2009-12-03 18:59:42 +01:00
|
|
|
/*
|
|
|
|
* Common Block IO controller cgroup interface
|
|
|
|
*
|
|
|
|
* Based on ideas and code from CFQ, CFS and BFQ:
|
|
|
|
* Copyright (C) 2003 Jens Axboe <axboe@kernel.dk>
|
|
|
|
*
|
|
|
|
* Copyright (C) 2008 Fabio Checconi <fabio@gandalf.sssup.it>
|
|
|
|
* Paolo Valente <paolo.valente@unimore.it>
|
|
|
|
*
|
|
|
|
* Copyright (C) 2009 Vivek Goyal <vgoyal@redhat.com>
|
|
|
|
* Nauman Rafique <nauman@google.com>
|
|
|
|
*/
|
|
|
|
#include <linux/ioprio.h>
|
2009-12-03 18:59:49 +01:00
|
|
|
#include <linux/seq_file.h>
|
|
|
|
#include <linux/kdev_t.h>
|
2009-12-04 16:36:41 +01:00
|
|
|
#include <linux/module.h>
|
2009-12-07 09:29:39 +01:00
|
|
|
#include <linux/err.h>
|
2010-04-02 00:01:41 +02:00
|
|
|
#include <linux/blkdev.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>
|
2010-04-13 10:05:49 +02:00
|
|
|
#include <linux/genhd.h>
|
2012-03-05 22:15:00 +01:00
|
|
|
#include <linux/delay.h>
|
|
|
|
#include "blk-cgroup.h"
|
2012-03-05 22:15:12 +01:00
|
|
|
#include "blk.h"
|
2009-12-04 16:36:42 +01:00
|
|
|
|
2010-04-09 08:31:19 +02:00
|
|
|
#define MAX_KEY_LEN 100
|
|
|
|
|
2009-12-04 16:36:42 +01:00
|
|
|
static DEFINE_SPINLOCK(blkio_list_lock);
|
|
|
|
static LIST_HEAD(blkio_list);
|
2009-12-03 18:59:47 +01:00
|
|
|
|
2012-03-05 22:15:13 +01:00
|
|
|
static DEFINE_MUTEX(all_q_mutex);
|
|
|
|
static LIST_HEAD(all_q_list);
|
|
|
|
|
2009-12-03 18:59:42 +01:00
|
|
|
struct blkio_cgroup blkio_root_cgroup = { .weight = 2*BLKIO_WEIGHT_DEFAULT };
|
2009-12-04 16:36:41 +01:00
|
|
|
EXPORT_SYMBOL_GPL(blkio_root_cgroup);
|
|
|
|
|
2012-03-05 22:15:04 +01:00
|
|
|
static struct blkio_policy_type *blkio_policy[BLKIO_NR_POLICIES];
|
|
|
|
|
2010-03-11 00:22:11 +01:00
|
|
|
static struct cgroup_subsys_state *blkiocg_create(struct cgroup_subsys *,
|
|
|
|
struct cgroup *);
|
2011-12-13 03:12:21 +01:00
|
|
|
static int blkiocg_can_attach(struct cgroup_subsys *, struct cgroup *,
|
|
|
|
struct cgroup_taskset *);
|
|
|
|
static void blkiocg_attach(struct cgroup_subsys *, struct cgroup *,
|
|
|
|
struct cgroup_taskset *);
|
2012-03-05 22:15:11 +01:00
|
|
|
static int blkiocg_pre_destroy(struct cgroup_subsys *, struct cgroup *);
|
2010-03-11 00:22:11 +01:00
|
|
|
static void blkiocg_destroy(struct cgroup_subsys *, struct cgroup *);
|
|
|
|
static int blkiocg_populate(struct cgroup_subsys *, struct cgroup *);
|
|
|
|
|
2010-09-15 23:06:33 +02:00
|
|
|
/* for encoding cft->private value on file */
|
|
|
|
#define BLKIOFILE_PRIVATE(x, val) (((x) << 16) | (val))
|
|
|
|
/* What policy owns the file, proportional or throttle */
|
|
|
|
#define BLKIOFILE_POLICY(val) (((val) >> 16) & 0xffff)
|
|
|
|
#define BLKIOFILE_ATTR(val) ((val) & 0xffff)
|
|
|
|
|
2010-03-11 00:22:11 +01:00
|
|
|
struct cgroup_subsys blkio_subsys = {
|
|
|
|
.name = "blkio",
|
|
|
|
.create = blkiocg_create,
|
2011-12-13 03:12:21 +01:00
|
|
|
.can_attach = blkiocg_can_attach,
|
|
|
|
.attach = blkiocg_attach,
|
2012-03-05 22:15:11 +01:00
|
|
|
.pre_destroy = blkiocg_pre_destroy,
|
2010-03-11 00:22:11 +01:00
|
|
|
.destroy = blkiocg_destroy,
|
|
|
|
.populate = blkiocg_populate,
|
|
|
|
.subsys_id = blkio_subsys_id,
|
|
|
|
.module = THIS_MODULE,
|
|
|
|
};
|
|
|
|
EXPORT_SYMBOL_GPL(blkio_subsys);
|
|
|
|
|
2009-12-03 18:59:42 +01:00
|
|
|
struct blkio_cgroup *cgroup_to_blkio_cgroup(struct cgroup *cgroup)
|
|
|
|
{
|
|
|
|
return container_of(cgroup_subsys_state(cgroup, blkio_subsys_id),
|
|
|
|
struct blkio_cgroup, css);
|
|
|
|
}
|
2009-12-04 16:36:41 +01:00
|
|
|
EXPORT_SYMBOL_GPL(cgroup_to_blkio_cgroup);
|
2009-12-03 18:59:42 +01:00
|
|
|
|
2011-05-16 15:24:08 +02:00
|
|
|
struct blkio_cgroup *task_blkio_cgroup(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
return container_of(task_subsys_state(tsk, blkio_subsys_id),
|
|
|
|
struct blkio_cgroup, css);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(task_blkio_cgroup);
|
|
|
|
|
2010-09-15 23:06:33 +02:00
|
|
|
static inline void
|
|
|
|
blkio_update_group_weight(struct blkio_group *blkg, unsigned int weight)
|
|
|
|
{
|
|
|
|
struct blkio_policy_type *blkiop;
|
|
|
|
|
|
|
|
list_for_each_entry(blkiop, &blkio_list, list) {
|
|
|
|
/* If this policy does not own the blkg, do not send updates */
|
|
|
|
if (blkiop->plid != blkg->plid)
|
|
|
|
continue;
|
|
|
|
if (blkiop->ops.blkio_update_group_weight_fn)
|
2012-03-05 22:15:03 +01:00
|
|
|
blkiop->ops.blkio_update_group_weight_fn(blkg->q,
|
2010-10-01 14:49:49 +02:00
|
|
|
blkg, weight);
|
2010-09-15 23:06:33 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-09-15 23:06:34 +02:00
|
|
|
static inline void blkio_update_group_bps(struct blkio_group *blkg, u64 bps,
|
|
|
|
int fileid)
|
|
|
|
{
|
|
|
|
struct blkio_policy_type *blkiop;
|
|
|
|
|
|
|
|
list_for_each_entry(blkiop, &blkio_list, list) {
|
|
|
|
|
|
|
|
/* If this policy does not own the blkg, do not send updates */
|
|
|
|
if (blkiop->plid != blkg->plid)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (fileid == BLKIO_THROTL_read_bps_device
|
|
|
|
&& blkiop->ops.blkio_update_group_read_bps_fn)
|
2012-03-05 22:15:03 +01:00
|
|
|
blkiop->ops.blkio_update_group_read_bps_fn(blkg->q,
|
2010-10-01 14:49:49 +02:00
|
|
|
blkg, bps);
|
2010-09-15 23:06:34 +02:00
|
|
|
|
|
|
|
if (fileid == BLKIO_THROTL_write_bps_device
|
|
|
|
&& blkiop->ops.blkio_update_group_write_bps_fn)
|
2012-03-05 22:15:03 +01:00
|
|
|
blkiop->ops.blkio_update_group_write_bps_fn(blkg->q,
|
2010-10-01 14:49:49 +02:00
|
|
|
blkg, bps);
|
2010-09-15 23:06:34 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-09-15 23:06:36 +02:00
|
|
|
static inline void blkio_update_group_iops(struct blkio_group *blkg,
|
|
|
|
unsigned int iops, int fileid)
|
|
|
|
{
|
|
|
|
struct blkio_policy_type *blkiop;
|
|
|
|
|
|
|
|
list_for_each_entry(blkiop, &blkio_list, list) {
|
|
|
|
|
|
|
|
/* If this policy does not own the blkg, do not send updates */
|
|
|
|
if (blkiop->plid != blkg->plid)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (fileid == BLKIO_THROTL_read_iops_device
|
|
|
|
&& blkiop->ops.blkio_update_group_read_iops_fn)
|
2012-03-05 22:15:03 +01:00
|
|
|
blkiop->ops.blkio_update_group_read_iops_fn(blkg->q,
|
2010-10-01 14:49:49 +02:00
|
|
|
blkg, iops);
|
2010-09-15 23:06:36 +02:00
|
|
|
|
|
|
|
if (fileid == BLKIO_THROTL_write_iops_device
|
|
|
|
&& blkiop->ops.blkio_update_group_write_iops_fn)
|
2012-03-05 22:15:03 +01:00
|
|
|
blkiop->ops.blkio_update_group_write_iops_fn(blkg->q,
|
2010-10-01 14:49:49 +02:00
|
|
|
blkg,iops);
|
2010-09-15 23:06:36 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-04-02 00:01:41 +02:00
|
|
|
/*
|
|
|
|
* Add to the appropriate stat variable depending on the request type.
|
|
|
|
* This should be called with the blkg->stats_lock held.
|
|
|
|
*/
|
2010-04-09 08:31:19 +02:00
|
|
|
static void blkio_add_stat(uint64_t *stat, uint64_t add, bool direction,
|
|
|
|
bool sync)
|
2010-04-02 00:01:41 +02:00
|
|
|
{
|
2010-04-09 08:31:19 +02:00
|
|
|
if (direction)
|
|
|
|
stat[BLKIO_STAT_WRITE] += add;
|
2010-04-02 00:01:41 +02:00
|
|
|
else
|
2010-04-09 08:31:19 +02:00
|
|
|
stat[BLKIO_STAT_READ] += add;
|
|
|
|
if (sync)
|
|
|
|
stat[BLKIO_STAT_SYNC] += add;
|
2010-04-02 00:01:41 +02:00
|
|
|
else
|
2010-04-09 08:31:19 +02:00
|
|
|
stat[BLKIO_STAT_ASYNC] += add;
|
2010-04-02 00:01:41 +02:00
|
|
|
}
|
|
|
|
|
2010-04-09 06:15:10 +02:00
|
|
|
/*
|
|
|
|
* Decrements the appropriate stat variable if non-zero depending on the
|
|
|
|
* request type. Panics on value being zero.
|
|
|
|
* This should be called with the blkg->stats_lock held.
|
|
|
|
*/
|
|
|
|
static void blkio_check_and_dec_stat(uint64_t *stat, bool direction, bool sync)
|
|
|
|
{
|
|
|
|
if (direction) {
|
|
|
|
BUG_ON(stat[BLKIO_STAT_WRITE] == 0);
|
|
|
|
stat[BLKIO_STAT_WRITE]--;
|
|
|
|
} else {
|
|
|
|
BUG_ON(stat[BLKIO_STAT_READ] == 0);
|
|
|
|
stat[BLKIO_STAT_READ]--;
|
|
|
|
}
|
|
|
|
if (sync) {
|
|
|
|
BUG_ON(stat[BLKIO_STAT_SYNC] == 0);
|
|
|
|
stat[BLKIO_STAT_SYNC]--;
|
|
|
|
} else {
|
|
|
|
BUG_ON(stat[BLKIO_STAT_ASYNC] == 0);
|
|
|
|
stat[BLKIO_STAT_ASYNC]--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
2010-04-09 06:15:35 +02:00
|
|
|
/* This should be called with the blkg->stats_lock held. */
|
|
|
|
static void blkio_set_start_group_wait_time(struct blkio_group *blkg,
|
|
|
|
struct blkio_group *curr_blkg)
|
|
|
|
{
|
|
|
|
if (blkio_blkg_waiting(&blkg->stats))
|
|
|
|
return;
|
|
|
|
if (blkg == curr_blkg)
|
|
|
|
return;
|
|
|
|
blkg->stats.start_group_wait_time = sched_clock();
|
|
|
|
blkio_mark_blkg_waiting(&blkg->stats);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This should be called with the blkg->stats_lock held. */
|
|
|
|
static void blkio_update_group_wait_time(struct blkio_group_stats *stats)
|
|
|
|
{
|
|
|
|
unsigned long long now;
|
|
|
|
|
|
|
|
if (!blkio_blkg_waiting(stats))
|
|
|
|
return;
|
|
|
|
|
|
|
|
now = sched_clock();
|
|
|
|
if (time_after64(now, stats->start_group_wait_time))
|
|
|
|
stats->group_wait_time += now - stats->start_group_wait_time;
|
|
|
|
blkio_clear_blkg_waiting(stats);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This should be called with the blkg->stats_lock held. */
|
|
|
|
static void blkio_end_empty_time(struct blkio_group_stats *stats)
|
|
|
|
{
|
|
|
|
unsigned long long now;
|
|
|
|
|
|
|
|
if (!blkio_blkg_empty(stats))
|
|
|
|
return;
|
|
|
|
|
|
|
|
now = sched_clock();
|
|
|
|
if (time_after64(now, stats->start_empty_time))
|
|
|
|
stats->empty_time += now - stats->start_empty_time;
|
|
|
|
blkio_clear_blkg_empty(stats);
|
|
|
|
}
|
|
|
|
|
|
|
|
void blkiocg_update_set_idle_time_stats(struct blkio_group *blkg)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
BUG_ON(blkio_blkg_idling(&blkg->stats));
|
|
|
|
blkg->stats.start_idle_time = sched_clock();
|
|
|
|
blkio_mark_blkg_idling(&blkg->stats);
|
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_set_idle_time_stats);
|
|
|
|
|
|
|
|
void blkiocg_update_idle_time_stats(struct blkio_group *blkg)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned long long now;
|
|
|
|
struct blkio_group_stats *stats;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
stats = &blkg->stats;
|
|
|
|
if (blkio_blkg_idling(stats)) {
|
|
|
|
now = sched_clock();
|
|
|
|
if (time_after64(now, stats->start_idle_time))
|
|
|
|
stats->idle_time += now - stats->start_idle_time;
|
|
|
|
blkio_clear_blkg_idling(stats);
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_idle_time_stats);
|
|
|
|
|
2010-04-13 19:59:17 +02:00
|
|
|
void blkiocg_update_avg_queue_size_stats(struct blkio_group *blkg)
|
2010-04-09 06:15:10 +02:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
struct blkio_group_stats *stats;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
stats = &blkg->stats;
|
|
|
|
stats->avg_queue_size_sum +=
|
|
|
|
stats->stat_arr[BLKIO_STAT_QUEUED][BLKIO_STAT_READ] +
|
|
|
|
stats->stat_arr[BLKIO_STAT_QUEUED][BLKIO_STAT_WRITE];
|
|
|
|
stats->avg_queue_size_samples++;
|
2010-04-09 06:15:35 +02:00
|
|
|
blkio_update_group_wait_time(stats);
|
2010-04-09 06:15:10 +02:00
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
}
|
2010-04-13 19:59:17 +02:00
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_avg_queue_size_stats);
|
|
|
|
|
2010-04-26 19:25:11 +02:00
|
|
|
void blkiocg_set_start_empty_time(struct blkio_group *blkg)
|
2010-04-14 11:22:38 +02:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
struct blkio_group_stats *stats;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
stats = &blkg->stats;
|
|
|
|
|
|
|
|
if (stats->stat_arr[BLKIO_STAT_QUEUED][BLKIO_STAT_READ] ||
|
|
|
|
stats->stat_arr[BLKIO_STAT_QUEUED][BLKIO_STAT_WRITE]) {
|
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2010-04-26 19:25:11 +02:00
|
|
|
* group is already marked empty. This can happen if cfqq got new
|
|
|
|
* request in parent group and moved to this group while being added
|
|
|
|
* to service tree. Just ignore the event and move on.
|
2010-04-14 11:22:38 +02:00
|
|
|
*/
|
2010-04-26 19:25:11 +02:00
|
|
|
if(blkio_blkg_empty(stats)) {
|
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2010-04-14 11:22:38 +02:00
|
|
|
stats->start_empty_time = sched_clock();
|
|
|
|
blkio_mark_blkg_empty(stats);
|
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_set_start_empty_time);
|
|
|
|
|
2010-04-13 19:59:17 +02:00
|
|
|
void blkiocg_update_dequeue_stats(struct blkio_group *blkg,
|
|
|
|
unsigned long dequeue)
|
|
|
|
{
|
|
|
|
blkg->stats.dequeue += dequeue;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_dequeue_stats);
|
2010-04-09 06:15:35 +02:00
|
|
|
#else
|
|
|
|
static inline void blkio_set_start_group_wait_time(struct blkio_group *blkg,
|
|
|
|
struct blkio_group *curr_blkg) {}
|
|
|
|
static inline void blkio_end_empty_time(struct blkio_group_stats *stats) {}
|
2010-04-09 06:15:10 +02:00
|
|
|
#endif
|
|
|
|
|
2010-04-13 19:59:17 +02:00
|
|
|
void blkiocg_update_io_add_stats(struct blkio_group *blkg,
|
2010-04-09 06:15:10 +02:00
|
|
|
struct blkio_group *curr_blkg, bool direction,
|
|
|
|
bool sync)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
blkio_add_stat(blkg->stats.stat_arr[BLKIO_STAT_QUEUED], 1, direction,
|
|
|
|
sync);
|
2010-04-09 06:15:35 +02:00
|
|
|
blkio_end_empty_time(&blkg->stats);
|
|
|
|
blkio_set_start_group_wait_time(blkg, curr_blkg);
|
2010-04-09 06:15:10 +02:00
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
}
|
2010-04-13 19:59:17 +02:00
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_io_add_stats);
|
2010-04-09 06:15:10 +02:00
|
|
|
|
2010-04-13 19:59:17 +02:00
|
|
|
void blkiocg_update_io_remove_stats(struct blkio_group *blkg,
|
2010-04-09 06:15:10 +02:00
|
|
|
bool direction, bool sync)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
blkio_check_and_dec_stat(blkg->stats.stat_arr[BLKIO_STAT_QUEUED],
|
|
|
|
direction, sync);
|
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
}
|
2010-04-13 19:59:17 +02:00
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_io_remove_stats);
|
2010-04-09 06:15:10 +02:00
|
|
|
|
2011-03-12 16:54:00 +01:00
|
|
|
void blkiocg_update_timeslice_used(struct blkio_group *blkg, unsigned long time,
|
|
|
|
unsigned long unaccounted_time)
|
2009-12-03 18:59:49 +01:00
|
|
|
{
|
2010-04-02 00:01:24 +02:00
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
blkg->stats.time += time;
|
2011-05-19 21:38:20 +02:00
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
2011-03-12 16:54:00 +01:00
|
|
|
blkg->stats.unaccounted_time += unaccounted_time;
|
2011-05-19 21:38:20 +02:00
|
|
|
#endif
|
2010-04-02 00:01:24 +02:00
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
2009-12-03 18:59:49 +01:00
|
|
|
}
|
2010-04-02 00:01:24 +02:00
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_timeslice_used);
|
2009-12-03 18:59:49 +01:00
|
|
|
|
2011-05-19 21:38:28 +02:00
|
|
|
/*
|
|
|
|
* should be called under rcu read lock or queue lock to make sure blkg pointer
|
|
|
|
* is valid.
|
|
|
|
*/
|
2010-04-09 08:31:19 +02:00
|
|
|
void blkiocg_update_dispatch_stats(struct blkio_group *blkg,
|
|
|
|
uint64_t bytes, bool direction, bool sync)
|
2010-04-02 00:01:41 +02:00
|
|
|
{
|
2011-05-19 21:38:28 +02:00
|
|
|
struct blkio_group_stats_cpu *stats_cpu;
|
2011-05-19 21:38:29 +02:00
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Disabling interrupts to provide mutual exclusion between two
|
|
|
|
* writes on same cpu. It probably is not needed for 64bit. Not
|
|
|
|
* optimizing that case yet.
|
|
|
|
*/
|
|
|
|
local_irq_save(flags);
|
2010-04-02 00:01:41 +02:00
|
|
|
|
2011-05-19 21:38:28 +02:00
|
|
|
stats_cpu = this_cpu_ptr(blkg->stats_cpu);
|
|
|
|
|
2011-05-19 21:38:29 +02:00
|
|
|
u64_stats_update_begin(&stats_cpu->syncp);
|
2011-05-19 21:38:28 +02:00
|
|
|
stats_cpu->sectors += bytes >> 9;
|
|
|
|
blkio_add_stat(stats_cpu->stat_arr_cpu[BLKIO_STAT_CPU_SERVICED],
|
|
|
|
1, direction, sync);
|
|
|
|
blkio_add_stat(stats_cpu->stat_arr_cpu[BLKIO_STAT_CPU_SERVICE_BYTES],
|
|
|
|
bytes, direction, sync);
|
2011-05-19 21:38:29 +02:00
|
|
|
u64_stats_update_end(&stats_cpu->syncp);
|
|
|
|
local_irq_restore(flags);
|
2010-04-02 00:01:41 +02:00
|
|
|
}
|
2010-04-09 08:31:19 +02:00
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_dispatch_stats);
|
2010-04-02 00:01:41 +02:00
|
|
|
|
2010-04-09 08:31:19 +02:00
|
|
|
void blkiocg_update_completion_stats(struct blkio_group *blkg,
|
|
|
|
uint64_t start_time, uint64_t io_start_time, bool direction, bool sync)
|
2010-04-02 00:01:41 +02:00
|
|
|
{
|
|
|
|
struct blkio_group_stats *stats;
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned long long now = sched_clock();
|
|
|
|
|
|
|
|
spin_lock_irqsave(&blkg->stats_lock, flags);
|
|
|
|
stats = &blkg->stats;
|
2010-04-09 08:31:19 +02:00
|
|
|
if (time_after64(now, io_start_time))
|
|
|
|
blkio_add_stat(stats->stat_arr[BLKIO_STAT_SERVICE_TIME],
|
|
|
|
now - io_start_time, direction, sync);
|
|
|
|
if (time_after64(io_start_time, start_time))
|
|
|
|
blkio_add_stat(stats->stat_arr[BLKIO_STAT_WAIT_TIME],
|
|
|
|
io_start_time - start_time, direction, sync);
|
2010-04-02 00:01:41 +02:00
|
|
|
spin_unlock_irqrestore(&blkg->stats_lock, flags);
|
|
|
|
}
|
2010-04-09 08:31:19 +02:00
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_completion_stats);
|
2010-04-02 00:01:41 +02:00
|
|
|
|
2011-05-23 10:02:19 +02:00
|
|
|
/* Merged stats are per cpu. */
|
2010-04-09 06:14:23 +02:00
|
|
|
void blkiocg_update_io_merged_stats(struct blkio_group *blkg, bool direction,
|
|
|
|
bool sync)
|
|
|
|
{
|
2011-05-23 10:02:19 +02:00
|
|
|
struct blkio_group_stats_cpu *stats_cpu;
|
2010-04-09 06:14:23 +02:00
|
|
|
unsigned long flags;
|
|
|
|
|
2011-05-23 10:02:19 +02:00
|
|
|
/*
|
|
|
|
* Disabling interrupts to provide mutual exclusion between two
|
|
|
|
* writes on same cpu. It probably is not needed for 64bit. Not
|
|
|
|
* optimizing that case yet.
|
|
|
|
*/
|
|
|
|
local_irq_save(flags);
|
|
|
|
|
|
|
|
stats_cpu = this_cpu_ptr(blkg->stats_cpu);
|
|
|
|
|
|
|
|
u64_stats_update_begin(&stats_cpu->syncp);
|
|
|
|
blkio_add_stat(stats_cpu->stat_arr_cpu[BLKIO_STAT_CPU_MERGED], 1,
|
|
|
|
direction, sync);
|
|
|
|
u64_stats_update_end(&stats_cpu->syncp);
|
|
|
|
local_irq_restore(flags);
|
2010-04-09 06:14:23 +02:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_update_io_merged_stats);
|
|
|
|
|
2012-03-05 22:15:06 +01:00
|
|
|
struct blkio_group *blkg_lookup_create(struct blkio_cgroup *blkcg,
|
|
|
|
struct request_queue *q,
|
|
|
|
enum blkio_policy_id plid,
|
|
|
|
bool for_root)
|
|
|
|
__releases(q->queue_lock) __acquires(q->queue_lock)
|
2011-05-19 21:38:28 +02:00
|
|
|
{
|
2012-03-05 22:15:06 +01:00
|
|
|
struct blkio_policy_type *pol = blkio_policy[plid];
|
|
|
|
struct blkio_group *blkg, *new_blkg;
|
2011-05-19 21:38:28 +02:00
|
|
|
|
2012-03-05 22:15:06 +01:00
|
|
|
WARN_ON_ONCE(!rcu_read_lock_held());
|
|
|
|
lockdep_assert_held(q->queue_lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This could be the first entry point of blkcg implementation and
|
|
|
|
* we shouldn't allow anything to go through for a bypassing queue.
|
|
|
|
* The following can be removed if blkg lookup is guaranteed to
|
|
|
|
* fail on a bypassing queue.
|
|
|
|
*/
|
|
|
|
if (unlikely(blk_queue_bypass(q)) && !for_root)
|
|
|
|
return ERR_PTR(blk_queue_dead(q) ? -EINVAL : -EBUSY);
|
|
|
|
|
|
|
|
blkg = blkg_lookup(blkcg, q, plid);
|
|
|
|
if (blkg)
|
|
|
|
return blkg;
|
|
|
|
|
2012-03-05 22:15:11 +01:00
|
|
|
/* blkg holds a reference to blkcg */
|
2012-03-05 22:15:06 +01:00
|
|
|
if (!css_tryget(&blkcg->css))
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate and initialize.
|
|
|
|
*
|
|
|
|
* FIXME: The following is broken. Percpu memory allocation
|
|
|
|
* requires %GFP_KERNEL context and can't be performed from IO
|
|
|
|
* path. Allocation here should inherently be atomic and the
|
|
|
|
* following lock dancing can be removed once the broken percpu
|
|
|
|
* allocation is fixed.
|
|
|
|
*/
|
|
|
|
spin_unlock_irq(q->queue_lock);
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
|
|
|
new_blkg = pol->ops.blkio_alloc_group_fn(q, blkcg);
|
|
|
|
if (new_blkg) {
|
|
|
|
new_blkg->stats_cpu = alloc_percpu(struct blkio_group_stats_cpu);
|
|
|
|
|
|
|
|
spin_lock_init(&new_blkg->stats_lock);
|
|
|
|
rcu_assign_pointer(new_blkg->q, q);
|
2012-03-05 22:15:11 +01:00
|
|
|
new_blkg->blkcg = blkcg;
|
2012-03-05 22:15:06 +01:00
|
|
|
new_blkg->plid = plid;
|
|
|
|
cgroup_path(blkcg->css.cgroup, new_blkg->path,
|
|
|
|
sizeof(new_blkg->path));
|
2012-03-05 22:15:11 +01:00
|
|
|
} else {
|
|
|
|
css_put(&blkcg->css);
|
2012-03-05 22:15:06 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
spin_lock_irq(q->queue_lock);
|
2009-12-03 18:59:42 +01:00
|
|
|
|
2012-03-05 22:15:06 +01:00
|
|
|
/* did bypass get turned on inbetween? */
|
|
|
|
if (unlikely(blk_queue_bypass(q)) && !for_root) {
|
|
|
|
blkg = ERR_PTR(blk_queue_dead(q) ? -EINVAL : -EBUSY);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* did someone beat us to it? */
|
|
|
|
blkg = blkg_lookup(blkcg, q, plid);
|
|
|
|
if (unlikely(blkg))
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
/* did alloc fail? */
|
|
|
|
if (unlikely(!new_blkg || !new_blkg->stats_cpu)) {
|
|
|
|
blkg = ERR_PTR(-ENOMEM);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* insert */
|
|
|
|
spin_lock(&blkcg->lock);
|
|
|
|
swap(blkg, new_blkg);
|
2009-12-03 18:59:42 +01:00
|
|
|
hlist_add_head_rcu(&blkg->blkcg_node, &blkcg->blkg_list);
|
2012-03-05 22:15:06 +01:00
|
|
|
pol->ops.blkio_link_group_fn(q, blkg);
|
|
|
|
spin_unlock(&blkcg->lock);
|
|
|
|
out:
|
|
|
|
if (new_blkg) {
|
|
|
|
free_percpu(new_blkg->stats_cpu);
|
|
|
|
kfree(new_blkg);
|
2012-03-05 22:15:11 +01:00
|
|
|
css_put(&blkcg->css);
|
2012-03-05 22:15:06 +01:00
|
|
|
}
|
|
|
|
return blkg;
|
2009-12-03 18:59:42 +01:00
|
|
|
}
|
2012-03-05 22:15:06 +01:00
|
|
|
EXPORT_SYMBOL_GPL(blkg_lookup_create);
|
2009-12-03 18:59:42 +01:00
|
|
|
|
2009-12-03 18:59:47 +01:00
|
|
|
static void __blkiocg_del_blkio_group(struct blkio_group *blkg)
|
|
|
|
{
|
|
|
|
hlist_del_init_rcu(&blkg->blkcg_node);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* returns 0 if blkio_group was still on cgroup list. Otherwise returns 1
|
|
|
|
* indicating that blk_group was unhashed by the time we got to it.
|
|
|
|
*/
|
2009-12-03 18:59:42 +01:00
|
|
|
int blkiocg_del_blkio_group(struct blkio_group *blkg)
|
|
|
|
{
|
2012-03-05 22:15:11 +01:00
|
|
|
struct blkio_cgroup *blkcg = blkg->blkcg;
|
2009-12-03 18:59:47 +01:00
|
|
|
unsigned long flags;
|
|
|
|
int ret = 1;
|
|
|
|
|
2012-03-05 22:15:11 +01:00
|
|
|
spin_lock_irqsave(&blkcg->lock, flags);
|
|
|
|
if (!hlist_unhashed(&blkg->blkcg_node)) {
|
|
|
|
__blkiocg_del_blkio_group(blkg);
|
|
|
|
ret = 0;
|
2009-12-03 18:59:47 +01:00
|
|
|
}
|
2012-03-05 22:15:11 +01:00
|
|
|
spin_unlock_irqrestore(&blkcg->lock, flags);
|
2010-05-03 14:28:55 +02:00
|
|
|
|
2009-12-03 18:59:47 +01:00
|
|
|
return ret;
|
2009-12-03 18:59:42 +01:00
|
|
|
}
|
2009-12-04 16:36:41 +01:00
|
|
|
EXPORT_SYMBOL_GPL(blkiocg_del_blkio_group);
|
2009-12-03 18:59:42 +01:00
|
|
|
|
|
|
|
/* called under rcu_read_lock(). */
|
2012-03-05 22:15:06 +01:00
|
|
|
struct blkio_group *blkg_lookup(struct blkio_cgroup *blkcg,
|
|
|
|
struct request_queue *q,
|
|
|
|
enum blkio_policy_id plid)
|
2009-12-03 18:59:42 +01:00
|
|
|
{
|
|
|
|
struct blkio_group *blkg;
|
|
|
|
struct hlist_node *n;
|
|
|
|
|
2012-03-05 22:15:03 +01:00
|
|
|
hlist_for_each_entry_rcu(blkg, n, &blkcg->blkg_list, blkcg_node)
|
|
|
|
if (blkg->q == q && blkg->plid == plid)
|
2009-12-03 18:59:42 +01:00
|
|
|
return blkg;
|
|
|
|
return NULL;
|
|
|
|
}
|
2012-03-05 22:15:06 +01:00
|
|
|
EXPORT_SYMBOL_GPL(blkg_lookup);
|
2009-12-03 18:59:42 +01:00
|
|
|
|
2012-03-05 22:15:00 +01:00
|
|
|
void blkg_destroy_all(struct request_queue *q)
|
|
|
|
{
|
|
|
|
struct blkio_policy_type *pol;
|
|
|
|
|
|
|
|
while (true) {
|
|
|
|
bool done = true;
|
|
|
|
|
|
|
|
spin_lock(&blkio_list_lock);
|
|
|
|
spin_lock_irq(q->queue_lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* clear_queue_fn() might return with non-empty group list
|
|
|
|
* if it raced cgroup removal and lost. cgroup removal is
|
|
|
|
* guaranteed to make forward progress and retrying after a
|
|
|
|
* while is enough. This ugliness is scheduled to be
|
|
|
|
* removed after locking update.
|
|
|
|
*/
|
|
|
|
list_for_each_entry(pol, &blkio_list, list)
|
|
|
|
if (!pol->ops.blkio_clear_queue_fn(q))
|
|
|
|
done = false;
|
|
|
|
|
|
|
|
spin_unlock_irq(q->queue_lock);
|
|
|
|
spin_unlock(&blkio_list_lock);
|
|
|
|
|
|
|
|
if (done)
|
|
|
|
break;
|
|
|
|
|
|
|
|
msleep(10); /* just some random duration I like */
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-05-19 21:38:30 +02:00
|
|
|
static void blkio_reset_stats_cpu(struct blkio_group *blkg)
|
|
|
|
{
|
|
|
|
struct blkio_group_stats_cpu *stats_cpu;
|
|
|
|
int i, j, k;
|
|
|
|
/*
|
|
|
|
* Note: On 64 bit arch this should not be an issue. This has the
|
|
|
|
* possibility of returning some inconsistent value on 32bit arch
|
|
|
|
* as 64bit update on 32bit is non atomic. Taking care of this
|
|
|
|
* corner case makes code very complicated, like sending IPIs to
|
|
|
|
* cpus, taking care of stats of offline cpus etc.
|
|
|
|
*
|
|
|
|
* reset stats is anyway more of a debug feature and this sounds a
|
|
|
|
* corner case. So I am not complicating the code yet until and
|
|
|
|
* unless this becomes a real issue.
|
|
|
|
*/
|
|
|
|
for_each_possible_cpu(i) {
|
|
|
|
stats_cpu = per_cpu_ptr(blkg->stats_cpu, i);
|
|
|
|
stats_cpu->sectors = 0;
|
|
|
|
for(j = 0; j < BLKIO_STAT_CPU_NR; j++)
|
|
|
|
for (k = 0; k < BLKIO_STAT_TOTAL; k++)
|
|
|
|
stats_cpu->stat_arr_cpu[j][k] = 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-04-02 00:01:24 +02:00
|
|
|
static int
|
2010-04-09 08:31:19 +02:00
|
|
|
blkiocg_reset_stats(struct cgroup *cgroup, struct cftype *cftype, u64 val)
|
2010-04-02 00:01:24 +02:00
|
|
|
{
|
|
|
|
struct blkio_cgroup *blkcg;
|
|
|
|
struct blkio_group *blkg;
|
2010-04-09 06:15:35 +02:00
|
|
|
struct blkio_group_stats *stats;
|
2010-04-02 00:01:24 +02:00
|
|
|
struct hlist_node *n;
|
2010-04-09 06:15:10 +02:00
|
|
|
uint64_t queued[BLKIO_STAT_TOTAL];
|
|
|
|
int i;
|
2010-04-09 06:15:35 +02:00
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
|
|
|
bool idling, waiting, empty;
|
|
|
|
unsigned long long now = sched_clock();
|
|
|
|
#endif
|
2010-04-02 00:01:24 +02:00
|
|
|
|
|
|
|
blkcg = cgroup_to_blkio_cgroup(cgroup);
|
|
|
|
spin_lock_irq(&blkcg->lock);
|
|
|
|
hlist_for_each_entry(blkg, n, &blkcg->blkg_list, blkcg_node) {
|
|
|
|
spin_lock(&blkg->stats_lock);
|
2010-04-09 06:15:35 +02:00
|
|
|
stats = &blkg->stats;
|
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
|
|
|
idling = blkio_blkg_idling(stats);
|
|
|
|
waiting = blkio_blkg_waiting(stats);
|
|
|
|
empty = blkio_blkg_empty(stats);
|
|
|
|
#endif
|
2010-04-09 06:15:10 +02:00
|
|
|
for (i = 0; i < BLKIO_STAT_TOTAL; i++)
|
2010-04-09 06:15:35 +02:00
|
|
|
queued[i] = stats->stat_arr[BLKIO_STAT_QUEUED][i];
|
|
|
|
memset(stats, 0, sizeof(struct blkio_group_stats));
|
2010-04-09 06:15:10 +02:00
|
|
|
for (i = 0; i < BLKIO_STAT_TOTAL; i++)
|
2010-04-09 06:15:35 +02:00
|
|
|
stats->stat_arr[BLKIO_STAT_QUEUED][i] = queued[i];
|
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
|
|
|
if (idling) {
|
|
|
|
blkio_mark_blkg_idling(stats);
|
|
|
|
stats->start_idle_time = now;
|
|
|
|
}
|
|
|
|
if (waiting) {
|
|
|
|
blkio_mark_blkg_waiting(stats);
|
|
|
|
stats->start_group_wait_time = now;
|
|
|
|
}
|
|
|
|
if (empty) {
|
|
|
|
blkio_mark_blkg_empty(stats);
|
|
|
|
stats->start_empty_time = now;
|
|
|
|
}
|
|
|
|
#endif
|
2010-04-02 00:01:24 +02:00
|
|
|
spin_unlock(&blkg->stats_lock);
|
2011-05-19 21:38:30 +02:00
|
|
|
|
|
|
|
/* Reset Per cpu stats which don't take blkg->stats_lock */
|
|
|
|
blkio_reset_stats_cpu(blkg);
|
2010-04-02 00:01:24 +02:00
|
|
|
}
|
2011-05-19 21:38:30 +02:00
|
|
|
|
2010-04-02 00:01:24 +02:00
|
|
|
spin_unlock_irq(&blkcg->lock);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:09 +01:00
|
|
|
static void blkio_get_key_name(enum stat_sub_type type, const char *dname,
|
|
|
|
char *str, int chars_left, bool diskname_only)
|
2010-04-02 00:01:24 +02:00
|
|
|
{
|
2012-03-05 22:15:09 +01:00
|
|
|
snprintf(str, chars_left, "%s", dname);
|
2010-04-02 00:01:24 +02:00
|
|
|
chars_left -= strlen(str);
|
|
|
|
if (chars_left <= 0) {
|
|
|
|
printk(KERN_WARNING
|
|
|
|
"Possibly incorrect cgroup stat display format");
|
|
|
|
return;
|
|
|
|
}
|
2010-04-09 08:31:19 +02:00
|
|
|
if (diskname_only)
|
|
|
|
return;
|
2010-04-02 00:01:24 +02:00
|
|
|
switch (type) {
|
2010-04-09 08:31:19 +02:00
|
|
|
case BLKIO_STAT_READ:
|
2010-04-02 00:01:24 +02:00
|
|
|
strlcat(str, " Read", chars_left);
|
|
|
|
break;
|
2010-04-09 08:31:19 +02:00
|
|
|
case BLKIO_STAT_WRITE:
|
2010-04-02 00:01:24 +02:00
|
|
|
strlcat(str, " Write", chars_left);
|
|
|
|
break;
|
2010-04-09 08:31:19 +02:00
|
|
|
case BLKIO_STAT_SYNC:
|
2010-04-02 00:01:24 +02:00
|
|
|
strlcat(str, " Sync", chars_left);
|
|
|
|
break;
|
2010-04-09 08:31:19 +02:00
|
|
|
case BLKIO_STAT_ASYNC:
|
2010-04-02 00:01:24 +02:00
|
|
|
strlcat(str, " Async", chars_left);
|
|
|
|
break;
|
2010-04-09 08:31:19 +02:00
|
|
|
case BLKIO_STAT_TOTAL:
|
2010-04-02 00:01:24 +02:00
|
|
|
strlcat(str, " Total", chars_left);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
strlcat(str, " Invalid", chars_left);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-04-09 08:31:19 +02:00
|
|
|
static uint64_t blkio_fill_stat(char *str, int chars_left, uint64_t val,
|
2012-03-05 22:15:09 +01:00
|
|
|
struct cgroup_map_cb *cb, const char *dname)
|
2010-04-09 08:31:19 +02:00
|
|
|
{
|
2012-03-05 22:15:09 +01:00
|
|
|
blkio_get_key_name(0, dname, str, chars_left, true);
|
2010-04-09 08:31:19 +02:00
|
|
|
cb->fill(cb, str, val);
|
|
|
|
return val;
|
|
|
|
}
|
2010-04-02 00:01:24 +02:00
|
|
|
|
2011-05-19 21:38:28 +02:00
|
|
|
|
|
|
|
static uint64_t blkio_read_stat_cpu(struct blkio_group *blkg,
|
|
|
|
enum stat_type_cpu type, enum stat_sub_type sub_type)
|
|
|
|
{
|
|
|
|
int cpu;
|
|
|
|
struct blkio_group_stats_cpu *stats_cpu;
|
2011-05-19 21:38:29 +02:00
|
|
|
u64 val = 0, tval;
|
2011-05-19 21:38:28 +02:00
|
|
|
|
|
|
|
for_each_possible_cpu(cpu) {
|
2011-05-19 21:38:29 +02:00
|
|
|
unsigned int start;
|
2011-05-19 21:38:28 +02:00
|
|
|
stats_cpu = per_cpu_ptr(blkg->stats_cpu, cpu);
|
|
|
|
|
2011-05-19 21:38:29 +02:00
|
|
|
do {
|
|
|
|
start = u64_stats_fetch_begin(&stats_cpu->syncp);
|
|
|
|
if (type == BLKIO_STAT_CPU_SECTORS)
|
|
|
|
tval = stats_cpu->sectors;
|
|
|
|
else
|
|
|
|
tval = stats_cpu->stat_arr_cpu[type][sub_type];
|
|
|
|
} while(u64_stats_fetch_retry(&stats_cpu->syncp, start));
|
|
|
|
|
|
|
|
val += tval;
|
2011-05-19 21:38:28 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return val;
|
|
|
|
}
|
|
|
|
|
|
|
|
static uint64_t blkio_get_stat_cpu(struct blkio_group *blkg,
|
2012-03-05 22:15:09 +01:00
|
|
|
struct cgroup_map_cb *cb, const char *dname,
|
|
|
|
enum stat_type_cpu type)
|
2011-05-19 21:38:28 +02:00
|
|
|
{
|
|
|
|
uint64_t disk_total, val;
|
|
|
|
char key_str[MAX_KEY_LEN];
|
|
|
|
enum stat_sub_type sub_type;
|
|
|
|
|
|
|
|
if (type == BLKIO_STAT_CPU_SECTORS) {
|
|
|
|
val = blkio_read_stat_cpu(blkg, type, 0);
|
2012-03-05 22:15:09 +01:00
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1, val, cb,
|
|
|
|
dname);
|
2011-05-19 21:38:28 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
for (sub_type = BLKIO_STAT_READ; sub_type < BLKIO_STAT_TOTAL;
|
|
|
|
sub_type++) {
|
2012-03-05 22:15:09 +01:00
|
|
|
blkio_get_key_name(sub_type, dname, key_str, MAX_KEY_LEN,
|
|
|
|
false);
|
2011-05-19 21:38:28 +02:00
|
|
|
val = blkio_read_stat_cpu(blkg, type, sub_type);
|
|
|
|
cb->fill(cb, key_str, val);
|
|
|
|
}
|
|
|
|
|
|
|
|
disk_total = blkio_read_stat_cpu(blkg, type, BLKIO_STAT_READ) +
|
|
|
|
blkio_read_stat_cpu(blkg, type, BLKIO_STAT_WRITE);
|
|
|
|
|
2012-03-05 22:15:09 +01:00
|
|
|
blkio_get_key_name(BLKIO_STAT_TOTAL, dname, key_str, MAX_KEY_LEN,
|
|
|
|
false);
|
2011-05-19 21:38:28 +02:00
|
|
|
cb->fill(cb, key_str, disk_total);
|
|
|
|
return disk_total;
|
|
|
|
}
|
|
|
|
|
2010-04-09 08:31:19 +02:00
|
|
|
/* This should be called with blkg->stats_lock held */
|
|
|
|
static uint64_t blkio_get_stat(struct blkio_group *blkg,
|
2012-03-05 22:15:09 +01:00
|
|
|
struct cgroup_map_cb *cb, const char *dname,
|
|
|
|
enum stat_type type)
|
2010-04-02 00:01:24 +02:00
|
|
|
{
|
|
|
|
uint64_t disk_total;
|
|
|
|
char key_str[MAX_KEY_LEN];
|
2010-04-09 08:31:19 +02:00
|
|
|
enum stat_sub_type sub_type;
|
|
|
|
|
|
|
|
if (type == BLKIO_STAT_TIME)
|
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1,
|
2012-03-05 22:15:09 +01:00
|
|
|
blkg->stats.time, cb, dname);
|
2011-03-22 21:26:54 +01:00
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
2011-03-12 16:54:00 +01:00
|
|
|
if (type == BLKIO_STAT_UNACCOUNTED_TIME)
|
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1,
|
2012-03-05 22:15:09 +01:00
|
|
|
blkg->stats.unaccounted_time, cb, dname);
|
2010-04-09 06:15:10 +02:00
|
|
|
if (type == BLKIO_STAT_AVG_QUEUE_SIZE) {
|
|
|
|
uint64_t sum = blkg->stats.avg_queue_size_sum;
|
|
|
|
uint64_t samples = blkg->stats.avg_queue_size_samples;
|
|
|
|
if (samples)
|
|
|
|
do_div(sum, samples);
|
|
|
|
else
|
|
|
|
sum = 0;
|
2012-03-05 22:15:09 +01:00
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1,
|
|
|
|
sum, cb, dname);
|
2010-04-09 06:15:10 +02:00
|
|
|
}
|
2010-04-09 06:15:35 +02:00
|
|
|
if (type == BLKIO_STAT_GROUP_WAIT_TIME)
|
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1,
|
2012-03-05 22:15:09 +01:00
|
|
|
blkg->stats.group_wait_time, cb, dname);
|
2010-04-09 06:15:35 +02:00
|
|
|
if (type == BLKIO_STAT_IDLE_TIME)
|
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1,
|
2012-03-05 22:15:09 +01:00
|
|
|
blkg->stats.idle_time, cb, dname);
|
2010-04-09 06:15:35 +02:00
|
|
|
if (type == BLKIO_STAT_EMPTY_TIME)
|
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1,
|
2012-03-05 22:15:09 +01:00
|
|
|
blkg->stats.empty_time, cb, dname);
|
2010-04-09 08:31:19 +02:00
|
|
|
if (type == BLKIO_STAT_DEQUEUE)
|
|
|
|
return blkio_fill_stat(key_str, MAX_KEY_LEN - 1,
|
2012-03-05 22:15:09 +01:00
|
|
|
blkg->stats.dequeue, cb, dname);
|
2010-04-09 08:31:19 +02:00
|
|
|
#endif
|
2010-04-02 00:01:24 +02:00
|
|
|
|
2010-04-09 08:31:19 +02:00
|
|
|
for (sub_type = BLKIO_STAT_READ; sub_type < BLKIO_STAT_TOTAL;
|
|
|
|
sub_type++) {
|
2012-03-05 22:15:09 +01:00
|
|
|
blkio_get_key_name(sub_type, dname, key_str, MAX_KEY_LEN,
|
|
|
|
false);
|
2010-04-09 08:31:19 +02:00
|
|
|
cb->fill(cb, key_str, blkg->stats.stat_arr[type][sub_type]);
|
2010-04-02 00:01:24 +02:00
|
|
|
}
|
2010-04-09 08:31:19 +02:00
|
|
|
disk_total = blkg->stats.stat_arr[type][BLKIO_STAT_READ] +
|
|
|
|
blkg->stats.stat_arr[type][BLKIO_STAT_WRITE];
|
2012-03-05 22:15:09 +01:00
|
|
|
blkio_get_key_name(BLKIO_STAT_TOTAL, dname, key_str, MAX_KEY_LEN,
|
|
|
|
false);
|
2010-04-02 00:01:24 +02:00
|
|
|
cb->fill(cb, key_str, disk_total);
|
|
|
|
return disk_total;
|
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
static int blkio_policy_parse_and_set(char *buf, enum blkio_policy_id plid,
|
|
|
|
int fileid, struct blkio_cgroup *blkcg)
|
2010-04-13 10:05:49 +02:00
|
|
|
{
|
2011-10-19 14:31:15 +02:00
|
|
|
struct gendisk *disk = NULL;
|
2012-03-05 22:15:07 +01:00
|
|
|
struct blkio_group *blkg = NULL;
|
2010-04-13 10:05:49 +02:00
|
|
|
char *s[4], *p, *major_s = NULL, *minor_s = NULL;
|
2011-09-21 10:22:10 +02:00
|
|
|
unsigned long major, minor;
|
2011-10-19 14:31:15 +02:00
|
|
|
int i = 0, ret = -EINVAL;
|
|
|
|
int part;
|
2010-04-13 10:05:49 +02:00
|
|
|
dev_t dev;
|
2011-09-21 10:22:10 +02:00
|
|
|
u64 temp;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
|
|
|
memset(s, 0, sizeof(s));
|
|
|
|
|
|
|
|
while ((p = strsep(&buf, " ")) != NULL) {
|
|
|
|
if (!*p)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
s[i++] = p;
|
|
|
|
|
|
|
|
/* Prevent from inputing too many things */
|
|
|
|
if (i == 3)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (i != 2)
|
2011-10-19 14:31:15 +02:00
|
|
|
goto out;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
|
|
|
p = strsep(&s[0], ":");
|
|
|
|
if (p != NULL)
|
|
|
|
major_s = p;
|
|
|
|
else
|
2011-10-19 14:31:15 +02:00
|
|
|
goto out;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
|
|
|
minor_s = s[0];
|
|
|
|
if (!minor_s)
|
2011-10-19 14:31:15 +02:00
|
|
|
goto out;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2011-10-19 14:31:15 +02:00
|
|
|
if (strict_strtoul(major_s, 10, &major))
|
|
|
|
goto out;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2011-10-19 14:31:15 +02:00
|
|
|
if (strict_strtoul(minor_s, 10, &minor))
|
|
|
|
goto out;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
|
|
|
dev = MKDEV(major, minor);
|
|
|
|
|
2011-10-19 14:31:15 +02:00
|
|
|
if (strict_strtoull(s[1], 10, &temp))
|
|
|
|
goto out;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2012-03-05 22:15:07 +01:00
|
|
|
disk = get_gendisk(dev, &part);
|
2012-03-05 22:15:08 +01:00
|
|
|
if (!disk || part)
|
2012-03-05 22:15:07 +01:00
|
|
|
goto out;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
spin_lock_irq(disk->queue->queue_lock);
|
|
|
|
blkg = blkg_lookup_create(blkcg, disk->queue, plid, false);
|
|
|
|
spin_unlock_irq(disk->queue->queue_lock);
|
2012-03-05 22:15:07 +01:00
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
if (IS_ERR(blkg)) {
|
|
|
|
ret = PTR_ERR(blkg);
|
|
|
|
goto out_unlock;
|
2011-09-21 10:22:10 +02:00
|
|
|
}
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2010-09-15 23:06:33 +02:00
|
|
|
switch (plid) {
|
|
|
|
case BLKIO_POLICY_PROP:
|
2011-09-21 10:22:10 +02:00
|
|
|
if ((temp < BLKIO_WEIGHT_MIN && temp > 0) ||
|
|
|
|
temp > BLKIO_WEIGHT_MAX)
|
2012-03-05 22:15:07 +01:00
|
|
|
goto out_unlock;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
blkg->conf.weight = temp;
|
|
|
|
blkio_update_group_weight(blkg, temp ?: blkcg->weight);
|
2010-09-15 23:06:34 +02:00
|
|
|
break;
|
|
|
|
case BLKIO_POLICY_THROTL:
|
2010-09-15 23:06:36 +02:00
|
|
|
switch(fileid) {
|
|
|
|
case BLKIO_THROTL_read_bps_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
blkg->conf.bps[READ] = temp;
|
|
|
|
blkio_update_group_bps(blkg, temp ?: -1, fileid);
|
2012-03-05 22:15:07 +01:00
|
|
|
break;
|
2010-09-15 23:06:36 +02:00
|
|
|
case BLKIO_THROTL_write_bps_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
blkg->conf.bps[WRITE] = temp;
|
|
|
|
blkio_update_group_bps(blkg, temp ?: -1, fileid);
|
2010-09-15 23:06:36 +02:00
|
|
|
break;
|
|
|
|
case BLKIO_THROTL_read_iops_device:
|
2012-03-05 22:15:07 +01:00
|
|
|
if (temp > THROTL_IOPS_MAX)
|
|
|
|
goto out_unlock;
|
2012-03-05 22:15:08 +01:00
|
|
|
blkg->conf.iops[READ] = temp;
|
|
|
|
blkio_update_group_iops(blkg, temp ?: -1, fileid);
|
2012-03-05 22:15:07 +01:00
|
|
|
break;
|
2010-09-15 23:06:36 +02:00
|
|
|
case BLKIO_THROTL_write_iops_device:
|
2011-09-21 10:22:10 +02:00
|
|
|
if (temp > THROTL_IOPS_MAX)
|
2012-03-05 22:15:07 +01:00
|
|
|
goto out_unlock;
|
2012-03-05 22:15:08 +01:00
|
|
|
blkg->conf.iops[WRITE] = temp;
|
|
|
|
blkio_update_group_iops(blkg, temp ?: -1, fileid);
|
2010-09-15 23:06:36 +02:00
|
|
|
break;
|
|
|
|
}
|
2010-09-15 23:06:33 +02:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
2011-10-19 14:31:15 +02:00
|
|
|
ret = 0;
|
2012-03-05 22:15:07 +01:00
|
|
|
out_unlock:
|
|
|
|
rcu_read_unlock();
|
2011-10-19 14:31:15 +02:00
|
|
|
out:
|
|
|
|
put_disk(disk);
|
2012-03-05 22:15:07 +01:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If queue was bypassing, we should retry. Do so after a short
|
|
|
|
* msleep(). It isn't strictly necessary but queue can be
|
|
|
|
* bypassing for some time and it's always nice to avoid busy
|
|
|
|
* looping.
|
|
|
|
*/
|
|
|
|
if (ret == -EBUSY) {
|
|
|
|
msleep(10);
|
|
|
|
return restart_syscall();
|
|
|
|
}
|
2011-10-19 14:31:15 +02:00
|
|
|
return ret;
|
2010-04-13 10:05:49 +02:00
|
|
|
}
|
|
|
|
|
2010-09-15 23:06:33 +02:00
|
|
|
static int blkiocg_file_write(struct cgroup *cgrp, struct cftype *cft,
|
|
|
|
const char *buffer)
|
2010-04-13 10:05:49 +02:00
|
|
|
{
|
|
|
|
int ret = 0;
|
|
|
|
char *buf;
|
2012-03-05 22:15:07 +01:00
|
|
|
struct blkio_cgroup *blkcg = cgroup_to_blkio_cgroup(cgrp);
|
2010-09-15 23:06:33 +02:00
|
|
|
enum blkio_policy_id plid = BLKIOFILE_POLICY(cft->private);
|
|
|
|
int fileid = BLKIOFILE_ATTR(cft->private);
|
2010-04-13 10:05:49 +02:00
|
|
|
|
|
|
|
buf = kstrdup(buffer, GFP_KERNEL);
|
|
|
|
if (!buf)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
ret = blkio_policy_parse_and_set(buf, plid, fileid, blkcg);
|
2010-04-13 10:05:49 +02:00
|
|
|
kfree(buf);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:10 +01:00
|
|
|
static const char *blkg_dev_name(struct blkio_group *blkg)
|
|
|
|
{
|
|
|
|
/* some drivers (floppy) instantiate a queue w/o disk registered */
|
|
|
|
if (blkg->q->backing_dev_info.dev)
|
|
|
|
return dev_name(blkg->q->backing_dev_info.dev);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
static void blkio_print_group_conf(struct cftype *cft, struct blkio_group *blkg,
|
|
|
|
struct seq_file *m)
|
2010-04-13 10:05:49 +02:00
|
|
|
{
|
2012-03-05 22:15:10 +01:00
|
|
|
const char *dname = blkg_dev_name(blkg);
|
2012-03-05 22:15:08 +01:00
|
|
|
int fileid = BLKIOFILE_ATTR(cft->private);
|
|
|
|
int rw = WRITE;
|
|
|
|
|
2012-03-05 22:15:10 +01:00
|
|
|
if (!dname)
|
|
|
|
return;
|
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
switch (blkg->plid) {
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_POLICY_PROP:
|
2012-03-05 22:15:08 +01:00
|
|
|
if (blkg->conf.weight)
|
2012-03-05 22:15:09 +01:00
|
|
|
seq_printf(m, "%s\t%u\n",
|
|
|
|
dname, blkg->conf.weight);
|
2010-09-15 23:06:34 +02:00
|
|
|
break;
|
|
|
|
case BLKIO_POLICY_THROTL:
|
2012-03-05 22:15:08 +01:00
|
|
|
switch (fileid) {
|
2010-09-15 23:06:36 +02:00
|
|
|
case BLKIO_THROTL_read_bps_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
rw = READ;
|
2010-09-15 23:06:36 +02:00
|
|
|
case BLKIO_THROTL_write_bps_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
if (blkg->conf.bps[rw])
|
2012-03-05 22:15:09 +01:00
|
|
|
seq_printf(m, "%s\t%llu\n",
|
|
|
|
dname, blkg->conf.bps[rw]);
|
2010-09-15 23:06:36 +02:00
|
|
|
break;
|
|
|
|
case BLKIO_THROTL_read_iops_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
rw = READ;
|
2010-09-15 23:06:36 +02:00
|
|
|
case BLKIO_THROTL_write_iops_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
if (blkg->conf.iops[rw])
|
2012-03-05 22:15:09 +01:00
|
|
|
seq_printf(m, "%s\t%u\n",
|
|
|
|
dname, blkg->conf.iops[rw]);
|
2010-09-15 23:06:36 +02:00
|
|
|
break;
|
|
|
|
}
|
2010-09-15 23:06:33 +02:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
}
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2010-09-15 23:06:33 +02:00
|
|
|
/* cgroup files which read their data from policy nodes end up here */
|
2012-03-05 22:15:08 +01:00
|
|
|
static void blkio_read_conf(struct cftype *cft, struct blkio_cgroup *blkcg,
|
|
|
|
struct seq_file *m)
|
2010-04-13 10:05:49 +02:00
|
|
|
{
|
2012-03-05 22:15:08 +01:00
|
|
|
struct blkio_group *blkg;
|
|
|
|
struct hlist_node *n;
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
spin_lock_irq(&blkcg->lock);
|
|
|
|
hlist_for_each_entry(blkg, n, &blkcg->blkg_list, blkcg_node)
|
|
|
|
if (BLKIOFILE_POLICY(cft->private) == blkg->plid)
|
|
|
|
blkio_print_group_conf(cft, blkg, m);
|
|
|
|
spin_unlock_irq(&blkcg->lock);
|
2010-09-15 23:06:33 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static int blkiocg_file_read(struct cgroup *cgrp, struct cftype *cft,
|
|
|
|
struct seq_file *m)
|
|
|
|
{
|
|
|
|
struct blkio_cgroup *blkcg;
|
|
|
|
enum blkio_policy_id plid = BLKIOFILE_POLICY(cft->private);
|
|
|
|
int name = BLKIOFILE_ATTR(cft->private);
|
|
|
|
|
|
|
|
blkcg = cgroup_to_blkio_cgroup(cgrp);
|
|
|
|
|
|
|
|
switch(plid) {
|
|
|
|
case BLKIO_POLICY_PROP:
|
|
|
|
switch(name) {
|
|
|
|
case BLKIO_PROP_weight_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
blkio_read_conf(cft, blkcg, m);
|
2010-09-15 23:06:33 +02:00
|
|
|
return 0;
|
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
break;
|
2010-09-15 23:06:34 +02:00
|
|
|
case BLKIO_POLICY_THROTL:
|
|
|
|
switch(name){
|
|
|
|
case BLKIO_THROTL_read_bps_device:
|
|
|
|
case BLKIO_THROTL_write_bps_device:
|
2010-09-15 23:06:36 +02:00
|
|
|
case BLKIO_THROTL_read_iops_device:
|
|
|
|
case BLKIO_THROTL_write_iops_device:
|
2012-03-05 22:15:08 +01:00
|
|
|
blkio_read_conf(cft, blkcg, m);
|
2010-09-15 23:06:34 +02:00
|
|
|
return 0;
|
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
break;
|
2010-09-15 23:06:33 +02:00
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int blkio_read_blkg_stats(struct blkio_cgroup *blkcg,
|
2011-05-19 21:38:28 +02:00
|
|
|
struct cftype *cft, struct cgroup_map_cb *cb,
|
|
|
|
enum stat_type type, bool show_total, bool pcpu)
|
2010-09-15 23:06:33 +02:00
|
|
|
{
|
|
|
|
struct blkio_group *blkg;
|
|
|
|
struct hlist_node *n;
|
|
|
|
uint64_t cgroup_total = 0;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
hlist_for_each_entry_rcu(blkg, n, &blkcg->blkg_list, blkcg_node) {
|
2012-03-05 22:15:10 +01:00
|
|
|
const char *dname = blkg_dev_name(blkg);
|
2012-03-05 22:15:09 +01:00
|
|
|
|
2012-03-05 22:15:10 +01:00
|
|
|
if (!dname || BLKIOFILE_POLICY(cft->private) != blkg->plid)
|
2012-03-05 22:15:09 +01:00
|
|
|
continue;
|
|
|
|
if (pcpu)
|
|
|
|
cgroup_total += blkio_get_stat_cpu(blkg, cb, dname,
|
|
|
|
type);
|
|
|
|
else {
|
|
|
|
spin_lock_irq(&blkg->stats_lock);
|
|
|
|
cgroup_total += blkio_get_stat(blkg, cb, dname, type);
|
|
|
|
spin_unlock_irq(&blkg->stats_lock);
|
2010-09-15 23:06:33 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
if (show_total)
|
|
|
|
cb->fill(cb, "Total", cgroup_total);
|
|
|
|
rcu_read_unlock();
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* All map kind of cgroup file get serviced by this function */
|
|
|
|
static int blkiocg_file_read_map(struct cgroup *cgrp, struct cftype *cft,
|
|
|
|
struct cgroup_map_cb *cb)
|
|
|
|
{
|
|
|
|
struct blkio_cgroup *blkcg;
|
|
|
|
enum blkio_policy_id plid = BLKIOFILE_POLICY(cft->private);
|
|
|
|
int name = BLKIOFILE_ATTR(cft->private);
|
|
|
|
|
|
|
|
blkcg = cgroup_to_blkio_cgroup(cgrp);
|
|
|
|
|
|
|
|
switch(plid) {
|
|
|
|
case BLKIO_POLICY_PROP:
|
|
|
|
switch(name) {
|
|
|
|
case BLKIO_PROP_time:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_TIME, 0, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_sectors:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_CPU_SECTORS, 0, 1);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_io_service_bytes:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_CPU_SERVICE_BYTES, 1, 1);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_io_serviced:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_CPU_SERVICED, 1, 1);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_io_service_time:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_SERVICE_TIME, 1, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_io_wait_time:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_WAIT_TIME, 1, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_io_merged:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-23 10:02:19 +02:00
|
|
|
BLKIO_STAT_CPU_MERGED, 1, 1);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_io_queued:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_QUEUED, 1, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
2011-03-22 21:26:54 +01:00
|
|
|
case BLKIO_PROP_unaccounted_time:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_UNACCOUNTED_TIME, 0, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_dequeue:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_DEQUEUE, 0, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_avg_queue_size:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_AVG_QUEUE_SIZE, 0, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_group_wait_time:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_GROUP_WAIT_TIME, 0, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_idle_time:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_IDLE_TIME, 0, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
case BLKIO_PROP_empty_time:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_EMPTY_TIME, 0, 0);
|
2010-09-15 23:06:33 +02:00
|
|
|
#endif
|
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
break;
|
2010-09-15 23:06:34 +02:00
|
|
|
case BLKIO_POLICY_THROTL:
|
|
|
|
switch(name){
|
|
|
|
case BLKIO_THROTL_io_service_bytes:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_CPU_SERVICE_BYTES, 1, 1);
|
2010-09-15 23:06:34 +02:00
|
|
|
case BLKIO_THROTL_io_serviced:
|
|
|
|
return blkio_read_blkg_stats(blkcg, cft, cb,
|
2011-05-19 21:38:28 +02:00
|
|
|
BLKIO_STAT_CPU_SERVICED, 1, 1);
|
2010-09-15 23:06:34 +02:00
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
break;
|
2010-09-15 23:06:33 +02:00
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
static int blkio_weight_write(struct blkio_cgroup *blkcg, int plid, u64 val)
|
2010-09-15 23:06:33 +02:00
|
|
|
{
|
|
|
|
struct blkio_group *blkg;
|
|
|
|
struct hlist_node *n;
|
|
|
|
|
|
|
|
if (val < BLKIO_WEIGHT_MIN || val > BLKIO_WEIGHT_MAX)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
spin_lock(&blkio_list_lock);
|
|
|
|
spin_lock_irq(&blkcg->lock);
|
|
|
|
blkcg->weight = (unsigned int)val;
|
|
|
|
|
2012-03-05 22:15:08 +01:00
|
|
|
hlist_for_each_entry(blkg, n, &blkcg->blkg_list, blkcg_node)
|
|
|
|
if (blkg->plid == plid && !blkg->conf.weight)
|
|
|
|
blkio_update_group_weight(blkg, blkcg->weight);
|
2010-09-15 23:06:33 +02:00
|
|
|
|
|
|
|
spin_unlock_irq(&blkcg->lock);
|
|
|
|
spin_unlock(&blkio_list_lock);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static u64 blkiocg_file_read_u64 (struct cgroup *cgrp, struct cftype *cft) {
|
|
|
|
struct blkio_cgroup *blkcg;
|
|
|
|
enum blkio_policy_id plid = BLKIOFILE_POLICY(cft->private);
|
|
|
|
int name = BLKIOFILE_ATTR(cft->private);
|
|
|
|
|
|
|
|
blkcg = cgroup_to_blkio_cgroup(cgrp);
|
|
|
|
|
|
|
|
switch(plid) {
|
|
|
|
case BLKIO_POLICY_PROP:
|
|
|
|
switch(name) {
|
|
|
|
case BLKIO_PROP_weight:
|
|
|
|
return (u64)blkcg->weight;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
blkiocg_file_write_u64(struct cgroup *cgrp, struct cftype *cft, u64 val)
|
|
|
|
{
|
|
|
|
struct blkio_cgroup *blkcg;
|
|
|
|
enum blkio_policy_id plid = BLKIOFILE_POLICY(cft->private);
|
|
|
|
int name = BLKIOFILE_ATTR(cft->private);
|
|
|
|
|
|
|
|
blkcg = cgroup_to_blkio_cgroup(cgrp);
|
|
|
|
|
|
|
|
switch(plid) {
|
|
|
|
case BLKIO_POLICY_PROP:
|
|
|
|
switch(name) {
|
|
|
|
case BLKIO_PROP_weight:
|
2012-03-05 22:15:08 +01:00
|
|
|
return blkio_weight_write(blkcg, plid, val);
|
2010-09-15 23:06:33 +02:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
BUG();
|
|
|
|
}
|
2010-04-13 10:05:49 +02:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2009-12-03 18:59:42 +01:00
|
|
|
struct cftype blkio_files[] = {
|
2010-04-13 10:05:49 +02:00
|
|
|
{
|
|
|
|
.name = "weight_device",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_weight_device),
|
|
|
|
.read_seq_string = blkiocg_file_read,
|
|
|
|
.write_string = blkiocg_file_write,
|
2010-04-13 10:05:49 +02:00
|
|
|
.max_write_len = 256,
|
|
|
|
},
|
2009-12-03 18:59:42 +01:00
|
|
|
{
|
|
|
|
.name = "weight",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_weight),
|
|
|
|
.read_u64 = blkiocg_file_read_u64,
|
|
|
|
.write_u64 = blkiocg_file_write_u64,
|
2009-12-03 18:59:42 +01:00
|
|
|
},
|
2009-12-03 18:59:49 +01:00
|
|
|
{
|
|
|
|
.name = "time",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_time),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2009-12-03 18:59:49 +01:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "sectors",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_sectors),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-02 00:01:24 +02:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "io_service_bytes",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_io_service_bytes),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-02 00:01:24 +02:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "io_serviced",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_io_serviced),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-02 00:01:24 +02:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "io_service_time",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_io_service_time),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-02 00:01:24 +02:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "io_wait_time",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_io_wait_time),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 08:31:19 +02:00
|
|
|
},
|
2010-04-09 06:14:23 +02:00
|
|
|
{
|
|
|
|
.name = "io_merged",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_io_merged),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 06:14:23 +02:00
|
|
|
},
|
2010-04-09 06:15:10 +02:00
|
|
|
{
|
|
|
|
.name = "io_queued",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_io_queued),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 06:15:10 +02:00
|
|
|
},
|
2010-04-09 08:31:19 +02:00
|
|
|
{
|
|
|
|
.name = "reset_stats",
|
|
|
|
.write_u64 = blkiocg_reset_stats,
|
2009-12-03 18:59:49 +01:00
|
|
|
},
|
2010-10-01 14:49:41 +02:00
|
|
|
#ifdef CONFIG_BLK_DEV_THROTTLING
|
|
|
|
{
|
|
|
|
.name = "throttle.read_bps_device",
|
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_THROTL,
|
|
|
|
BLKIO_THROTL_read_bps_device),
|
|
|
|
.read_seq_string = blkiocg_file_read,
|
|
|
|
.write_string = blkiocg_file_write,
|
|
|
|
.max_write_len = 256,
|
|
|
|
},
|
|
|
|
|
|
|
|
{
|
|
|
|
.name = "throttle.write_bps_device",
|
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_THROTL,
|
|
|
|
BLKIO_THROTL_write_bps_device),
|
|
|
|
.read_seq_string = blkiocg_file_read,
|
|
|
|
.write_string = blkiocg_file_write,
|
|
|
|
.max_write_len = 256,
|
|
|
|
},
|
|
|
|
|
|
|
|
{
|
|
|
|
.name = "throttle.read_iops_device",
|
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_THROTL,
|
|
|
|
BLKIO_THROTL_read_iops_device),
|
|
|
|
.read_seq_string = blkiocg_file_read,
|
|
|
|
.write_string = blkiocg_file_write,
|
|
|
|
.max_write_len = 256,
|
|
|
|
},
|
|
|
|
|
|
|
|
{
|
|
|
|
.name = "throttle.write_iops_device",
|
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_THROTL,
|
|
|
|
BLKIO_THROTL_write_iops_device),
|
|
|
|
.read_seq_string = blkiocg_file_read,
|
|
|
|
.write_string = blkiocg_file_write,
|
|
|
|
.max_write_len = 256,
|
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "throttle.io_service_bytes",
|
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_THROTL,
|
|
|
|
BLKIO_THROTL_io_service_bytes),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "throttle.io_serviced",
|
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_THROTL,
|
|
|
|
BLKIO_THROTL_io_serviced),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
|
|
|
},
|
|
|
|
#endif /* CONFIG_BLK_DEV_THROTTLING */
|
|
|
|
|
2009-12-03 18:59:49 +01:00
|
|
|
#ifdef CONFIG_DEBUG_BLK_CGROUP
|
2010-04-09 06:15:10 +02:00
|
|
|
{
|
|
|
|
.name = "avg_queue_size",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_avg_queue_size),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 06:15:10 +02:00
|
|
|
},
|
2010-04-09 06:15:35 +02:00
|
|
|
{
|
|
|
|
.name = "group_wait_time",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_group_wait_time),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 06:15:35 +02:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "idle_time",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_idle_time),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 06:15:35 +02:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = "empty_time",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_empty_time),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 06:15:35 +02:00
|
|
|
},
|
2010-04-09 06:15:10 +02:00
|
|
|
{
|
2009-12-03 18:59:49 +01:00
|
|
|
.name = "dequeue",
|
2010-09-15 23:06:33 +02:00
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_dequeue),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
2010-04-09 06:15:10 +02:00
|
|
|
},
|
2011-03-22 21:26:54 +01:00
|
|
|
{
|
|
|
|
.name = "unaccounted_time",
|
|
|
|
.private = BLKIOFILE_PRIVATE(BLKIO_POLICY_PROP,
|
|
|
|
BLKIO_PROP_unaccounted_time),
|
|
|
|
.read_map = blkiocg_file_read_map,
|
|
|
|
},
|
2009-12-03 18:59:49 +01:00
|
|
|
#endif
|
2009-12-03 18:59:42 +01:00
|
|
|
};
|
|
|
|
|
|
|
|
static int blkiocg_populate(struct cgroup_subsys *subsys, struct cgroup *cgroup)
|
|
|
|
{
|
|
|
|
return cgroup_add_files(cgroup, subsys, blkio_files,
|
|
|
|
ARRAY_SIZE(blkio_files));
|
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:11 +01:00
|
|
|
static int blkiocg_pre_destroy(struct cgroup_subsys *subsys,
|
|
|
|
struct cgroup *cgroup)
|
2009-12-03 18:59:42 +01:00
|
|
|
{
|
|
|
|
struct blkio_cgroup *blkcg = cgroup_to_blkio_cgroup(cgroup);
|
2009-12-03 18:59:47 +01:00
|
|
|
unsigned long flags;
|
|
|
|
struct blkio_group *blkg;
|
2012-03-05 22:15:03 +01:00
|
|
|
struct request_queue *q;
|
2009-12-04 16:36:42 +01:00
|
|
|
struct blkio_policy_type *blkiop;
|
2009-12-03 18:59:47 +01:00
|
|
|
|
|
|
|
rcu_read_lock();
|
2012-03-05 22:15:11 +01:00
|
|
|
|
2010-05-03 14:28:55 +02:00
|
|
|
do {
|
|
|
|
spin_lock_irqsave(&blkcg->lock, flags);
|
2009-12-03 18:59:47 +01:00
|
|
|
|
2010-05-03 14:28:55 +02:00
|
|
|
if (hlist_empty(&blkcg->blkg_list)) {
|
|
|
|
spin_unlock_irqrestore(&blkcg->lock, flags);
|
|
|
|
break;
|
|
|
|
}
|
2009-12-03 18:59:47 +01:00
|
|
|
|
2010-05-03 14:28:55 +02:00
|
|
|
blkg = hlist_entry(blkcg->blkg_list.first, struct blkio_group,
|
|
|
|
blkcg_node);
|
2012-03-05 22:15:03 +01:00
|
|
|
q = rcu_dereference(blkg->q);
|
2010-05-03 14:28:55 +02:00
|
|
|
__blkiocg_del_blkio_group(blkg);
|
2009-12-03 18:59:42 +01:00
|
|
|
|
2010-05-03 14:28:55 +02:00
|
|
|
spin_unlock_irqrestore(&blkcg->lock, flags);
|
2009-12-03 18:59:47 +01:00
|
|
|
|
2010-05-03 14:28:55 +02:00
|
|
|
/*
|
|
|
|
* This blkio_group is being unlinked as associated cgroup is
|
|
|
|
* going away. Let all the IO controlling policies know about
|
2010-10-01 14:49:44 +02:00
|
|
|
* this event.
|
2010-05-03 14:28:55 +02:00
|
|
|
*/
|
|
|
|
spin_lock(&blkio_list_lock);
|
2010-10-01 14:49:44 +02:00
|
|
|
list_for_each_entry(blkiop, &blkio_list, list) {
|
|
|
|
if (blkiop->plid != blkg->plid)
|
|
|
|
continue;
|
2012-03-05 22:15:03 +01:00
|
|
|
blkiop->ops.blkio_unlink_group_fn(q, blkg);
|
2010-10-01 14:49:44 +02:00
|
|
|
}
|
2010-05-03 14:28:55 +02:00
|
|
|
spin_unlock(&blkio_list_lock);
|
|
|
|
} while (1);
|
2010-04-13 10:05:49 +02:00
|
|
|
|
2009-12-03 18:59:47 +01:00
|
|
|
rcu_read_unlock();
|
2012-03-05 22:15:11 +01:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void blkiocg_destroy(struct cgroup_subsys *subsys, struct cgroup *cgroup)
|
|
|
|
{
|
|
|
|
struct blkio_cgroup *blkcg = cgroup_to_blkio_cgroup(cgroup);
|
|
|
|
|
2010-03-11 00:22:11 +01:00
|
|
|
if (blkcg != &blkio_root_cgroup)
|
|
|
|
kfree(blkcg);
|
2009-12-03 18:59:42 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static struct cgroup_subsys_state *
|
|
|
|
blkiocg_create(struct cgroup_subsys *subsys, struct cgroup *cgroup)
|
|
|
|
{
|
2010-05-07 08:57:00 +02:00
|
|
|
struct blkio_cgroup *blkcg;
|
|
|
|
struct cgroup *parent = cgroup->parent;
|
2009-12-03 18:59:42 +01:00
|
|
|
|
2010-05-07 08:57:00 +02:00
|
|
|
if (!parent) {
|
2009-12-03 18:59:42 +01:00
|
|
|
blkcg = &blkio_root_cgroup;
|
|
|
|
goto done;
|
|
|
|
}
|
|
|
|
|
|
|
|
blkcg = kzalloc(sizeof(*blkcg), GFP_KERNEL);
|
|
|
|
if (!blkcg)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
|
|
|
|
blkcg->weight = BLKIO_WEIGHT_DEFAULT;
|
|
|
|
done:
|
|
|
|
spin_lock_init(&blkcg->lock);
|
|
|
|
INIT_HLIST_HEAD(&blkcg->blkg_list);
|
|
|
|
|
|
|
|
return &blkcg->css;
|
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:12 +01:00
|
|
|
/**
|
|
|
|
* blkcg_init_queue - initialize blkcg part of request queue
|
|
|
|
* @q: request_queue to initialize
|
|
|
|
*
|
|
|
|
* Called from blk_alloc_queue_node(). Responsible for initializing blkcg
|
|
|
|
* part of new request_queue @q.
|
|
|
|
*
|
|
|
|
* RETURNS:
|
|
|
|
* 0 on success, -errno on failure.
|
|
|
|
*/
|
|
|
|
int blkcg_init_queue(struct request_queue *q)
|
|
|
|
{
|
2012-03-05 22:15:13 +01:00
|
|
|
int ret;
|
|
|
|
|
2012-03-05 22:15:12 +01:00
|
|
|
might_sleep();
|
|
|
|
|
2012-03-05 22:15:13 +01:00
|
|
|
ret = blk_throtl_init(q);
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
mutex_lock(&all_q_mutex);
|
|
|
|
INIT_LIST_HEAD(&q->all_q_node);
|
|
|
|
list_add_tail(&q->all_q_node, &all_q_list);
|
|
|
|
mutex_unlock(&all_q_mutex);
|
|
|
|
|
|
|
|
return 0;
|
2012-03-05 22:15:12 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* blkcg_drain_queue - drain blkcg part of request_queue
|
|
|
|
* @q: request_queue to drain
|
|
|
|
*
|
|
|
|
* Called from blk_drain_queue(). Responsible for draining blkcg part.
|
|
|
|
*/
|
|
|
|
void blkcg_drain_queue(struct request_queue *q)
|
|
|
|
{
|
|
|
|
lockdep_assert_held(q->queue_lock);
|
|
|
|
|
|
|
|
blk_throtl_drain(q);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* blkcg_exit_queue - exit and release blkcg part of request_queue
|
|
|
|
* @q: request_queue being released
|
|
|
|
*
|
|
|
|
* Called from blk_release_queue(). Responsible for exiting blkcg part.
|
|
|
|
*/
|
|
|
|
void blkcg_exit_queue(struct request_queue *q)
|
|
|
|
{
|
2012-03-05 22:15:13 +01:00
|
|
|
mutex_lock(&all_q_mutex);
|
|
|
|
list_del_init(&q->all_q_node);
|
|
|
|
mutex_unlock(&all_q_mutex);
|
|
|
|
|
2012-03-05 22:15:12 +01:00
|
|
|
blk_throtl_exit(q);
|
|
|
|
}
|
|
|
|
|
2009-12-03 18:59:42 +01:00
|
|
|
/*
|
|
|
|
* We cannot support shared io contexts, as we have no mean to support
|
|
|
|
* two tasks with the same ioc in two different groups without major rework
|
|
|
|
* of the main cic data structures. For now we allow a task to change
|
|
|
|
* its cgroup only if it's the only owner of its ioc.
|
|
|
|
*/
|
2011-12-13 03:12:21 +01:00
|
|
|
static int blkiocg_can_attach(struct cgroup_subsys *ss, struct cgroup *cgrp,
|
|
|
|
struct cgroup_taskset *tset)
|
2009-12-03 18:59:42 +01:00
|
|
|
{
|
2011-12-13 03:12:21 +01:00
|
|
|
struct task_struct *task;
|
2009-12-03 18:59:42 +01:00
|
|
|
struct io_context *ioc;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
/* task_lock() is needed to avoid races with exit_io_context() */
|
2011-12-13 03:12:21 +01:00
|
|
|
cgroup_taskset_for_each(task, cgrp, tset) {
|
|
|
|
task_lock(task);
|
|
|
|
ioc = task->io_context;
|
|
|
|
if (ioc && atomic_read(&ioc->nr_tasks) > 1)
|
|
|
|
ret = -EINVAL;
|
|
|
|
task_unlock(task);
|
|
|
|
if (ret)
|
|
|
|
break;
|
|
|
|
}
|
2009-12-03 18:59:42 +01:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2011-12-13 03:12:21 +01:00
|
|
|
static void blkiocg_attach(struct cgroup_subsys *ss, struct cgroup *cgrp,
|
|
|
|
struct cgroup_taskset *tset)
|
2009-12-03 18:59:42 +01:00
|
|
|
{
|
2011-12-13 03:12:21 +01:00
|
|
|
struct task_struct *task;
|
2009-12-03 18:59:42 +01:00
|
|
|
struct io_context *ioc;
|
|
|
|
|
2011-12-13 03:12:21 +01:00
|
|
|
cgroup_taskset_for_each(task, cgrp, tset) {
|
2012-01-15 21:24:45 +01:00
|
|
|
/* we don't lose anything even if ioc allocation fails */
|
|
|
|
ioc = get_task_io_context(task, GFP_ATOMIC, NUMA_NO_NODE);
|
|
|
|
if (ioc) {
|
|
|
|
ioc_cgroup_changed(ioc);
|
2012-02-07 07:51:30 +01:00
|
|
|
put_io_context(ioc);
|
2012-01-15 21:24:45 +01:00
|
|
|
}
|
2011-12-13 03:12:21 +01:00
|
|
|
}
|
2009-12-03 18:59:42 +01:00
|
|
|
}
|
|
|
|
|
2012-03-05 22:15:13 +01:00
|
|
|
static void blkcg_bypass_start(void)
|
|
|
|
__acquires(&all_q_mutex)
|
|
|
|
{
|
|
|
|
struct request_queue *q;
|
|
|
|
|
|
|
|
mutex_lock(&all_q_mutex);
|
|
|
|
|
|
|
|
list_for_each_entry(q, &all_q_list, all_q_node) {
|
|
|
|
blk_queue_bypass_start(q);
|
|
|
|
blkg_destroy_all(q);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void blkcg_bypass_end(void)
|
|
|
|
__releases(&all_q_mutex)
|
|
|
|
{
|
|
|
|
struct request_queue *q;
|
|
|
|
|
|
|
|
list_for_each_entry(q, &all_q_list, all_q_node)
|
|
|
|
blk_queue_bypass_end(q);
|
|
|
|
|
|
|
|
mutex_unlock(&all_q_mutex);
|
|
|
|
}
|
|
|
|
|
2009-12-04 16:36:42 +01:00
|
|
|
void blkio_policy_register(struct blkio_policy_type *blkiop)
|
|
|
|
{
|
2012-03-05 22:15:13 +01:00
|
|
|
blkcg_bypass_start();
|
2009-12-04 16:36:42 +01:00
|
|
|
spin_lock(&blkio_list_lock);
|
2012-03-05 22:15:04 +01:00
|
|
|
|
|
|
|
BUG_ON(blkio_policy[blkiop->plid]);
|
|
|
|
blkio_policy[blkiop->plid] = blkiop;
|
2009-12-04 16:36:42 +01:00
|
|
|
list_add_tail(&blkiop->list, &blkio_list);
|
2012-03-05 22:15:04 +01:00
|
|
|
|
2009-12-04 16:36:42 +01:00
|
|
|
spin_unlock(&blkio_list_lock);
|
2012-03-05 22:15:13 +01:00
|
|
|
blkcg_bypass_end();
|
2009-12-04 16:36:42 +01:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkio_policy_register);
|
|
|
|
|
|
|
|
void blkio_policy_unregister(struct blkio_policy_type *blkiop)
|
|
|
|
{
|
2012-03-05 22:15:13 +01:00
|
|
|
blkcg_bypass_start();
|
2009-12-04 16:36:42 +01:00
|
|
|
spin_lock(&blkio_list_lock);
|
2012-03-05 22:15:04 +01:00
|
|
|
|
|
|
|
BUG_ON(blkio_policy[blkiop->plid] != blkiop);
|
|
|
|
blkio_policy[blkiop->plid] = NULL;
|
2009-12-04 16:36:42 +01:00
|
|
|
list_del_init(&blkiop->list);
|
2012-03-05 22:15:04 +01:00
|
|
|
|
2009-12-04 16:36:42 +01:00
|
|
|
spin_unlock(&blkio_list_lock);
|
2012-03-05 22:15:13 +01:00
|
|
|
blkcg_bypass_end();
|
2009-12-04 16:36:42 +01:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkio_policy_unregister);
|