_zone_page_state(struct page *page, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
mod_state(page_zone(page), item, 1, 1);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(inc_zone_page_state);
|
|
|
|
|
|
|
|
void dec_zone_page_state(struct page *page, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
mod_state(page_zone(page), item, -1, -1);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(dec_zone_page_state);
|
|
|
|
#else
|
|
|
|
/*
|
|
|
|
* Use interrupt disable to serialize counter updates
|
|
|
|
*/
|
|
|
|
void mod_zone_page_state(struct zone *zone, enum zone_stat_item item,
|
|
|
|
int delta)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
|
|
|
__mod_zone_page_state(zone, item, delta);
|
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(mod_zone_page_state);
|
|
|
|
|
2006-06-30 10:55:44 +02:00
|
|
|
void inc_zone_state(struct zone *zone, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
|
|
|
__inc_zone_state(zone, item);
|
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
|
|
|
|
2006-06-30 10:55:33 +02:00
|
|
|
void inc_zone_page_state(struct page *page, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
struct zone *zone;
|
|
|
|
|
|
|
|
zone = page_zone(page);
|
|
|
|
local_irq_save(flags);
|
2006-06-30 10:55:44 +02:00
|
|
|
__inc_zone_state(zone, item);
|
2006-06-30 10:55:33 +02:00
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(inc_zone_page_state);
|
|
|
|
|
|
|
|
void dec_zone_page_state(struct page *page, enum zone_stat_item item)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
2006-09-01 06:27:34 +02:00
|
|
|
__dec_zone_page_state(page, item);
|
2006-06-30 10:55:33 +02:00
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(dec_zone_page_state);
|
2010-12-14 17:28:46 +01:00
|
|
|
#endif
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2013-09-11 23:21:31 +02:00
|
|
|
static inline void fold_diff(int *diff)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
|
|
|
|
if (diff[i])
|
|
|
|
atomic_long_add(diff[i], &vm_stat[i]);
|
|
|
|
}
|
|
|
|
|
2006-06-30 10:55:33 +02:00
|
|
|
/*
|
2013-09-11 23:21:30 +02:00
|
|
|
* Update the zone counters for the current cpu.
|
2008-02-05 07:29:16 +01:00
|
|
|
*
|
2007-05-09 11:35:14 +02:00
|
|
|
* Note that refresh_cpu_vm_stats strives to only access
|
|
|
|
* node local memory. The per cpu pagesets on remote zones are placed
|
|
|
|
* in the memory local to the processor using that pageset. So the
|
|
|
|
* loop over all zones will access a series of cachelines local to
|
|
|
|
* the processor.
|
|
|
|
*
|
|
|
|
* The call to zone_page_state_add updates the cachelines with the
|
|
|
|
* statistics in the remote zone struct as well as the global cachelines
|
|
|
|
* with the global counters. These could cause remote node cache line
|
|
|
|
* bouncing and will have to be only done when necessary.
|
2006-06-30 10:55:33 +02:00
|
|
|
*/
|
2013-09-11 23:21:32 +02:00
|
|
|
static void refresh_cpu_vm_stats(void)
|
2006-06-30 10:55:33 +02:00
|
|
|
{
|
|
|
|
struct zone *zone;
|
|
|
|
int i;
|
2008-02-05 07:29:16 +01:00
|
|
|
int global_diff[NR_VM_ZONE_STAT_ITEMS] = { 0, };
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2009-04-01 00:19:31 +02:00
|
|
|
for_each_populated_zone(zone) {
|
2013-09-11 23:21:32 +02:00
|
|
|
struct per_cpu_pageset __percpu *p = zone->pageset;
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2013-09-11 23:21:32 +02:00
|
|
|
for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++) {
|
|
|
|
int v;
|
2006-06-30 10:55:33 +02:00
|
|
|
|
2013-09-11 23:21:32 +02:00
|
|
|
v = this_cpu_xchg(p->vm_stat_diff[i], 0);
|
|
|
|
if (v) {
|
2008-02-05 07:29:16 +01:00
|
|
|
|
|
|
|
atomic_long_add(v, &zone->vm_stat[i]);
|
|
|
|
global_diff[i] += v;
|
2007-05-09 11:35:14 +02:00
|
|
|
#ifdef CONFIG_NUMA
|
|
|
|
/* 3 seconds idle till flush */
|
2013-09-11 23:21:32 +02:00
|
|
|
__this_cpu_write(p->expire, 3);
|
2007-05-09 11:35:14 +02:00
|
|
|
#endif
|
2006-06-30 10:55:33 +02:00
|
|
|
}
|
2013-09-11 23:21:32 +02:00
|
|
|
}
|
2008-04-28 11:13:37 +02:00
|
|
|
cond_resched();
|
2007-05-09 11:35:14 +02:00
|
|
|
#ifdef CONFIG_NUMA
|
|
|
|
/*
|
|
|
|
* Deal with draining the remote pageset of this
|
|
|
|
* processor
|
|
|
|
*
|
|
|
|
* Check if there are pages remaining in this pageset
|
|
|
|
* if not then there is nothing to expire.
|
|
|
|
*/
|
2013-09-11 23:21:32 +02:00
|
|
|
if (!__this_cpu_read(p->expire) ||
|
|
|
|
!__this_cpu_read(p->pcp.count))
|
2007-05-09 11:35:14 +02:00
|
|
|
continue;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We never drain zones local to this processor.
|
|
|
|
*/
|
|
|
|
if (zone_to_nid(zone) == numa_node_id()) {
|
2013-09-11 23:21:32 +02:00
|
|
|
__this_cpu_write(p->expire, 0);
|
2007-05-09 11:35:14 +02:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2013-09-11 23:21:32 +02:00
|
|
|
|
|
|
|
if (__this_cpu_dec_return(p->expire))
|
2007-05-09 11:35:14 +02:00
|
|
|
continue;
|
|
|
|
|
2013-09-11 23:21:32 +02:00
|
|
|
if (__this_cpu_read(p->pcp.count))
|
2014-06-05 01:07:56 +02:00
|
|
|
drain_zone_pages(zone, this_cpu_ptr(&p->pcp));
|
2007-05-09 11:35:14 +02:00
|
|
|
#endif
|
2006-06-30 10:55:33 +02:00
|
|
|
}
|
2013-09-11 23:21:31 +02:00
|
|
|
fold_diff(global_diff);
|
2006-06-30 10:55:33 +02:00
|
|
|
}
|
|
|
|
|
2013-09-11 23:21:30 +02:00
|
|
|
/*
|
|
|
|
* Fold the data for an offline cpu into the global array.
|
|
|
|
* There cannot be any access by the offline cpu and therefore
|
|
|
|
* synchronization is simplified.
|
|
|
|
*/
|
|
|
|
void cpu_vm_stats_fold(int cpu)
|
|
|
|
{
|
|
|
|
struct zone *zone;
|
|
|
|
int i;
|
|
|
|
int global_diff[NR_VM_ZONE_STAT_ITEMS] = { 0, };
|
|
|
|
|
|
|
|
for_each_populated_zone(zone) {
|
|
|
|
struct per_cpu_pageset *p;
|
|
|
|
|
|
|
|
p = per_cpu_ptr(zone->pageset, cpu);
|
|
|
|
|
|
|
|
for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
|
|
|
|
if (p->vm_stat_diff[i]) {
|
|
|
|
int v;
|
|
|
|
|
|
|
|
v = p->vm_stat_diff[i];
|
|
|
|
p->vm_stat_diff[i] = 0;
|
|
|
|
atomic_long_add(v, &zone->vm_stat[i]);
|
|
|
|
global_diff[i] += v;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-09-11 23:21:31 +02:00
|
|
|
fold_diff(global_diff);
|
2013-09-11 23:21:30 +02:00
|
|
|
}
|
|
|
|
|
2013-04-30 00:08:38 +02:00
|
|
|
/*
|
|
|
|
* this is only called if !populated_zone(zone), which implies no other users of
|
|
|
|
* pset->vm_stat_diff[] exsist.
|
|
|
|
*/
|
2012-10-09 01:33:39 +02:00
|
|
|
void drain_zonestat(struct zone *zone, struct per_cpu_pageset *pset)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
|
|
|
|
if (pset->vm_stat_diff[i]) {
|
|
|
|
int v = pset->vm_stat_diff[i];
|
|
|
|
pset->vm_stat_diff[i] = 0;
|
|
|
|
atomic_long_add(v, &zone->vm_stat[i]);
|
|
|
|
atomic_long_add(v, &vm_stat[i]);
|
|
|
|
}
|
|
|
|
}
|
2006-06-30 10:55:33 +02:00
|
|
|
#endif
|
|
|
|
|
2006-06-30 10:55:44 +02:00
|
|
|
#ifdef CONFIG_NUMA
|
|
|
|
/*
|
|
|
|
* zonelist = the list of zones passed to the allocator
|
|
|
|
* z = the zone from which the allocation occurred.
|
|
|
|
*
|
|
|
|
* Must be called with interrupts disabled.
|
2011-03-23 00:33:12 +01:00
|
|
|
*
|
|
|
|
* When __GFP_OTHER_NODE is set assume the node of the preferred
|
|
|
|
* zone is the local node. This is useful for daemons who allocate
|
|
|
|
* memory on behalf of other processes.
|
2006-06-30 10:55:44 +02:00
|
|
|
*/
|
2011-03-23 00:33:12 +01:00
|
|
|
void zone_statistics(struct zone *preferred_zone, struct zone *z, gfp_t flags)
|
2006-06-30 10:55:44 +02:00
|
|
|
{
|
2008-04-28 11:12:14 +02:00
|
|
|
if (z->zone_pgdat == preferred_zone->zone_pgdat) {
|
2006-06-30 10:55:44 +02:00
|
|
|
__inc_zone_state(z, NUMA_HIT);
|
|
|
|
} else {
|
|
|
|
__inc_zone_state(z, NUMA_MISS);
|
2008-04-28 11:12:14 +02:00
|
|
|
__inc_zone_state(preferred_zone, NUMA_FOREIGN);
|
2006-06-30 10:55:44 +02:00
|
|
|
}
|
2011-03-23 00:33:12 +01:00
|
|
|
if (z->node == ((flags & __GFP_OTHER_NODE) ?
|
|
|
|
preferred_zone->node : numa_node_id()))
|
2006-06-30 10:55:44 +02:00
|
|
|
__inc_zone_state(z, NUMA_LOCAL);
|
|
|
|
else
|
|
|
|
__inc_zone_state(z, NUMA_OTHER);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2010-05-24 23:32:25 +02:00
|
|
|
#ifdef CONFIG_COMPACTION
|
2010-10-26 23:22:04 +02:00
|
|
|
|
2010-05-24 23:32:25 +02:00
|
|
|
struct contig_page_info {
|
|
|
|
unsigned long free_pages;
|
|
|
|
unsigned long free_blocks_total;
|
|
|
|
unsigned long free_blocks_suitable;
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Calculate the number of free pages in a zone, how many contiguous
|
|
|
|
* pages are free and how many are large enough to satisfy an allocation of
|
|
|
|
* the target size. Note that this function makes no attempt to estimate
|
|
|
|
* how many suitable free blocks there *might* be if MOVABLE pages were
|
|
|
|
* migrated. Calculating that is possible, but expensive and can be
|
|
|
|
* figured out from userspace
|
|
|
|
*/
|
|
|
|
static void fill_contig_page_info(struct zone *zone,
|
|
|
|
unsigned int suitable_order,
|
|
|
|
struct contig_page_info *info)
|
|
|
|
{
|
|
|
|
unsigned int order;
|
|
|
|
|
|
|
|
info->free_pages = 0;
|
|
|
|
info->free_blocks_total = 0;
|
|
|
|
info->free_blocks_suitable = 0;
|
|
|
|
|
|
|
|
for (order = 0; order < MAX_ORDER; order++) {
|
|
|
|
unsigned long blocks;
|
|
|
|
|
|
|
|
/* Count number of free blocks */
|
|
|
|
blocks = zone->free_area[order].nr_free;
|
|
|
|
info->free_blocks_total += blocks;
|
|
|
|
|
|
|
|
/* Count free base pages */
|
|
|
|
info->free_pages += blocks << order;
|
|
|
|
|
|
|
|
/* Count the suitable free blocks */
|
|
|
|
if (order >= suitable_order)
|
|
|
|
info->free_blocks_suitable += blocks <<
|
|
|
|
(order - suitable_order);
|
|
|
|
}
|
|
|
|
}
|
2010-05-24 23:32:26 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* A fragmentation index only makes sense if an allocation of a requested
|
|
|
|
* size would fail. If that is true, the fragmentation index indicates
|
|
|
|
* whether external fragmentation or a lack of memory was the problem.
|
|
|
|
* The value can be used to determine if page reclaim or compaction
|
|
|
|
* should be used
|
|
|
|
*/
|
2010-05-24 23:32:30 +02:00
|
|
|
static int __fragmentation_index(unsigned int order, struct contig_page_info *info)
|
2010-05-24 23:32:26 +02:00
|
|
|
{
|
|
|
|
unsigned long requested = 1UL << order;
|
|
|
|
|
|
|
|
if (!info->free_blocks_total)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/* Fragmentation index only makes sense when a request would fail */
|
|
|
|
if (info->free_blocks_suitable)
|
|
|
|
return -1000;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Index is between 0 and 1 so return within 3 decimal places
|
|
|
|
*
|
|
|
|
* 0 => allocation would fail due to lack of memory
|
|
|
|
* 1 => allocation would fail due to fragmentation
|
|
|
|
*/
|
|
|
|
return 1000 - div_u64( (1000+(div_u64(info->free_pages * 1000ULL, requested))), info->free_blocks_total);
|
|
|
|
}
|
2010-05-24 23:32:30 +02:00
|
|
|
|
|
|
|
/* Same as __fragmentation index but allocs contig_page_info on stack */
|
|
|
|
int fragmentation_index(struct zone *zone, unsigned int order)
|
|
|
|
{
|
|
|
|
struct contig_page_info info;
|
|
|
|
|
|
|
|
fill_contig_page_info(zone, order, &info);
|
|
|
|
return __fragmentation_index(order, &info);
|
|
|
|
}
|
2010-05-24 23:32:25 +02:00
|
|
|
#endif
|
|
|
|
|
|
|
|
#if defined(CONFIG_PROC_FS) || defined(CONFIG_COMPACTION)
|
2008-10-06 02:13:52 +02:00
|
|
|
#include <linux/proc_fs.h>
|
2006-06-30 10:55:32 +02:00
|
|
|
#include <linux/seq_file.h>
|
|
|
|
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
static char * const migratetype_names[MIGRATE_TYPES] = {
|
|
|
|
"Unmovable",
|
|
|
|
"Reclaimable",
|
|
|
|
"Movable",
|
|
|
|
"Reserve",
|
2011-12-29 13:09:50 +01:00
|
|
|
#ifdef CONFIG_CMA
|
|
|
|
"CMA",
|
|
|
|
#endif
|
2013-02-23 01:33:58 +01:00
|
|
|
#ifdef CONFIG_MEMORY_ISOLATION
|
add "Isolate" migratetype name to /proc/pagetypeinfo
In a5d76b54a3f3a40385d7f76069a2feac9f1bad63 (memory unplug: page isolation by
KAMEZAWA Hiroyuki), "isolate" migratetype added. but unfortunately, it
doesn't treat /proc/pagetypeinfo display logic.
this patch add "Isolate" to pagetype name field.
/proc/pagetype
before:
------------------------------------------------------------------------------------------------------------------------
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 1 2 2 2 1 2 2 1 1 0 0
Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 2 3 3 1 3 3 2 0 0 0 0
Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 0 1
Node 0, zone DMA, type <NULL> 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Unmovable 1 9 7 4 1 1 1 1 0 0 0
Node 0, zone Normal, type Reclaimable 5 2 0 0 1 1 0 0 0 1 0
Node 0, zone Normal, type Movable 0 1 1 0 0 0 1 0 0 1 60
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 1
Node 0, zone Normal, type <NULL> 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone HighMem, type Unmovable 0 0 1 1 1 0 1 1 2 2 0
Node 0, zone HighMem, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone HighMem, type Movable 236 62 6 2 2 1 1 0 1 1 16
Node 0, zone HighMem, type Reserve 0 0 0 0 0 0 0 0 0 0 1
Node 0, zone HighMem, type <NULL> 0 0 0 0 0 0 0 0 0 0 0
Number of blocks type Unmovable Reclaimable Movable Reserve <NULL>
Node 0, zone DMA 1 0 2 1 0
Node 0, zone Normal 10 40 169 1 0
Node 0, zone HighMem 2 0 283 1 0
after:
------------------------------------------------------------------------------------------------------------------------
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 1 2 2 2 1 2 2 1 1 0 0
Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 2 3 3 1 3 3 2 0 0 0 0
Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 0 1
Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Unmovable 0 2 1 1 0 1 0 0 0 0 0
Node 0, zone Normal, type Reclaimable 1 1 1 1 1 0 1 1 1 0 0
Node 0, zone Normal, type Movable 0 1 1 1 0 1 0 1 0 0 196
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 1
Node 0, zone Normal, type Isolate 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone HighMem, type Unmovable 0 1 0 0 0 1 1 1 2 2 0
Node 0, zone HighMem, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone HighMem, type Movable 1 0 1 1 0 0 0 0 1 0 200
Node 0, zone HighMem, type Reserve 0 0 0 0 0 0 0 0 0 0 1
Node 0, zone HighMem, type Isolate 0 0 0 0 0 0 0 0 0 0 0
Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
Node 0, zone DMA 1 0 2 1 0
Node 0, zone Normal 8 4 207 1 0
Node 0, zone HighMem 2 0 283 1 0
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-15 23:34:42 +02:00
|
|
|
"Isolate",
|
2013-02-23 01:33:58 +01:00
|
|
|
#endif
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
};
|
|
|
|
|
2006-06-30 10:55:32 +02:00
|
|
|
static void *frag_start(struct seq_file *m, loff_t *pos)
|
|
|
|
{
|
|
|
|
pg_data_t *pgdat;
|
|
|
|
loff_t node = *pos;
|
|
|
|
for (pgdat = first_online_pgdat();
|
|
|
|
pgdat && node;
|
|
|
|
pgdat = next_online_pgdat(pgdat))
|
|
|
|
--node;
|
|
|
|
|
|
|
|
return pgdat;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void *frag_next(struct seq_file *m, void *arg, loff_t *pos)
|
|
|
|
{
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
|
|
|
|
(*pos)++;
|
|
|
|
return next_online_pgdat(pgdat);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void frag_stop(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
/* Walk all the zones in a node and print using a callback */
|
|
|
|
static void walk_zones_in_node(struct seq_file *m, pg_data_t *pgdat,
|
|
|
|
void (*print)(struct seq_file *m, pg_data_t *, struct zone *))
|
2006-06-30 10:55:32 +02:00
|
|
|
{
|
|
|
|
struct zone *zone;
|
|
|
|
struct zone *node_zones = pgdat->node_zones;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
for (zone = node_zones; zone - node_zones < MAX_NR_ZONES; ++zone) {
|
|
|
|
if (!populated_zone(zone))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&zone->lock, flags);
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
print(m, pgdat, zone);
|
2006-06-30 10:55:32 +02:00
|
|
|
spin_unlock_irqrestore(&zone->lock, flags);
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
}
|
|
|
|
}
|
2010-05-24 23:32:25 +02:00
|
|
|
#endif
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
|
2011-09-15 01:21:05 +02:00
|
|
|
#if defined(CONFIG_PROC_FS) || defined(CONFIG_SYSFS) || defined(CONFIG_NUMA)
|
2011-05-25 02:11:28 +02:00
|
|
|
#ifdef CONFIG_ZONE_DMA
|
|
|
|
#define TEXT_FOR_DMA(xx) xx "_dma",
|
|
|
|
#else
|
|
|
|
#define TEXT_FOR_DMA(xx)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef CONFIG_ZONE_DMA32
|
|
|
|
#define TEXT_FOR_DMA32(xx) xx "_dma32",
|
|
|
|
#else
|
|
|
|
#define TEXT_FOR_DMA32(xx)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef CONFIG_HIGHMEM
|
|
|
|
#define TEXT_FOR_HIGHMEM(xx) xx "_high",
|
|
|
|
#else
|
|
|
|
#define TEXT_FOR_HIGHMEM(xx)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#define TEXTS_FOR_ZONES(xx) TEXT_FOR_DMA(xx) TEXT_FOR_DMA32(xx) xx "_normal", \
|
|
|
|
TEXT_FOR_HIGHMEM(xx) xx "_movable",
|
|
|
|
|
|
|
|
const char * const vmstat_text[] = {
|
|
|
|
/* Zoned VM counters */
|
|
|
|
"nr_free_pages",
|
mm: page_alloc: fair zone allocator policy
Each zone that holds userspace pages of one workload must be aged at a
speed proportional to the zone size. Otherwise, the time an individual
page gets to stay in memory depends on the zone it happened to be
allocated in. Asymmetry in the zone aging creates rather unpredictable
aging behavior and results in the wrong pages being reclaimed, activated
etc.
But exactly this happens right now because of the way the page allocator
and kswapd interact. The page allocator uses per-node lists of all zones
in the system, ordered by preference, when allocating a new page. When
the first iteration does not yield any results, kswapd is woken up and the
allocator retries. Due to the way kswapd reclaims zones below the high
watermark while a zone can be allocated from when it is above the low
watermark, the allocator may keep kswapd running while kswapd reclaim
ensures that the page allocator can keep allocating from the first zone in
the zonelist for extended periods of time. Meanwhile the other zones
rarely see new allocations and thus get aged much slower in comparison.
The result is that the occasional page placed in lower zones gets
relatively more time in memory, even gets promoted to the active list
after its peers have long been evicted. Meanwhile, the bulk of the
working set may be thrashing on the preferred zone even though there may
be significant amounts of memory available in the lower zones.
Even the most basic test -- repeatedly reading a file slightly bigger than
memory -- shows how broken the zone aging is. In this scenario, no single
page should be able stay in memory long enough to get referenced twice and
activated, but activation happens in spades:
$ grep active_file /proc/zoneinfo
nr_inactive_file 0
nr_active_file 0
nr_inactive_file 0
nr_active_file 8
nr_inactive_file 1582
nr_active_file 11994
$ cat data data data data >/dev/null
$ grep active_file /proc/zoneinfo
nr_inactive_file 0
nr_active_file 70
nr_inactive_file 258753
nr_active_file 443214
nr_inactive_file 149793
nr_active_file 12021
Fix this with a very simple round robin allocator. Each zone is allowed a
batch of allocations that is proportional to the zone's size, after which
it is treated as full. The batch counters are reset when all zones have
been tried and the allocator enters the slowpath and kicks off kswapd
reclaim. Allocation and reclaim is now fairly spread out to all
available/allowable zones:
$ grep active_file /proc/zoneinfo
nr_inactive_file 0
nr_active_file 0
nr_inactive_file 174
nr_active_file 4865
nr_inactive_file 53
nr_active_file 860
$ cat data data data data >/dev/null
$ grep active_file /proc/zoneinfo
nr_inactive_file 0
nr_active_file 0
nr_inactive_file 666622
nr_active_file 4988
nr_inactive_file 190969
nr_active_file 937
When zone_reclaim_mode is enabled, allocations will now spread out to all
zones on the local node, not just the first preferred zone (which on a 4G
node might be a tiny Normal zone).
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Paul Bolle <paul.bollee@gmail.com>
Cc: Zlatko Calusic <zcalusic@bitsync.net>
Tested-by: Kevin Hilman <khilman@linaro.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-09-11 23:20:47 +02:00
|
|
|
"nr_alloc_batch",
|
2011-05-25 02:11:28 +02:00
|
|
|
"nr_inactive_anon",
|
|
|
|
"nr_active_anon",
|
|
|
|
"nr_inactive_file",
|
|
|
|
"nr_active_file",
|
|
|
|
"nr_unevictable",
|
|
|
|
"nr_mlock",
|
|
|
|
"nr_anon_pages",
|
|
|
|
"nr_mapped",
|
|
|
|
"nr_file_pages",
|
|
|
|
"nr_dirty",
|
|
|
|
"nr_writeback",
|
|
|
|
"nr_slab_reclaimable",
|
|
|
|
"nr_slab_unreclaimable",
|
|
|
|
"nr_page_table_pages",
|
|
|
|
"nr_kernel_stack",
|
|
|
|
"nr_unstable",
|
|
|
|
"nr_bounce",
|
|
|
|
"nr_vmscan_write",
|
2011-11-01 01:07:59 +01:00
|
|
|
"nr_vmscan_immediate_reclaim",
|
2011-05-25 02:11:28 +02:00
|
|
|
"nr_writeback_temp",
|
|
|
|
"nr_isolated_anon",
|
|
|
|
"nr_isolated_file",
|
|
|
|
"nr_shmem",
|
|
|
|
"nr_dirtied",
|
|
|
|
"nr_written",
|
|
|
|
|
|
|
|
#ifdef CONFIG_NUMA
|
|
|
|
"numa_hit",
|
|
|
|
"numa_miss",
|
|
|
|
"numa_foreign",
|
|
|
|
"numa_interleave",
|
|
|
|
"numa_local",
|
|
|
|
"numa_other",
|
|
|
|
#endif
|
2014-04-03 23:47:51 +02:00
|
|
|
"workingset_refault",
|
|
|
|
"workingset_activate",
|
mm: keep page cache radix tree nodes in check
Previously, page cache radix tree nodes were freed after reclaim emptied
out their page pointers. But now reclaim stores shadow entries in their
place, which are only reclaimed when the inodes themselves are
reclaimed. This is problematic for bigger files that are still in use
after they have a significant amount of their cache reclaimed, without
any of those pages actually refaulting. The shadow entries will just
sit there and waste memory. In the worst case, the shadow entries will
accumulate until the machine runs out of memory.
To get this under control, the VM will track radix tree nodes
exclusively containing shadow entries on a per-NUMA node list. Per-NUMA
rather than global because we expect the radix tree nodes themselves to
be allocated node-locally and we want to reduce cross-node references of
otherwise independent cache workloads. A simple shrinker will then
reclaim these nodes on memory pressure.
A few things need to be stored in the radix tree node to implement the
shadow node LRU and allow tree deletions coming from the list:
1. There is no index available that would describe the reverse path
from the node up to the tree root, which is needed to perform a
deletion. To solve this, encode in each node its offset inside the
parent. This can be stored in the unused upper bits of the same
member that stores the node's height at no extra space cost.
2. The number of shadow entries needs to be counted in addition to the
regular entries, to quickly detect when the node is ready to go to
the shadow node LRU list. The current entry count is an unsigned
int but the maximum number of entries is 64, so a shadow counter
can easily be stored in the unused upper bits.
3. Tree modification needs tree lock and tree root, which are located
in the address space, so store an address_space backpointer in the
node. The parent pointer of the node is in a union with the 2-word
rcu_head, so the backpointer comes at no extra cost as well.
4. The node needs to be linked to an LRU list, which requires a list
head inside the node. This does increase the size of the node, but
it does not change the number of objects that fit into a slab page.
[akpm@linux-foundation.org: export the right function]
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Rik van Riel <riel@redhat.com>
Reviewed-by: Minchan Kim <minchan@kernel.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Bob Liu <bob.liu@oracle.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Metin Doslu <metin@citusdata.com>
Cc: Michel Lespinasse <walken@google.com>
Cc: Ozgun Erdogan <ozgun@citusdata.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Roman Gushchin <klamm@yandex-team.ru>
Cc: Ryan Mallon <rmallon@gmail.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-03 23:47:56 +02:00
|
|
|
"workingset_nodereclaim",
|
2011-05-25 02:11:28 +02:00
|
|
|
"nr_anon_transparent_hugepages",
|
2012-10-09 01:32:02 +02:00
|
|
|
"nr_free_cma",
|
2011-05-25 02:11:28 +02:00
|
|
|
"nr_dirty_threshold",
|
|
|
|
"nr_dirty_background_threshold",
|
|
|
|
|
|
|
|
#ifdef CONFIG_VM_EVENT_COUNTERS
|
|
|
|
"pgpgin",
|
|
|
|
"pgpgout",
|
|
|
|
"pswpin",
|
|
|
|
"pswpout",
|
|
|
|
|
|
|
|
TEXTS_FOR_ZONES("pgalloc")
|
|
|
|
|
|
|
|
"pgfree",
|
|
|
|
"pgactivate",
|
|
|
|
"pgdeactivate",
|
|
|
|
|
|
|
|
"pgfault",
|
|
|
|
"pgmajfault",
|
|
|
|
|
|
|
|
TEXTS_FOR_ZONES("pgrefill")
|
2012-04-26 01:01:48 +02:00
|
|
|
TEXTS_FOR_ZONES("pgsteal_kswapd")
|
|
|
|
TEXTS_FOR_ZONES("pgsteal_direct")
|
2011-05-25 02:11:28 +02:00
|
|
|
TEXTS_FOR_ZONES("pgscan_kswapd")
|
|
|
|
TEXTS_FOR_ZONES("pgscan_direct")
|
2012-08-01 01:44:39 +02:00
|
|
|
"pgscan_direct_throttle",
|
2011-05-25 02:11:28 +02:00
|
|
|
|
|
|
|
#ifdef CONFIG_NUMA
|
|
|
|
"zone_reclaim_failed",
|
|
|
|
#endif
|
|
|
|
"pginodesteal",
|
|
|
|
"slabs_scanned",
|
|
|
|
"kswapd_inodesteal",
|
|
|
|
"kswapd_low_wmark_hit_quickly",
|
|
|
|
"kswapd_high_wmark_hit_quickly",
|
|
|
|
"pageoutrun",
|
|
|
|
"allocstall",
|
|
|
|
|
|
|
|
"pgrotated",
|
|
|
|
|
2014-04-03 23:48:19 +02:00
|
|
|
"drop_pagecache",
|
|
|
|
"drop_slab",
|
|
|
|
|
2012-11-02 15:52:48 +01:00
|
|
|
#ifdef CONFIG_NUMA_BALANCING
|
|
|
|
"numa_pte_updates",
|
2013-11-13 00:08:32 +01:00
|
|
|
"numa_huge_pte_updates",
|
2012-11-02 15:52:48 +01:00
|
|
|
"numa_hint_faults",
|
|
|
|
"numa_hint_faults_local",
|
|
|
|
"numa_pages_migrated",
|
|
|
|
#endif
|
2012-10-19 11:46:20 +02:00
|
|
|
#ifdef CONFIG_MIGRATION
|
|
|
|
"pgmigrate_success",
|
|
|
|
"pgmigrate_fail",
|
|
|
|
#endif
|
2011-05-25 02:11:28 +02:00
|
|
|
#ifdef CONFIG_COMPACTION
|
2012-10-19 13:00:10 +02:00
|
|
|
"compact_migrate_scanned",
|
|
|
|
"compact_free_scanned",
|
|
|
|
"compact_isolated",
|
2011-05-25 02:11:28 +02:00
|
|
|
"compact_stall",
|
|
|
|
"compact_fail",
|
|
|
|
"compact_success",
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef CONFIG_HUGETLB_PAGE
|
|
|
|
"htlb_buddy_alloc_success",
|
|
|
|
"htlb_buddy_alloc_fail",
|
|
|
|
#endif
|
|
|
|
"unevictable_pgs_culled",
|
|
|
|
"unevictable_pgs_scanned",
|
|
|
|
"unevictable_pgs_rescued",
|
|
|
|
"unevictable_pgs_mlocked",
|
|
|
|
"unevictable_pgs_munlocked",
|
|
|
|
"unevictable_pgs_cleared",
|
|
|
|
"unevictable_pgs_stranded",
|
|
|
|
|
|
|
|
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
|
|
|
|
"thp_fault_alloc",
|
|
|
|
"thp_fault_fallback",
|
|
|
|
"thp_collapse_alloc",
|
|
|
|
"thp_collapse_alloc_failed",
|
|
|
|
"thp_split",
|
2012-12-12 22:51:09 +01:00
|
|
|
"thp_zero_page_alloc",
|
|
|
|
"thp_zero_page_alloc_failed",
|
2011-05-25 02:11:28 +02:00
|
|
|
#endif
|
2014-01-21 23:33:16 +01:00
|
|
|
#ifdef CONFIG_DEBUG_TLBFLUSH
|
2013-09-11 23:20:24 +02:00
|
|
|
#ifdef CONFIG_SMP
|
2013-09-11 23:20:23 +02:00
|
|
|
"nr_tlb_remote_flush",
|
|
|
|
"nr_tlb_remote_flush_received",
|
2014-01-21 23:33:16 +01:00
|
|
|
#endif /* CONFIG_SMP */
|
2013-09-11 23:20:23 +02:00
|
|
|
"nr_tlb_local_flush_all",
|
|
|
|
"nr_tlb_local_flush_one",
|
2014-01-21 23:33:16 +01:00
|
|
|
#endif /* CONFIG_DEBUG_TLBFLUSH */
|
2011-05-25 02:11:28 +02:00
|
|
|
|
2014-06-05 01:06:46 +02:00
|
|
|
#ifdef CONFIG_DEBUG_VM_VMACACHE
|
|
|
|
"vmacache_find_calls",
|
|
|
|
"vmacache_find_hits",
|
|
|
|
#endif
|
2011-05-25 02:11:28 +02:00
|
|
|
#endif /* CONFIG_VM_EVENTS_COUNTERS */
|
|
|
|
};
|
2011-09-15 01:21:05 +02:00
|
|
|
#endif /* CONFIG_PROC_FS || CONFIG_SYSFS || CONFIG_NUMA */
|
2011-05-25 02:11:28 +02:00
|
|
|
|
|
|
|
|
2010-05-24 23:32:25 +02:00
|
|
|
#ifdef CONFIG_PROC_FS
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
static void frag_show_print(struct seq_file *m, pg_data_t *pgdat,
|
|
|
|
struct zone *zone)
|
|
|
|
{
|
|
|
|
int order;
|
|
|
|
|
|
|
|
seq_printf(m, "Node %d, zone %8s ", pgdat->node_id, zone->name);
|
|
|
|
for (order = 0; order < MAX_ORDER; ++order)
|
|
|
|
seq_printf(m, "%6lu ", zone->free_area[order].nr_free);
|
|
|
|
seq_putc(m, '\n');
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This walks the free areas for each zone.
|
|
|
|
*/
|
|
|
|
static int frag_show(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
walk_zones_in_node(m, pgdat, frag_show_print);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void pagetypeinfo_showfree_print(struct seq_file *m,
|
|
|
|
pg_data_t *pgdat, struct zone *zone)
|
|
|
|
{
|
|
|
|
int order, mtype;
|
|
|
|
|
|
|
|
for (mtype = 0; mtype < MIGRATE_TYPES; mtype++) {
|
|
|
|
seq_printf(m, "Node %4d, zone %8s, type %12s ",
|
|
|
|
pgdat->node_id,
|
|
|
|
zone->name,
|
|
|
|
migratetype_names[mtype]);
|
|
|
|
for (order = 0; order < MAX_ORDER; ++order) {
|
|
|
|
unsigned long freecount = 0;
|
|
|
|
struct free_area *area;
|
|
|
|
struct list_head *curr;
|
|
|
|
|
|
|
|
area = &(zone->free_area[order]);
|
|
|
|
|
|
|
|
list_for_each(curr, &area->free_list[mtype])
|
|
|
|
freecount++;
|
|
|
|
seq_printf(m, "%6lu ", freecount);
|
|
|
|
}
|
2006-06-30 10:55:32 +02:00
|
|
|
seq_putc(m, '\n');
|
|
|
|
}
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Print out the free pages at each order for each migatetype */
|
|
|
|
static int pagetypeinfo_showfree(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
int order;
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
|
|
|
|
/* Print header */
|
|
|
|
seq_printf(m, "%-43s ", "Free pages count per migrate type at order");
|
|
|
|
for (order = 0; order < MAX_ORDER; ++order)
|
|
|
|
seq_printf(m, "%6d ", order);
|
|
|
|
seq_putc(m, '\n');
|
|
|
|
|
|
|
|
walk_zones_in_node(m, pgdat, pagetypeinfo_showfree_print);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void pagetypeinfo_showblockcount_print(struct seq_file *m,
|
|
|
|
pg_data_t *pgdat, struct zone *zone)
|
|
|
|
{
|
|
|
|
int mtype;
|
|
|
|
unsigned long pfn;
|
|
|
|
unsigned long start_pfn = zone->zone_start_pfn;
|
2013-02-23 01:35:23 +01:00
|
|
|
unsigned long end_pfn = zone_end_pfn(zone);
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
unsigned long count[MIGRATE_TYPES] = { 0, };
|
|
|
|
|
|
|
|
for (pfn = start_pfn; pfn < end_pfn; pfn += pageblock_nr_pages) {
|
|
|
|
struct page *page;
|
|
|
|
|
|
|
|
if (!pfn_valid(pfn))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
page = pfn_to_page(pfn);
|
2009-05-13 18:34:48 +02:00
|
|
|
|
|
|
|
/* Watch for unexpected holes punched in the memmap */
|
|
|
|
if (!memmap_valid_within(pfn, page, zone))
|
2008-08-14 12:10:14 +02:00
|
|
|
continue;
|
2009-05-13 18:34:48 +02:00
|
|
|
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
mtype = get_pageblock_migratetype(page);
|
|
|
|
|
2008-08-14 12:10:14 +02:00
|
|
|
if (mtype < MIGRATE_TYPES)
|
|
|
|
count[mtype]++;
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Print counts */
|
|
|
|
seq_printf(m, "Node %d, zone %8s ", pgdat->node_id, zone->name);
|
|
|
|
for (mtype = 0; mtype < MIGRATE_TYPES; mtype++)
|
|
|
|
seq_printf(m, "%12lu ", count[mtype]);
|
|
|
|
seq_putc(m, '\n');
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Print out the free pages at each order for each migratetype */
|
|
|
|
static int pagetypeinfo_showblockcount(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
int mtype;
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
|
|
|
|
seq_printf(m, "\n%-23s", "Number of blocks type ");
|
|
|
|
for (mtype = 0; mtype < MIGRATE_TYPES; mtype++)
|
|
|
|
seq_printf(m, "%12s ", migratetype_names[mtype]);
|
|
|
|
seq_putc(m, '\n');
|
|
|
|
walk_zones_in_node(m, pgdat, pagetypeinfo_showblockcount_print);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This prints out statistics in relation to grouping pages by mobility.
|
|
|
|
* It is expensive to collect so do not constantly read the file.
|
|
|
|
*/
|
|
|
|
static int pagetypeinfo_show(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
|
2008-04-30 09:52:13 +02:00
|
|
|
/* check memoryless node */
|
2012-12-12 22:51:37 +01:00
|
|
|
if (!node_state(pgdat->node_id, N_MEMORY))
|
2008-04-30 09:52:13 +02:00
|
|
|
return 0;
|
|
|
|
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
seq_printf(m, "Page block order: %d\n", pageblock_order);
|
|
|
|
seq_printf(m, "Pages per block: %lu\n", pageblock_nr_pages);
|
|
|
|
seq_putc(m, '\n');
|
|
|
|
pagetypeinfo_showfree(m, pgdat);
|
|
|
|
pagetypeinfo_showblockcount(m, pgdat);
|
|
|
|
|
2006-06-30 10:55:32 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-10-06 02:13:52 +02:00
|
|
|
static const struct seq_operations fragmentation_op = {
|
2006-06-30 10:55:32 +02:00
|
|
|
.start = frag_start,
|
|
|
|
.next = frag_next,
|
|
|
|
.stop = frag_stop,
|
|
|
|
.show = frag_show,
|
|
|
|
};
|
|
|
|
|
2008-10-06 02:13:52 +02:00
|
|
|
static int fragmentation_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
return seq_open(file, &fragmentation_op);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations fragmentation_file_operations = {
|
|
|
|
.open = fragmentation_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = seq_release,
|
|
|
|
};
|
|
|
|
|
2008-10-06 02:15:36 +02:00
|
|
|
static const struct seq_operations pagetypeinfo_op = {
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
.start = frag_start,
|
|
|
|
.next = frag_next,
|
|
|
|
.stop = frag_stop,
|
|
|
|
.show = pagetypeinfo_show,
|
|
|
|
};
|
|
|
|
|
2008-10-06 02:15:36 +02:00
|
|
|
static int pagetypeinfo_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
return seq_open(file, &pagetypeinfo_op);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations pagetypeinfo_file_ops = {
|
|
|
|
.open = pagetypeinfo_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = seq_release,
|
|
|
|
};
|
|
|
|
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
|
|
|
|
struct zone *zone)
|
2006-06-30 10:55:32 +02:00
|
|
|
{
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
int i;
|
|
|
|
seq_printf(m, "Node %d, zone %8s", pgdat->node_id, zone->name);
|
|
|
|
seq_printf(m,
|
|
|
|
"\n pages free %lu"
|
|
|
|
"\n min %lu"
|
|
|
|
"\n low %lu"
|
|
|
|
"\n high %lu"
|
2009-06-17 00:32:30 +02:00
|
|
|
"\n scanned %lu"
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
"\n spanned %lu"
|
2012-12-12 22:52:12 +01:00
|
|
|
"\n present %lu"
|
|
|
|
"\n managed %lu",
|
mm: page allocator: adjust the per-cpu counter threshold when memory is low
Commit aa45484 ("calculate a better estimate of NR_FREE_PAGES when memory
is low") noted that watermarks were based on the vmstat NR_FREE_PAGES. To
avoid synchronization overhead, these counters are maintained on a per-cpu
basis and drained both periodically and when a threshold is above a
threshold. On large CPU systems, the difference between the estimate and
real value of NR_FREE_PAGES can be very high. The system can get into a
case where pages are allocated far below the min watermark potentially
causing livelock issues. The commit solved the problem by taking a better
reading of NR_FREE_PAGES when memory was low.
Unfortately, as reported by Shaohua Li this accurate reading can consume a
large amount of CPU time on systems with many sockets due to cache line
bouncing. This patch takes a different approach. For large machines
where counter drift might be unsafe and while kswapd is awake, the per-cpu
thresholds for the target pgdat are reduced to limit the level of drift to
what should be a safe level. This incurs a performance penalty in heavy
memory pressure by a factor that depends on the workload and the machine
but the machine should function correctly without accidentally exhausting
all memory on a node. There is an additional cost when kswapd wakes and
sleeps but the event is not expected to be frequent - in Shaohua's test
case, there was one recorded sleep and wake event at least.
To ensure that kswapd wakes up, a safe version of zone_watermark_ok() is
introduced that takes a more accurate reading of NR_FREE_PAGES when called
from wakeup_kswapd, when deciding whether it is really safe to go back to
sleep in sleeping_prematurely() and when deciding if a zone is really
balanced or not in balance_pgdat(). We are still using an expensive
function but limiting how often it is called.
When the test case is reproduced, the time spent in the watermark
functions is reduced. The following report is on the percentage of time
spent cumulatively spent in the functions zone_nr_free_pages(),
zone_watermark_ok(), __zone_watermark_ok(), zone_watermark_ok_safe(),
zone_page_state_snapshot(), zone_page_state().
vanilla 11.6615%
disable-threshold 0.2584%
David said:
: We had to pull aa454840 "mm: page allocator: calculate a better estimate
: of NR_FREE_PAGES when memory is low and kswapd is awake" from 2.6.36
: internally because tests showed that it would cause the machine to stall
: as the result of heavy kswapd activity. I merged it back with this fix as
: it is pending in the -mm tree and it solves the issue we were seeing, so I
: definitely think this should be pushed to -stable (and I would seriously
: consider it for 2.6.37 inclusion even at this late date).
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Reported-by: Shaohua Li <shaohua.li@intel.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Nicolas Bareil <nico@chdir.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: <stable@kernel.org> [2.6.37.1, 2.6.36.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-01-14 00:45:41 +01:00
|
|
|
zone_page_state(zone, NR_FREE_PAGES),
|
2009-06-17 00:32:12 +02:00
|
|
|
min_wmark_pages(zone),
|
|
|
|
low_wmark_pages(zone),
|
|
|
|
high_wmark_pages(zone),
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
zone->pages_scanned,
|
|
|
|
zone->spanned_pages,
|
2012-12-12 22:52:12 +01:00
|
|
|
zone->present_pages,
|
|
|
|
zone->managed_pages);
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
|
|
|
|
for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
|
|
|
|
seq_printf(m, "\n %-12s %lu", vmstat_text[i],
|
|
|
|
zone_page_state(zone, i));
|
|
|
|
|
|
|
|
seq_printf(m,
|
2014-08-07 01:07:14 +02:00
|
|
|
"\n protection: (%ld",
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
zone->lowmem_reserve[0]);
|
|
|
|
for (i = 1; i < ARRAY_SIZE(zone->lowmem_reserve); i++)
|
2014-08-07 01:07:14 +02:00
|
|
|
seq_printf(m, ", %ld", zone->lowmem_reserve[i]);
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
seq_printf(m,
|
|
|
|
")"
|
|
|
|
"\n pagesets");
|
|
|
|
for_each_online_cpu(i) {
|
|
|
|
struct per_cpu_pageset *pageset;
|
|
|
|
|
2010-01-05 07:34:51 +01:00
|
|
|
pageset = per_cpu_ptr(zone->pageset, i);
|
2008-02-05 07:29:19 +01:00
|
|
|
seq_printf(m,
|
|
|
|
"\n cpu: %i"
|
|
|
|
"\n count: %i"
|
|
|
|
"\n high: %i"
|
|
|
|
"\n batch: %i",
|
|
|
|
i,
|
|
|
|
pageset->pcp.count,
|
|
|
|
pageset->pcp.high,
|
|
|
|
pageset->pcp.batch);
|
2006-09-01 06:27:35 +02:00
|
|
|
#ifdef CONFIG_SMP
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
seq_printf(m, "\n vm stats threshold: %d",
|
|
|
|
pageset->stat_threshold);
|
2006-09-01 06:27:35 +02:00
|
|
|
#endif
|
2006-06-30 10:55:32 +02:00
|
|
|
}
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
seq_printf(m,
|
|
|
|
"\n all_unreclaimable: %u"
|
2008-10-19 05:26:34 +02:00
|
|
|
"\n start_pfn: %lu"
|
|
|
|
"\n inactive_ratio: %u",
|
2013-09-11 23:22:36 +02:00
|
|
|
!zone_reclaimable(zone),
|
2008-10-19 05:26:34 +02:00
|
|
|
zone->zone_start_pfn,
|
|
|
|
zone->inactive_ratio);
|
Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo
This patch provides fragmentation avoidance statistics via /proc/pagetypeinfo.
The information is collected only on request so there is no runtime overhead.
The statistics are in three parts:
The first part prints information on the size of blocks that pages are
being grouped on and looks like
Page block order: 10
Pages per block: 1024
The second part is a more detailed version of /proc/buddyinfo and looks like
Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
Node 0, zone DMA, type Unmovable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reclaimable 1 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Movable 0 0 0 0 0 0 0 0 0 0 0
Node 0, zone DMA, type Reserve 0 4 4 0 0 0 0 1 0 1 0
Node 0, zone Normal, type Unmovable 111 8 4 4 2 3 1 0 0 0 0
Node 0, zone Normal, type Reclaimable 293 89 8 0 0 0 0 0 0 0 0
Node 0, zone Normal, type Movable 1 6 13 9 7 6 3 0 0 0 0
Node 0, zone Normal, type Reserve 0 0 0 0 0 0 0 0 0 0 4
The third part looks like
Number of blocks type Unmovable Reclaimable Movable Reserve
Node 0, zone DMA 0 1 2 1
Node 0, zone Normal 3 17 94 4
To walk the zones within a node with interrupts disabled, walk_zones_in_node()
is introduced and shared between /proc/buddyinfo, /proc/zoneinfo and
/proc/pagetypeinfo to reduce code duplication. It seems specific to what
vmstat.c requires but could be broken out as a general utility function in
mmzone.c if there were other other potential users.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:26:02 +02:00
|
|
|
seq_putc(m, '\n');
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Output information about zones in @pgdat.
|
|
|
|
*/
|
|
|
|
static int zoneinfo_show(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
walk_zones_in_node(m, pgdat, zoneinfo_show_print);
|
2006-06-30 10:55:32 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-10-06 02:19:42 +02:00
|
|
|
static const struct seq_operations zoneinfo_op = {
|
2006-06-30 10:55:32 +02:00
|
|
|
.start = frag_start, /* iterate over all zones. The same as in
|
|
|
|
* fragmentation. */
|
|
|
|
.next = frag_next,
|
|
|
|
.stop = frag_stop,
|
|
|
|
.show = zoneinfo_show,
|
|
|
|
};
|
|
|
|
|
2008-10-06 02:19:42 +02:00
|
|
|
static int zoneinfo_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
return seq_open(file, &zoneinfo_op);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations proc_zoneinfo_file_operations = {
|
|
|
|
.open = zoneinfo_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = seq_release,
|
|
|
|
};
|
|
|
|
|
2010-10-26 23:21:36 +02:00
|
|
|
enum writeback_stat_item {
|
|
|
|
NR_DIRTY_THRESHOLD,
|
|
|
|
NR_DIRTY_BG_THRESHOLD,
|
|
|
|
NR_VM_WRITEBACK_STAT_ITEMS,
|
|
|
|
};
|
|
|
|
|
2006-06-30 10:55:32 +02:00
|
|
|
static void *vmstat_start(struct seq_file *m, loff_t *pos)
|
|
|
|
{
|
2006-06-30 10:55:33 +02:00
|
|
|
unsigned long *v;
|
2010-10-26 23:21:36 +02:00
|
|
|
int i, stat_items_size;
|
2006-06-30 10:55:32 +02:00
|
|
|
|
|
|
|
if (*pos >= ARRAY_SIZE(vmstat_text))
|
|
|
|
return NULL;
|
2010-10-26 23:21:36 +02:00
|
|
|
stat_items_size = NR_VM_ZONE_STAT_ITEMS * sizeof(unsigned long) +
|
|
|
|
NR_VM_WRITEBACK_STAT_ITEMS * sizeof(unsigned long);
|
2006-06-30 10:55:32 +02:00
|
|
|
|
2006-06-30 10:55:45 +02:00
|
|
|
#ifdef CONFIG_VM_EVENT_COUNTERS
|
2010-10-26 23:21:36 +02:00
|
|
|
stat_items_size += sizeof(struct vm_event_state);
|
2006-06-30 10:55:45 +02:00
|
|
|
#endif
|
2010-10-26 23:21:36 +02:00
|
|
|
|
|
|
|
v = kmalloc(stat_items_size, GFP_KERNEL);
|
2006-06-30 10:55:33 +02:00
|
|
|
m->private = v;
|
|
|
|
if (!v)
|
2006-06-30 10:55:32 +02:00
|
|
|
return ERR_PTR(-ENOMEM);
|
2006-06-30 10:55:33 +02:00
|
|
|
for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
|
|
|
|
v[i] = global_page_state(i);
|
2010-10-26 23:21:36 +02:00
|
|
|
v += NR_VM_ZONE_STAT_ITEMS;
|
|
|
|
|
|
|
|
global_dirty_limits(v + NR_DIRTY_BG_THRESHOLD,
|
|
|
|
v + NR_DIRTY_THRESHOLD);
|
|
|
|
v += NR_VM_WRITEBACK_STAT_ITEMS;
|
|
|
|
|
2006-06-30 10:55:45 +02:00
|
|
|
#ifdef CONFIG_VM_EVENT_COUNTERS
|
2010-10-26 23:21:36 +02:00
|
|
|
all_vm_events(v);
|
|
|
|
v[PGPGIN] /= 2; /* sectors -> kbytes */
|
|
|
|
v[PGPGOUT] /= 2;
|
2006-06-30 10:55:45 +02:00
|
|
|
#endif
|
2010-11-03 18:56:49 +01:00
|
|
|
return (unsigned long *)m->private + *pos;
|
2006-06-30 10:55:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static void *vmstat_next(struct seq_file *m, void *arg, loff_t *pos)
|
|
|
|
{
|
|
|
|
(*pos)++;
|
|
|
|
if (*pos >= ARRAY_SIZE(vmstat_text))
|
|
|
|
return NULL;
|
|
|
|
return (unsigned long *)m->private + *pos;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int vmstat_show(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
unsigned long *l = arg;
|
|
|
|
unsigned long off = l - (unsigned long *)m->private;
|
|
|
|
|
|
|
|
seq_printf(m, "%s %lu\n", vmstat_text[off], *l);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void vmstat_stop(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
kfree(m->private);
|
|
|
|
m->private = NULL;
|
|
|
|
}
|
|
|
|
|
2008-10-06 02:17:48 +02:00
|
|
|
static const struct seq_operations vmstat_op = {
|
2006-06-30 10:55:32 +02:00
|
|
|
.start = vmstat_start,
|
|
|
|
.next = vmstat_next,
|
|
|
|
.stop = vmstat_stop,
|
|
|
|
.show = vmstat_show,
|
|
|
|
};
|
|
|
|
|
2008-10-06 02:17:48 +02:00
|
|
|
static int vmstat_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
return seq_open(file, &vmstat_op);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations proc_vmstat_file_operations = {
|
|
|
|
.open = vmstat_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = seq_release,
|
|
|
|
};
|
2006-06-30 10:55:32 +02:00
|
|
|
#endif /* CONFIG_PROC_FS */
|
|
|
|
|
2006-09-01 06:27:35 +02:00
|
|
|
#ifdef CONFIG_SMP
|
2007-05-09 11:35:12 +02:00
|
|
|
static DEFINE_PER_CPU(struct delayed_work, vmstat_work);
|
2007-05-09 11:35:13 +02:00
|
|
|
int sysctl_stat_interval __read_mostly = HZ;
|
2007-05-09 11:35:12 +02:00
|
|
|
|
|
|
|
static void vmstat_update(struct work_struct *w)
|
|
|
|
{
|
2013-09-11 23:21:32 +02:00
|
|
|
refresh_cpu_vm_stats();
|
2014-06-05 01:07:56 +02:00
|
|
|
schedule_delayed_work(this_cpu_ptr(&vmstat_work),
|
2009-04-03 01:56:39 +02:00
|
|
|
round_jiffies_relative(sysctl_stat_interval));
|
2007-05-09 11:35:12 +02:00
|
|
|
}
|
|
|
|
|
2013-06-19 20:53:51 +02:00
|
|
|
static void start_cpu_timer(int cpu)
|
2007-05-09 11:35:12 +02:00
|
|
|
{
|
2009-10-29 14:34:13 +01:00
|
|
|
struct delayed_work *work = &per_cpu(vmstat_work, cpu);
|
2007-05-09 11:35:12 +02:00
|
|
|
|
2012-08-21 22:18:23 +02:00
|
|
|
INIT_DEFERRABLE_WORK(work, vmstat_update);
|
2009-10-29 14:34:13 +01:00
|
|
|
schedule_delayed_work_on(cpu, work, __round_jiffies_relative(HZ, cpu));
|
2007-05-09 11:35:12 +02:00
|
|
|
}
|
|
|
|
|
2013-11-13 00:08:13 +01:00
|
|
|
static void vmstat_cpu_dead(int node)
|
|
|
|
{
|
|
|
|
int cpu;
|
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
for_each_online_cpu(cpu)
|
|
|
|
if (cpu_to_node(cpu) == node)
|
|
|
|
goto end;
|
|
|
|
|
|
|
|
node_clear_state(node, N_CPU);
|
|
|
|
end:
|
|
|
|
put_online_cpus();
|
|
|
|
}
|
|
|
|
|
2006-09-01 06:27:35 +02:00
|
|
|
/*
|
|
|
|
* Use the cpu notifier to insure that the thresholds are recalculated
|
|
|
|
* when necessary.
|
|
|
|
*/
|
2013-06-19 20:53:51 +02:00
|
|
|
static int vmstat_cpuup_callback(struct notifier_block *nfb,
|
2006-09-01 06:27:35 +02:00
|
|
|
unsigned long action,
|
|
|
|
void *hcpu)
|
|
|
|
{
|
2007-05-09 11:35:12 +02:00
|
|
|
long cpu = (long)hcpu;
|
|
|
|
|
2006-09-01 06:27:35 +02:00
|
|
|
switch (action) {
|
2007-05-09 11:35:12 +02:00
|
|
|
case CPU_ONLINE:
|
|
|
|
case CPU_ONLINE_FROZEN:
|
2010-09-10 01:38:14 +02:00
|
|
|
refresh_zone_stat_thresholds();
|
2007-05-09 11:35:12 +02:00
|
|
|
start_cpu_timer(cpu);
|
2010-01-05 07:34:51 +01:00
|
|
|
node_set_state(cpu_to_node(cpu), N_CPU);
|
2007-05-09 11:35:12 +02:00
|
|
|
break;
|
|
|
|
case CPU_DOWN_PREPARE:
|
|
|
|
case CPU_DOWN_PREPARE_FROZEN:
|
2010-12-14 16:21:17 +01:00
|
|
|
cancel_delayed_work_sync(&per_cpu(vmstat_work, cpu));
|
2007-05-09 11:35:12 +02:00
|
|
|
per_cpu(vmstat_work, cpu).work.func = NULL;
|
|
|
|
break;
|
|
|
|
case CPU_DOWN_FAILED:
|
|
|
|
case CPU_DOWN_FAILED_FROZEN:
|
|
|
|
start_cpu_timer(cpu);
|
|
|
|
break;
|
2006-12-07 05:33:08 +01:00
|
|
|
case CPU_DEAD:
|
2007-05-09 11:35:10 +02:00
|
|
|
case CPU_DEAD_FROZEN:
|
2006-12-07 05:33:08 +01:00
|
|
|
refresh_zone_stat_thresholds();
|
2013-11-13 00:08:13 +01:00
|
|
|
vmstat_cpu_dead(cpu_to_node(cpu));
|
2006-12-07 05:33:08 +01:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
break;
|
2006-09-01 06:27:35 +02:00
|
|
|
}
|
|
|
|
return NOTIFY_OK;
|
|
|
|
}
|
|
|
|
|
2013-06-19 20:53:51 +02:00
|
|
|
static struct notifier_block vmstat_notifier =
|
2006-09-01 06:27:35 +02:00
|
|
|
{ &vmstat_cpuup_callback, NULL, 0 };
|
2008-10-06 02:13:52 +02:00
|
|
|
#endif
|
2006-09-01 06:27:35 +02:00
|
|
|
|
2007-10-16 10:26:27 +02:00
|
|
|
static int __init setup_vmstat(void)
|
2006-09-01 06:27:35 +02:00
|
|
|
{
|
2008-10-06 02:13:52 +02:00
|
|
|
#ifdef CONFIG_SMP
|
2007-05-09 11:35:12 +02:00
|
|
|
int cpu;
|
|
|
|
|
2014-03-10 21:42:27 +01:00
|
|
|
cpu_notifier_register_begin();
|
|
|
|
__register_cpu_notifier(&vmstat_notifier);
|
2007-05-09 11:35:12 +02:00
|
|
|
|
2013-11-13 00:08:12 +01:00
|
|
|
for_each_online_cpu(cpu) {
|
2007-05-09 11:35:12 +02:00
|
|
|
start_cpu_timer(cpu);
|
2013-11-13 00:08:12 +01:00
|
|
|
node_set_state(cpu_to_node(cpu), N_CPU);
|
|
|
|
}
|
2014-03-10 21:42:27 +01:00
|
|
|
cpu_notifier_register_done();
|
2008-10-06 02:13:52 +02:00
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_PROC_FS
|
|
|
|
proc_create("buddyinfo", S_IRUGO, NULL, &fragmentation_file_operations);
|
2008-10-06 02:15:36 +02:00
|
|
|
proc_create("pagetypeinfo", S_IRUGO, NULL, &pagetypeinfo_file_ops);
|
2008-10-06 02:17:48 +02:00
|
|
|
proc_create("vmstat", S_IRUGO, NULL, &proc_vmstat_file_operations);
|
2008-10-06 02:19:42 +02:00
|
|
|
proc_create("zoneinfo", S_IRUGO, NULL, &proc_zoneinfo_file_operations);
|
2008-10-06 02:13:52 +02:00
|
|
|
#endif
|
2006-09-01 06:27:35 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
module_init(setup_vmstat)
|
2010-05-24 23:32:25 +02:00
|
|
|
|
|
|
|
#if defined(CONFIG_DEBUG_FS) && defined(CONFIG_COMPACTION)
|
|
|
|
#include <linux/debugfs.h>
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Return an index indicating how much of the available free memory is
|
|
|
|
* unusable for an allocation of the requested size.
|
|
|
|
*/
|
|
|
|
static int unusable_free_index(unsigned int order,
|
|
|
|
struct contig_page_info *info)
|
|
|
|
{
|
|
|
|
/* No free memory is interpreted as all free memory is unusable */
|
|
|
|
if (info->free_pages == 0)
|
|
|
|
return 1000;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Index should be a value between 0 and 1. Return a value to 3
|
|
|
|
* decimal places.
|
|
|
|
*
|
|
|
|
* 0 => no fragmentation
|
|
|
|
* 1 => high fragmentation
|
|
|
|
*/
|
|
|
|
return div_u64((info->free_pages - (info->free_blocks_suitable << order)) * 1000ULL, info->free_pages);
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
static void unusable_show_print(struct seq_file *m,
|
|
|
|
pg_data_t *pgdat, struct zone *zone)
|
|
|
|
{
|
|
|
|
unsigned int order;
|
|
|
|
int index;
|
|
|
|
struct contig_page_info info;
|
|
|
|
|
|
|
|
seq_printf(m, "Node %d, zone %8s ",
|
|
|
|
pgdat->node_id,
|
|
|
|
zone->name);
|
|
|
|
for (order = 0; order < MAX_ORDER; ++order) {
|
|
|
|
fill_contig_page_info(zone, order, &info);
|
|
|
|
index = unusable_free_index(order, &info);
|
|
|
|
seq_printf(m, "%d.%03d ", index / 1000, index % 1000);
|
|
|
|
}
|
|
|
|
|
|
|
|
seq_putc(m, '\n');
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Display unusable free space index
|
|
|
|
*
|
|
|
|
* The unusable free space index measures how much of the available free
|
|
|
|
* memory cannot be used to satisfy an allocation of a given size and is a
|
|
|
|
* value between 0 and 1. The higher the value, the more of free memory is
|
|
|
|
* unusable and by implication, the worse the external fragmentation is. This
|
|
|
|
* can be expressed as a percentage by multiplying by 100.
|
|
|
|
*/
|
|
|
|
static int unusable_show(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
|
|
|
|
/* check memoryless node */
|
2012-12-12 22:51:37 +01:00
|
|
|
if (!node_state(pgdat->node_id, N_MEMORY))
|
2010-05-24 23:32:25 +02:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
walk_zones_in_node(m, pgdat, unusable_show_print);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct seq_operations unusable_op = {
|
|
|
|
.start = frag_start,
|
|
|
|
.next = frag_next,
|
|
|
|
.stop = frag_stop,
|
|
|
|
.show = unusable_show,
|
|
|
|
};
|
|
|
|
|
|
|
|
static int unusable_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
return seq_open(file, &unusable_op);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations unusable_file_ops = {
|
|
|
|
.open = unusable_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = seq_release,
|
|
|
|
};
|
|
|
|
|
2010-05-24 23:32:26 +02:00
|
|
|
static void extfrag_show_print(struct seq_file *m,
|
|
|
|
pg_data_t *pgdat, struct zone *zone)
|
|
|
|
{
|
|
|
|
unsigned int order;
|
|
|
|
int index;
|
|
|
|
|
|
|
|
/* Alloc on stack as interrupts are disabled for zone walk */
|
|
|
|
struct contig_page_info info;
|
|
|
|
|
|
|
|
seq_printf(m, "Node %d, zone %8s ",
|
|
|
|
pgdat->node_id,
|
|
|
|
zone->name);
|
|
|
|
for (order = 0; order < MAX_ORDER; ++order) {
|
|
|
|
fill_contig_page_info(zone, order, &info);
|
2010-05-24 23:32:30 +02:00
|
|
|
index = __fragmentation_index(order, &info);
|
2010-05-24 23:32:26 +02:00
|
|
|
seq_printf(m, "%d.%03d ", index / 1000, index % 1000);
|
|
|
|
}
|
|
|
|
|
|
|
|
seq_putc(m, '\n');
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Display fragmentation index for orders that allocations would fail for
|
|
|
|
*/
|
|
|
|
static int extfrag_show(struct seq_file *m, void *arg)
|
|
|
|
{
|
|
|
|
pg_data_t *pgdat = (pg_data_t *)arg;
|
|
|
|
|
|
|
|
walk_zones_in_node(m, pgdat, extfrag_show_print);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct seq_operations extfrag_op = {
|
|
|
|
.start = frag_start,
|
|
|
|
.next = frag_next,
|
|
|
|
.stop = frag_stop,
|
|
|
|
.show = extfrag_show,
|
|
|
|
};
|
|
|
|
|
|
|
|
static int extfrag_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
return seq_open(file, &extfrag_op);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations extfrag_file_ops = {
|
|
|
|
.open = extfrag_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = seq_release,
|
|
|
|
};
|
|
|
|
|
2010-05-24 23:32:25 +02:00
|
|
|
static int __init extfrag_debug_init(void)
|
|
|
|
{
|
2012-05-30 00:06:22 +02:00
|
|
|
struct dentry *extfrag_debug_root;
|
|
|
|
|
2010-05-24 23:32:25 +02:00
|
|
|
extfrag_debug_root = debugfs_create_dir("extfrag", NULL);
|
|
|
|
if (!extfrag_debug_root)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
if (!debugfs_create_file("unusable_index", 0444,
|
|
|
|
extfrag_debug_root, NULL, &unusable_file_ops))
|
2012-05-30 00:06:22 +02:00
|
|
|
goto fail;
|
2010-05-24 23:32:25 +02:00
|
|
|
|
2010-05-24 23:32:26 +02:00
|
|
|
if (!debugfs_create_file("extfrag_index", 0444,
|
|
|
|
extfrag_debug_root, NULL, &extfrag_file_ops))
|
2012-05-30 00:06:22 +02:00
|
|
|
goto fail;
|
2010-05-24 23:32:26 +02:00
|
|
|
|
2010-05-24 23:32:25 +02:00
|
|
|
return 0;
|
2012-05-30 00:06:22 +02:00
|
|
|
fail:
|
|
|
|
debugfs_remove_recursive(extfrag_debug_root);
|
|
|
|
return -ENOMEM;
|
2010-05-24 23:32:25 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
module_init(extfrag_debug_init);
|
|
|
|
#endif
|